fever' - dm5migu4zj3pb.cloudfront.netjapanese (4). more recently, hullinghorst and 1 this work...

THE RENAL LESION IN EPIDEMIC HEMORRHAGICFEVER'

By JEAN OLIVER AND MURIEL MACDOWELL

(From the Renal Research Unit, Overlook Hospital, Summit, N. J.)

(Submitted for publication June 21, 1956; accepted September 5, 1956)

Structural and functional derangements in thekidney during the course of Epidemic Hemor-rhagic Fever run as a continuing thread of causeand effect through a complex of physiological andbiochemical disturbances that has perhaps noparallel among the acute infectious diseases. Soclosely knit are these renal aberrations in the fabricof the clinical syndrome that investigators havefound it impractical to disentangle the element of"renal failure" from the related phenomena ofshock, hypotension or hypertension and electrolyteor water imbalance, and so simply state as theirconclusion that "renal complications" are presentin all examples of the disease and at every stage ofits progress (1).

The general nature of the renal lesion has beenestablished. In its clinical aspects, it is an ex-ample of "acute renal failure" with the typicalcharacteristics of proteinuria, oliguria and late diu-resis. From the physiological viewpoint, the in-vestigations of Froeb and McDowell (2) and ofSyner and Markels (3) have demonstrated similardisturbances of renal blood flow which occur eitheras a part, or at times independently, of the generalcirculatory collapse which so frequently ensues inthe early phases of the disease.

The pathological alterations that occur in thekidneys have been described by the Russian in-vestigators who first recognized the disease ineastern Siberia (4),2 and in later studies by theJapanese (4). More recently, Hullinghorst and

1 This work was done under the auspices of the Com-mission on Hemorrhagic Fever, Armed Forces Epidemio-logical Board, and was supported by the Office of theSurgeon General of the Army, and in part by a researchgrant (H-1515-C2) from the National Heart Institute, ofthe National Institutes of Health, Public Health Service,and the Life Insurance Medical Research Fund.

2 Though the earlier reported cases of EHF were lim-ited to Manchuria and eastern Siberia, later observationshave shown that a very similar disease occurs in theBalkans and Czechoslovakia. The mild "epidemic ne-phropathy" observed in Finland, Sweden and Norway re-sembles EHF in some ways, but an identity has not beenproved (5).

Steer (6), Steer (7) and Lukes (8) have re-ported on material obtained during epidemicsamong American troops in Korea. Though theresemblance of the histological picture to thatwhich is found in other forms of the acute renalnecrosis associated with traumatic and toxic in-jury has impressed pathologists, certain charac-teristics of the structural lesions of EHF are sodistinctive that in the absence of a demonstrableetiological agent they have become the final cri-terion for its diagnosis.

A general discussion and detailed analysis ofthe clinical, physiological and biochemical de-rangements that occur in EHF have appeared ina Symposium published under the editorial di-rection of Dr. David P. Earle (9). Against thisbackground an attempt will now be made to inte-grate the development of the structural aspects ofthe renal lesion with the progress of its functionaldisturbances.

MATERIAL AND METHODS

The clinical data and pathological material of thisstudy have been taken from the records of the Commis-sion of Epidemic Diseases and the collections of theArmed Forces Institute of Pathology.

Thirty-nine fatal cases of EHF which occurred dur-ing the Spring and Fall epidemics of 1952 and 1953among American troops in Korea have been examined.An analysis of thirty-four of these appears in the Sym-posium (9) ; the case numbers of this earlier study havebeen retained, and though much of its substance has beenincorporated into our report further pertinent informa-tion may be derived from this source, in particular fromTable VIII, page 636.

The distribution of the cases in regard to the phaseof the disease in which the individual died was asfollows:

Hypotensive phase:(Primary shock 12)(Transition shock 7)

Oliguric phase:Diuretic phase, 1 to 16 days of diuresis:Convalescence, 149 days from onset:

19

1091

3999

JEAN OLIVER AND MURIEL MACDOWELL

The method of study was as follows. The autopsyprotocols and clinical records were first reviewed. Sec-tions of all the tissues from each autopsy were then ex-amined for whatever relation they might bear to therenal lesion and the sections of kidney studied in detailby the conventional methods of histological examination.In the material of this preliminary survey 10 per centformalin and Zenker's solution had been the fixative,and hematoxylin and eosin the stain; in subsequent ex-aminations the Masson stain, iron hematoxylin and eosin,Acid Fast Green and certain histochemical procedureswere used.

After this examination had established the general na-ture of the renal lesion, microdissection of the kidneytissue as previously described (10) was done to observethe detail and topographical relations of abnormalities inthe nephrons. The dissected nephrons were stainedwith iron hematoxylin, and camera lucida drawings orphotomicrographs were made for permanent record. Aswill appear in the descriptions of the structural altera-tions, a continuous comparison was made during theprogress of the work between the appearance of the lesionsas they were revealed by the two methods of examination.

Selected examples illustrating the course and develop-ment of the pathological alterations appear in our il-lustrations. It is perhaps unnecessary to point out thatconsiderations of expense and space have made necessarya considerable restriction in the presentation of our find-ings. During the dissection of each case hundreds ofnephrons were examined, many were drawn by means ofthe camera lucida, and a lesser number photographed atmagnifications of up to 200 times. Of the latter onlytypical examples can be published, and at great reduction;the original of Plate I, for example, is composed of 97individual 4 by 5 microphotographs which, when assem-bled to show the complete nephron, cover a space of 4 by9 feet; much detail is therefore lost in the process ofpublication.

In the exposition of our findings the nomenclature usedby clinical investigators to designate the course of thedisease has been adopted. A brief review of the syndromeof EHF may orient the reader for the more detailed de-scription of the physiological, clinical and pathologicalcorrelations that are to follow.

THE CLINICAL SYNDROMEOF EHF

During all but the first of the five successivephases, Febrile, Hypotensive, Oliguric, Diureticand Convalescent, into which the clinical courseof EHF has been divided, disturbances of renalactivity are prominent. The Febrile Phase is char-acterized by the usual fever, chills, headache andnausea of an acute infectious disease, and endsaround the fifth day by defervescence.

With passage into the Hypotensive Phase, theearlier manifestations of disturbance in the periph-eral circulation, such as the intense flush of the

face and the injection of the pharynx, become de-finitive in the appearance of petechiae. Concomi-tantly, there occurs a rise in the hematocrit which,along with direct determinations of decreased cir-culating volume, suggests that plasma fluid is leak-ing through damaged capillaries into tissue spaces.As a result, the blood pressure falls and shock en-sues.

These conditions persist during the HypotensivePhase, in which death from primary shock ac-counted for one-third of all fatalities. Proteinuriais now massive and the urinary output is irregu-larly decreased. Blood urea nitrogen rises and,with passage from the Hypotensive Phase, theoliguria, previously irregularly manifested, be-comes permanently established to mark the thirdstage of the disease.

The elevated hematocrit of the HypotensivePhase has now decreased toward normal, at timesabruptly, but more usually requiring one to threedays before it becomes stabilized at its originallevel. During this Transition Period from theHypotensive to the Oliguric Phase, a circulatorycollapse of a nature somewhat different from the"primary shock" of the earlier period may occur;in "transition shock" the observation that totalserum proteins do not change as the hematocritfalls, and direct measurements of an increasingcirculatory volume, suggest that fluid previouslylost to the extravascular spaces is returning.

During the Phase of Established Oliguria whichfollows, hypertension is commonly observed, but amore remarkable circulatory disturbance duringthis period is the development in certain cases ofa relative hypervolemia along with hemorrhagicphenomena during which the complications of pul-monary edema, hemorrhage or convulsions may befatal.

After 3 to 5 days of oliguria, the Diuretic Phasebegins; a daily output as high as 18 L. has beenrecorded. During this phase abnormalities inelectrolyte and water balance may develop and acondition of "limited homeostasis" be establishedin which secondary shock and pulmonary edemaoccur. Judging from the evidence of increasingclearances there then follows an improvement inthe renal circulation which gradually reaches nor-mal in the following weeks. The BUN has fal-len, proteinuria has disappeared, and in the Con-valescent Period most patients after two months

100

RENAL LESION IN EPIDEMIC HEMORRHAGICFEVER

PHYSIOLOGICAL DISTURBANCESIN EPIDEMIC HEMORRAGICFEVER

PHASE FEBRILE TCNSIV OLIGURIC | DIURETIC

DAYS I, ,3, 5S 7, 9, 11, 13, ,IS, ,17, ,9

FEVER__

SHOCK̂aa

Petechide |5HEMORRHAGES

RELATIVEHYPERVOLEMIA

FLUID & ELECTRO-LYTE IMBALANCE

RENAL FAILURE

BLOODPRESSURE

URINE FLOW _

CLEARANCESPAN-CR

BLOODUREAN ITROGEN

PROTEINURIA

HEMATOCRIT

PLATLETS

SERUMPROTEINS

RENALVASCULAR

DISTURBANCESSUBCORTICALCONGESTION

STASIS-HEMORRHAGE

X TUtJULAR NECROSIS

RESOLUTION

RECOVERY

TEXT FIG. 1. COURSEOF CLINICAL AND LABORATORYOBSERVATIONS INSEVERE EPIDEMIC HEMORRHAGICFEVER

Modified from Sheedy and his associates (1).

can concentrate their urine to a specific gravity of1.023 or better.

Text Figure 1, which is a modification of Fig-ures 2 and 3 of Sheedy and his co-workers (9),shows the clinical manifestations and laboratorymeasurements of EHFin graphic form. The low-est level illustrates the occurrence and relative in-tensity of certain pathological phenomena that wereobserved in the kidney, and will be explained byour later descriptions.

Although, as previously stated, the renal lesionis to be described as it develops in the consecutiveclinical stages of the disease, the procedure of pre-senting the individual cases has been to arrange

them in a series based on the evolution of thepathological lesions. As these lesions are of variednature, involving vascular, interstitial and paren-chymal reactions which may develop at differingrates of intensity, such a series can be only ap-proximate in arrangement.

THE RENALLESION IN THE FEBRILE AND

HYPOTENSIVEPHASE

(The Prehemorrhagic Stage of SubcorticalCongestion)

The earliest view of the renal lesion in EHFencountered in this series was obtained in Case

101


1 ~~~~~~~.;. . . . 0. . .. . .l . ...//////////////////////TIME IN HOURS 5 10 15

TEXT FIG. 2. CASE 38-DEATH IN PRIMARY SHOCKPHASEOF HEMORRHAGICFEVER

Note initial response to continuous intravenous pressor therapy but increasing requirements during sub-sequent course, and death after 18 hours' observation despite 400 cc. concentrated human albumin, Pitressin®and adrenal cortical extract. Note also shock at eleventh hour when L-arterenol infusion infiltrated intosubcutaneous tissue. Blood pressure could be measured only by palpatory method after fifth hour. Markedretroperitoneal edema was present at autopsy.

Text Figures 2, 4, 6, 7 and 9 and their legends are from the Symposium on Epidemic Hemorrhagic Fe-ver (9) and are published by permission of author and publisher from the American Journal of Medicine,1954, 16, 617.

38, who died on the fourth day of his illness.After three days of severe headache, chills, fever,nausea and vomiting, he entered the hospital witha temperature of 103°, a blood pressure of 100/70mm. Hg, a 1 + proteinuria and a hematocrit of 59per cent. Physical examination showed intenseinjection of the conjunctivae and soft palate, butno petechiae were present. During the examina-tion he suddenly vomited and collapsed; his bloodpressure fell to 58/0 mm. Hg. In spite of continu-ous treatment,3 including the infusion of 4 units ofserum albumin which produced intermittent re-

3 In this and the cases which follow, only that thera-peutic procedure has been given for which there is evidencethat it may produce a structural alteration in the kidney,namely, the administration of concentrated human serumalbumin (1 unit= 100 cc. 25 per cent salt-poor serumalbumin). Many individuals also received whole bloodtransfusions, glucose and nor-adrenalin infusions andoccasionally Dextran. A summary of these proceduresmay be found elsewhere (9).

establishment of blood pressure (Text Figure 2),he expired in shock 18 hours after admission.During this period, with a fluid intake of 2864 cc.he passed 415 cc. of urine; the last sample beforedeath showed 2 + protein and the terminal BUNwas 42.3 mg. per cent.

At autopsy no petechiae were present. Therewas a considerable amount of retroperitonealedema lightly stained with the dye, T-1824, thathad been injected for determination of blood vol-ume. The kidneys were not swollen (175 to 200gm.) and the prosector noted that though the cor-tex was more sharply differentiated from themedulla than normally, there was not the intensecongestive zone which he was accustomed to findas typical of EHF. There were, however, theusual hemorrhages in the right atrium and an ex-tensive hemorrhagic infiltration of the anteriorlobe of the pituitary.

On histological examination the glomerular

102


tufts were essentially normal; the capillaries con-tained red blood cells in normal numbers andthere was precipitated granular material in Bow-man's space. Cross sections of proximal convolu-tions in the cortex were lined with their typicalepithelium which showed some evidence of apicalswelling of its cells (Figures 1 and 2); the cyto-plasm of these cells was normal. The lumens ofcross sections of ascending limbs in the medullaryrays and the distal convolutions in the cortexwere dilated and appeared empty (Figure 2).

In the subcortical medullary zone were scatteredareas of dilatation of intertubular vessels whichwere crowded with discrete red blood cells (Fig-ure 3). The thin outlines of the intact walls ofthese vessels were clearly evident and there wasno escape of blood into the interstitial tissue (Fig-ure 4). In these areas of congestion the epi-thelium and basement membranes of the "straighttubules", terminal proximal convolutions and as-cending limbs of Henle's loop, though somewhatcompressed, were well preserved. Throughoutthe deeper portions of the medulla were irregu-larly distributed patches of intertubular edema,but little evidence of congestion and none of hem-orrhage (Figure 5). The tubules of the collect-ing system were normal and contained few casts.

Microdissections of the renal tissue showed es-sentially normal configurations in the nephrons,except for some slight pressure effect on thoseterminal segments of the proximal convolutionsand ascending limbs of Henle's loop which passedthrough areas of congestion.

