endocrine cells tumour-bearing - thorax

Thorax 1996;51:721-726

Endocrine cells in tumour-bearing lungs

J D H Sheard, J R Gosney

Department ofPathology,University ofLiverpool,PO Box 147,Liverpool L69 3BX,UKJ D H SheardJ R Gosney

Correspondence to:Dr J R Gosney,University Department ofPathology,Duncan Building,Royal Liverpool UniversityHospital,Daulby Street,Liverpool, L7 8XW, UK.

Received 23 June 1995Retumed to authors6 September 1995Revised version received17 January 1996Accepted for publication25 January 1996

AbstractBackground - When hormones are de-tected in the serum of patients withbronchial carcinoma they are generallyconsidered to originate from the tumour,but this may be not the only explanation.Pulmonary endocrine cells proliferate inlungs affected by non-neoplastic diseaseand their products are often demonstrablein the serum. The aim of this study was toexamine the pulmonary endocrine systemsof a series of tumour-bearing lungs to see

whether any changes in them could pos-sibly account for raised levels of pul-monary peptides in the blood.Methods - The morphology, number, dis-tribution, and content ofpulmonary endo-crine cells in 30 pairs of tumour-bearinglungs from patients coming to necropsywith bronchial carcinoma were examined.These features were related to the path-ology of the tumour and to other patho-logical changes present in the lungs, andcompared with pulmonary endocrine cellsin 10 pairs of control lungs from patientswithout pulmonary disease.Results - Increased numbers of endocrinecells, often in the form of large abnormalaggregates, were present in 17 pairs oftumour-bearing lungs where they were

associated not with the tumour but withnon-tumoral pathology, especially in-flammation and changes associated withcardiac failure. Appropriate and in-appropriate peptides were identifiedwithin them.Conclusion - The possibility is raised that,in some subjects with bronchial carcinomawho have raised serum hormone levels,the source of these substances might bethe endocrine cells in the diseased lungaround the tumour.(Thorax 1996;51:721-726)

Keywords: pulmonary endocrine cells, ectopic se-cretion, bronchial carcinoma.

When increased levels of amine, peptide, or

protein hormones are found in the blood ofpatients with bronchial tumours, they are gen-erally assumed to be products of the neoplasm.However, it has often proved difficult or im-possible to localise the hormone in question tothe tumour allegedly producing it, irrespectiveof how it is sought and even when the wholeof the tumour has been available for study.'2This is usually attributed to the kinetics ofsynthesis and secretion or inadequacies of theavailable technology,' but an alternative ex-

planation has been proposed'4 - namely, that

the substance or substances in question mayarise not from the tumour itself, but fromproliferating endocrine cells in the diseasedlung around it. With this hypothesis in mind,we have investigated the nature and extent ofany changes that might have developed in thepulmonary endocrine system of lungs con-taining primary bronchial malignancies, andhave considered whether they could possiblyexplain the increased hormone levels oc-casionally found in the serum of such patientswhich are usually attributed to the tumouritself.

MethodsThirty pairs of lungs from patients comingto necropsy with primary bronchial carcinomawere studied, together with 10 pairs of controllungs from age matched subjects with no clin-ical or necroscopic evidence of pulmonary dis-ease. Cases were included only if the necropsyhad been performed within 48 hours of thedeath of the patient. Necropsy material wasused because it allowed extensive sampling,not only of the tumour and the lung adjacentto it, but also of the remaining ipsilateral andthe contralateral lung. This was essential forstudying the entire endocrine cell populationirrespective of its topographical relationship tothe neoplasm. Lungs were distended with 10%neutral buffered formalin, fixed for 24-48 hoursand extensively sampled, 10 blocks of tissuebeing taken from the primary tumour, 5-10from the pulmonary tissue immediately ad-jacent to it (that is, from the lobe into whichit had grown or within which it had arisen),and 5-10 blocks from each of the remainingfour pulmonary lobes.

