397

Beck, M. D., Eaton, M. D., O’Donnell, R. (1944) J. exp. Med. 79,65.

Heilman, F. R., Herrell, W. E. (1944) Proc. Mayo Clin. 19, 204.Nigg, C. (1942) Science, 93, 49. Parker, R. F., Diefendorf, H. W. (1944) Proc. Soc. exp. Biol., N.Y.

57, 351.Sorsby, A. (1945) Brit. med. J. i, 903. Todd, E. W. (1945) Lancet, i, 74. Turgassen, F. E. (1944) J. Amer. med. Ass. 126, 1150.

ASPIRATION LIVER BIOPSY

TECHNIQUE AND DIAGNOSTIC APPLICATION

SHEILA SHERLOCK, M B EDIN., M R C PBEIT MEMORIAL RESEARCH FELLOW

Department of Medicine, British Postgraduate Medical School

DESPITE the battery of laboratory tests available forstudying hepatic diseases, diagnosis is often difficult.Any safe, sure diagnostic method is therefore to bewelcomed. When Iversen and Roholm (1939) publishedtheir series of 160 aspiration liver biopsies with nomishaps, it was believed that this method might becomegenerally applicable. However, early results obtainedby Dible, McMichael, and Sherlock (1943) indicated thatthe method was not without risk. Later modificationsin the selection of cases and in technique have increasedthe safety of the procedure, and we now believe thataspiration liver biopsy has a definite place in the elucida-tion of obscure hepatic diseases. Therefore we publishthe technique in detail with some indications of its use indiagnosis. ,

Technique of Aspiration Liver BiopsyPREPARATION OF PATIENT

To a jaundiced patient it is advisable to give a vitamin-Kpreparation for three days before the puncture. If thejaundice is non-obstructive (urobilin present in the urine),two 5 mg. tablets of ’ Kapilon are given by mouth threetimes a day. If the jaundice is obstructive, the vitaminmust be administered intramuscularly ; 5 mg. daily is asuitable dose.In every case, before biopsy is performed, the patient’s

blood-group should be known, and two pints of compat-ible blood should be readily available. Aspiration liverbiopsy should not be performed unless adequate facilitiesfor blood-transfusion are available in case of complicatinghaemorrhage. A sedative is unnecessary before puncture.

HEPATIC PUNCTURE

The patient lies supine in bed with the right side asnear the edge of the bed as possible. A firm pillow maybe placed under the left side in the hollow of the bed, sothat the body is slightly tilted to the right. The rightarm is placed behind the head. A site is chosen in theninth or tenth intercostal space in the middle or anterioraxillary line. After cleansing, the skin is anaesthetisedwith 2% procaine solution. A long fine-bore needle isused to infiltrate the pleura and is then passed throughthe diaphragm to anaesthetise the peritoneum and thecapsule of the liver. At least 10 ml. of local anaestheticis needed. If the skin is tough, a preliminary -nick maythen be made with a scalpel. The cannula used is 15 cm.long and- 1 mm. in bore. It is fitted with a handledtrocar. The instrument is passed through the skin, andthe patient is then instructed to

" take a deep breath in,let it out, and then hold your breath." This displacesthe lung upwards and ensures apposition of diaphragm-atic and costal pleura. The trocar and cannula arenow pushed through the diaphragm into the right lobeof the liver. The trocar is not withdrawn until theinstrument is fully’half an inch inside the liver substance.The cylinder of liver tissue is then punched out by push-ing the cannula on a further 4-5 cm. A20-ml.’Record’

syringe is connected to the cannula, and suction is appliedand maintained while the cannula is withdrawn. Thepuncture wound in the skin is sealed with collodion. Thefragment of liver is usually found in the barrel ofthe syringe ; occasionally it remains in the cannula.The aspiration of blood with the biopsy specimen neednot occasion alarm.

AFTERCARE

As a little local pain may follow puncture, morphinegr. or gr. is given subcutaneously, according to thesize and type of patient. This allays discomfort andprevents restlessness. If necessary, a further sedative,such asbarbitone soluble gr. 10, may be given in theevening. The pulse is charted hourly for the first 24hours after biopsy ; the physician should be called if thepulse-rate shows a rise. Routine visits should be paid 4and 8 hours after biopsy. A very careful watch must bekept on the patient, and if there is any sign of heamor-rhage the cross-matched blood should be administered.Absolute rest in bed is essential for 24 hours. Thepatient can then be up and about and if desired can leavehospital 48 hours after the liver puncture.The procedure is attended with very little disturbance

to the patient. During the puncture there may be acomplaint of a drawing feeling across the epigastrium.Afterwards some patients have a slight ache in the rightside for about 24 hours, and some complain of painreferred from the diaphragm to the right shoulder.Most patients agree that the discomfort compares favour-ably with that associated with sternal or lumbar punc-ture. Thirty patients had more than one biopsy ; onehad four.