Plate I, A to F, shows a complete nephronstained with iron hematoxylin. Apart from thelocalized effects just noted, the general configura-tions of this nephron are so slightly disturbedthat it may serve as a control example of the ap-pearance of a normal nephron prepared by ourtechnique for comparison with the damaged speci-mens that are to follow. In all the preparations,the intense black staining of the proximal con-volution is the reaction of the mitochondria whichfill the cytoplasm of its epithelial cells in decreas-ing amount as one departs from the glomerulus;these cellular organelles no longer exist in theiroriginal rod-like form but, apparently as a resultof the maceration in HCl, are resolved into fine"granules". As in histological preparationsstained with iron hematoxylin, the nuclei do not

stain heavily but appear as light round spots onthe dark background of the cytoplasm, except inthe premitotic phases of regenerative proliferationwhen their excessive and hyperchromatic chro-matin makes them clearly visible. It should berecalled that the tubule is being viewed throughits entire thickness, 60 to 70,u in the case of theproximal convolution, so that cellular detail is ofnecessity somewhat clouded; on the other handthis increase in the absolute amount of tissue un-der observation at times permits the recognitionof cytoplasmic lesions which are not appreciablein thin histological sections. Since a familiaritywith the microscopic appearance of dissectednephrons is essential to the recognition and theinterpretation of the alterations that are to be seenin the abnormal specimens, a detailed descriptionof Plate I will be given.

The glomerulus, due to the bulk of its tissues,shows little cellular detail; its size and shape arenormal, however, and the relation of the moreopaque tuft to Bowman's capsule is evident.

The proximal convolution is of normal configur-ation and diameter, except in its terminal medul-lary portion, which, passing into an area of con-gestion in the subcortical outer stripe of the outerzone of the medulla, is somewhat compressed.The cytoplasmic pattern is in general well pre-served; the normal mitochondrial gradient isshown by the gradual decrease in the intensity ofthe reaction to iron hematoxylin, and the clear,vacuolar, round nuclei, more visible as the inten-sity of the mitochondrial staining lessens, arenormally distributed in the tubule wall. At ain Plate IB, the medullary segment of the con-volution entered an area of intense intertubularcongestion which lay just below the cortex. Thetubule is definitely narrowed, and although itscells are in general well preserved, slight irregu-larities are noted toward the end of the segment,and a scattering of intracellular detritus indicatessome epithelial damage.

The greater part of the thin portion of Henle'sloop, though successfully dissected and mounted,was swept away during the process of staining(Plate ID); a short remnant, somewhat dilated,remains (Plate IE) and passes abruptly into thethick portion of the limb which, continuing todescend, forms the loop and turns towards thecortex. In the area of congestion, lying adjacent

103


to the compressed terminal segment of the prox-imal convolution, the cellular pattern of the as-cending limb is slightly disturbed and there issome deeply stained intracellular detritus. Abovethis point and extending high into the cortex theascending limb is moderately dilated; its cellularcomponents appear essentially normal.

The distal convolution (Plate IF) is much moreand very unevenly dilated, so that there is a con-sequent patchy thinning of its wall which accountsfor the irregularity of its staining; there are nocoagula or casts in its distended lumen. The junc-ture of its terminal portion, the connecting tubule,with that of a neighboring nephron is not dilated.

Dissected collecting tubules, from cortex todeep medulla, were entirely normal; those whichpassed through the areas of congestion were un-affected by the dilatation of the intertubular ves-sels which surrounded them.

The nephron shown in Plate I was typical ofthose which were involved in an area of subcorti-cal congestion, and none of these showed moresevere alterations. Those which passed outsideor between the scattered areas of congestion, at arough estimate perhaps one-half of all nephrons,showed no visible abnormalities in their structure.

Text Figure 3 is a graphic representation of arenal lobule from this kidney. The nephronsshown are camera lucida tracings of dissected spec-imens. They and the areas of intertubular con-gestion are arranged as they were observed to liein the kidney during the course of dissection andas similar structures and relations appeared in his-tological sections. For the sake of clarity only afew nephrons are shown and these are widelyspaced, but otherwise the figure presents not aninterpretive diagram but a reconstruction ofactuality.

It is clear that the architecture of the kidneyand the topographical relation of its constituentnephrons have not been greatly affected by thevascular disturbance in the subcortical medulla.The dilatation of the distal portions of the neph-ron lying in the cortex, including ascending limbsof Henle's loop and distal convolutions, is, how-ever, definite if not striking (Figure 2 and PlateIF); this dilatation was associated with a zoneof moderate pressure effect from the congestionin the sub-cortical medulla, a relation that will

assume greater significance with the progress ofthe renal lesion.

* * * * *

Since the renal lesions become increasingly com-plex with the progress of the disease, a summaryof the structural findings and a discussion of theirsignificance along with a reference to pertinentliterature will be given at the end of each clinicalperiod. This procedure may make for reiteration,but will perhaps be less burdensome to the readerthan the continuing back reference to minutiaelong since described which would be required ifcritical considerations were postponed to a finalconclusion.

The kidney of this individual in whom defer-vescence was interrupted by sudden death fromprimary shock may be taken as illustrative of therenal status during the late Febrile Phase at itspassage into the Hypotensive Phase. The lesioncan be briefly summarized. There is a markedcongestive hyperemia localized to the subcorticalzone of the medulla, occurring not diffusely butin irregular patchy areas, with no intertubularhemorrhage. There are some slight localizedpressure effects of this congestion, both of com-pression and dilatation, on the tubules of neph-rons that by chance pass through it, but no struc-tural evidence of general cellular damage in anyof the tubules. The tubular passages are clear;there are few casts.

The correlation of these structural changes withthe clinical and physiological data of the periodseems reasonably clear; all lines of evidence in-dicate the beginning of widespread disturbances inthe peripheral vascular bed which are character-ized first by dilatation and then by increased per-meability of capillaries and venules.

In the kidneys, the functional effects of thesedisturbances are reflected in the clearance exami-nations of Froeb and McDowell (2) and of Synerand Markels (3), who found not a lessening ofrenal blood flow, as is observed in classical acuterenal failure, but either its full maintenance or ac-tual increase. The histological sections show thedistribution of this abundant blood flow in thecongestive hyperemia of the subcortical medulla.To a consideration of the reason for this localiza-tion we shall return in a later description of theprogress of the vascular lesion. Neither histo-

104

TEXT FIG. 3. A RENAL LOBULE FROMTHE EARLY HYPOTENSIVE PHASE OF EPIDEMIC HEMORRHAGICFEVER-CASE 38

Except for some dilatation of the distal convolutions the topography of the nephrons is normal. There is an ir-regular intense congestion of the subcortical medulla but no intertubular hemorrhage. Magnification 18 X.


logical sections nor dissections showed frankstructural alterations in the nephrons, a findingwhich correlates with the clinical observation thatthere was only a moderate degree of proteinuriaand elevation of BUN.

It may be concluded, therefore, from both thestructural and functional findings, that the vascu-lar bed of the kidney shares in the general disturb-ance of the peripheral circulation, the clinical evi-dences of which were vividly present in the in-tense flush of the skin of the face and chest andmucous membranes. All these phenomena to-gether constitute the first evidences of generalizedcapillary damage, and in the case described theyterminated abruptly in primary shock when plasmaescaped into the extravascular spaces, a phenome-non indicated by the clinical observation of a ris-ing hematocrit and confirmed by the pathologicalfinding of massive retroperitoneal edema. It isnoteworthy that the ultimate lesion of vasculardamage was as yet not developed in either theskin or the kidney, for neither petechiae werepresent nor intertubular hemorrhage. This grav-est aspect of the vascular lesion was, however,present in two areas, the right auricle and theanterior hypophysis; the small vessels of theseregions are apparently peculiarly susceptible sincein every case in this series they were the seat ofextensive hemorrhage.

* * * * *

The remaining cases of those dying in primaryshock may be considered to have definitely en-tered the Hypotensive Phase, a conclusion sup-ported not only by the evolution of the clinicalsymptomatology but by the progress of the renallesion.

The first of these, Case 9, entered hospital witha 3 + proteinuria 3% days after a typical febrileonset. The next morning the hematocrit roseto 61.9 per cent, the blood pressure fell, and thepatient went into shock; in spite of the administra-tion of 1200 cc. of blood and 3 units of serumalbumin, he died in shock the morning of the fol-lowing day. There was no oral intake of fluid,and there was a urinary output of 760 cc.; theterminal temperature was 1060; the BUNon theday before death was 16.8 mg. per cent. At au-topsy, marked retroperitoneal edema, hemorrhagesin the right auricle and an extensive hemorrhagic

infiltration in the anterior pituitary were present.The kidneys were swollen; their medullae aredescribed as "reddish purple".

Histological examination showed an intense,patchy engorgement of the subcortical zone ofthe medulla. In certain areas the capillaries be-tween the terminal proximal convolutions andascending limbs of Henle's loop were dilated tothe diameter of the adjacent tubules and crowdedwith red blood cells which in some vessels werediscrete and in others packed and fused into ahyaline mass (Figure 6). In spite of this dis-tention the walls of the vessels were not greatlythinned and their intact endothelial cells wereplainly visible. The tubules surrounded by thesevessels were compressed by the resulting tensionand their epithelium showed a moderate degreeof protoplasmic disturbance.

Not only were the small vessels of the outermedulla distended with red cells, but the capil-laries of the swollen glomerular tufts throughoutthe cortex were greatly dilated and packed withred blood cells (Figure 7). There was con-siderable precipitated granular material in Bow-man's space. Between the cross sections of thecortical proximal convolutions the intertubularcapillaries were not particularly prominent.Both the large veins and the arteries of the cortico-medullary junction were, however, engorged withblood.

Whereas the epithelium of the proximal con-volutions of the previously described case was es-sentially normal, there was now present diffuselythroughout the cortex a definite swelling of theepithelial cells, and their protoplasm showed ir-regular variations in density without, however, theoccurrence of definite vacuoles. Although the epi-thelial protoplasm was increased in its granu-larity, no large, discrete hyaline droplets werepresent.

* * * * *

In this individual, who died somewhat later inthe course of the disease than the previous case,a definite increase in renal structural change isapparent. The subcortical congestion is more in-tense, though the excess blood is quite clearly con-tained within the dilated straight vessels of theouter medulla. There is also a marked congestionof glomerular capillaries and somewhat less of the

106


cortical intertubular network, and both the largearteries and veins are greatly distended. One hasthe impression, since there was no general visceralcongestion observed post mortem to account forthe excess blood in the kidney, that dilatation ofan atonic intrarenal vascular bed had, by a reduc-tion in peripheral resistance, flooded the entireorgan.

The other advance in the development of therenal lesion concerned the epithelium of the proxi-mal convolutions. Throughout their length therewas noted a frank protoplasmic disturbance.

* * * * *

Case 14 was a man admitted on the third dayof his illness with a temperature of 1050 and ablood pressure of 128/70 mm. Hg; on the thirdday in the hospital it fell suddenly to 90/74.Three units of serum albumin were given. Theurine showed a 3 + proteinuria; on an intake of800 cc. his urinary output was 2280 cc.; the BUNwas 24 mg. per cent.

The renal lesion may be briefly summarized assimilar to the preceding case, except that the con-gestion was more closely limited to the subcorticalmedulla, and the protoplasmic swelling of the cellsof the proximal convolutions was more pro-nounced.

A further development of the disturbance in thecortex of the kidney was noted in Case 20. Thisman died of primary shock on the seventh day ofhis illness in spite of the repeated administrationof infusions over a period of 3 days of serum al-bumin to a total amount of 8 units. There wasno fluid intake other than by these infusions, andthere was a urinary output in 4 days of 1060 cc.His urine was negative for protein for the first 5days; then a trace appeared increasing in a fewhours to 4 +. BUN on entrance was 22.9 mg.per cent, and on the day before death, 85.0 mg.per cent.

Histological examination showed the subcorti-cal congestion and in the cells of the proximalconvolutions the protoplasmic disturbance that haspreviously been described; the latter had advancedto the formation of definite vacuoles. In addition,irregular areas of edema were scattered through-out the cortex, the cross sections of tubule beingwidely separated by collections of fluid (Figure8). The cells of the proximal convolutions in

the congested subcortical zone were filled withlarge hyaline droplets.

Case 26 was admitted on the third day of hisillness with a temperature of 1050; his urine wasfree of protein. Shock intervened almost im-mediately and continued in spite of the ad-ministration of 16 units of serum albumin over aperiod of 3 days. On the fifth day of his illness,the urine showed a 4 + proteinuria. His urinaryoutput during hospitalization was 460 cc.; theBUN rose to 69 mg. per cent; he died in shockon the sixth day of the disease.

At autopsy, massive retroperitoneal edema wasnoted. The prosector was impressed by the"dark appearance" of the subcortical congestion.Histological examination showed this congestionin the form of dilated but intact intertubular ves-sels; congestion was not excessive in the cortex.The glomeruli were normal but the capsular spacescontained considerable amounts of precipitatedprotein. The most pronounced abnormalitynoted was the marked swelling of the tubular epi-thelium in the cortical portions of the proximalconvolutions. Vacuolization was extreme, large,clear, rounded spaces extending through the en-tire depth of the epithelial cells (Figure 9).The terminal segments of proximal convolutionsthat lay in the subcortical zone of dilated intertu-bular vessels were less swollen; their lumens con-tained considerable amounts of clear, coagulatedprotein in which were enclosed bubble-like spacesand many desquamated cells with pyknotic nu-clei. The cells which still lined the compressedtubule were filled with hyaline droplets.

The increasing alteration of the epithelium ofthe proximal convolution described in the pre-ceding three cases appeared at a maximum in anindividual (Case C) who, entering the hospitalwith a temperature of 105.80 on the third day ofhis illness, went into shock on the followingmorning. In spite of treatment, including theadministration of 3 units of serum albumin and 3units of Dextran, he succumbed with pulmonaryedema on the eighth day. The maximum hemato-crit on the first day of shock was 64.2 per cent,falling on the seventh day to 32.9 per cent; in 5days the intake of fluid was 10,505 cc., the uri-nary output, 1300 cc. A 4 + proteinuria de-veloped and the BUN rose from 21.4 mg. per

107


cent on the fourth day reaching 250 mg. per centon the day preceding death.

At autopsy the kidneys showed a moderatedegree of cortico-medullary congestion. Retro-peritoneal edema was present.

On histological examination the usual dilationof the intertubular capillaries in the cortico-medul-lary zone was observed. The most striking alter-ation was an extreme swelling of the epitheliumof the proximal convolutions. The finely granu-lar cytoplasm of the cells appeared as if in-flated, resulting in a protrusion of their apicesinto the lumen of the tubule; the brush border was,however, preserved, as were the nuclei which weredisplaced towards the base of the cells. This in-tense swelling of tubular epithelium was sharplylimited in its distribution to the proximal convolu-tion, the epithelium of the ascending limbs beingessentially normal (Figure 10). It extended intoBowman's space, which was in part obliteratedby the protrusion of the foamy cytoplasm of thecells of its parietal layer; the epithelium coveringthe tuft was not involved (Figure 11). No hya-line droplets were observed in the altered epi-thelium of the proximal convolutions.