Tissue was embedded in paraffin wax. Sec-tions from all blocks were stained with haema-toxylin and eosin and examined by lightmicroscopy to allow classification of thetumour5 and histopathological scrutiny of bothlungs. Adjacent sections were immunolabelledby the avidin-biotin complex technique6 forneuron specific enolase (NSE) and protein geneproduct 9 5 (PGP 9 5), general markers of cellsof the diffuse endocrine system,' as well asfor a range of secretory products of humanpulmonary endocrine cells including gastrinreleasing peptide (GRP), calcitonin, calcitoningene-related peptide (CGRP), and 5-hydroxy-tryptamine (5-HT; serotonin), all of which arefound in the endocrine cells of normal lungs,and certain substances which have been de-scribed in pulmonary endocrine cells in dis-eased lungs'78 - namely, adrenocorticotrophin(ACTH), arginine vasopressin (AVP), leucine(leu-) enkephalin, somatostatin, human growth

721

on February 22, 2022 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.721 on 1 July 1996. D

ownloaded from

http://thorax.bmj.com/

Sheard, Gosney

Antisera used for immunohistochemistry

Antisera* Source Dilution Incubation Tissue controls

PGP 9 5 Biogenesis 1:2000 18 hours Human pancreas(Bournemouth, UK)

NSE DAKO Ltd 1:200 45 minutes Human pancreas(High Wycombe, UK)

Calcitonin DAKO Ltd 1:2000 45 minutes Medullary carcinoma of human thyroid gland(High Wycombe, UK)

GRP DAKO Ltd 1:2000 18 hours Human fetal lung(High Wycombe, UK)

CGRP Peninsula Labs 1:4800 45 minutes Medullary carcinoma of human thyroid gland(St Helens, UK)

5-HT Peninsula Labs 1:1000 45 minutes Human ileum(St Helens, UK)

ACTH DAKO Ltd 1:250 45 minutes Human pituitary(High Wycombe, UK)

HGH DAKO Ltd 1:800 45 minutes Human pituitary(High Wycombe, UK)

AVP ICN Biomedicals Ltd 1:800 45 minutes Human pituitary(High Wycombe, UK)

Somatostatin Seralab 1:400 45 minutes Human pituitary(Crawley Down, Sussex, UK)

Leu Peninsula Labs 1:800 45 minutes Human adrenal medulla(St Helens, UK)

VIP Various Various Various Various

PGP=protein gene product; NSE=neuron-specific enolase; GRP=gastrin-releasing peptide; CGRP=calcitonin gene-relatedpeptide; 5-HT =5-hydroxytryptamine (serotonin); ACTH = adrenocorticotrophin; HGH =human growth hormone; AVP=arginine-vasopressin; Leu=leucine-enkephalin; VIP=vasoactive intestinal polypeptide.

hormone (HGH), and vasoactive intestinalpolypeptide (VIP).

Following treatment with hydrogen peroxidein methanol to inactivate endogenous per-oxidase and with normal swine serum to pre-vent non-specific binding of antibody, sectionswere incubated with the primary anti-serumfor a variable period of time depending on itsnature and concentration. This was then linkedto the avidin-biotin complex by biotin con-jugated donkey anti-rabbit immunoglobulin(for polyclonal antisera) or donkey anti-mouseimmunoglobulin (for monoclonal antisera).The chromogen used was 3',3'-diamino-benzidine tetrahydrochloride dihydrate. Ap-propriate positive and negative tissue controlswere employed in all cases. The source, work-ing dilution, incubation period, and the natureofthe control tissue used for each ofthe primaryantisera are shown in the table. Immuno-labelled sections were examined by light micro-scopy and the presence or otherwise of theabove substances was sought in the tumour aswell as in the endocrine cells of the lung, theirmorphology, distribution, and content all beingnoted and topographically interrelated.There are various ways in which pulmonary

