[CANCER RESEARCH 40, 395-402, February19800008-5472/80/0040-0000$02.OO

Cyclophosphamide-induced Changes in the Cellular Composition of aMethyicholanthrene-induced Tumor and Their Relation to BoneMarrow and Blood Leukocyte Level&

Robert Evans,2 Laird D. Madison, and Denise M. Eidlen

The Jackson Laboratory, Bar Harbor, Maine 04609

ABSTRACT

A single i.p. injection of cyclophosphamide(CV) (6 mg)induced regression of a methylcholanthrene-induced fibrosarcoma (MCA/76-9) frequently to a nonpalpable mass. Theextent of regression was dependent on dose of drug and timeof administration in relation to tumor size, but even underoptimal conditions few (<2%) permanent regressions wererecorded. Well-marked histological changes were observedafter CY injection, involving an initial increase in macrophagenumbers followed by an accumulation of mature and immaturegranulocytes. These cells largely replaced neoplastic cells by7 to i 0 days after CV injection. Changes observed histologically were verified and quantitated after enzymic disaggregation of tumors at various times before and after CV injection. Abioassay was used to assess the number of tumorigenic cellsassociated with the tumor mass after CV injection by injectinggraded doses of cells from enzymically disaggregated tumorsinto normal recipients. Within 1 hr after CV injection, more than1 0@ tumor cells were required for a take (compared with 102

cells from control tumors). However, by 3 days fewer cells(10@)were required, and by i 0 days only 102 cells wereneeded. Bone marrow and peripheral blood leukocyte numbersdecreased rapidly after CV injection, but absolute numbersreturned to control values by 7 to 10 days. However, differentialcounts on blood smears indicated that lymphocyte countsrequired at least 21 days to return to control values, whereasthe number of monocytes and granulocytes was elevated abovethose of control mice by Days 7 to 10. The possibility isdiscussed that the observed intratumor cellular changes mightreflect a variety of different reactions unrelated to tumor regression or recurrence.

INTRODUCTION

During the last few years, there have been numerous reportson the presence of host-derived leukocytes within progressingand regressing tumors. The relevance of these cells to thenatural history of the tumor is still obscure, although it hasbeen reported that macrophages and lymphocytes isolatedfrom progressing and regressing tumors are able to destroytarget cells or to stimulate proliferation, depending on theconditions (5, 7). Whether such leukocytes perform eitheractivity in situ is uncertain. Thus, one of the major problemsduring drug-induced antitumor action is not knowing what the

I This research was supported by NIH Grant CA 20920.2 To whom requests for reprints should be addressed.

3 The abbreviations used are: CY, cyclophosphamide, Fc@ cells, Fc-receptor

positive cells; EA, antibody-coated sheep red blood cells; BMC, bone marrowcells; PMN. polymorphonuclear cell(s).

Received June 6, 1979; accepted November 6, 1979.

precise mechanism of successful antitumor action is or whatevents are actually occurring at the tumor site. Histologicalevaluation and electron microscopy (2, 20) have made someinroads by showing that certain events, such as the accumulation of macrophages, may occur during defined forms oftreatment, but quantitative information is difficult to obtainunder such conditions. Clearly, under conditions where leukocyte numbers may increase or decrease at the tumor site,such as during some forms of drug treatment (4, 22), quantitative and qualitative information regarding their precise functions would be of value.

In the present series of experiments, we used CY3as a probeto investigate the possible interrelationship of host cells in thetumor mass and leukocytes in bone marrow and peripheralblood and whether any such relationship may correlate withdrug-mediated antitumor effects. CV has well-defined in vivoeffects. The drug may immunosuppress recipients; it destroysdividing cells, normal as well as tumor, and has a profoundeffect on hemopoietic tissue; it destroys suppressor T- and B-lymphocytes (9, 1i ), and there are reports that CY inhibits thecytotoxicity of T-lymphocytes by destroying the cells responsible for their generation (14, i 6). Its overall mechanism ofaction during therapy is, therefore, complex, on the one hand,facilitating destruction of neoplastic cells by the direct actionof its metabolites on DNA synthesis and by removing suppressor cells and, on the other hand, preventing the generation ofpotent effector cells. Although the duration of CV action islimited to only a few hr, it is evident that a fine balance iscreated between the reduction of tumor burden, subsequentproliferation of residual neoplastic cells, and the generation ofantitumor effector mechanisms.

