December I953 PINNIGER: Cytological Examination of the Sputum and Pleural Effusion 609

contraindications to pneumonectomy. The mostimportant is the presence of contralateral secondarydeposits. This is, in fact, not a very commonfinding and may be suggested if one or several3 to 30 mm. circular shadows are seen. An isolatedcircular shadow may, of course, be due to atuberculous caseous area or even an old healedinfarct, and such lesions cannot always be excludedunless previous radiographs show they are of recentorigin. Secondary deposits in ribs, clavicles,vertebrae, etc., will be very discouraging. Onthe other hand, local rib erosion by direct extensionof the tumour mass is only a relative contraindica-tion; and on some occasions the neoplasm anderoded rib have been removed at the same opera-tion with apparent success.

Restriction of movement of one side of thediaphragm is unimportant, since it may be theresult of distal inflammation. On the other haiatrue paradoxical movement, often best seen whenthe patient sniffs, is very suggestive of phrenicnerve involvement by glandular secondary de-posits. However, it may be possible to remove theoffending gland if it is on the left side near thepericardium (Tubbs, I951).

When pneumonectomy is under considerationfor the treatment of a carcinoma of the bronchusthe oesophagus should be examined radiologicallyby means of a 'barium swallow.' A fairly thickpaste of barium sulphate is made up with waterand its passage down the oesophagus is observedby fluoroscopy, the patient being rotated so that itcan be seen from several angles. Any abnormalindentations or local deviation should be observedand radiographed. Deviation or indentation un-associated with the atelectasis or the bulk of theneoplasm itself will suggest the presence of medi-astinal secondary deposits. Local irregular fillingdefects, or localized alterations in the mucosalpattern, will strongly suggest invasion of the oeso-phageal wall by the carcinomatous deposits andwill preclude successful resection. The finding ofsuch changes is, fortunately, not common, but this*,is.no rea f g lcn-ep l reure.

BIBLIOGRAPHYBROCK, R. C. (I950), Thorax, 5, 5.DAVIDSON, M. (I95I), 'The Diagnosis and Treatment of Intra-

thoracic New Growths,' London & Oxford University Press.LIVINGSTONE (1952), 'Modem Trends in Tuberculosis,'

Butterworth, London.TUBBS, 0. S. (I95I), 'The Diagnosis and Treatment of Intra-

thoracic New Growths,' London & Oxford University Press.

' CYTOLOGICAL EXAMINATION OF THESPUTUM AND PLEURAL EFFUSION INCARCINOMA OF THE BRONCHUS

By J. L. PINNIGER, M.A., D.M., M.R.C.P.Physician-in-Charge of Clinical Laboratories, St. Thomas's Hospital, London

IntroductionCytological and histological procedures are the

most precise pathological methods available to theclinician in his attempt to make a diagnosis ofcarcinoma of the bronchus. The former have theadvantage that surgery is not required in order toobtain the specimens and on this score alone thereis a strong case for developing the specializedtechnique that is required in their use.

Historically the method has been evolved alongtwo independent channels, one in England havingDudgeon as the fount and inspiration and theother in the United States with Papanicolaou as thepioneer. Sporadic reports of the presence of neo-plastic cells in sputum appeared in the Continental-literature early in the century (von Hoesslin, I92I),

but Dudgeon and his co-workers (Dudgeon andWrigley, I935; Barrett, 1938) and his successorBamforth (1946) were the first to make this diag-nostic method the subject of intensive study. Theygained their initial experience by making scrapingsfrom solid tissues such as the breast (Dudgeon andPatrick, I927; Dudgeon and Barrett, 1934).Papanicolaou's prime interest was in the cytologyof the female genital tract (Papanicolaou, 1933),and his studies were the stimulus for the release,a decade or so after Dudgeon's original publica-tion, of a flood of papers in America on the applica-tion of cytological, technique to malignant con-ditions all over the body, culminating in a thoroughand well illustrated monograph by Graham andher co-workers (Graham, 1950).