* * * * *

The evolution of the renal lesion as shown bythe development of structural alterations in thepreceding four cases is evidenced by two phenom-ena; increase in the intertubular congestion of thecortico-medullary zone, and a swelling of theproximal convolutions. The latter has the histo-logical appearance of an increased hydration ofthe protoplasm of the epithelial cells which reachedthe point of vacuolization.

The increase in subcortical congestion took theform of a spread in the extent of involved areasrather than in a progressive dilatation of indi-vidual channels. In the less far advanced lesion,the groups of dilated intertubular vessels werewidely spaced on a background of less affectedmedulla and had therefore the normal distributionof the horsetail-like strands of capillaries whichresult from the subdivision of each of the large,straight efferents from juxta-medullary glomeruliin the lower cortex (11). Originally formingisolated clusters, the addition of more dilatedcapillary radicles at their periphery resulted intheir fusion to a more or less continuous band

which in the end separated cortex from medulla.The congested zone thus occupied the entire outerstripe and extended into the inner stripe of theouter zone of the medulla. The altered vascularpattern is therefore that which might result froma continuing flow of blood into abnormal inter-tubular vessels that had lost their tonicity; thelocalized distinctiveness of this pattern will beconsidered later.

Swelling of the tubular epithelium and proto-plasmic disturbances in its cytoplasm have not re-ceived great attention in considerations of thepathogenesis of acute renal failure and are fre-quently dismissed under the stock description of"cloudy swelling" with the implication that theimportance of this at times dubious histologicallesion is not great. Funck-Brentano has correctedthis arbitrary view in his thesis on the physiopath-ological mechanisms of anuria in the acute neph-ropathies (12). In experiments on rabbits, oli-guria and anuria were found to follow the infu-sion of serum containing 50 per cent glucose,whereas no such effect was produced by infusionsof serum containing 9.9 per cent NaCl. Sectionsof the anuric kidneys showed a marked swellingof the epithelial cells of the proximal convolutionsvery similar to the illustrations of the swollenepithelia in Figures 10 and 11. He concludes thatthe swelling is due to cellular hyperhydrationwhich is derived as an effect of alterations in theosmotic pressure of the plasma; this he demon-strates by a corrected freezing point determina-tion. Similar changes in the freezing point ofthe plasma were noted in human clinical examplesof the oliguric phase in acute renal failure. Thereare no data available on plasma osmolarity inEHF, but the results of Funck-Brentano's experi-ments in the production of epithelial lesions bymodification of osmotic factors and consequentshifts in hydration of cells are remarkably similarto those seen in EHF.

In Case 20 of the preceding group there is notedfor the first time in our description of the renallesion the occurrence of hyaline droplets in theepithelial cells of the proximal convolutions, andthese objects will be found in many examples thatfollow. A consideration of their significance inthis report would lead to a considerable digression.We shall therefore only state that in our opinionthese droplets do not represent an effect of the

108


disease processes, though these do modify theconditions of their formation, but rather are theresult of a therapeutic procedure, namely, the ad-ministration of large amounts of human serum al-bumin. The hyaline droplets are in other wordsregarded not as the products of some hypotheticalcellular "degeneration" but as droplets of ab-sorbed modified protein similar to those whichform in the cells of the proximal convolutions ofexperimental animals which have been givensimilarly large injections of homologous plasmaproteins (13). Since the droplets are thereforenot considered an essential feature of the renallesion of EHF, the data concerning them will bepresented elsewhere in a general discussion of hy-aline droplet formation in human renal diseaseand its significance in the metabolism of the plasmaproteins.

* * * * *

The three remaining cases from the Hypoten-sive Phase of EHF show increasing structuralchange of the nature so far described and the de-velopment of another aspect of the renal lesionthat is to assume predominating proportions andresult in grave renal damage in later stages of thedisease, namely, the ill effects of the vascular dis-turbance on the tubules of the nephrons.

Case A developed fever and nausea the day af-ter his discharge from the hospital following anacute appendectomy from which he had appar-ently recovered with an uneventful convalescence.Two days after readmission to the hospital aDouglas' pouch abscess perforated and drainedspontaneously; he then went into shock. Anintra-abdominal hemorrhage was suspected butnone was found on laparotomy. Transferred tothe EHF hospital in shock, he received 5 units ofalbumin and other treatment with little effect, anddied on the fifth day. He was anuric during hislast hospitalization; his BUN rose to 64 mg. percent.

At autopsy the kidneys presented the usual sub-cortical congestion and the anterior lobe of thepituitary was hemorrhagic.

Histological examination showed that in therenal cortex the epithelium of the convoluted tu-bules was greatly swollen and vacuolated, resem-bling the alterations illustrated in Figure 9;there were few if any hyaline droplets in their

epithelial cells except as the convolutions ap-proached the cortico-medullary boundary. In thearea of subcortical congestion the excess bloodwas still contained within widely dilated capillaryspaces. The straight tubules in this region werefilled with great numbers of what appeared atfirst glance to be "hyaline casts"; these also ex-tended upwards in the medullary rays (Figure12). Collecting tubules in the deeper parts of themedulla and papillae showed only moderate num-bers of casts.

The terminal medullary segments of the proxi-mal convolutions which lay in the congested sub-cortical zone were compressed and crowded to-gether though their lumens were irregularlydilated and filled with hyaline material. These col-lections on closer examination did not have the ap-pearance of solid casts, the great majority beingcomposed of an albuminous fluid which had, un-der the influence of the fixative, coagulated inbubble-like configurations containing debris anddesquamated cells (Figure 13). The epithelialwall lying contiguous to the entrapped fluid wasinfiltrated with eosinophilic material; its cellswere disarranged, in part desquamated, andshowed pyknotic nuclei. The protoplasm of thebetter preserved cells was crowded with hyalinedroplets of varying size.

Similar changes were noted in Case 21, who wasadmitted with a typical 3-day history of EHFanda temperature of 1030, a blood pressure of 118/58mm. Hg, and 1 + proteinuria. After an unevent-ful 24 hours, his hematocrit abruptly rose to 59.0per cent, and his blood pressure fell to 86/64 mm.Hg with a pulse of 140. In spite of 5 units ofserum albumin and other supportive treatment,he died in primary shock on the next day, thesixth day of the disease.

At autopsy there was marked retroperitonealedema; the kidneys presented the typical appear-ance of subcortical congestion and there was agross hemorrhage in the anterior pituitary.

Histological examination showed the usualswelling and vacuolization of the proximal con-volutions in the cortex and scattered areas of in-tertubular edema similar to those illustrated inFigure 8. In the markedly congested subcorti-cal medullary zone the terminal proximal convo-lutions and ascending limbs were compressed andfilled with hyaline material and desquamated cells.

109


The intact epithelium of proximal convolutionswas filled with hyaline droplets; these and the en-trapped fluid are shown in the dissected speci-mens of Plate II. Droplet formation was notlimited to this region but extended throughoutthe entire length of the proximal convolutions upto the glomeruli, so that every cross section ofthem in the cortex was filled with large Grampositive droplets (Figure 14). The integrity ofthe cells, even when crowded almost to bursting,was apparently preserved for, when visible amongthe massed droplets, their nuclei appeared normal(Figure 15).

Essentially similar alterations were observed inthe kidneys of Case 8. This man was admittedwith a typical 5-day history and a temperature of1030. On the sixth day he went into shock andin spite of treatment, including the administrationof 4 units of albumin in 24 hours, he died on thenext day. His urinary output was 1410 cc.; therewas a 3 + proteinuria.

At autopsy the usual subcortical congestion andpituitary hemorrhage were present. The histo-logical appearances were very similar to those de-scribed in the previous case, hyaline droplets fillingthe cells in cross sections of proximal convolu-tion throughout the cortex as well as in the termi-nal segments where entrapped albuminous ma-terial was concentrated.

* * * * *

A resume may now be given of the developmentof the renal lesion as it is seen during the firstweek of EHF, a period including the Febrile andthe Hypotensive Phases of the disease, in indi-viduals where its further evolution was arrestedby death from circulatory collapse. Combiningthe physiological, clinical and pathological data,the following would appear to be the course ofevents.

In the Febrile Phase, which initiates the clini-cal syndrome, the physiological evidence of highclearances indicates that renal blood flow is eithernormal or increased. In the transition from theFebrile to the Hypotensive Phase, various clinicalphenomena appear which indicate widespreaddamage to small vessels, capillaries and venules;the intense and increasing flush of the skin andmucous membranes and, more certainly, widelydilated nail bed capillaries detectable by micro-

scopic examination ( 14) are indications of lossof arteriolar tone and vasodilatation. It is note-worthy that skin hemorrhages in the form ofpetechiae develop somewhat later. A rising he-matocrit without corresponding increase in serumprotein concentration and the post-mortem find-ing of T-1824 which has escaped into massiveretroperitoneal edema are conclusive demonstra-tions that plasma is now leaking from the dam-aged vessels. The abrupt occurrence of a heavyproteinuria, which in the Febrile Phase had beeninsignificant or absent, shows that the renal capil-laries as well are involved in this general vascularlesion.

Renal blood flow now decreases as indicated bya depression of clearances (2, 3); this may occurindependently of general circulatory collapse andis exaggerated when primary shock intervenes, asit did in all the cases of this series. The result ofthis decreased flow through the kidney is, how-ever, quite different in case of EHF than it is inthe individual with normal renal vessels who suf-fers a similar circulatory depression as a result,say, of surgical shock. In both instances thereis one factor in common-the distinctive anatomi-cal pattern of the vascular bed in the subcorticalzone of the medulla; in the kidney of EHF, how-ever, this peculiar vascular bed is abnormal sincethe vessels which compose it are both atonic andpermeable.

The vascularization of the medulla has been theobject of extensive anatomical and functionalstudy in recent years. The older descriptions ofthe arteria rectae which supply it from efferentsof glomeruli in the juxta-medullary zone of thecortex have been amplified by the studies ofTrueta, Barclay, Daniel, Franklin, and Prichard(11), which demonstrate their relatively largediameter. In extensive anatomical and physio-logical experiments, Block, Wakim, and Mann(15) have revealed the complexity of the reactionsthat occur in the subcortical zone to various ab-normal situations, in particular those of shock andrelated conditions. Barrie, Klebanoff and Cateshave described this vascular bed most aptly as the"arcuate sponge" (16) and have pointed out thepossibility of free communication by sinusoids be-tween the arcuate arteries, the medullary networkand the veins.

Although there have been some differences in

110


interpretations of the functional effects of theseanatomical peculiarities, our examination of thedisturbed renal circulation in the acute renal fail-ure associated with traumatic and toxic injury(10) agrees in both its experimental and patho-logical aspects with the more detailed and exactstudies of Block,Wakim, and Mann (15). Theseinvestigators point out three constantly occurringphenomena in their varied experiments in whichthe renal circulation was disturbed by electricalstimulation of the renal nerves, sudden clampingof the renal artery, injection of substances such asepinephrine, and the reduction by hemorrhage ofthe mean arterial pressure: 1) when renal vaso-constriction operates in a renal vascular disturb-ance it appears to be limited to the cortex alone;2) under conditions of sharply decreased bloodpressure and flow in the kidney the medulla maybecome congested even though blood is not flow-ing through it; such an appearance, therefore, doesnot of necessity indicate an actual "shunting" ofthe renal circulation past the cortex through themedulla into the renal veins; and 3) in all situa-tions, as the blood flow fails, it is best preservedin the cortico-medullary zone of the kidney. Theyfurther point out that from the anatomical stand-point two paths are present for the flow of bloodinto the medulla-the long arteriolae rectae andthe general network of peritubular capillarieswhich supply the cortico-medullary region. Whenrenal blood flow is reduced, most of the flow fromthe cortex into the medulla is through the peritu-bular network into the cortico-medullary regionwith a corresponding decrease through the longerarteriolae rectae which pass to the deeper medulla.Blood flow may, therefore, continue into but notthrough the cortico-medullary region under cir-cumstances in which flow is reduced to a mini-mumelsewhere in the kidney.

If, in the light of these considerations, one con-siders the situation in EHF in which after the in-creased influx of blood of the Febrile Phase, areduction of flow ensues in a renal vascular bedwhich is atonic, dilated and permeable to the es-cape of plasma, it is evident how and why theintertubular vessels of the "medullary sponge"become so intensely engorged with blood in spiteof the decrease in RBF which characterizes theHypotensive Phase. As is illustrated in Figure6 the contents of the subcortical capillaries be-

come an almost solid mass of entrapped red bloodcells. Plasma which has escaped through the ab-normally permeable vessel walls is found in areasof intertubular edema which may extend into themedulla (Figure 5) and, by means of medullaryrays, into the cortex (Figure 8).

The effect of these circulatory disturbances onthe nephrons is already apparent in the studies ofindividuals who died in primary shock. In theearliest example (Case 38, Figures 1 and 2) theywere so minimal as to be of doubtful significanceas a factor in the renal lesion. If death from cir-culatory collapse had not occurred, it is reasonableto suppose that a return of renal blood flow wouldhave resulted in complete restitution of a normalrenal status. In most of the examples of the Hy-potensive Phase described, however, there wasdefinite and increasing evidence of damage tonephrons, so that here again the structural changesand the physiological aspects of the renal lesioncorrelate with the clinical indications of a begin-ning "renal failure" that were shown in the rise ofBUN.

The earliest tubular alterations to develop, pres-ent indeed in slight degree in Case 38, were swell-ing and subsequent vacuolization of both the cor-tical and medullary portions of the proximal con-volutions. Except for the increase in volume ofthe renal cells and a decrease in the density of theircytoplasm which reached its maximum in Case C(Figures 10 and 11), there was no great altera-tion of their configuration; their nuclei appearednormal. One might suppose, therefore, that hereagain the renal lesion is reversible.

The similarity in appearance of the swollen epi-thelium of the Hypotensive Phase to the changesproduced by Funck-Brentano by experimentallyaltering the osmolarity of the plasma has beennoted. Whereas in his experiments hyperhydra-tion of the epithelium may be considered the chieffactor in the cause of oliguria, acting either directlyby luminal compression or indirectly through amodification of renal blood flow from increasedintrarenal tension, the relations are less simple inthe case of EHF. Decreased urine flow was ir-regularly observed during the Hypotensive Phasebut not the definitive oliguria to be described later,so that incidental factors, such as variations incardiac output, fall in blood pressure, decreases inrenal blood flow and consequent glomerular fil-

illl


tration would seem to be alternative explanationsfor the variation in renal output in the earlierphases of the disease.