The present report is a confirmation and extension of previous studies (6, 7). The system to be described involves amurine fibrosarcoma that responded predictably to a single i.p.injection of CV by regressing initially to a small or nonpalpablemass, and then in most cases by regrowing, and finally byoverwhelming the host. The purpose of this investigation wasto further our understanding of the nature of the events associated with the antitumor effects mediated by CV and with thesubsequent recovery and proliferation of cells.

MATERIALSAND METHODS

Mice. Throughout the experiments, female C57BL/6J mice,6 to 8 weeks old, (i 8 to 20 g) were used (5 mice/group).

Tumors. The MCA/76-9 fibrosarcoma was induced by Dr.Liisa Prehn using 5% methylcholanthrene. Injection of i 02tumor cells consistently produced tumors in at least 60% ofcontrol mice. The tumor was immunogenic, immunized micerejecting an i.m. injection of up to 106 cells. Experiments

FEBRUARY1980 395

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

R. Evans et a!.

described in this paper were spread over about 9 months,utilizing tumors from animal passages 3 through i 2. Tumorswere maintained by injecting i O@or i 06 cells i.m. at regularintervals. ‘‘Daysof tumor growth' ‘given in the text refers to thetime elapsing between injection of tumor cells and measurement of a particular parameter.

Disaggregation Procedure. This has been described fullyelsewhere (3). The popliteal lymph node was removed beforeexcision of the tumor mass. The excised tumors were mincedand totally disaggregated by using trypsin (250 j@g/ml,Sigmatype III, twice crystallized; Sigma Chemical Company, St. Louis,Mo.), collagenase (250 @tg/mI;Sigma type I), and DNase (1rig/mI; Sigma). Cells were washed twice by centnfugation at1000 rpm and resuspended at i 06 cells/mI of Roswell Park

Memorial Institute Tissue Culture Medium 1640. Counts weremade in a hemocytometer using phase-contrast optics. Viability(usually >90%) was assessed by combined trypan blue exclusion and phase-contrast microscopy. The number of Fc@cellswas assessed by mixing the tumor cell suspension with EA atan EA:tumor cell ratio of 20: 1. Rosetting was estimated on thebasis of hemocytometer counts. In addition, cytocentrifugepreparations were made using a Perkin-Elmer cytospin (Coleman Instruments Division, Oak Brook, III.) before and after EArosetting. Cytochemical stains were used to identify nonspecific esterase in macrophages or monocytes and specific esterase in granulocytes, using a-naphthylbutyrate and naphtholAS-D chloroacetate (Sigma), respectively, as substrates. During the series of experiments, tumors underwent a change; cellsuspensions derived from tumor passages 3 to 6 were shownto contain 19 ±2.3% (S.D.) Fc@cells (predominantly macrophages), but by passage 7 and through to passage i 2, thispercentage increased to 32 ±i .2%. At the same time, tumorgrowth rates appeared to increase when a standard inoculumof 106 cells was used. The percentage of granulocytes (abouti 1 %) did not show any apparent changes. The reasons for the

changes were not investigated. Passage i 2 tumors were immunologically similar to earlier passages as tested by immunogenicity experiments.

CV. Cytoxan(Mead Johnson,Evansville,Ind.)was dissolvedin distilled water at a concentration of 60 mg/mI. In all experiments, except those cited in Chart 1, 0.1 ml (6 mg) CV wasinjected i.p. This was equivalent to approximately 300 mg/kg.Mice were given injections when tumor diameters had reachedabout iO mm.

BMC. These were obtainedby flushingthroughthe tibiawith2 ml of Roswell Park Memorial Institute Medium 1640 containing DNase, 1 @sg/ml.

Blood. This was obtained from the retroorbital sinus. Bothperipheral blood and bone marrow leukocytes were counted ina Model ZBI Coulter Counter. Blood smears were prepared byusing the Perkin-Elmer cytospin and stained with Wright'sstain. Differential counts were performed routinely. Resultswere expressed as the number of cells per ml of blood on thebasis of the counts from the Coulter Counter.

Histology. Sections of tumors were stained by the conventional hematoxylin-eosin method.