Protected by copyright.

on Decem

ber 30, 2019 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.29.338.609 on 1 Decem

ber 1953. Dow

nloaded from

6io POSTGRADUATE MtDICAL JOURNAL beceber I95

Indications for UseSputum. Early diagnosis of bronchial carcinoma

is essential if the remote chance of radical cure isto be achieved. However, a positive cytologicaldiagnosis is so rare before clinical suspicion isaroused that there is no place for making use of thetechnique as part of a routine screening procedurein all people of middle age having chronic coughs.In the group of patients with clinical signs andsymptoms sputum examination should not beundertaken in preference to bronchoscopy unlessthe clinical state of the patient renders the latter adangerous procedure. It has, however, an im-portant place in those patients who have positiveradiological signs, and in whom the tumour isbeyond the range of bronchoscopic vision. In suchcases investigation of many specimens is warrantedif no positive findings are at first obtained. Forthose patients who have a bronchial carcinomawhich is clearly inoperable on clinical grounds,positive cytological confirmation is often valuable.Objective evidence is always more satisfactory thaninference and, if palliative radiotherapy is beingconsidered, is usually required by the radio-therapist before treatment is started.

Pleural fluid. The question of early diagnosisof malignant disease does not arise here, but inthose patients in whom there is uncertainty as towhether an effusion is tuberculous or malignant, anearly decision in favour of tuberculosis is obviouslyof benefit to the patient. As in the case of sputumexamination, a positive cytological diagnosis pro-vides satisfactory objective evidence of the correctdiagnosis in patients with clinically malignantpleural effusions. It is perhaps more importanthere in view of the greater difficulty in obtainingbiopsy evidence.

Other material. Cytological diagnostic methodshave been applied to examination of other materialin patients suspected of bronchial carcinoma.Perhaps the most useful application is in the ex-amination of tissue aspirated at lung puncture fora peripherally situated radio-opaque pulmonarynodule. Examination of bronchial aspirationstaken at the time of bronchoscopy has proved dis-appointing and misleading in our hands having,except in isolated cases, no advantage over sputumpreparations from the same patient (Bamforth,1953; Wetherley-Mein, 1953). Scrapings takenfrom lymph nodes or skin nodules removed atbiopsy can often lead to a confident diagnosis ofcarcinoma within halfan hour of operative removal.This may occasionally be desirable when ananswer is required clinically before a histologicalpreparation is ready. The tissue from which thescrapings are made should still be in the fresh, andnot formalin fixed, state.

TechniqueThe techniques used by most authors are basic-

ally similar, differing only as to the fixatives andstains used. In order to get good preparations it isof cardinal importance to obtain fresh specimens.Sputa must be genuine expectorations so oftensaliva is uselessly presented for examination-andcare must be taken in spreading samples on to theslide, streaky white and especially bloodstainedareas being the most likely sources of tumourcells. With pleural fluids the centrifuged depositis used for preparing the slides. Technical detailsof preparation can be found elsewhere (Dudgeonand Wrigley, I935; Barrett, I938; Graham,1950). Briefly, the slides are placed while wetinto a fixative such as Schaudinn's mercuric chlor-ide solution, stained with haemalum and eosin orcorresponding dyes and finally mounted. At leastthree slides should be prepared from each sample,and these are examined microscopically, using lowand high power dry objectives, with the oilimmersion lens for detail.