If the effects of cellular swelling and the pres-sure of engorged intertubular capillaries had notgreatly interfered with urine flow through the lu-men of the tubule, they nevertheless had producedtheir deleterious effect on the integrity of thenephrons. This was apparent in the irregularcompression, dilatation and entrapment of protein-aceous urine and desquamated epithelial cells inthe terminal medullary segments of the proximalconvolutions, as illustrated in Figure 13 and PlateII.

In EHF, these tubular changes in the subcorti-cal zone are the first evidences of a destructivedamage to the nephrons which increases to irre-versible and irreparable alterations in later stagesof the renal lesion.

THE RENALLESION IN THE PERIOD OF TRANSITION

(The Incidence of Intertubular Hemorrhage)

The transition from the Hypotensive Phase tothat of Oliguria is characterized by physiologicaland clinical phenomena which indicate that es-caped plasma is returning to the circulation: thehematocrit falls, retroperitoneal edema decreasesand direct measurements of plasma volume arehigher. In spite of these evidences of circula-tory recovery, fatal "transition shock" may oc-cur, apparently due to arteriolar dysfunction (9,p. 695). Eight fatalities occurring under thesecircumstances appear in this series; all of theseshowed the development of a structural alterationpreviously unencountered.

This characteristic lesion of the Transition Pe-riod was the appearance of hemorrhage in thecongested subcortical zone. Whereas in the previ-ous Hypotensive Phase the excess of blood wascontained within greatly dilated intact intertubu-lar vessels (Figures 4 and 6), foci of exudativehemorrhage were now seen scattered widespreadthrough the area of congestion.

Case 39 was admitted on the fourth day of hisillness with a temperature of 1010, a blood pres-sure of 100/80 mm. Hg and a 4 + proteinuria.He went into shock shortly after admission andremained in circulatory collapse, with remissions,in spite of 9 units of serum albumin and other

treatment until his death on the seventh day. Onthe fifth day, his hematocrit was 49.5 per cent.His urinary output was 150 cc. the day beforedeath and the BUNwas 50 mg. per cent. Cer-tain details of his clinical course are shown inText Figure 4.

At autopsy there was only a moderate retro-peritoneal edema; the kidneys showed the typicalcongestion of the cortico-medullary zone, andthe usual extensive hemorrhage was present inthe anterior pituitary.

The general condition of the kidneys was anexaggeration of what has been observed in themore severe cases during the Hypotensive Phase.Swelling and epithelial alteration of the proximalconvolutions and scattered focal areas of inter-tubular edema were present in the cortex (Figure16); entrapment of albuminous fluid and desqua-mated epithelial cells was seen in the straight tu-bules of the subcortical zone of congestion (Fig-ure 17), which, along with irregular areas ofintertubular edema, extended into the deepermedulla reaching the papillae. The ducts ofBellini were filled with renal failure casts (Fig-ure 18).

In the subcortical zone of congestion, manycapillaries, though packed with red blood cells,were intact; though the outlines of the vesselswere still visible, the intertubular spaces betweencertain tubules contained free red blood cells.

Three other cases, 8, 4 and 27, were so similarin their clinical behavior and in the nature of thepathological lesions found at autopsy that a de-scription of them would be needlessly reiterative.All showed a somewhat higher elevation of BUN;on admission the BUNfigures were 27, 49 and 12mg. per cent, respectively, and on the day beforedeath, 117, 117 and 122 mg. per cent. Proteinuriashowed a similar increase from 1 + on admissionto 4 + on the day of death, which occurred on theninth day of illness in the first case and the eighthday in the other two.

Case 28, who died of transition shock on thetenth day, illustrates the acme of intensity in therenal lesion present in this series. He was ad-mitted on the fifth day of his illness with manypetechiae and a 3 + proteinuria. He immediatelywent into shock. In spite of temporary improve-ment under the usual treatment, which includedthe administration of 16 units of concentrated se-

112


TEXT FIG. 4. CASE 39- EXAMPLEOF TRANSITION SHOCKIN HEMORRHAGICFEVERNote the dependency of blood pressure on pressor therapy and the relative ineffectiveness of serum al-

bumin therapy despite minimal increase in hematocrit.

rum albumin in 72 hours, he passed from oneepisode of shock to another, 14 in all, to death onthe tenth day. The hematocrit was 53.5 per centon the first day of hospitalization, and 51.3 percent on the day before death. A urinary outputof 200 cc. was noted on the first day and "lessthan 50 cc." in the remaining period. Blood ureanitrogen was 58.5 mg. per cent on the day pre-ceding death.

At autopsy only a moderate retroperitonealedema was observed; the kidney showed extremesubcortical congestion and there was a hemorrhagein the anterior pituitary.

Histological examination showed the subcorti-cal zone to be the seat of widespread infiltrativehemorrhage; the intertubular capillaries, which inexamples of earlier stages of the disease werewidely distended but whose endothelium was stillintact, were now no longer visible as distinctchannels. The fibrillar remnants of their walls,however, could be seen scattered through the ir-regular spaces which, densely packed with red

blood cells, isolated each tubular cross sectionfrom its neighbor. The enclosed tubules werecompressed and distorted and their epitheliumnecrotic and desquamated so that the character-istic cellular pattern of the various tubular seg-ments was lost (Figure 19).

This intense vascular disturbance separatedthe cortex by a broad band from the deepermedulla; it was not, however, confined to the sub-cortical zone. Extending into the cortex with de-creasing intensity as the capsule was approachedwere extensions of the same extreme congestionand frank intertubular hemorrhage (Figure 20).In the areas free of congestion and hemorrhage,the epithelial cells of the proximal convolutionsshowed the usual swelling and vacuolization of theearlier periods; in tubules surrounded by infil-trating hemorrhage, necrosis of portions of theentire wall including the basement membrane wasapparent (Figure 21). The intact epithelial cellsof the proximal convolutions were crowded withgreat numbers of hyaline droplets, both in the

113


cortical convolutions and in those few terminalstraight segments in the subcortical zone whichwere better preserved.

Congestion, hemorrhage and tubular necrosisalso extended deep through the medulla into thepapillae. Straight tubules, including collectingducts, were thus isolated and compressed; theirepithelium and even the basement membraneswere necrotic over great stretches of their extent(Figure 22). The large ducts of Bellini were sur-rounded by widely dilated vascular spaces andscattered areas of hemorrhage, and contained re-nal failure casts (Figure 23).

Dissection of nephrons showed the nature andwidespread distribution of the tubular damage,and in particular its topographic relation to thesubcortical zone of intertubular hemorrhage.

In Plate III, A to D, all that remained of aproximal convolution is shown; the degree andextent of irregularly scattered tubular disruptionas observed when a continuous tubule is examined,rather than in the interrupted cross sections of ahistological section, are apparent. Beginning atthe origin of the convolution, its cellular patternis markedly disturbed as can be appreciated bycomparison with the more normal convolution ofPlate I; nuclei are obscured in the disintegratingcytoplasm of the tubular epithelium, which formsirregular masses of deeply stained material alter-nating with areas of more lightly stained disruptedtubular wall. The lesion increases in intensity(Plate III, B and C) as the straight segmentleaves the cortex and enters the subcortical zone.

Here the tubule thins to a bare basement mem-brane which ultimately disappears as its course isinterrupted by complete disintegration (PlateIII D). An exact determination of the percentageof proximal convolutions thus physically inter-rupted was impossible since the lesion consists ofa gradual attenuation of the tubule in which a fewfibrillar remnants indicative of its former courseremain. Such remnants are so fragile that theycannot be dissected, and though in a sense theymight be considered still existent, they certainlycan have none of the functional attributes of a tu-bule, even the mechanical effect of constituting aconfining channel for fluid. Regarded from thissimplest of tubular functions, only a few intactterminal proximal convolutions could be foundin the zone of hemorrhage.

As previously mentioned, the hemorrhagic areaformed an almost continuous band between thecortex and deeper medulla, and extended to thepapillae by irregular infiltrations. Within theseextensions of the hemorrhage all tubules showedin varying degree the deleterious effect of pressureand anoxia. In Plate IV a fairly intact loop ofHenle lying within the hemorrhagic zone is shown,throughout the course of which irregular stretchesof damages are evidenced by cellular detritus andthe obliteration of nuclear detail; at the bend ofthe loop the tubule is filled with a dense coagulumwhich stains black.

Dissected collecting tubules passing throughthe zone of hemorrhage, though they had pre-served their external configuration, were in greatpart entirely necrotic, showing nothing of the dis-tinctive clear cellular pattern which characterizesthese structures; their luminal space was filledwith debris and densely black stained casts (PlateV).

Returning to a description of the cortex in whichlay portions of nephrons which, in relation to thedirection of urine flow, lay both before and beyondthe interrupting band of subcortical hemorrhage,it was evident that the tubules distal to the zonewere not only much better preserved than theseverely damaged terminal proximal convolutionswhich lay in it, but in general were less severelyaffected than the first portions of the cortical prox-imal convolution which led directly from theglomeruli.

In Plate VI, the origin of a collecting duct isshown with four ascending limbs which, passinginto distal convolutions, join by way of connectingtubules to form the peripheral origin of the col-lecting system. All these structures lay in theouter cortex beneath the capsule and were not in-volved by the irregularly scattered hemorrhagicinfiltration that extended from the subcorticalzone. It will be noted that all these tubules are,compared to the severe damage illustrated in thecortical portions of proximal convolutions (PlateIII), relatively well preserved. There is, there-fore, as in all examples of acute tubular necrosisand renal failure, no preferential damage to the"lower nephron" but rather the reverse. Thenormal cellular pattern of the ascending limbsthat lie in the cortex is visible although some aredilated and contain coagulated material and debris.

114

TEXT FIG. 5. A RENAL LOBULEFROMCASE28 WHODIED OF TRANSITION SHOCKON THE TENTH DAYNote the extensive subcortical intertubular hemorrhage with extension into the cortex. Necrosis of tubules within

the involved areas. Heavy black, glomeruli and casts within tubules; cross-hatching, necrosis of tubules; smallcircles within tubules, desquamated cells and debris. Many terminal proximal convolutions end by disruption inthe zone of hemorrhage. Magnification 18 X.


TEXT FIG. 6. CASE42-EXAMPLE OF FATALITY IN SEVERE PRIMARY SHOCKIN HEMORRHAGICFEVER DESPITE CONTINUOUSL-ARTERENOL INFUSIONThis patient was in severe shock when admitted and blood pressure was

unobtainable until given both plasma volume expander and an infusion con-taining a large amount of L-arterenol. Accidental stopping or slowing ofthe infusion resulted in marked decrease in blood pressure. The severeshock that followed the second episode did not respond to further plasmavolume expanders or heroic pressor therapy.

The distal convolutions are also dilated, andthough cellular detail is obscured by the intenseblack staining of the solid coagulum that distendstheir lumens, occasionally a flattened but intactepithelium is visible. The connecting tubules arenot dilated and in most instances are free ofblack stained coagula; this dense material appearsagain in the peripheral collecting tubules and, ob-scuring cellular detail, fills the remaining portionof the collecting duct.

Text Figure 5 shows a reconstruction of a lobulefrom camera lucida drawings of dissected neph-rons; it may be compared with Text Figure 3, asimilar reconstruction of the earlier stage whichshowed so little departure from normal topographi-cal relations. The important features now notedare: 1) the presence of an irregular band of hemor-rhage separating cortex from medulla; 2) the ex-tension of the congestion and hemorrhage to thepapillae on the one hand and, in diminishing scat-tered areas, to the surface of the kidney on theother; 3) the interruption by necrosis of the ma-jority of all the tubules, terminal medullary seg-ments of proximal convolutions, descending andascending limbs-both broad and thin-and col-lecting tubules, which pass through the zone of

hemorrhage; and 4) the dilatation with coagulatedproteinaceous material of those tubules which liein the cortex distal to the zone of hemorrhage.

The clinical course of Case 42 is shown in TextFigure 6; admitted in a state of shock on the fifthday of illness, he died in 10 hours in spite of treat-ment that included the administration of 31/2 unitsof serum albumin and 500 cc. Dextran; 400 cc. ofurine were passed. At autopsy a moderate retro-peritoneal edema was present as well as the usualsubcortical congestion in the kidney and hemor-rhage in the pituitary.

Histological examination showed an intensehemorrhagic infiltration throughout the cortico-medullary zone. The outlines of the intertubularcapillaries could not be made out, red blood cellsfilling the spaces between the cross sections of tu-bules in densely packed masses. Congestion andhemorrhage extended deep into the medulla andalso into the cortex. Many of the proximal con-volutions in the cortex were filled with hyalinedroplets; the nuclei of the cells containing themwere well preserved. In the areas of subcorticalintertubular hemorrhage the tubules were com-pressed and distorted and their epithelial wallsdestroyed.

116


The two remaining cases, which showed mas-sive intertubular hemorrhage in the congestedcortico-medullary zone, and which were consid-ered clinically to have died in the late Hypoten-sive Phase or in the Transition Period of shock,may be briefly mentioned, as the renal lesions wereessentially similar to those which have been de-scribed. Case 17 survived for 10 days, the last 5days of which he was in more or less continuousshock in spite of intensive treatment which in-cluded the administration of 12% units of serumalbumin during the last 4 days. During this periodhe passed 190 cc. of urine; it is of interest that hisurine had been free of protein during the FebrilePhase until the fifth day, when, with the onset ofthe Hypotensive Phase, it suddenly showed a 4 +proteinuria.

Case 1 was admitted on the fifth day of illnesswith a blood pressure of 90/60 mm. Hg, and diedthe next morning in shock. Two units of serumalbumin were given; with a total fluid intake of1500 cc., his urinary output was 390 cc.

The structural changes were essentially similarin these two cases. Intense subcortical intertubu-lar hemorrhage was present with extension intothe papillae and less to the cortex, with tubulardestruction not only in the medullary zone butscattered irregularly throughout the cortex anddeep medulla. In the individual who had received12% units of human serum albumin, the cells ofthe better preserved convoluted tubules were filledwith droplets; these were present but sparse in theother, who had been given 2 units.

* * * * *

Since the period covered by the group of casesjust described is by definition transitional from theHypotensive to the Oliguric Phase, a discussion ofthe structural findings can be postponed until thefull development of the renal lesion in the nextperiod has been given. An important correlationbetween clinical and pathological aspects shouldbe noted in passing; as the clinical phenomena re-quired the special recognition of an interim be-tween two major phases in the term TransitionPeriod, so the finding of a distinctive structuralelement, intertubular hemorrhage, sets this periodoff pathologically in the evolution of renal lesion.There has been no need for such a subdivision indescriptions of general Acute Renal Failure, for

no such structural distinction in the form of sub-cortical hemorrhage occurs.