RESULTS

Relationship between CV Dose and Tumor Regression.Tumor-bearing mice were given i.p. injections of graded doses

of CV on Day 13 after i.m. implantation of i O@cells. Chart ishows a direct correlation between dose, extent of tumorregression, and subsequent tumor regrowth. Doses of 6 and 8mg were not significantly different in their effects on tumorgrowth. However, 8 mg caused i 5% of the mice to die (with orwithout tumor). In subsequent experiments, a dose of 6 mgwas used. Very few permanent regressions (<2%) have beenseen in 5 experiments including nearly 400 CY-injected tumorbearers. In most cases, tumors regrew 2 to 3 weeks after CVinjection, some tumors appearing sooner than did others,hence the observed wide range in variation seen from Day 30onwards. The growth rate of recurring tumors appeared to besomewhat faster than that of untreated control tumors. A second injection of CV given when tumor recurrence was evidentusually had only a slight effect on tumor growth and often noapparent effect at all. This was not due to the emergence ofCY-resistant tumor lines since CV treatment of mice aftertransplantation of these tumors into normal recipients resultedin the same type of result seen in Chart i.

Tumor burden also was a limiting factor in the antitumoreffect of CV, since CV treatment after Day 13 resulted in lessof an effect than when CV was given before this time. Anotherlimiting factor related to the rate of tumor growth. The experiments summarized in Chart i were representative of a seriescarried out on tumors derived from animal passages 3 to 6.However, passage 7 tumors were observed to increase in sizemore rapidly from a standard inoculum of cells, and it wasnecessary in subsequent experiment (passage 7 through i 2)to inject CV at earlier times to reproduce the sequence ofevents seen in Chart i . Injection of CV at a later time eitherproduced only temporary growth inhibition or no apparentchange in growth rate.

Cellular Compositionof Tumors. A summaryof the numberof Fc@cells and PMN in cell suspensions obtained by enzymicdisaggregation of tumors at various times before and after CVinjection is given in Table 1. The salient features are: (a) controltumors showed a progressive increase in both Fc@cells andPMN, which formed about 20 and 10%, respectively,of thetotal cell population throughout. Control tumor suspensionscontained about 60 to 70% neoplastic cells; (b) total cell yieldsdecreased progressively after CV injection until about Day i 8when tumors were very small or nonpalpable. At this time,enzyme disaggregation produced relatively few cells; (C) Fc@cells, which were mostly macrophages, showed an increase byDay 4 but after Day 7 they declined in parallel with the decreasein tumor size. Numbers increased again on reemergence of thetumors; (d) PMN numbers showed a fairly sharp increase fromDays i 0 to i 4 followed by a decrease thereafter. Cytocentrifuge preparations showed that many of the PMN were immaturering forms; (e) neoplastic cells, identified in cytocentrifugepreparations, became fewer from Day 4 until tumors regrew.From Days 4 to 7, there were many large multinucleate cellsand few typical neoplastic cells. By Day 10, these large cellshad more or less disappeared. By Day i 8, neoplastic cellswere estimated to form <1 5% of the total cell population. Therewere also unidentified cells (<20% of the total) of differingmorphology. These probably included vascular endothelialcells, fibroblasts (especially from Days i 0 to i 8) and somelymphocytes.

In essence, the above changes were seen throughout thecourse of the experiments using tumor passages 3 to i 2.

396 CANCERRESEARCHVOL. 40

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Analysis of the host-derived Ieukocytes found associated with regressingandrecurringMCA/76-9tumorsTumor

bearers, 5/group, were given Injections of 6 mg CY i.p. on Day 9oftumorgrowth.Total

cell counts/tumor mess (x 10@@)aDays

of TotalcellTumorsgrowth yield Fc@cellsPMNControls

7 31 ± 70b 78(25)C3.3(11)979±14.4 14.2(18)4.6(6)CY-treated

0(—Day9) 79±14.4 14.2(18)4.6(6)464± 8.5 25.6(40)3.2(5)777±10.5 25.4(33)7.7(10)1044 ± 5.5 13.2(30) 15.4(35)18

4± 2.5 1.1(26)1.6(40)25234±77 46.8(20) 19.2(8)

CY-induced !ntratumor Changes

Chart 1. RelationshIp between dose of CYand antl-MCA/76-9 fibrosarcoma effect. Eachpoint represents the mean of up to 10 tumors.Bars, S.D.