HistologyThe identification of malignant cells by this

method requires not only an understanding ofhistological criteria for malignancy but also asound knowledge of the varied cytology of theparticular material examined (Barrett, 1938;Graham, I950). The latter needs many hours ofoften tedious microscopic study, but this is theonly sure way of giving reliable reports as to thepresence of malignant cells. In the sputum thecommon carcinoma cells seen are the small ana-plastic, or ' oat,' cells and squamous cells. Onoccasions the neoplastic cells may show consider-able morphological variation in a single specimen.' Oat' cells are characteristically in small clustersor streaks, appearing as naked nuclei with fineirregular chromatin stippling (Fig. i), whereassquamous cells may be identified either as smallclumps or isolated atypical cells, often having thecharacteristic' bird's-eye ' feature (Fig. 2). Mucussecreting (columnar cell) carcinoma is rare here,but in contrast it is the common form seen in thepleural fluid (Figs. 3 and 4). Carcinoma cellswhen seen in the sputum nearly always come froma primary tumour of bronchus, whereas in thepleural fluid secondary carcinoma cells are as fre-quent as those of primary carcinoma of thebronchus. In the latter group mucus secreting(columnar) cells are most common, oat cells aremuch less frequent and squamous cells are rare(Bamforth, 1946).

ResultsIt is reasonable to give one of three reports on

the examination of prepared slides-that carcinoma

Protected by copyright.

on Decem

ber 30, 2019 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.29.338.609 on 1 Decem

ber 1953. Dow

nloaded from

December I953 PINNIGER: Cytological Examination of the Sputum and Pleural Effusion 6ii

.. ...... ....... .... ........

FIG. i.-Oat cell carcinoma in sputum showing thecharacteristic lack of cytoplasm and stippling ofnucleus. X 500.

... ......:

... .. ...

... .. ..

..

FIG. 2.-Squamous cell carcinoma in sputum showinga 'birds-eye' cell lying at the top left hand cornerseparate from the main mass. X 500.

.sS4 1!

..... ......

W..SS.S..S.'.......

s eWFIG. 3.-Low power view of sediment from pleural

fluid showing the characteristic clumps of mucussecreting (columnar cell) carcinoma. X I 20.

.... .....

............. . ....

FIG. 4.-High power view of one of the clumps takenfrom Fig. 3. X 500.

cells either are, or are not, present or that -someatypical cells, which defy precise classification, areobserved. This last report must in no way be takenas indicating carcinoma, but implies that a searchof further specimens is warranted. In good handsa positive report is as definite as a positive histo-logic finding. False positives decrease with in-creasing experience, but even in expert handsare not entirely unknown, owing to the occasionalatypical cellular proliferation which occurs ininflammatory states such as bronchiectasis (Dud-geon and Wrigley, I935; Bamforth, 1946;Woolner and McDonald, I949). It is easier to becertain of a positive diagnosis in sputum than inpleural fluid, for in the latter case endothelialcell proliferation can at times be so marked insome non-malignant states that cellular clumps,closely simulating carcinoma, are produced in thefluid. The cells in Fig. 5 are taken from aneffusion following a pulmonary infarct. Errors inpleural fluid will be minimized if positive diagnosesare limited to those which show a number of large

clumps of hyperchromatic and atypical cells in thecentrifuged deposit, as in Figs. 3 and 4. Indeed itis a good working rule both for sputum andpleural fluid, in view of the seriousness of a falsepositive report, that carcinoma should never bediagnosed unless the findings are unequivocal. Ifany element of doubt exists a report should in-dicate that atypical cells have been observed,inviting the submission of further specimens forexamination.

Clinical action on a negative cytological reportmust take notice of the patient's symptoms andsigns. If the latter indicate carcinoma as a likelydiagnosis-and no other information is forth-coming, such as by the bronchoscope or by lymphnode or other secondary biopsy-then, as statedearlier, repeated examination of many samples ofsputum is warranted. Ir the case of pleuraleffusion, sputa should also always be examinedand, of course, the same applies to any subsequentpleural fluids if further samples become available.