As will become apparent when the functional ef-fects of this new development are described, itsimportance in the ultimate fate of the kidney as acollocation of nephrons is critically determinative.

THE RENALLESION IN THE PHASEOF ESTABLISHED

OLIGURIA

(Infiltrative Intertubular Hemorrhage and Stasis;Tubular Necrosis and Intrarenal

Obstruction)

Intertubular hemorrhage resulting from increas-ing congestion that proceeds to ultimate stasis indamaged terminal vessels is not, by the nature ofits origin, a sudden event but a gradual infiltrationthat develops at varying rates depending on theseverity of both congestion and vascular damage.It is not surprising, therefore, to find its structuraland functional effects spread over three phases ofthe renal lesion, the late Hypotensive, the Transi-tion Period and the Oliguric Phase.

Ten cases, each of whomdied on or between theeighth to eleventh days, presented the clinical char-acteristics of an established oliguria and showedthe full development of the renal transformationof EHF. The physiological disturbances, in par-ticular the detail of fluid balance, of a typical ex-ample, Case 23, are shown in Text Figure 7. Allcases showed the lesions which have been notedabove in our description of those who died in lateprimary or transition shock, i.e., a subcortical zoneof intertubular congestion and hemorrhage withextension of both to the deep medulla and in lesserdegree to the cortex, with tubular disruption ofthose portions of nephrons which were includedin the affected areas.

All ten cases showed in addition, however, anadded structural change which characterized thePhase of Established Oliguria pathologically,namely, a marked dilatation of the cortical tubulesof the nephrons. This is illustrated in Case 31, anindividual who entered hospital on the fourth dayof his illness, went through the Hypotensive Phasewith shock, for which he received 2 units of albu-min, and passed into the Oliguric Phase in whichhe died on the eleventh day with a BUN of 184mg. per cent. His fluid intake during the last five

117


DAYS OF ILLNESS 4 I 5 I 6 1 7 1 8

TEXT FIG. 7. CASE23-EXAMPLE OF DEATH IN SECONDARYSHOCKDURING OLIGURIC PHASE OF

HEMORRHAGICFEVER

Primary shock on fifth day during Hypotensive Phase was mild. Subsequent slight increase of hemato-crit was presumably due to dehydration rather than to continued capillary leakage of plasma. On the seventhday the patient developed right lower lobe pneumonia and hyperpyrexia that precipitated shock not respon-sive to therapy. At autopsy characteristic findings of hemorrhagic fever were present, although there wasno retroperitoneal edema. In addition, a patchy pneumonia chiefly in the right lower lobe was noted.

days was 4184 cc. and his urinary output was 360cc.

Histological sections showed a wide dilatation ofall tubules in the cortex. In the case of the dilatedproximal convolutions, the distention reachedBowman's space with a resulting compression ofthe glomerular tuft (Figure 24). Not only thoseportions of the nephrons lying proximal to thezone of medullary hemorrhage were distended butthe cortical ascending limbs and distal convolu-tions, the latter containing cellular debris, werealso dilated. The relation of this tubular dilatation

to the compression of the subcortical zone ofhemorrhage is shown in Figure 25.

The effects of the dilatation of all the corticalportions of the nephron are shown in dissectedspecimens in Plate VII, A to D. The wide lumenof the proximal convolution is clearly apparentthroughout its entire length, as can be seen bycomparison with Plate I; the stretching of theepithelial wall increases the space between the nu-clei, and so exaggerates the cellular pattern. Theepithelial cells are, however, fairly well preserved.As its terminal segment approaches and enters the

118


subcortical region of congestion it narrows to one-third its previous diameter and is lost in the inter-tubular hemorrhage. A portion of ascending limband the distal convolution from the cortex, whichwas originally entwined with the dilated proximal,and which was therefore known to be the continu-ation of the same nephron, show an equal dilata-tion; there is some dark stained material in thedistal convolution.

In Case 25 there was more than the usual cellu-lar damage in proximal convolutions, and evi-dences of beginning repair were visible both inregenerative mitotic proliferation of epithelium(Figure 26) and in the ingrowth of interstitialgranulation tissue into the disrupted tubule lu-mens. Case 41 showed somewhat less dilatationof the lumens of the cortical tubules, and scatteredareas of intertubular edema (Figure 27). Theremaining cases showed the general picture ofcortico-medullary hemorrhage and dilatation ofcortical tubules above it, but added nothing furthertowards the elucidation of the renal lesion.

* * * * *

It is in the Phase of Established Oliguria thatthe structural and functional lesions of EHF reachthe height of their evolution. When the structuralchanges in the nephron are thus fully establishedthey appear to constitute an irreversible status,for though a "kidney" of sorts might evolve fromthose nephrons which by the fortunate circum-stance of their location had escaped the effects ofcongestion and hemorrhage, the reconstitution ofnephrons whose middle half has been destroyedwould seem impossible. We shall see in laterdescriptions that the attempt is made, but the ap-propriate anastomosis of four frayed ends of fragiletubule lying loose in a mass of interstitial hemor-rhage under considerable pressure is on the sheerbasis of probability unlikely, even when the po-tentialities of repair are present. It would seem,therefore, that the renal lesions we have seen inthe fatal cases of established oliguria could nothave existed in similar quantitative relations in thegreat majority of those suffering from EHF, forthey survived and went on to eventual recovery.

In the earlier stages of the syndrome tubulardamage was evident in the general swelling andvacuolization of the epithelium of the proximalconvolution, whereas in the Transition Period and

Oliguric Phase the tubular alteration consists oflocalized physical disruption; the two lesions aretherefore quite different in their structural aspect.Though the early epithelial changes are of thesort that have been classically derived from "toxicaction," there has been no demonstration of atoxin in EHF other than the hypothetical sub-stance which acts on the small blood vessels. Dis-turbances of circulation, however, are clearly visi-ble at this time, so that it seems reasonable to ac-cept them as the cause of the early epithelialdamage.

Structural interruption of the nephrons involvesproximal convolutions, ascending and descendinglimbs of Henle's loops and collecting tubules indis-criminately. The damage is confined to the areasof congestion and hemorrhage, and so lies chieflyin the medulla, but is also present extensivelythroughout the cortex when the circulatory dis-turbance has spread to this area. Here again, itwould be difficult to explain this random distribu-tion in terms of the direct action on the renal epi-thelium of a nephrotoxic poison derived, hypo-thetically, from the infectious agent of the disease.The histological appearance of the epithelial lesionis also not that of nephrotoxic damage, in whichepithelial cells, absorbing the poisonous substance,die and desquamate within a tubule still maintainedby an intact basement membrane. There has oc-curred, rather, a disruption of the entire tubulewall, the tubulorhexic lesion characteristic of thekidney of Acute Renal Failure (10). The in-ference seems certain, therefore, that the tubularlesions of EHF, both early and late, are the re-sult of circulatory disturbances. This conclusiondoes not, of course, rule out the primacy of aninfectious agent or its products as the cause of thevascular disturbances, for recent experiments byThal (17) in the production of bilateral corticalnecrosis by means of staphylococcus toxin showhow this may be produced in the kidney and soresult indirectly in extensive tubular damage.There are many similarities between the natureof the tubular lesion in his experiments and whatis seen in the kidney of EHF; the differences intopographical distribution seem less important,for, as Thal points out, "cortical necrosis" is notso restricted as the descriptive term would sug-gest, nor are the tubular lesions of EHF limitedto the medulla.

119


There is evidence that the circulatory disturb-ance acts on the tubule by means of two deleteriousforces; the first, and perhaps less important, isphysical pressure, and the second is anoxia fol-lowing the congestive stasis and the infiltrativeintertubular hemorrhage which has produced acessation of circulation in the involved areas.

Evidences of pressure, especially on the moresensitive proximal convolutions, were seen in theearliest stages of the renal lesion where there wasonly a simple and moderate congestion of theterminal vessels of the subcortical zone. With anincrease in vascular distention the effect of in-creasing intertubular pressure was evident in thecompression and consequent entrapment of pro-teinaceous fluid in the lumens of the terminalproximals that lay in the involved area (Figure 13and Plate II), and its final effect is demonstratedby the "intrarenal hydronephrotic" tubular dila-tation which ultimately develops in the definitivelyoliguric kidneys (Figure 24). This dilatation in-volved all the tubules lying in the cortex abovethe subcortical zone of pressure and included notonly proximal convolutions which are, in regardto the direction of urine flow, proximal to thezone of compression in the medulla but also thedistal convolutions, cortical ascending limbs andperipheral collecting tubules which lie distal to it.

This apparent paradox of postobstructional dila-tation arises from current misconceptions of thetopographical arrangement in the human kidneyof nephrons of varying length. In most diagrams(Peter, 18, Figure 63) two types of nephrons areshown, short-looped and long-looped, both ofwhich pass through the outer stripe of the outerzone which in EHF is the zone of hemorrhage.If this were an accurate and complete pictureof the human kidney, the distal convolutions andcollecting tubules would be cut off from the sourceof glomerular filtrate and could not dilate. As amatter of fact, the human kidney contains a veryconsiderable number of nephrons similar to thoseshown in Peter's diagram of the pig's kidney (18)which turn within the cortex and never reach themedulla. An example is given in our Plate VIII,A to C; many have been shown elsewhere (19).4

4Work in progress in this laboratory on quantitative as-pects of the morphology of nephrons shows that theirclassical subdivision in the human kidney into two arbi-trary groups of "long-looped" and "short-looped" neph-

In a kidney with a zone of subcortical obstruc-tion glomerular filtrate can therefore flow throughuninterrupted cortical nephrons to the point oforigin of the collecting system (cf. Plate IXA) andfrom thence to all distal portions of the neighbor-ing nephrons which are connected with it, evenwhen they have been shut off from their normalsource of filtrate by the zone of medullary hemor-rhage. This retrograde flow is in fact favored bythe obstruction in the lower collecting systemwhich itself passes through the obstructing zoneof hemorrhage.

The question may be asked why then do notall kidneys which suffer extensive disruption oftheir nephrons, such as, for example, Case 28 ofthe Transition Period, show extensive tubulardilatation in their cortices. The answer wouldseem to be that for dilatation to occur there mustbe an adequate source of filtrate; a comparison ofFigures 20 and 24 suggests that in the formercase the extreme cortical congestion and hemor-rhage is incompatible with a sufficient glomerularcirculation and consequent formation of filtrate.Thus a lesion which interrupts tubules in themedulla yet spares the circulation in glomeruli isnecessary for the development of the tubular pat-tern observed in the Phase of Established Oliguria.

The discussion of these minutiae concerning fluidflow in damaged nephrons would seem to warrantattention not so much for an importance inherentin their functional effect-for in both the situationsdescribed above urinary output was similarly les-sened-but as illustrating how the same or simi-lar pathological processes can by variation in theirtemporal or spatial relations modify or evenchange the direction of the evolution of the renallesion. In the examples cited it was a spatial re-lation; in the ensuing Phase of Diuresis it is tem-poral. In any case, the simple statement that thislesion or that is present in a section of kidneyis meaningless as a description of the structural-functional renal status.

The structural evidence of physical impedanceto the flow of urine through the nephrons is of

rons is a considerable and, as in the present case, at timesmisleading oversimplification. Our findings show thatthe loops of nephrons are indeed of varying length butthat this variation can be described by a normal fre-quency distribution curve rather than by absolute cate-gories; in short, some loops are longer than others.

120


interest to others in addition to the theoretical pa-thologist. The clinical observers who noted vary-ing degrees of reduction in urinary output in theearlier Hypotensive Phase recognized the distinc-tive and definitive character of the oliguria whichensued in the later period, and so designated it tothe Phase of Established Oliguria. The structuralfindings show that the distinction is real, since itis based on a difference in causal mechanisms.In the earlier Hypotensive Phase no structurallesion other than evidences of disturbed circulationwas present in the kidney to account for decreasesin urinary volume, and it is therefore reasonableto suppose that factors such as variations in tis-sue hydration, cardiac output, renal blood flow andglomerular filtration were responsible for thisvariable oliguria. In the later phase, what "es-tablished" the oliguria is clearly evident in thecompression and destruction of tubules and theobstructional dilatation of nephrons that followed.

Of greater importance than pressure effects inthe production of tubular damage is the anoxiathat results from the stagnation and cessation ofblood flow in the widely dilated and abnormallypermeable intertubular capillaries of the sub-cortical zone. It would appear likely that theblood is moving with some difficulty under suchcircumstances as are shown in an earlier periodin Figure 6, where crowded red blood cells arepacked to a solid hyaline mass. Any doubts thatcirculation may cease entirely under these condi-tions have been answered by the conclusive ex-perimental demonstration of Thal (17), who in-jected India ink into the general circulation of theliving rabbit with an analogous renal "stagnationhyperemia" and found that it did not penetrateinto vascular channels which were similarly dis-tended and packed with red blood cells. Thefurther evolution of the congestive lesion in EHFthrough diapedesis to frank, diffuse intertubularhemorrhage ends, therefore, in stasis.

Under the conditions here described we aredealing with what Ricker, in his classical study ofthe nature and effects of local circulatory disturb-ances (20, p. 96 et seq.), would have called "rotestasis," and from the late irreversible stage of thisrubrostasis he would have derived the "Seques-trationsnekrose" that interrupted the course of thenephrons. It is of particular interest in our prob-lem that in his examination of these effects invarious tissues and organs he chose, in the case of

the kidney, the necrosis that is produced experi-mentally by injection of vinylamin. This toxicagent differs from other renal poisons by produc-ing necrosis of tubules in the medulla rather thanin the proximal convolutions of the cortex.

Ricker points out that the lesion of vinylaminpoisoning begins in the papilla with a "prdstatisch"hyperemia followed by diapedesis of red bloodcells, and progresses to "Stasis" and eventual"Dauerstasis." The vascular disturbance in hisexperiments extended to the subcortical zone, andin one instance to the cortex, with resulting ne-crosis of tubules in the involved areas. His con-clusion that the tubular damage cannot be the re-sult of either a functional secretion or absorptionof the poison by epithelial cells, as was suggestedby Oka (21), or due to specific "affinity" for thepoison as Levaditi (22) assumed, has been con-vincingly confirmed by subsequent evidences thatproximal convolutions, Henle's loops and collect-ing tubules cannot be considered functionallysimilar; hence it would not seem likely that theyshould all absorb a poison similarly or, much less,be given to the same "affinities." Wecan accept,therefore, Ricker's explanation that the tubularnecrosis is a phenomenon of ischemic seques-tration due to the anoxia of the vascular disturb-ance; 5 with his broader conclusion that local cir-culatory disturbances are mediated through thenervous system we need not here be concerned.