endocrine cells can be quantitated.' The mostaccurate is expression of their number in termsof either the total epithelial population (endo-crine cells/1 03 epithelial cells) or of the unitlength of epithelium (endocrine cells/cm epi-thelium). However, in this study detailedquantitation proved impossible to apply withany meaning because the changes in endocrinecells observed were not only present in par-enchyma as well as in the airways, but werefrequently gross and focal and of an irregulardistribution. For the same reasons, definingendocrine cell subpopulations by their contentof secretory products in this way also provedunworkable. Changes in endocrine cells were

therefore expressed semiquantitatively ac-cording to the sequence of morphologicalchanges they have been shown to pass throughwhen they increase in number and becomegradually disorganised. The first of these is anincrease in their number so that, instead ofbeing evenly scattered throughout the bronchialand bronchiolar mucosa as is normally the case,they line up to form interrupted rows (IRs)along the basement membrane. The next stageinvolves the formation of nodular aggregates(NAs) of more than 10 cells arranged in adisorganised fashion basally and suprabasally,sometimes ensheathing the entire lumen of thesmall airways or arising in the parenchyma.The degree of endocrine cell proliferation

was expressed according to the number of IRsand/or NAs in the tissue examined as follows:0-5 IRs per five sections = +; 5-10 IRs perfive sections = 2 +; > 10 IRs per five sections =3 +; 0-5 NAs per five sections = 4 +; > 5 NAsper five sections = 5 +.

ResultsAll 10 pairs of control lungs (mean age 75years, range 58-89) contained endocrine cellswith a morphology, distribution, and contenttypical of the endocrine system of healthyhuman lungs (fig 1).9 Thus, the majority weresolitary with a regular distribution in intra-pulmonary bronchi and bronchioles and alwaysbasally located in the epithelium. The onlythree clusters of endocrine cells identified inthese control lungs were small (comprising 2-3cells), basal, compact and regular, the typicalappearance of neuroepithelial bodies'0 whichare normally present in adult human lungs butin very small numbers.9 About two thirds ofthe pulmonary endocrine cells revealed withantisera to NSE and PGP 9 5 contained GRPand most ofthe remainder contained calcitonin

722



j.com/


ownloaded from



t....; .. Al..A:,

P.'...i--.

:. ..

V..

Figure 1 One solitary endocrine cell and one cluster,probably a neuroepithelial body, in a terminal bronchiolein one of the control lungs. This normal arrangement ofendocrine cells was seen also in 13 of the 30 pairs oftumour-bearing lungs in which there was no non-tumoralpathology. ABC method for PGP 9 5; originalmagnification x 600.

or CGRP. Occasional cells were weaklyimmunoreactive for 5-HT, but none of theother secretory products sought was identified.Ofthe 30 bronchial neoplasms, 12 were small

cell and 18 were non-small cell tumours. Nineof the latter were squamous, five adeno-carcinomas, two mixed adenosquamous, andtwo morphologically undifferentiated. Eight ofthe 12 small cell tumours were immunoreactivefor NSE and/or PGP and there was focal im-munopositivity for 5-HT in three. Two con-tained ACTH or a substance crossreacting withit. None of the other substances sought wasidentified in the group of small cell lesions,although five of the non-small tumours showedimmunoreactivity for NSE and/or PGP.

In 13 pairs of tumour-bearing lungs the pul-monary endocrine cells appeared entirely nor-mal in number, distribution and content, withcharacteristics identical to the equivalent popu-lations in the 10 pairs of control lungs. In noneof these lungs was there any significant non-tumoral pathology apart from focal collapse inthe region of the tumour. In particular, therewas no evidence of consolidation, cardiac fail-ure (intra-alveolar haemorrhage or oedema),or fibrosis.

In the remaining 17 pairs of tumour-bearinglungs, however, proliferation ofpulmonary endo-crine cells had occurred to varying degreesin the form of IRs (fig 2), NAs (fig 3), or both,and in these lungs there was significant non-tumoral pathology, the most common ofwhichwas infective consolidation. This was a featureof 1 0 ofthe peritumoral and 11 ofthe ipsilateral/contralateral specimens. It was usually as-sociated with the most marked form of endo-crine cell proliferation, namely NAs (fig 3).These structures were seen in five of the 10specimens of peritumoral lung (proliferation4 +) and eight of the 11 specimens of ipsilateraland contralateral lung (proliferation 4 +)affected by consolidation. In the nine remainingspecimens of pulmonary tissue showing in-fective consolidation, pulmonary endocrinecells had proliferated to form IRs in six (pro-liferation + in two, 2+ in one, and 3+ inthree; fig 2), being normal in just three. The

.....r...