It is difficult to state precisely what percentageof negative reports are false ones, as the long term

D

Protected by copyright.

on Decem

ber 30, 2019 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.29.338.609 on 1 Decem

ber 1953. Dow

nloaded from

6I2 POSTGRADUATE MEDICAL JOURNAL December I953

*F'

iL

FIG. 5.-High power view of endothelial cell clumpsin a pleural fluid from a case of pulmonary in-farction. x 500.

follow-up of all such cases investigated is well-nigh impossible in a hospital having many out-patients. It is certain that such false negativesexist and form an appreciable percentage of thewhole. There are various reasons for this.Specimens may be unsatisfactory because they arenot fresh or because in the case of sputa they con-sist mainly of saliva. Some tumours desquamateless easily than others, figures given by someauthors for sputum indicating that these formapproximately 30 per cent. of the total (Dudgeonand Wrigley, I935; Gowzr, I943; Woolner andMcDonald, I949). When occasional exfoliatedcells are present singly, they may not be sufficietitlydistinctive to warrant a positive diagnosis. Finally,not the least important factor in many laboratoriesis the falling off of acuity of observation if themicroscopist is faced with a large number of slidesrequiring attention. In many centres in theUnited States cytological diagnostic proceduresare undertaken by staffs devoting their whole time

to the study, and consequently screening of pre-pared slides can be a leisured and thorough pro-cess. In this country, by contrast, cytologicaldiagnosis is but one of the many jobs that fall to aclinical pathologist, with the result that the caregiven to the individual specimen must vary in-versely with the total number submitted.

SummaryIt can be said that cytological examination of

sputa and pleural fluids is a very useful tool in thediagnosis of carcinoma of the bronchus, and aftera long trial period has come to stay. Considerableexperience is necessary before reliable interpreta-tion of results can be achieved, and the numbXer ofpositive diagnoses will be increased by an in-telligent selection of clinical material and by thesubmission of adequate freshly obtained specimensfor cytological analysis, whether they be sputum,pleural fluid or biopsy tissue.

AcknowledgmentsI wish to acknowledge the help I have received

from Drs. J. Bamforth, D. S. Cadman and G.Wetherley-M\,ein in preparing this article. Mythanks are also due to Mr. A. E. Clark for thephotomicrographs.

BIBLIOGRAPHYBAMFORTH, J. (1946), Thorax, I, II8.BAMFORTH, J. (I953), Practitioner, in the press.BARRETT, N. R. (1938), Y. Thorac. Surg., 8, I69.DUDGEON, L. S., and PATRICK, C. V. (I927), Brit. J. Surg.,

IS, 250.DUDGEON, L. S., and BARRETT, N. R. (934), Ibid., 22, 4.DUDGEON, L. S., and WRIGLEY, C. H. (I935), J. Laryng.,

50, 752.GOWAR, F. J. S. (I943), Brit. J. Surg., 30, 193.GRAHAM, R. M., and other members of the Vincent Memorial

Laboratory staff (1950), 'The Cytologic Diagnbsis of Cancer,'W. B. Saunders Co., Philadelphia and London.

VON HOESSLIN, H. (I92I), 'Das Sputum,' Julius Springer,Berlin.

PAPAN1COLAOU, G. N. (1933), Amer. J. Anat. Supplement, 52,519.

WETHERLEY-MEIN, G. (1953), Personal communication.WOOLNI;R. L. B., and McDONALD, J. R. (949), Amer. j. Clin.

Path., 19, 765.

RUTHIN CASTLE,NORTH WALESA Clinic for the diagnosis and treatment of Internal Diseases (except Mental or Infectious Diseases). The

Clinic is provided with a staff of doctors, technicians and nurses.The surroundings are beautiful. The climate is mild. There is central heating throughout. The annual

rainfall is 30.5 inches. that is, less than the average for England.The Fees are inclusive and vary according to the room occupied.

For particulars apply to THE SECRETARY, Ruthin Castle, North Wales.Telegrams: Castle, Ruthin. Telephone: Ruthin 66

Protected by copyright.

on Decem

ber 30, 2019 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.29.338.609 on 1 Decem

ber 1953. Dow

nloaded from