DAYS AFTER iMPLANTATION

siderin. By 7 days (Fig. 3), the outstanding feature of sectionswas the presence of large numbers of multinucleate and giantneoplastic cells that emphasized the presence of numerousmononuclear cells and to a lesser extent PMN. By I 0 days,these giant cells had more or less disappeared leaving fewdemonstrable residual neoplastic cells. Although mononuclearcells appeared to comprise the bulk of the tumor mass (Fig. 4),distinct, and frequently large, focal points of darkly staininghyperchromatic cells and granulocytes were clearly visible(Figs. 5 and 6). These appeared to be localized in the capsularregion and in the muscle. The groups of hyperchromatic cellswere sometimes bounded by vascular endothelium. Large numbers of mature and immature granulocytes (ring forms) wereassociated with these areas (Fig. 7). On Day 14, when regression had evidently ceased, sections showed much fibrous,birefringent material with which fibroblasts were associated.By Day 2i , when tumor growth recurred, increase in tumorsize was accompanied by an increase in the number of mitoticfigures and identifiable neoplastic cells. The pigmented macrophages became localized in the capsular region, which became relatively thick compared with the capsule of tumors seenbefore CV treatment.

Bone Marrow and Peripheral Leukocyte Levels. Charts 2and 3 summarize the overall data on BMC and WBC when micewere given injections of CV 8 days after tumor implantation.Charts 4 to 6 summarize the blood leukocyte differentials. It isseen that BMC and WBC of control, non-tumor-bearing micereturned to normal values 7 to i 0 days after CV injection, andthis was essentially similar to the reaction seen in tumor bearers. The differential count showed that monocyte numbers(Chart 4) fell rapidly after CV injection of normal and tumorbearers, but by Day 7 (15 days after tumor implantation) thenumbers had risen to above those of the untreated counterparts. Although monocyte levels of treated normal mice returned to untreated control values soon after this, those oftumor bearers returned to normal at a slower rate. The elevatednumber of monocytes seen in untreated tumor bearers on the25th day of tumor growth was not seen in the CY-treatedtumor-bearing mice. This difference has been seen in 3 separate experiments. Granulocyte numbers (Chart 5) showed asimilar decline and rise after CV treatment. However, unlike

Table 1

U Mean value of 5 tumors.

b Mean ±S.D.C Numbers In parentheses, percentages of each cell type within the total cell

population.

However, when tumors from passages 7 to i 2 were used inCY-induced regression experiments, it was not uncommon forthe macrophage content to reach levels of 60 to 70% duringthe 7 days immediately after CV injection. The following histologicaldescriptionis concernedwith tumorsfrom passages7through 12.

HistologicalEvaluation.Thehistologicalappearanceof control tumors (Fig. 1) was typically fibrosarcomatous. They weretypified by having a relatively high density of mononuclear cellsin the subcapsular region (not shown here) with only a modestand scattered infiltration of mononuclear cells elsewhere. PMNwere rarely seen outside areas of necrosis and then mainly inlow numbers in the muscle. Within 24 hr after CV injection,tumors underwent profound morphological changes, and by 3days intratumor damage was visible (Fig. 2). Nuclei of neoplastic cells were frequently pyknotic or swollen, and the cytoplasmwas often vacuolated. Overall, there was a decrease in stainingintensity of sections, presumably due to damage or loss ofcells. Mitotic figures were seen, but often these were of abizarre nature. Mononuclear cells became especially prominent throughout the tumor section. Large cells containing pigment were often seen associatedwith areas of hemorrhage.These were considered to be macrophages containing hemo

FEBRUARY1980 397

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

-,:1@@@ ‘‘‘‘‘‘

.—..TUMORWNTROL0—a TUMOR + CV

0- -0 6ONTRCL CV

P. Evans et a!.

0.2

0.

-J

0

0.‘0

@ 06

‘0

@ 0.5

8 0 2 4 6 8 20 22 24 26DAYS AFTER IMPLANTATION

Chart 2. Effect of CY on BMC from tumor-bearing (MCA/76-9) mice. Six mgCY were injected p. on Day 8 of tumor growth.

Chart 3. Effect of CY on peripheral blood leukocyte counts of tumor-bearing(MCA/76-9) mice.