4'

Figure 2 An interrupted row of endocrine cells in atumour-bearing lung affected by bronchopneumonicconsolidation. ABC method for PGP 9 5; originalmagnification x 600.

Figure 3 A nodular aggregate of endocrine cells in atumour-bearing lung affected by bronchopneumonicconsolidation. ABC method for PGP 9.5; originalmagnification x 600.

most florid proliferation of all was seen in asubject with candidal infection in whom NAswere numerous and contained very large num-bers of cells (proliferation 5+; fig 4). Of theremaining five specimens in which NAs werefound but in which there was no consolidation,evidence of cardiac failure was present. In allof these proliferation was 4 +.There was no suggestion of any association

between endocrine cell proliferation andtumoral pathology. Of the 12 lungs with NAsof pulmonary endocrine cells, six containedsmall cell and six non-small cell tumours (NS,Fisher's exact test). There was no suggestioneither that NAs were more prevalent closer tothe tumour and, in fact, most were found inipsilateral lung distant from the tumour or incontralateral lung rather than in a peritumorallocation. In two samples of lung which ap-peared entirely normal, the endocrine cells werealso normal. The endocrine cells were alsonormal in five samples of lung where therewere pathological changes - two in which therewas consolidation alone, one in which therewas consolidation as well as evidence of cardiacfailure, and two in which evidence of cardiacfailure was accompanied by changes of chronicbronchitis or bone marrow embolism.

723

AL

....



j.com/


ownloaded from


Sheard, Gosney

Figure 4 A very large nodular aggregate of endocrine cellsin a lung affected by a severe candidal pneumonia. ABCmethod for PGP 9 5; original magnification x 600.

bFigure 5 Immunoreactivity ofproliferating endocrine cellsfor (a) ACTH and (b) HGH in candidal pneumonia.ABC method for ACTH and HGH; originalmagnification x 600.

In areas oftumour-bearing lung where endo-crine cells were normal, their content was

predominantly of GRP and calcitonin, as inthe control lungs. In IRs calcitonin often pre-dominated but, apart from one case in whichthere was candidal infection, aberrant sub-stances - that is, products not found in normalhuman pulmonary endocrine cells - were notgenerally seen. In this particular case, however,in which proliferation of endocrine cells was

particularly florid (5 +), ACTH and HGH were

both seen (fig 5). In three further cases ACTHwas demonstrable but HGH was not. In one

further case HGH was found but immuno-reactivity for ACTH was absent. AVP, somato-statin, and leu-enkephalin were not seen inany cells. Labelling for VIP was abandonedbecause of inadequate results with control tis-sue and despite using a variety of antisera andlabelling conditions. In a number of IRs andNAs none of the secretory products sought -

normal or aberrant - could be demonstrated.There was no obvious relationship between

the substances synthesised by any of thetumours and those found in the proliferating

endocrine cells in the surrounding lung, al-though in one case ACTH was present in boththe neoplasm (a small cell carcinoma) and theNAs in the lung around it where there wasevidence of cardiac failure.