THE RENAL LESION IN THE PHASEOF DIURESIS

(Intertubular Hemorrhage; Tubular Disruptionwith Evidences of Resolution, Regeneration

and Repair)

The remaining cases in this series show struc-tural changes that should be considered as illustra-tive of the effects of the passage of time on the

5 Ricker also compares the action of vinylamin withthat of mercury, which damages the proximal convolu-tions; this effect he also derives from circulatory disturb-ances and not from direct nephrotoxic action on the epi-thelium. Here, modern studies, histochemical and phar-macological, do not support his contentions, though itshould be noted that he anticipated a later demonstration(10) that the vascular element of ischemia may playan important if subsidiary role in the production of tubu-lar damage (tubulorhexis) in renal lesions of poisoning.A recent study by Lindegard and Lofgren (23) has con-firmed and extended these observations by demonstratingan irregular cortical ischemia not only in the anuricphase of mercuric poisoning but also in its diuretic phase.

121


fully developed lesions of the Oliguric Phase ratherthan as representative of the exact picture of theDiuretic Phase observed by the clinician duringthe recovery of the great majority of his cases.If the individual in the Oliguric Phase does notdie of secondary shock, dehydration, electrolyteimbalance or infection by the eighth to tenth days,but survives another week or two, the processesof resolution, regeneration and repair, as in allpathological complexes, begin their automatic op-eration. A part of these mechanisms of restitu-tion is the functional phenomenon of diuresis.As the pathologist sees them, the progress of theseprocesses, even if in the right direction, is blindand in the end futile, and the individual dies of arenal failure often obscured by some complicationsuch as the infection of a bronchopneumonia, thehyperkalemia of electrolyte and fluid imbalance, ora cerebral accident. It would seem unlikely thatthese complications can be merely incidental to therenal lesion, but rather it seems that in some man-ner, either directly or indirectly, renal insuffi-ciency is the basis of their development; for in allthe examples studied, though the complicationswere various, the renal lesion, with its gravestructural damage and evidence of functional in-adequacy in a BUNelevated at times to over 400mg. per cent, was constant.

What happens then in the great majority, 95per cent of all cases of EHF, has not been seen;the description of the typical kidney in the survi-vor must therefore be deduced and reconstructedon a hypothetical basis, and will be further con-sidered in our discussion.

* * b * *

As happens not infrequently in the descriptionsof passage from one phase of a disease to the next,our first case, D, might well have been classifiedas an example of a renal lesion at the end of theOliguric Phase; he survived, however, 15 daysand as the urinary output, though continuing low,increased somewhat on the last day but one of hislife, the description was clinically placed underthe title of the Diuretic Phase and will be hereso considered. As will be seen, there is no evi-dence of restitution of disrupted nephrons and ingeneral the renal lesion resembles that observedin the previous phase.

This patient was admitted on the fourth day of

his illness with a typical history of the FebrilePhase. The next day he abruptly developed a 4 +proteinuria. From the seventh to the tenth day hewas in the Hypotensive Phase with repeatedshock that responded poorly to treatment; an es-tablished oliguria followed but the urine volumeincreased from around 15 cc. in 24 hours to 250cc. on the day before his death. On the twelfthday dialysis by the artificial kidney reduced theserum K from 6.9 to 4.2 mEq. per L. but the pa-tient lapsed into coma, his temperature rose andcoughing with bloody purulent sputum developed;a bronchopneumonia was confirmed by X-ray ex-amination. On the day before his death, the fif-teenth day of illness, the BUNwas 484 mg. percent.

At autopsy the kidneys weighed 300 gm. andshowed the typical and fully developed lesionof subcortical congestive hemorrhage with deepmedullary involvement. There was a bilateralbronchopneumonia with multiple small abscessescontaining Gram positive cocci, scattered hemor-rhages in the right atrium, gastrointestinal tract,the cerebral hemispheres and pons and the an-terior pituitary.

Histological examination showed that congestivehemorrhage was extreme in the subcortical zoneand had extended irregularly and deeply into themedulla, reaching the papillae (Figure 28). Inthe outer zone there was an almost complete de-struction by necrosis of all tubules contained in theareas of hemorrhage, and these, coalescing, sepa-rated cortex from medulla (Figures 29 and 30).Extending downward into the infarct-like area ofhemorrhagic infiltration could be seen wisp-likeunoriented remnants of straight tubules; theirwalls were disrupted and in great part bare, butscattered along the denuded basement membraneswere huge, atypical nuclei with intensely stainedchromatin (Figure 31). Mitotic figures were fre-quent. The intensity of this excessive regenera-tive proliferation distorted the tubular configura-tion so that it was impossible to accurately identifythe parts of the nephron concerned in histologicalsection.

In the cortex the characteristic dilatation of thePhase of Established Oliguria was still present(Figure 32), involving Bowman's space and bothproximal and distal convolutions; it will be re-called that in this patient's "diuresis" only 250

122


cc. per 24 hours was draining from the renal tu-bules. The epithelium of the proximal convolu-tions was not only thin but cellularly atypical; itsnuclei were not round, but oval, and were irregu-larly distributed in the tubule wall, in some placesforming islands of excessive proliferation (Figure33); the appearance was that of a regenerated epi-thelium. Moreover, the tubules were more widelyspaced than normal (cf. Figure 1) and the inter-stitial tissue was infiltrated with mononuclear cellsand fibroblasts. The glomerular tufts were largewith widely dilated capillaries that contained dis-crete, apparently circulating red blood cells.

The topographical aspect of the lesions just de-scribed is shown in the dissected nephrons. PlateVIII, A to C, shows a complete cortical nephronwith a short loop that did not extend into the sub-cortical zone of intertubular hemorrhage and istherefore uninterrupted. As in all examples ofacute renal failure associated with traumatic andtoxic injury, the tubule is damaged from glomeru-lus to collecting tubule. As is usual, the damageis more severe in the proximal convolution than inthe so-called "lower nephron," consisting of scat-tered areas of tubulorhexic disruption with thelumen containing desquamated epithelial cells.Where disruption has not occurred, the wall iscomposed of a heavily stained, irregularly thinlayer of atypical, regenerated epithelial cells quitedifferent from the plump, evenly contoured cellsof the normal tubule (cf. Plate I). The transitionto the loop of Henle is abrupt and there is, as inall cortical nephrons, only a suggestion of a thinportion. From here on through the loop and as-cending limb and distal convolution the tubuleis greatly dilated with clear tubule fluid and con-tains desquamated cells and debris. As is sooften the case, the short connecting tubule is lessdistended.

Plate IX, A to E, shows the peripheral collect-ing system from its origin in a junction of con-necting tubules of five nephrons that lay beneaththe surface of the kidney to its destruction in thehemorrhagic zone of the subcortical medulla.The connecting tubules of 9 nephrons in all, filledwith black stained material that obscures theirepithelial pattern, join to form a collecting tubulewhich is empty. In many instances the collectingtubules contained similar material, but this ex-ceptional example was chosen for illustration be-

cause it shows the cellular pattern more clearly.It will be observed that the epithelium is notdisturbed until the proximity of the subcorticalzone of hemorrhage is approached; scattered par-ticles of cellular debris then appear in increasingnumbers until, in the substance of the hemor-rhagic zone, a complete necrosis and coagulationof the entire tubular wall is evident. The tuning-fork juncture with a neighboring collecting tubulewhich is similarly affected is well shown; belowthis point the single common duct contains black-stained material. The appearance on histologicalsection in other cases of similar, completely ne-crotic collecting tubules which have neverthelessmaintained their external configuration is shownin Figures 22 and 29. The general architecturalpattern of the kidney is shown in Text Figure 8 inaxreconstruction of the renal lobule composed ofcamera lucida tracings of dissected nephrons.

In Case 43 a frank diuresis had been establishedfor the last five days of the patient's life, and therenal lesion has special interest because, of all theexamples of this phase of the disease that came toautopsy, in this one alone resolution of whateverstructural damage had occurred had reached atleast the point at which one might imagine a chanceof eventual functional restitution to be possible.The resulting "kidney" of such a problematicalrestitution would, however, have been severelydeficient in its "reserve" of nephrons.

The patient entered hospital on the third day ofhis illness in the Febrile Phase with a temperatureof 103.60, and on the fifth day went into typicalshock of the Hypotensive Phase with the hemato-crit reaching 66 per cent. With appropriate treat-ment he passed through this episode to the Oligu-ric Phase, in which his average daily urinary out-put was approximately 100 cc. with a BUNrisingto 280 mg. per cent. His blood pressure alsorose (170/106 mm. Hg), and the hypervolemicsyndrome accompanied by 5 convulsions followed.Daily urinary output increased rather abruptly in24 hours on the twelfth day from 190 to 1000 cc.and reached 4600 cc. the day before his death.During this terminal period he presented the diffi-cult problem in water and electrolyte balance of a"limited homeostasis" illustrated in Text Figure 9,so that the margin between dehydration with thethreat of secondary shock and ample hydrationwith impending pulmonary edema was represented

123

1,s er

L:

:,1

,p

-;/1 U

.'Fk4.'

L:

*;A

.1.

.1

I~~~~~~~~~~~~~~~

.

I-.

TEXT FIG. 8. A RENAL LOBULE FROMCASE D WHODIED ON THE FIRST DAY OF THE DIURETIC PHASE

Note the persisting subcortical hemorrhage with destruction of tubules. Heavy lined proximals indicate re-

generated atypical epithelium (cf. Figure 33 and Plate VIII A); other symbols as in Text Figure 5. Magnification18X.

V

.&

9

XA

-4

&"N ."-, ;..I'..-,..

A


IN: TOTAL FLUID + 5660 hilOUT: Urine - 4930 ml

SUMMARYOF Vomitus - 570 mlFLUID. BALANCE TOTAL - 5500 mlFOR 37 HOURS GROSSBALANCE + 160 ml

INSENSIBLE LOSS - 750 mlNET BALANCE - 590 ml

TEXT FIG. 9. CASE 43-EXAMPLE OF LIMITED HOMEOSTAsIS DURING DIURETIC PHASE OF HEMORRHAGICFEVER

These observations were made on fourteenth and fifteenth days of illness. Death occurred on the sixteenthday, 24 hours after study. The patient was very dehydrated at the start of the observations, having beenin negative fluid balance for many days. Note the response of shock and blood pressure to increased fluidsbetween the eleventh and fifteenth hours. Also note decrease in blood pressure between the thirtieth andthirty-fifth hours when output was allowed to exceed intake. Pulmonary edema occurred in last periodwhen intravenous fluids were increased even though net balance was just barely positive during this time.(Note: Right-hand column, bottom line in 30 to 35-hour box should be cross-hatched.)

by a net negative fluid balance of 600 cc. Duringepisodes of hypotension and shock the diuresis dis-appeared and urinary output was at oliguric lev-els, to rise again to 4000 cc. when arterial pressurewas restored by nor-adrenalin. The last BUNtaken the day before his death had decreased to173 mg. per cent, and there was only a trace ofprotein in a urine of 1.011 specific gravity.

At post mortem there was found a confluentbronchopneumonia, hemorrhages in the rightatrium, cerebral hemispheres and basal ganglia,and in the anterior pituitary. The gross appear-ance of the renal lesion is summarized in the au-topsy protocol as "lower nephron nephrosis,slight."

Histological examination showed a marked con-trast to the pattern of a general dilatation of allcortical tubules seen in the previous stage of es-tablished oliguria; the lumens of the proximalconvolutions and other cortical tubules were notwidely dilated but contained some granular ma-terial, as did Bowman's spaces of glomeruli whichwere otherwise normal (Figure 34). The proxi-mal convolutions were lined with an epitheliumof the original normal type; there was little evi-dence of any regenerative renewal of cells, such asthe irregularity in thickness and nuclear size andshape and staining affinity that characterized theepithelium of proximal convolutions in most casesat this stage of the disease. The general appear-

125


ance was therefore of a cortex which had neverbeen greatly damaged.

Contrasting, was the alteration in the subcor-tical outer stripe of medulla. Extending down-ward from the cortex into this zone, which hadthe appearance of a region of resolution of hemor-rhage and replacement fibrosis, were the thick-ened terminal segments of proximal convolutions;their epithelium was quite atypical, varied inthickness and containing many irregularly dis-tributed large nuclei (Figure 35). In the broadband of the affected region irregular foci oftypical hemorrhagic infiltration persisted, alternat-ing and blending with ill-defined areas where thetubules were separated not by masses of red bloodcells but by cellular fibrous tissue (Figure 36) ; aconsiderable increase in collagenous fibrils sur-rounded atrophic tubules and fused with theirthickened basement membranes (Figure 37). Ex-tension of the fibrosis could be traced upwardsinto the lower cortex in the form of cellular scarswhich replaced tubules and surrounded glomeruli(Figure 38).

The renal lesions in five cases-two of which, 16and 46, died on the fourth day of diuresis, twoothers, 11 and 19, on the ninth day, and one, 29,on the eleventh day-showed essentially similarlesions to those just described but with more evi-dence of previous damage and a lesser degree ofrestitution. In the subcortical zone of the medulla,intertubular hemorrhage still persisted in all in-stances with intermingled areas of resolution andinterstitial reaction (Figure 39). In two of thefive cases, 16 and 46, the epithelium of the corticalproximal convolutions was of the original maturetype; in three, 11, 19 and 29, it resembled a re-placement by atypical regenerated epithelium simi-lar to that shown in Figure 40; in none were therethe marked distention and dilatation of tubule lu-mens observed in the preceding Phase of Estab-lished Oliguria.

The last two cases were of the longest durationthat occurred in this series and therefore presentthe maximum effect of the passage of time on therenal lesion that was available for observation.

The first of these, Case 33, died on the nine-teenth day of his disease in the tenth day of theDiuretic Phase. He had passed through the Hy-potensive Phase with two episodes of shock whichresponded to appropriate therapy, and an Oligu-

ric Phase of 4 days in which his urinary outputaveraged 250 cc. in 24 hours when the BUNreached a maximum of 286 mg. per cent. Duringthis period he was hypertensive (158/112 mm.Hg). On the ninth day of his illness, diuresis be-gan abruptly and continued until his death onthe nineteenth day, the urinary output rangingfrom 1410 to 4400 cc. in 24 hours. The lastdays of his life were an example of homeostaticinstability with repeated episodes of dehydration,hypotensive shock, fluid replacement and threat-ening pulmonary edema. During this terminalperiod he received 11 units of serum albumin.Cerebral involvement became apparent, his tem-perature reached 1040 and he died on the nine-teenth day of his illness. The BUNon the daypreceding death had decreased to 145 mg. percent.