Chart 4. The effect of CY on blood monocyte levels of tumor-bearing (MCA/76-9) mice.

monocyte counts, those of PMN in untreated and treated turhorbearers showed a substantial increase above the control background. Lymphocyte numbers (Chart 6) were largely unaffected by the presence of the growing tumor. After CV injection

DAYS AFTER IMPLANTATION

Chart 5. The effect of CY on blood grenulocyte levels of tumor-bearing (MCA/76-9) mice.

Chart 6. The effect of CY on blood lymphocyte levels of tumor-bearing (MCA/76-9) mice.

of control mice or tumor bearers, overall lymphocyte numbersdeclined and did not return to untreated control values for 2idays.

Tumorigenicity of Neoplastic Cells Recovered after CVTreatment. To assess what proportion of the total cell yieldwas able to induce tumors in normal recipients, tumors wereexcised at intervals before and after CV injection and enzymically disaggregated. Serial dilutions of the total cell yield wereprepared, and 0.i ml was injected into each of 5 mice, whichwere then observed for tumor growth. Table 2 summarizes thedata in terms of the incidence of tumors after injecting gradeddoses of cells. Within i hr of CV injection and up to 24 hrfollowing, there was a significant reduction in the number of

2 4 6 8 20 22 24 26DAYSAFTERIMPLANTATION

-J

‘Ua-

I-

0DAYS AFTER IMPLANTATION

DAYS AFTER IMPLANTATION

398 CANCERRESEARCHVOL. 40

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table2Tumorigenicityof MCA/ 76-9 tumor cells after injection of mice withCVMice

were given injections of 6 mg CY 9 days after i.m. injection of 106MCA/76-9fibrosarcome cells.Lowest

dose of cells inducing a tuTime after CY injectionmor1

hr 1064hr2.5x106(2O)l2hr

2.5x106(2O)1day 2.0 x106(20)3days 1o@(40)7days 10'(60)1

0 days 102(60)14 days102(60)Control

tumor cells102(60)a

Numbers In parentheses, percentage of incidence of tumors.

CY-induced Intratumor Changes

ending with the change from a densely packed cellular mass toone consisting of fibrous material and few cells. During theseextremes, there was an initial increase in the proportion of Fc@mononuclear cells, mainly macrophages, as also observed byothers (22). This was followed by an increase in the number ofmature and immature granulocytes. At this stage, therefore,the tumor mass was essentially granulation tissue containingrelatively low numbers of neoplastic cells. In view of the reportsthat granuloma formation at the site of tumor growth oftenleads to tumor regression, e.g. , after intralesional injection ofBacil!us Ca!mette-Guërin(20), these, together with the increase in the ratio of macrophages and PMN to residual neoplastic cells, might have led one to expect eradication ofresidual neoplastic cells. This was rarely seen. Moreover, datafrom the bioassay indicated that the number of tumorigeniccells was increasing almost in parallel with the increase inmacrophages and PMN even though the tumor mass wasdecreasing in size. These results, therefore, would seem toindicate that the tumor-associated host cells present at thistime were not responsible for the observed tumor regressionand that regression was probably mediated by the direct effectof CV metabolites. That the antitumor action was of relativelyshort duration was indicated by the increase in the number oftumorigenic cells within 3 days of CV injection. It would followthat the mechanism required for total tumor eradication waseither absent in the present system or inhibited in its action.

The enzymic disaggregation of tumors yielded data thatconfirmed the histological appearance of tumors in that relatively high proportions of macrophages and PMN were seen atthe appropriate times. Cytocentrifuge preparations showed themorphologically different cell types in proportions similar tothose expected from the histological appearance. This wouldindicate that the enzymic disaggregation procedure did notselectively kill cells, although a certain amount of cell deathmust be expected when this technique is used. Whether theenzymes used affected the physiology, function, or structureof the various cell types was not evaluated. The Fc receptorsof the macrophages were clearly intact, although whether theavidity for binding EA had been modified is not known. Thefailure of most PMN to bind EA, as reported previously (3),might indicate a change induced by enzymic treatment. Theneoplastic cells from control-tumor were not dramaticallychanged in that a high percentage of takes was always obtamed from low doses of cells. Although higher doses of tumorcells were required for a take after CV treatment (see Table 2),it seems unlikely on the basis of the histological appearance of

tumors that the take rate was unduly influenced by the actionof enzymes on the isolated tumor cells. The pattern of eventssuggested that tumorigenic cell numbers were rapidly reducedto low levels by the antitumor action of CV and that a certainamount of time was required before proliferation occurred andtumors became palpable again. What has yet to be establishedis whether the macrophages or the PMN contained in the tumorcell inocula were responsible for the poor take rates whentumors were disaggregated soon after CV injection, i.e. , within24 hr. Conversely, the contribution of the high proportion ofmacrophages and PMN in the inocula of tumors taken 7 to 14days after CV treatment to the high take rate from smallnumbers of cells also needs evaluation.