DiscussionSynthesis and secretion of a wide range ofamine, peptide, and protein hormones is awell recognised feature ofbronchial carcinoma,especially small cell carcinoma.' These sub-stances are detectable in a significant pro-portion of patients with the disease andoccasionally cause clinical effects, Cushing'ssyndrome due to release of ACTH-like sub-stances being the classical example. So ac-cepted has this association become thatincreased levels of such hormones in patientswith pulmonary tumours are automatically as-sumed to be due to secretion by the neoplasm,despite the fact that the substance in questionis often undetectable in tumour tissue. Ourfindings suggest that an alternative mechanism,namely leakage of these substances from pro-liferating endocrine cells in diseased lung out-with the tumour, might be responsible.Although the functions of pulmonary endo-

crine cells remain unclear, there is considerableevidence now to suggest that they are almostcertainly involved in the pathogenesis of orresponse to a variety of pulmonary diseases,including such conditions as bronchial asthma,respiratory infections, chronic bronchitis andemphysema, cystic fibrosis, hypertensive pul-monary vascular disease, eosinophilic gran-uloma, and several respiratory disorders ofearlylife.'78"' 18 In all of these conditions endocrinecells increase in number in what appears to bea fairly stereotyped way, their normal evenlyspaced sparse distribution being first replacedby interrupted rows of cells (IRs) and then,probably only if the insult persists, by nodularaggregates of cells (NAs) which pile up toprotrude into airways or alveoli.' In tandemwith these changes, calcitonin often replacesGRP as their predominant secretory productand, occasionally, aberrant products appear.'The observation of this pattern of endocrinecell proliferation has been made repeatedlyin lungs affected by the wide range of non-neoplastic conditions outlined above,'78"1-8and we consider this response to pulmonaryinjury to be well established in the literature.There is evidence also to suggest that the

products of proliferating endocrine cells some-times leak into the circulation. Most attentionhas focused on calcitonin, with hyper-calcitoninaemia and hypercalcitoninuria welldocumented accompaniments of a range ofinflammatory pulmonary diseases,'920 butACTH is also sometimes increased in patientswith non-neoplastic pulmonary disease2122 inwhom it is predominantly in the "big" formand is non-suppressible, features which suggestthat it is not of pituitary origin. The results ofinvestigations such as these, and of the presentstudy, suggest that increased serum levels ofa variety of hormones should be frequentlydetectable in the blood of patients with non-

724

...-Amkla..,,,..,..c

diihh.lnolL iM-m-



j.com/


ownloaded from



neoplastic pulmonary disease. Surveying theavailable literature reveals that this is, in fact,the case.' Indeed, one of the problems whichhas confounded the use of numerous initiallypromising hormonal serum markers of bron-chial carcinoma over the years is the prevalenceof increased levels in the blood in patients withinflammatory disorders of the lungs.' It is thislack of specificity, rather than sensitivity, whichhas led to eventual disappointment in theirpractical use as markers of pulmonary malig-nancy.The characteristics of the endocrine cell pro-

liferation described in this study are typical ofthat described in previous work from this'2 1317and other78"'4- 1618 laboratories and raise cur-rently unanswered questions about their pos-sible role in the development of or responseto a range of pulmonary disorders. The clearassociation between proliferation of these cellsand non-tumoral pathology gives support forsuch a role, whereas the absence of any linkwith the tumour itself suggests that there is nodirect association between them. A graduallyincreasing body of evidence from studies ofnormal pulmonary development as well as dis-eased lungs points strongly towards a rolefor peptides such as GRP in the growth, de-velopment, and repair of pulmonary tissues,1 23an hypothesis supported by the findings of thepresent study. Why the most marked pro-liferation (NAs) should be associated with con-solidation of the parenchyma or its filling withoedema fluid or haemosiderin-laden macro-phages due to cardiac failure is unclear, but itis interesting to note that pulmonary infarctiondue to thromboembolism, in which the alveoliare flooded with blood, has also been associatedwith proliferation of pulmonary endocrinecells'7 (and, indeed, with "ectopic" secretionof ACTH22). Perhaps the severity of the par-enchymal damage is an important factor.The lungs we studied were taken at necropsy,