At autopsy a confluent bronchopneumonia inaddition to hemorrhages in the right auricle, pitui-tary, and interventricular septum were present.The renal lesion is described as showing "a fairlymarked degree of congestion and focal hemorrhagein the renal medulla."

In the histological sections the proximal con-volutions in the cortex were lined with an atypicalepithelium. The lumens were therefore irregu-larly widened and contained a moderate amountof coagulated material and desquamated cells(Figure 40). The contrast in appearance betweenthese tubules whose lumens were somewhat in-creased by the irregularity of their epithelium andthe frank and even distention of a tubule dilatedby internal pressure, as occurred in the OliguricPhase, may be seen in a comparison of Figures40 and 24, and of Plates X and VII. Bowman'sspace was not dilated and the glomerular tufts ap-peared normal. The tubules of the cortex werewidely spaced and the interstices filled with aloose fibrous connective tissue in which wereclusters of round cell infiltration. In the sub-cortical zone of the medulla and extending to thepapillae extensive areas of persisting intertubularhemorrhage were seen; in these areas the smallertubules were widely separated and apparentlyreduced in number, but the larger collecting tu-bules were not only intact but showed an extremeand irregular proliferation of their distinctiveepithelium, which formed irregular plaque-likeformations distorting the normally even contours

126


of their lumens (Figure 41). The resulting dis-tortions of the terminal collecting tubules and theducts of Bellini entering the pelvis, the latter filledin part with renal failure casts, are shown inPlate XI.

Case 18 survived 27 days; he had passedthrough the typical Febrile and HypotensivePhases of the disease with only moderate shock.On the seventh day of his illness he entered a 6-day period of Established Oliguria with hyper-tension and a rising BUNthat continued withoutremission until his death, reaching a final figure of329 mg. per cent. On the twelfth day his urinaryoutput increased from 195 to 2175 cc. in 24 hoursand a moderate but irregular diuresis, at times of3000 cc., continued to his death. During thisterminal period the usual problem of maintenanceof fluid balance between dehydration and pulmo-nary edema was present with repeated episodes ofsecondary shock. In this period he received noserum albumin.

At autopsy, a marked hypostatic congestion andedema of the lungs were present with the usualhemorrhages in right auricle and pituitary. Therenal lesion is described as showing the "classicalmedullary hemorrhage with some evidence ofresorption."

Histological examination showed extensive in-tertubular hemorrhage in the subcortical regionextending deep into the medulla. The areas ofhemorrhage were not continuous, but formed anirregular zone in which regions of tubular col-lapse and dilatation alternated with definite fibro-sis of the intertubular tissue. The dilated tubuleslying in the fibrous areas extended by medullaryrays into the cortex and were either empty orfilled with hyaline material. Proximal convolu-tions in the cortex were irregularly dilated; theirepithelium was of a normal mature type. Theglomeruli, save for coagulated material in Bow-man's space, were not remarkable.

* * * * *

All of the individuals surviving the phase ofoliguria for periods varying from 1 to 16 daysshowed widespread evidence of the usual renalreparative processes of epithelial regeneration andreplacement fibrosis; the general dilatation of cor-tical tubules was not present. The structural re-pair concerned the kidney tissue rather than the

nephrons, for there had occurred no restitution oforgan structure in those elements whose continu-ity had been destroyed. In Case 43 the histologi-cal pattern of the cortex with its glomeruli andproximal convolutions appeared not greatly ab-normal (Figure 34) and, at least from the struc-tural viewpoint, had the appearance of a possiblefunctional adequacy even though renal insuffi-ciency was still evident in the BUN which haddecreased from a maximum of 280 to 173 mg. percent. Judging from the amount of scattered scartissue and distorted tubules in both the subcorticalzone and in the cortex, a considerable number ofnephrons must have been in part destroyed.

* * * * *

Further evidence of the nature of the repair isfound in an individual who, having recoveredfrom EHF, died of causes other than renal in-sufficiency 149 days later.

This patient was first admitted to hospital onthe third day of his illness with the typical clini-cal picture of EHF. On the next day, with ablood pressure of 110/90 mm. Hg, a pulse of 112and a hematocrit of 60 per cent, he was given oneunit of serum albumin; from then on the bloodpressure was stable. On the sixth day, the uri-nary output dropped to 25 cc.; on the seventhit was 125 cc.; he then became hypertensive (150/100; 170/100 mm. Hg). On the tenth day diu-resis of 1700 cc. ensued; the BUN, which hadrisen to 70 mg. per cent, fell to 43 on the twelfthday. He was transferred to a hospital in theStates where, 79 days after the onset of his illness,a test of his maximum concentration capacityshowed a specific gravity of 1.010. During thisstay in hospital the patient presented symptoms ofadrenal insufficiency, cardiac irregularity and oc-casional "clonic seizures." In one of these he died,149 days after the onset of the attack of EHF.

At autopsy the anterior pituitary was found tobe almost entirely destroyed by an old hemorrhagicinfarction; the adrenals weighed 6.2 gm.; in histo-logical section there was a thinning of the cortexwith a marked lipoid depletion and areas of necro-sis in the zona glomerulosa; the medulla was wellpreserved. The other endocrine glands, savefor atrophy and maturation arrest of spermato-genesis in the testicles, were not remarkable.

127


The kidneys weighed 370 gm.; the surface of one

was described as smooth, the other as finelygranular.

Histological examination showed similar lesionsin both kidneys, but the cellular scars to be de-scribed were more prominent in one. The sub-cortical zone of the medulla, which from the his-tory may be assumed to have been the seat ofcirculatory disturbances, showed an increase inintertubular fibrous tissue and a consequent widespacing of the tubules passing through it. Themajority of these straight tubules appeared nor-

mal, but scattered through the denser areas offibrosis were collecting tubules which showed a

hyperplastic proliferation of their epithelium (Fig-ure 42). The deeper parts of the medulla, in-cluding the papillae, were free of fibrosis and ap-

peared normal. Rather different from the acel-lular fibrous scarring were other more recentlyappearing wedge-shaped areas of monocytic infil-tration and fibrous proliferation which extendedfrom mid-medulla through the cortex to the sur-

face of the kidney. Within the confines of thesescars remnants of atrophic tubules and glomeruliwere visible (Figure 43).

* * * * *

In the kidneys of this individual who had re-

covered from the acute episode of EHF there are

two types of fibrous scarring and tubular destruc-tion. The one, present in the subcortical zone, iscomposed of an acellular fibrosis; the other, ex-

tending from medulla to cortex, is relativelyafibrous and is filled with an infiltration of inflam-matory monocytic cells. The former correspondsin location and in its fibrous components with thatseen in the reparative process in cases of shorterduration. The frankly inflammatory nature ofthe latter type of fibrosis seems to be either a new

element in the pathological complex or a greatlyexaggerated one, since in the cases formerly de-scribed cellular infiltration of an inflammatory re-

action, though occasionally present, was at a

minimum. In the present case the noncommittaldiagnosis of sub-acute pyelonephritis is at leastwarranted; it is a matter of speculation what re-

lation it bears to the original lesion of EHF whichappears to have healed with a benign fibrosis andthe destruction of a certain number of nephrons.

RESUME

Concurrent with our description of the struc-tural changes in the kidneys during the succeed-ing phases of EHF have run a discussion of theirsignificance as factors in the renal status and acritical consideration of the pertinent literature.What now follows is an uninterrupted recapitula-tion of the course of the disease, attempting achronological synthesis in structural-functionalcorrelations that will integrate the renal lesion intothe clinical syndrome.

During the first hours of the Febrile Phase theindividual shows clinical evidence in the intenseflush of the skin and visible mucous membranesof vasodilatation of peripheral small vessels. Thecapillaries of the nail folds are widely dilated; thesame is apparently true of the small vessels in therenal vascular bed, as ERBF is either normal orincreased. It can be assumed, therefore, that thekidney during this first stage is flooded with arapid circulation of blood through vessels whichare, save for vasodilatation, essentially normal.Since only traces of protein are found in theurine, it would seem that the glomerular capillar-ies have not as yet suffered the characteristic andcatastrophic lesion of the disease which is to de-velop in the next stage.

This vascular disturbance is revealed in thesucceeding Hypotensive Phase by the abrupt leak-age of plasma from the capillaries, and resultsin a reduced circulating blood volume. It is as-sociated with decrease in arteriolar tone so thathypotension and shock are the result. The struc-tural-functional correlation at this point is clear,for at autopsy the escaped plasma is found in theretroperitoneal spaces.

Lowered arterial pressure of shock is a suffi-cient cause of decreased renal blood flow, butthere are also local disturbances in the renal vascu-lar system. Proteinuria abruptly increases, oftenin the course of a few hours, as evidence of leak-age of plasma through glomerular capillaries thathave undergone damage similar to that which isso widely spread throughout the tissues; the cor-relative structural aspect of the lesion is seen inthe precipitate which fills Bowman's space and inthe areas of intertubular edema, analogous to theretroperitoneal edema, that are found in medullaand cortex. Clearances show a sharp reduction in

128


ERBF even in the absence of clinical shock; ifthe latter concomitantly or subsequently occurs,as it did in most of the fatal cases, a summation ofgeneral and local effects operates to reduce renalblood flow.

It is the failure of renal blood flow, previouslyeither abundant or excessive in an engorged vas-cular bed the arterioles of which have lost theirtone and whose capillaries and venules are perme-able to plasma but not to red blood cells, that isresponsible for the origin of the characteristicrenal lesion of the disease. The distinctive locali-zation of this circulatory disturbance is due to theanatomical and functional peculiarities of the vas-cular system of the kidney which have been previ-ously described. Even with a reduced blood flow,the atonic and permeable vessels of the "medul-lary sponge" in the subcortical zone of the me-dulla, a region shown to be the last to sufferischemia, are flooded and distended to the point oftubular compression. At this stage the circula-tion in cortex and deep medulla is not greatlydisturbed.

The status of the renal circulation is now thatwhich Ricker (20) would have called a "prestatic"congestion, for, judging from histological appear-ances, the blood within the intact dilated inter-tubular capillaries is circulating. There is littlestructural or functional effect other than moderatepressure on the tubules in the region of subcorti-cal congestion. Urinary output has been irregu-larly and variably reduced and follows, in themain, alterations of blood pressure and consequentrenal blood flow. In clinical terms, oliguria hasnot been "established."

In the Transition Period and Oliguric Phase,clinical evidences accumulate of an intensificationof the general damage to small vessels and the con-sequent escape of more than plasma. Petechiaein the skin increase to a maximum on the fifth orsixth day; ecchymoses or hematomata at the siteof trivial trauma, hematemesis, melena and he-moptysis all appear toward the end of the Hy-potensive Phase along with a reduction in plate-lets. So in the subcortical zone of the medulla,hemorrhage, at first by diapedesis and in the endby capillary disruption, infiltrates the renal paren-chyma. All tubules in the areas of involvementnow lie widely separated in a mass of non-circu-lating red blood cells.

The circulatory disturbance has now passedfrom Ricker's prestatic phase to that of rubrostasisand is accompanied by necrosis of those portionsof the nephrons which lie in the anoxic areas ofcongestive stasis. This interruption of the intra-renal passages which tends to isolate cortex frommedulla is not complete, however, for it is formedby the coalescence of focal areas of congestivehemorrhage which arose from the peculiar horse-tail configuration of the arteriae rectae of the lowerglomerular efferents. The tubules of many neph-rons and of the collecting system may be preservedeven in what appears in section to be a severelyaffected kidney. The importance of intact chan-nels through the area of destruction for the con-tinuation of some flow of urine in the OliguricPhase and for its increase in the Diuretic Phasewhich may follow, is apparent.

The clinical observation of the development ofan established oliguria during this period corre-lates with the pathological finding of compressedand interrupted intrarenal channels. Not onlyare the compression and disruption of tubulesseen in the subcortical zone of intertubular hemor-rhage, but, more conclusive, the effect of that ob-struction is evident in the consequent dilatation ofcortical tubules, both proximal and distal, thatlie above it. Such retrograde alterations in thecourse of urine flow through the kidney are madepossible by the presence of intracortical nephronswhich have escaped medullary disruption. It isnoteworthy that this distention disappears withthe onset of diuresis, for it is not present in thekidney of those who have passed through a simi-lar period and died in the Diuretic Phase.

Tubular destruction is not limited to the sub-cortical zone of congestion; stasis in the arterioerectae reaches the papillae and the tissue lesions ofhemorrhage and necrosis follow. Extension intothe cortex is less obvious, yet nephrons from se-verely damaged kidneys regularly showed thetubulorhexic, disruptive lesions of anoxia through-out the proximal convolutions.

Those who died in the Phase of EstablishedOliguria and in the following Phase of Diuresisshowed such extensive damage to both nephronsand collecting system that any adequate struc-tural restitution in their kidneys would seem tohave been impossible. Wehave previously con-sidered in detail the mechanisms of both struc-

129


tural and functional recovery after the tubularnecrosis of acute renal failure (24) ; it is unlikely,therefore, considering the broad extent of tubulethat may be destroyed by the tubulorhexic lesion,that the repair of the damaged nephrons in EHFis possible. Healing, as has been demonstratedin the case of acute renal necrosis, is due to forma-tion of scars that contain afunctional remnants ofdestroyed nephrons.

Survival of the individual and ultimate re-covery therefore depends on the escape of an ade-quate number of nephrons from anoxic necrosis.Since 95 per cent of the individuals with EHFsurvive and, in routine clinical examination dur-ing convalescence, appear to have recovered nor-mal renal function, it follows that the great ma-jority never could have developed the extensivestructural damage that was present in those whodied in oliguria or later diuresis. In the typicalcase of EHF the turning point towards recoveryor exitus comes in the Transition Period; if thevascular disturbance remains at the level of a pre-static congestion with little or only moderate in-tertubular hemorrhage and consequent destruc-tion of a few nephrons, then a return of adequateblood flow is the major requisite for the restitu-tion of the renal status. These relations are il-lustrated graphically in the lowest subdivision ofText Figure 1. Individuals recover from the renallesion of EHF not by repair of nephrons but be-cause a great number of their nephrons have notbeen irreparably damaged.