The interrelationship of blood and tumor-associated leukocytes is far from clear. Little is known about the kinetics of the

tumorigenic cells compared with control tumors which at allstages tested (from 7 to 20 days) showed at least a 60% takerate after injection of 102 cells. By 3 to 7 days after CVinjection, however, the number of cells required for a take wasclearly decreasing, and by Day 14 there was no differencefrom that needed for a take in control mice.

DISCUSSION

In the above tumor system, we used CV as a probe toinvestigate the cellular changes induced peripherally and withinthe tumor mass. The intention was to further our understandingof the nature of the changes occurring during CV-inducedtumor regression and tumor recurrence. The approach wasdesigned to cover 2 important aspects of tumor-host relationships, namely, the relationship between peripheral and intratumor leukocytes and the effect of the latter on the behavior ofthe tumor, whether this involved regression or progression.

CV is a well-establishedantitumorreagent, the mechanismof action of which is mediated basically through the effect of itsmetabolites on DNA synthesis of the neoplastic cells. That thismechanism alone is insufficient to account for tumor eradication is evident from a number of reports (1, 13, 15, i 9, 2i),which suggest that immunity directed towards the neoplasmmay effectively deal with the tumor burden when this has beenreduced. Whether immune reactivity is involved or necessaryin order to obtain only temporary tumor regression is still anopen question. Radov et a!. (19) reported that melphalan orCV-induced tumor regression was dependent on establishedantitumor immunity. However, a comparison between the antitumor effects of CV in a nonimmunogenic and immunogenictumor model system (6) indicated that there was no apparentcorrelation between CV-induced tumor regression, induction ofimmunity, or the immunogenicity of the tumor. Similarly, in thepresent system, when mice were thymectomized and irradiatedbefore injection of the MCA/76-9 fibrosarcoma cells, theyresponded to the antitumor effects of CV as readily as controlmice, although none of the groups showed permanent regression.4 The fact that few permanent regressions of MCA/76-9tumors were obtained suggests that there was a distinctionbetweenthe mechanismsinvolvingtumorregressionandthoserequired for total and permanent eradication of neoplastic cells.

Injection of CV into MCA/76-9 tumor-bearing mice resultedin the tumors regressing, often to nonpalpable masses. Histologically, well defined and reproducible events were observedbeginning with immediate changes in neoplastic cells and

4R.Evans,D.M.Eidlen,andG.J.Baigent,unpublisheddata.

FEBRUARY1980 399

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

R. Evans et a!.

accumulation of tumor-associated host cells (8), how long theypersist within the tumor, whether they proliferate in situ, orwhether they move away from the tumor at any given time. Ithas been suggested that accumulation of monocytes in theblood during tumor growth is the result of a tumor product thatinhibits migration to sites of inflammation (17, 18). However,there is no direct evidence that monocytes fail to migrate to thetumor site, even though a monocytosis may be demonstrable.Similarly, a granulocytosis is sometimes seen (10, 12), as inthe current experiments, but there is no evidence that this isindicative of a loss of migratory potential to the tumor mass. Inthe present experiments, there was no evidence to support theview that the observed monocytosis or the granulocytosisresulted in decreased levels of these cells at the tumor site.Unfortunately, until there is more information on the kinetics ofinfiltration of tumors by host leukocytes, it is not possible todraw conclusions concerning the precise relationship betweenperipheral and tumor-associated leukocytes. In the above experiments, data were presented for an active accumulation ofboth macrophages and PMN at the tumor site up to 10 daysafter CV injection of mice. The accumulation was evidenthistologically and demonstrable by cell counts on enzymedisaggregated tumors. The distribution of macrophages andPMN was, however, quite different; the former spread throughout the regressing tumor, and the latter more or less localizedto the muscle and capsule. The significance of this is unknown.What is also not clear is whether this accumulation was due toproliferation of immature forms in situ or to extravasation fromthe blood capillaries, or to both. Mitotic figures were seenamong the hyperchromatic cell population with which PMNrather than macrophages were usually associated, but as yetthere is no evidence that the PMN were derived from thesecells. Autoradiography studies should indicate whether DNAsynthesis is occurring at these hyperchromatic focal points,and injection of tumor cell suspensions into lethally irradiatedmice might indicate whether these regressing tumors containprecursor cells for the macrophage and PMN lineage.