at the end point of the natural history of bron-chial carcinoma when the associated non-tumoral pathology described is common. How-ever, earlier in the course of the disease, suchas at initial presentation, these changes areless often present, so that the endocrine cellproliferation we describe might be less markedand less likely to account for the increasedserum levels of hormones sometimes seen insuch patients at this earlier stage. We have noway of knowing whether significant endocrinecell proliferation is present at this time, a ques-tion which could be adequately answered onlyby studying surgically resected lungs. Evidencefrom studies of the endocrinology of acuteinflammatory pulmonary disease, however,suggests that the proliferation ofendocrine cellsin such circumstances is a rapid event and oftenextremely florid12 20 so that a causal relationshipwith increased serum hormone levels early inthe course of bronchial carcinoma seems by nomeans implausible.Why aberrant substances such as ACTH

and HGH should appear at a time of markedproliferation of endocrine cells is uncertain, butit is interesting to note how both substanceshave been demonstrated in tumourlets,24 the

locally invasive aggregates of pulmonary endo-crine cells first described by Whitwell25 whichprobably represent a further stage in their pro-liferation beyond that of NAs and which areseen only when the pulmonary damage stim-ulating it is particularly severe or prolonged.'The appearance of these substances is probablyjust a functional accompaniment of increasingmorphological disorder.The undisclosed products that might be pres-

ent in the IRs and NAs which immunolabelledonly for NSE and PGP 9 5 is a matter forspeculation. A range of substances beyondthose we were able to seek has been described inhuman pulmonary endocrine cells in occasionalstudies,26 but whether they were present insome of the proliferating cells in the presentinvestigation we cannot say.Our hypothesis is not unprecedented. Recent

studies of hepatobiliary disease describe neuro-endocrine differentiation in and elaboration ofparathyroid hormone-related protein by biliaryepithelium, suggesting that the paraneoplastichypercalcaemia sometimes seen in patientswith hepatocellular carcinoma may be a con-sequence of changes in the bile ducts ratherthan inappropriate secretion by the tumour.2728Because we were confined to necropsy material,a consequence of needing to sample both lungsextensively, we were unable to measure serumlevels of the substances we localised by im-munolabelling so the mechanism we suggestremains unproven and based on only cir-cumstantial evidence. A logical next step wouldbe to extend the study to surgically resectedpulmonary tumours and the endocrine cells ofthe pulmonary tissue removed along with them,and to relate the findings to serum levels ofhormones measured in blood taken before sur-gery.

1 Gosney JR. Pulnionary endocrine pathology: endocrine cellsand endocrine tuniours of the lung. Oxford: Butterworth-Heinemann, 1992.

2 Coates PJ, Doniach I, Howlett TA, Rees LH, Besser GM.Immunocytochemical study of 18 tumours causing ectopicCushing's syndrome. 7 Clin Pathol 1986;39:955-60.

3 Blackman MR, Rosen SW, Weintraub BD. Ectopic hor-mones. Adv Intern Med 1978;23:85-113.

4 Gosney MA, Gosney JR, Lye MDW. Serum adreno-corticotrophin and calcitonin in non-small cell carcinomaof the bronchus. Life Sci Adv 1991;10:53-7.

5 World Health Organization. The World Health Organizationhistological typing of lung tumours. 2nd edition. An _7Clin Pathol 1982;77:123-36.

6 Hsu SM, Raine L, Fanger H. Use of avidin-biotin peroxidasecomplex (ABC) in immunoperoxidase techniques: a com-parison between ABC and unlabelled antibody (PAP)procedures. 7 Histochenm Cytocheni 1981;29:577-80.

7 Gould VE, Linnoila RI, Memoli VA, Warren WH. Neuro-endocrine components of the bronchopulmonary tract:hyperplasias, dysplasias and neoplasms. Lab Invest 1983;49:519-37.

8 Tsutsumi Y, Osamura RY, Watanabe K, Yanaihara N.Immunohistochemical studies on gastrin-releasing peptideand adrenocorticotropic hormone-containing cells in thehuman lung. Lab Invest 1983;43:623-32.

9 Gosney JR. Neuroendocrine cell populations in postnatalhuman lungs: minimal variation from childhood to oldage. Anat Rec 1993;236:177-80.

10 Lauweryns JM, Peuskens JC. Neuroepithelial bodies(neuroreceptor or secretory organs?) in human infantbronchial epithelium. Anat Rec 1972;172:471-82.