How frequently future difficulties in renal func-tion are to be anticipated in those who have sur-vived an attack of EHF remains uncertain untilexact measurements by clearances of suitablecases have been accumulated. In all but themilder cases some nephrons have most likely beendestroyed, and it is known that after most formsof acute renal failure, in spite of some hypertrophyand hyperplasia of the survivors (24), this loss isdemonstrable for a considerable time (25).Whether a kidney with such a lessened "reserve"is a potential hazard remains speculative; the onecase available for examination in this seriesshowed the lesions of a sub-acute pyelonephritis.

Since the Transition Period and the OliguricPhase is the time of renal crisis, it would be mosthelpful if some distinguishing clinical character-istic or laboratory procedure might be discovered

that would differentiate the passing "renal in-sufficiency" occurring in the Oliguric Phase inthe individual who ultimately is to recover, fromthe definitive "renal failure" of one who has anirreparable destruction of nephrons and is to die.Can the former temporary insufficiency be con-sidered, for example, a "functional" phenomenonin some nature different from an "organic," ir-reversible failure, and if so cannot the two sortsof disturbance be distinguished by a refinementof clinical or laboratory technique?

These questions of differential diagnosis which,though they certainly have metaphysical impli-cations that need not here concern us, would, ifanswerable, be eminently practical. The prob-lem has been examined under experimental condi-tions in the perfused frog's kidney, where a sim-plification of the factors involved, if extreme, atleast makes its consideration possible (26). Ifthe conclusions of these experiments are ac-cepted, that which has been assumed to be twodifferent sorts of disturbed renal activity, "func-tional" and "organic," are not similar; they arein fact one and identical and no elaboration oftechnique can make them two. It is true that be-cause of present limitations in the scope ofmorphological techniques the visible structuralaspect of the situation at times differs; if onecan see a structural alteration, then the differ-entiation between the evanescent and the ir-reversible renal disturbance may be possible.For these reasons it is understandable how an"established oliguria" in the clinical sense wasobserved in individuals who recovered; in theirkidneys some nephrons may have undergone theirreversible structural lesions that characterizedthe phenomenon in the pathological sense, but not,as in the fatal case, a significant number to pre-clude recovery of an adequate renal status. Todistinguish between the two situations the struc-tural alterations evident in a renal biopsy, whichin the present case was impractical, would berequired.

In general, the mechanisms of the two varia-tions in urinary output, oliguria and diuresis, aresimilar in EHF to those which operate in theclassical example of Acute Renal Failure associ-ated with various forms of traumatic or toxicinjury. In both, a decrease and subsequent restor-ation of renal blood flow would seem to be im-

130


mediate factors in the production of decreased andaugmented urine flow. As accessory mechanismsthat reduce tubule flow in both forms of renaldamage, increased intra-renal tension from inter-stitial edema or swelling of osmotic origin of theepithelium of the proximal convolutions must beconsidered, and any "casts" that may be presentmust act as deterrents of tubular flow.

In contrast to these similarities between the na-ture and causes of decreased urinary output inAcute Renal Failure and EHF one differencehas been previously noted; the establishment ofthe oliguria in the latter. Though the factorswhich establish the oliguria, tubular disruption,are present in all forms of ischemic renal damage,the concentration of its obstructive effect by a zoneof subcortical hemorrhage is present only in EHF.

The sudden onset of diuresis, at times from nourine excretion to over several liters in 24 hours,makes it certain that a circulatory phenomenon isconcerned and not the restitution of some struc-tural element.

It should be noted that the diuresis observedin the fatal cases was not the flood that occurredin those who recovered. In the latter instancesthe urinary output commonly stabilized at 6 to 8liters per day with an exceptional output of 18liters. In the fatal cases here reported, only anoccasional individual passed as much as 4 litersand the more common daily output was around 2.This depression in fatal cases of a diuresis thatwas ''normal" in the recovering case is explicableby the predominating and persisting effect of thestructural alterations making for decreased flowin the Phase of Established Oliguria, namely, in-tertubular hemorrhage and tubular disruption.Here again is evidence that the cases which re-covered did not have the grave disruptive lesionin the nephrons that was observed in the fatalcases.

Variations in urinary volume in Acute RenalFailure and EHF are both "glomerular" and "tu-bular" in their origins. The glomerular functionalmechanism is relatively simple, the tuft filteringmore or less in accord with the flow and pres-sure of blood circulating through its capillaries.Its functional response to the return of circulationis immediate, since its structure has not beengreatly altered and filtration is a simple process.Return of tubular function, absorption of elec-

trolyte and water, is slow because of the need ofelaborate reconstitutions of cellular mechanisms,such as the mitochondrial apparatus (24). Henceit is the temporal relation of varying responses,prompt filtration and delayed reabsorption, to oneevent, the return of circulation, that determines thefinal effect, i.e., diuresis and loss of elements whichare normally conserved.

The tubular element in abnormal variation ofurine output is more complex, even if we ignorethe obligatory phase of water absorption which ispresumed to occur in the distal portions of thenephrons or collecting ducts. To do so in ourproblem would seem permissible, since the magni-tude of output during the Diuretic Phase pointsto trouble in that portion of the nephron wherewe know by direct observation (27) that 80 percent of the glomerular filtrate is absorbed, namely,the proximal convolution.

The anomalous and paradoxical effects of dis-turbances in tubular function are apparent enoughwhen attempts are made to examine them indi-rectly by means of "clearances" done in mam-malian experiments; it is perhaps equally opti-mistic to pass to the other extreme of the situation,in which the experimental oversimplification ofexamining the problem in the perfused frog's kid-neys may seem excessive. However, certain phe-nomena become apparent under these conditionswhich are at least suggestive in a hypotheticalconsideration of the problem in man.

If a frog's kidney, lying in situ, is perfused bythe renal artery and renal-portal vein with modi-fied Locke's solution containing glucose, a nor-mal "urine" is formed in which the effects of tu-bular function are evident in the hyposthenuria,the absence of sugar and, if present in the perfu-sate, the secretion and concentration in the urineof a dye, neutral red. If a poison, urethane orHgCl,, is administered in low dosage to the tu-bules alone and their functions are thus moderatelydepressed, there develops a marked increase in vol-ume output, a fall in rate of dye excretion, an in-crease in total electrolyte elimination, and the ap-pearance of sugar in the urine. If the dosage ofpoison is increased the urine volume decreases,with no return of sugar absorption or dye secre-tion, ultimately reaching zero, and the kidney isseen to be swollen and edematous. Histologicalexamination of such perfused kidneys shows no

131


frank epithelial structural lesions in the "stage ofdiuresis," and marked epithelial destruction andintertubular edema in the later "stage of oliguria"(28, 29).

In a tentative manner one might imagine analo-gous tubular disturbances operating in the com-plexity of the renal lesions of Acute Renal Fail-ure and EHF. Under clinical conditions the se-quence of events in the two diseases is reversed,oliguria and structural damage of the grave renalinsult preceding a phase of restitution of renalfunction that is indicated by diuresis when tu-bules, still inadequate, fail to absorb the fluid ofan increasing glomerular filtration that has fol-lowed a return of renal circulation.

As this discussion has progressed comparisonsand analogies have continuously presented them-selves relating the structural and functional as-pects of the renal lesion of EHF to those of therenal lesion of Acute Renal Failure which is as-sociated with all the various forms of traumaticand toxic injury. In another place (30) theserenal lesions have been compared, in their merg-ing patterns that lie between two types of struc-tural alteration, nephrotoxic and tubulorhexic, tothe gradation of spectral bands that, individual,yet blend to a continuum; in this spectrum eachcase presents its characteristic signature which isderived from the peculiar qualities of its clinicalorigin. On such a spectrum the distinctive ele-ments of the renal lesion in EHF, such as itsprominent Transition Period and the EstablishedOliguria Phase with its causal relation to theinterruption of urinary channels by the charac-teristic subcortical zone of congestive hemorrhage,stand out in bright lined contrast; appearancesboth functional and structural, are very similarand yet sharply different from those of AcuteRenal Failure. It would seem possible to re-solve this apparent anomaly and rationalize ourmetaphor by the recognition that the renal le-sion in EHF is Acute Renal Failure in an indi-vidual whose peripheral vascular bed, includingthe renal, is atonic and permeable as a result ofan infectious disease. Thal's experiments (17)have removed all doubts that infectious noxa canproduce the vascular disturbance of renal ischemiaand its distinctive tubulorhexic necrosis.

There still remains the task of establishing whatthe exact nature of the infectious noxa in EHF

may be. It has been shown that it does not havethe characteristics of a histamine-like substance,and preliminary evidences of a specific vasodilatorsubstance that circulates in the plasma have sofar proved inconclusive (31). As in the morefundamental problem of the nature of the causa-tive agent in the disease, it would seem that herelies hidden what might prove to be the definitivecontribution to the understanding of the renal le-sion in Epidemic Hemorrhagic Fever.

REFERENCES

1. Giles, R. B., Sheedy, J. A., Ekman, C. N., Froeb,H. F., Conley, C. C., Stockard, J. L., Cugell, D.W., Vester, J. W., Kiyasu, R. K., Entwisle, G.,and Yoe, R. H., The sequelae of epidemic hemor-rhagic fever: with a note on causes of death. Am.J. Med., 1954, 16, 629.

2. Froeb, H. F., and McDowell, M. E., Renal func-tion in epidemic hemorrhagic fever. Am. J. Med.,1954, 16, 671.

3. Syner, J. C., and Markels, R. A., Continuous renalclearance studies in epidemic hemorrhagic fever.U. S. Armed Forces Med. J., 1955, 6, 807.

4. Mayer, C. F., Epidemic hemorrhagic fever of theFar East (EHF) or endemic hemorrhagic ne-phroso-nephritis. Lab. Invest., 1952, 1, 291.

5. Mayer, C. F., European outbreaks of the Far-Eastern(Korean) type of epidemic hemorrhagic fever(EHF). Mil. Med., 1955, 117, 502.

6. Hullinghorst, R. L., and Steer, A., Pathology of epi-demic hemorrhagic fever. Ann. Int. Med., 1953,38, 77.

7. Steer, A., Pathology of hemorrhagic fever. A com-parison of the findings-1951 and 1952. Am. J.Path., 1955, 31, 201.

8. Lukes, R. J., The pathology of thirty-nine fatal casesof epidemic hemorrhagic fever. Am. J. Med.,1954, 16, 639.

9. Symposium on Epidemic Hemorrhagic Fever, DavidP. Earle, Ed. Am. J. Med., 1954, 16, 617.

10. Oliver, J., MacDowell, M., and Tracy, A., The patho-genesis of acute renal failure associated with trau-matic and toxic injury. Renal ischemia, nephro-toxic damage and the ischemuric episode. J. Clin.Invest., 1951, 30, 1305.

11. Trueta, J., Barclay, A. E., Daniel, P. M., Franklin,K. J., and Prichard, M. M. L., Studies of theRenal Circulation. Oxford, Blackwell ScientificPublications, 1947.

12. Funck-Brentano, J. L., Contribution a L'Atude duMechanism Physiopathologique de L'Anurie auCours des Nephropathies Aigues. Paris, Impri-merie Cario, 1953.

13. Oliver, J., Straus, W., Kretchmer, N., Lee, Y. C.,Dickerman, H. W., and Cherot, F., The histo-chemical characteristics of absorption droplets in thenephron. J. Histochem. & Cytochem., 1955, 3, 277.

132


14. Greisman, S. E., Nailfold Capillary Observations inHemorrhagic Fever. Report to the Commission on

Hemorrhagic Fever, August 25, 1954.15. Block, M. A., Wakim, K. G., and Mann, F. C., Cer-

tain features of the vascular beds of the cortico-medullary and medullary regions of the kidney.Arch. Path., 1952, 53, 437.

16. Barrie, H. J., Klebanoff, S. J., and Cates, G. W.,Direct medullary arterioles and arteriovenousanastomoses in the arcuate sponges of the kidney.Lancet, 1950, 1, 23.

17. Thal, A., Selective renal vasospasm and ischemicrenal necrosis produced experimentally withstaphylococcal toxin. Observations on the patho-genesis of bilateral cortical necrosis. Am. J. Path.,1955, 31, 233.

18. Peter, K., Untersuchungen uber Bau und Entwick-lung der Niere. Gustave Fischer, Jena, 1927.

19. Oliver, J., The Architecture of the Kidney inChronic Bright's Disease. New York, Paul B.Hoeber, 1939.

20. Ricker, G., Pathologie als Naturwissenschaft, Rela-tionspathologie; fur Pathologen, Physiologen,Mediziner und Biologen. Berlin, J. Springer,1924, p. 104, et seq.

21. Oka, K. A., Zur Histologie der Vinylaminnephritis.Virch. Arch., 1913, 214, 149.

22. Levaditi, C., Eperimentelle Untersuchungen uber dieNekrose der Nierenpapille. Arch. internat. depharmacodyn. et de therap., 1901, 8, 45.

23. Lindega'rd, B., and Lofgren, F., The circulatory pat-tern of the kidney in mercuric chloride poisoning.Kingl. Fysiograf. Sallskapet. Lund Forhand.,1953, 23, 39.

24. Oliver, J., Correlations of structure and functionand mechanisms of recovery in acute tubularnecrosis. Am. J. Med., 1953, 15, 535.

25. Lowe, K. G., The late prognosis in acute tubular ne-

crosis: an interim follow-up report on 14 patients.Lancet, 1952, 1, 1086.

26. Oliver, J., When seeing is believing. Stanford M.Bull., 1948, 6, 7.

27. Walker, A. M., Bott, P. A., Oliver, J., and Mac-Dowell, M. C., The collection and analysis offluid from single nephrons of the mammalian kid-ney. Am. J. Physiol., 1941, 134, 580.

28. Oliver, J., and Shevky, E., Experimental nephritis inthe frog. II. Perfusion methods of testing thekidney by dissociation of its functions. J. Exper.Med., 1931, 53, 763.

29. Oliver, J., and Smith, P., Experimental nephritis inthe frog. III. The extravital production of ana-

tomical lesions in the kidney and their correlationwith the functional aspects of damage. J. Exper.Med., 1931, 53, 785.

30. Oliver, J., The spectrum of structural change inAcute Renal Failure. Mil. Med., 1955, 117, 229.

31. Greisman, S., and Moe, G. K., A Vasodilator Sub-stance in the Plasma of Hemorrhagic Fever Pa-tients. Report to the Commission on HemorrhagicFever, Jan. 8, 1954.

133

fever' - dm5migu4zj3pb.cloudfront.netjapanese (4). more recently, hullinghorst and 1 this work...

Documents