Since the results of the above experiments suggest that thetumor-associated host cells were not directly involved in theoverall mechanism of CV-induced regression, the questionarises as to the reasons for their presence. A speculativediscussion on this aspect has been presented elsewhere (7).An aspect that should not be ignored is that, during the processof regression, other basic biological reactions were presumablytaking place. For example, the localized degeneration of themajority of neoplastic cells implies that there was a need toeliminate debris or dead cells from the site. Perhaps the observed granuloma formation fulfilled this requirement. In addition, the initial growth of the i.m. tumor and its subsequentregression must have involved severe local tissue damage.Since mice regained the use of the limb in which the tumor hadregressed, wound healing and formation of scar tissue musthave ensued. Such reactions are characterized initially by thepresence of granulation tissue. Moreover, the possibility thatthe accumulating leukocytes contributed towards the successful proliferation of tumorigenic cells during the period of granuloma formation cannot be ignored.

Protocols involving the use of chemotherapeutic agents,such as CV, are largely empirical and based on observationsderived from a combination of in vitro data and in vivo trialexperiments that have been designed to obtain maximal ef

fects. Often a high degree of success is obtained, if not intotally eradicating the tumor mass, then in temporarily checkingthe tumor growth. Little is known about the nature of the eventsoccurring at the level of the tumor mass, which must ultimatelybe the focal point of investigation. This report demonstrateschanges occurring within the tumor and in the blood. It formsthe beginning of a series of experiments that will determine thefunctional capacity of tumor-associated host cells duringregression and tumor recurrence and also whether the findingscan be utilized and extended to rationalize certain practicalaspects of adjuvant therapy.

REFERENCES

1. Burkltt, D. P. Host defence mechanism In Burkitt's lymphoma and Kaposi'ssarcoma. Br. Med. J., 4: 424—426, 1970.

2. Dlluzzio, N. R., McNemee, R., Jones, E., Cook, J. A., end Hoffman, E. 0.The employment of glucan and glucan-activated macrophages In the enhancement of host resistance to malignancies in experimental animals. In:M. Fink (ed), The Macrophage in Neoplasia pp. 181—198. New York:Academic Press Inc., 1976.

3. Evans, R. The effect of X-irrediation on host cell infiltration and growth of amurine fibrosercome. Br. J. Cancer, 35: 557—566,1977.

4. Evans, R. The effect of ezathioprine on host cell infiltration and growth of amurine fibrosercome. mt. J. Cancer, 20: 120—128, 1977.

5. Evans, R. Macrophages In solid tumors. In: K. James, B. McBride, and A.Stuart (eds.), The Mecrophage end Cancer, pp. 321—329,EdInburgh:Econoprint, 1977.

6. Evans, R. Failure to relate the anti-tumor action of cyclophosphamlde withthe immunogenicity of two murine fibrosercomes. Int. J. Cancer, 21: 611—616, 1978.

7. Evens, R. Host cells in transplanted murine tumors end their possiblerelevance to tumor growth. J. Reticuloendothel. Soc., 26: 427—437,1979.

8. Evans, R. Cellular basis for regulation of tumor growth, In: I. Wltz and M. G.Henna (eds.), Contemporary Topics in Immunobiology, New York: PlenumPublishing Co., In press, 1980.

9. Gill, H. K., and Liew, F. Y. Regulation of delayed-type hypersensitivity. Ill.Effect of cylcophosphamide on the suppressor cells for delayed-type hypersensitivity to sheep erythrocytes in mice. Eur. J. Immunol., 8: 172—175.1978.

10. Hibberd, A. D., and Metcalf. D. Proliferation of mecrophage and granulocyteprecursors in response to primary and transplanted tumors. lsr. J. Med. Sci.,7:202—210, 1971.

11. Katz, S. I., Parker, D., end Turk, J. L B-cell suppression of delayedhypersensitivity reactions. Nature (Lond.), 25 1: 550—551. 1974.