11 Stanislawski EC, Hernandez-Garcia J, Mora-Torres MA,Abrajan-Polanco E. Lung neuroendocrine structures.Topography, morphology, composition and relation withintrinsic asthma (non-immune). Arch Invest Med 198 1; 12:559-77.

12 Allibone RO, Gosney JR. Rapid changes in pulmonaryendocrine cells in acute bronchopneumonic consolidation._7Pathol 1990;161:347.

13 Gosney JR, Sissons MCJ, Allibone RO, Blakey AF. Pul-monary endocrine cells in chronic bronchitis and em-physema. _7 Pathol 1989;157:127-33.

725



j.com/


ownloaded from


Sheard, Gosney

14 Fukayama M, Hayashi Y, Shiozawa Y, Furukawa E, FunataN, Koike M. Human chorionic gonadotropin alpha-sub-unit in endocrine cells of fibrotic and neoplastic lung. LabInvest 1990;62:444-51.

15 Johnson DE, Wobken JD, Landrum BG. Changes inbombesin, calcitonin and serotonin immunoreactive pul-monary neuroendocrine cells in cystic fibrosis and afterprolonged mechanical ventilation. Amn Rev Respir Dis 1988;137:123-31.

16 Aguayo SM, King TE, Waldron JA, Sherritt KM, KaneAAA, Miller YE. Increased pulmonary neuroendocrinecells with bombesin-like immunoreactivity in adultpatients with eosinophilic granuloma. Clin Invest 1990;86:383-44.

17 Heath D, Yacoub M, Gosney JR, Madden B, Caslin AW,Smith P. Pulmonary endocrine cells in hypertensive pul-monary vascular disease. Histopathology 1990;16:21-8.

18 Johnson DE, Georgieff MK. Pulmonary neuroendocrinecells. Their secretory products and their potential role inhealth and chronic lung disease in infancy. Am Rev RespirDis 1989;140:1807-12.

19 Becker KL, Nash D, Silva OL, Snider RH, Moore CF.Increased serum and urinary calcitonin levels in patientswith pulmonary disease. Chest 1981;79:211-6.

20 Becker KL, Silva OL, Snider RH, Moore CF, GeelhoedGW, Nash D, et al. The pathophysiology of pulmonarycalcitonin. In: Becker KL, Gazdar AF, eds. The endocnine

lung in health and disease. Philadelphia: WB Saunders,1984:277-99.

21 Ayvazian LF, Schneider B, Gerwirtz G, Yalow RS. Ectopicproduction of big ACTH in carcinoma of the lung: itsclinical usefulness as a biologic marker. Anm Rev, Respir Dis1975;111:279-87.

22 Dupont AG, Somers G, VanSteirteghem AC, Watson F,Vanhaelst L. Ectopic adrenocorticotropin production: dis-appearance after removal of inflammatory tissue. ClinEndocrinol Metab 1984;58:654-8.

23 Miller YE. Pulmonary neuroendocrine cells and lung de-velopment: dim outlines emerge. 7 Clin Invest 1994;91:1861.

24 Gosney JR, Green ART, Taylor W. Appropriate and in-appropriate neuroendocrine products in pulmonarytumourlets. Thorax 1990;45:679-83.

25 Whitwell F. Tumourlets of the lung. Pathol Bacteriol 1955;70:529-41.

26 Polak J, Becker KL, Becker E, Cutz E, Gail DB, Gon-iakowska-Witalinska L, et al. Lung endocrine cell markers,peptides and amines. Anat Rec 1993;236:169-71.

27 Roskams T, Willems, M, Campos RV, Drucker DJ, YapSH, Desmet VJ. Parathyroid hormone-related peptideexpression in primary and metastatic liver tumours. Histo-pathology 1993;23:519-25.

28 Burt AD, MacSween RNM. Bile duct proliferation - itstrue significance? Histopathology 1993;23: 599-602.

726



j.com/


ownloaded from


endocrine cells tumour-bearing - thorax

Documents