12. Lappat, E. J., and Cewein, M. A study of the leukemold response totransplantable A-280 tumor In mice. Cancer Res., 24: 302—311, 1964.

13. Mathé,G.. Halle-Panneko, 0., end Bourut, C. Immune manipulation by BCGadministered before or after cyclophosphemide chemo-immunotherapy ofLi 21 0 leukemia. Eur. J. Cancer, 10: 661 —666,1974.

14. Miller. R. G., Schilling, R. M., end Philips, R. A. Requirement for non-T cellsIn the generation of cytotoxic T lymphocytes (CTL) in vitro. II. Charecterizetionof the activecellsinthe spleenof nudemice.J. Immunol.,118: 166—174, 1977.

15. Moore, M., and Williams, D. E. Contribution of host Immunity to cyclophosphamide therapy of a chemically-induced murine fibrosercome. Int. J. Cancer, 11:358-368, 1973.

16. Neta, R., end Salvin, S. B. T and B lymphocytes in the regulation of delayedhypersensitivity. J. Immunol.. 117: 20i 4—2020,1976.

17. Normann, S. J., Schardt, M., and Sorkin, E. Do tumors escape surveillanceby depression of macrophage inflammation? In: K. James, B. McBride, endA. Stuart (eds.), The Mecrophege and Cancer pp. 247—257.EdInburgh:Econoprint, 1977.

18. Pike, M. C.. and Synderman, R. Depression of mecrophege function by afactor produced by neoplasms: a mechanism for abrogation of immunesurveillance. J. Immunol., 117: 1243—1249, 1976.

19. Radov, L. A., Haskill, J. S., and Kom, J. H. Host immune potentiation ofdrug responses to a murine mammary adenocarcinoma. mt. J. Cancer, 17:773—779,1976.

20. Snodgrass, M. J., and Henna, M. G. Ultrastructural studies of histocytetumor cell interactions during tumor regression after intralesional Injectionof Mycobacterium bovis. Cancer Res.. 33: 701-71 6. 1973.

21. Steele, G., and Pierce, G. E. Effects of cyclophosphamide on Immunityagainst chemically induced syngeneic murine flbrosarcomes. Int. J. Cancer,13: 572-578, 1974.

22. Szymanlec. S.. and James, K. Studies on the Fc receptor bearing cells in atransplanted methylcholanthrene induced mouse fibrosarcome. Br. J. Cancer, 33: 36-50, 1976.

400 CANCERRESEARCHVOL. 40

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CY-induced Intratumor Changes

Of •«•£X-4Õ!

t •-<w. ••.^/ 1" » «»•.»Sbd

Fig. 1. Histological section of a control MCA/76-9 fibrosarcoma (passage 9). x 200.

Fig. 2. Histological section of MCA/76-9 fibrosarcoma (passage 9) 3 days after mice had received 6 mg CY ¡.p.X 200.

Fig. 3. Mistological section of MCA/76-9 fibrosarcoma (passage 9) 7 days after CY injection. Multinucleate, giant cells, and smaller mononuclear cells are visible.X 200.

Fig. 4. Histological section of MCA/76-9 fibrosarcoma (passage 9) 10 days after CY injection. Few neoplastic cells are visible among the prominent mononuclearcells, x 200.

FEBRUARY 1980 401

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

R. Evans et a!.

- ±I@,.@

Fig. 5. Same as Fig. 4, showing dense hyperchrometic foci in the region of the capsule end muscle. x 50. The circled area is enlarged in Fig. 6.Fig. 6. Enlargement of circled area of Fig. 5 showing densely stained cells and granulocytes in capsule. X 200.Fig. 7. Histological appearance of the MCA/76-9 fibrosarcoma 10 days after CY injection. The capsule end muscle show the intensive accumulation of mature

and immature granulocytes. x 200.

402 CANCER RESEARCH VOL. 40

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1980;40:395-402. Cancer Res   Robert Evans, Laird D. Madison and Denise M. Eidlen  Relation to Bone Marrow and Blood Leukocyte LevelsComposition of a Methylcholanthrene-induced Tumor and Their Cyclophosphamide-induced Changes in the Cellular

  Updated version

  http://cancerres.aacrjournals.org/content/40/2/395

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

on April 13, 2017. © 1980 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from