DIFFICULTIES

There may be failure to get an adequate sample of liver.Hoffbauer (quoted by Watson 1944) had 40% failureswith the Tripoli and Fader (1941) technique. Iversenand Roholm (1939) reported a 10-15% failure-rate ;van Beek and Haex (1943), however, using the samemethod, state that the puncture failed but rarely. Inour first 126 biopsies there were 10% of failures ; in thenext 138 only 2%. Difficulties arise most often in hepaticcirrhosis, especially if there is associated ascites. Incirrhosis the tough liver is difficult to pierce and a fewliver-cells may be extracted leaving the fibrous frame-work.behind. In ascites the liver is very " ballottable "and is difficult to transfix. A paracentesis abdominisshould be undertaken before the liver biopsy is attempted,and the patient should lie well over on the right sideduring the puncture. This brings the liver into contactwith the’chest wall.. Another source of difficulty may bepulmonary emphysema ; the liver is pushed downwardsby the low diaphragm. It is very easy for the trocar topass above the liver. If a low diaphragm is suspectedbefore biopsy, the chest should be radiographed, and ifnecessary the puncture can be made through a lowerintercostal space.

TABLE I-MORTALITY OF ASPIRATION LIVER BIOPSY

RISKS

The fatality-rates in published cases are shown in tableI, the combined rate being 0-67%. In the first 126punctures in this series (Dible et al. 1943) there were twodeaths, one of them in a patient already moribund withsubacute necrosis of the liver. Since then a further 138punctures have been performed with no evidence ofhaemorrhage. This lessening of risk may be attributedto changes in technique and to more careful selection ofcases.

The trocar used in the original series of 126 punctureswas 2 mm. in diameter and 10 cm. long ; the new instru-ment is longer and narrower. In one of the fatal casesthe blood had leaked from a cylindrical hole on the liversurface. The narrower cannula makes a smaller woundin the liver ; the longer instrument allows the sharp trocar

398

to be pushed well into the liver before it is withdrawn ;the incision in the capsule is then clean-cut and can healeasily, while the small cylinder of liver is taken deeply inthe liver substance.Accidents will be prevented if the biopsy is confined topatients who are cooperative. It is dangerous if thepatient breathes with the trocar in the liver, as a longi-tudinal rent can then be produced. For this reasonspeed in puncture is essential. Any modification in

TABLE II-INCIDENCE OF HEMORRHAGE AFTER ASPIRATIONLIVER BIOPSY IN JAUNDICED SUBJECTS

technique, such as the injection of an anticoagulantthrough the trocar or the intercostal use of a more com-plicated needle, such as that of Tripoli and Fader, willadd to the time taken for the biopsy and hence increasethe risk of tearing the liver. The risk of haemorrhage isgreatest in the severely jaundiced, especially if the jaun-dice is due to acute parenchymatous liver disease. Wehave never met clinical evidence of haemorrhage in thenon-jaundiced group. Table II shows a comparison ofthe results in the previous series with those of the newtechnique in jaundiced patients.

-

INSPECTION AND FIXATION OF BIOPSY SPECIMEN

Some information may be gained by inspection of thebiopsy specimen. Fatty livers have a characteristicgreasy look. Biopsy specimens of livers with excessfibrous tissue tend to crumble into fragments with alobulated contour. If a malignant deposit has beenpunctured, the dull white appearance of the biopsyspecimen is characteristic.

’

The most useful routine fixative is absolute alcohol.Though it dissolves out red cells, it preserves the glycogenin the liver-cells and enables Best’s carmine stain to beapplied. To demonstrate the elements of blood in thehepatic sinusoids, formol-saline is the more satisfactoryfixative. Fat can be shown by fixing the material in

,

Bouin’s solution followed by 2% osmic acid.’ Afterfixation the specimen is embedded, sectioned longitudin-ally along the line of the cylinder, and stained in theusual way. If desired, the sample can be examinedbacteriologically.

Reliability of Biopsy Specimens in AssessingLiver Histology

The use of these small biopsy specimens as representa-tive of the pathology of the whole organ may be ques-tioned. Pathologists, including Stewart (1917), Millerand Rutherford (1923), Bergstrand (1930), and Lucke(1944), recorded that in massive hepatic necrosis in manthe left lobe of the liver was more damaged than theright. Himsworth and Glynn (1944) mentioned similarfindings in experimental " trophopathic hepatitis." Inour cases of acute hepatitis the histological damage inthe biopsy specimens corresponded well with the clinicalseverity of the disease, and there was little variation fromlobule to lobule. Excluding the obviously localisedconditions-e.g., malignant metastases, abscesses, andcysts-most other examples of human liver disease have areasonably uniform histology. Thirty cases came event-ually to necropsy ; in each the histology of the biopsy.specimen was a fair sample of the liver as a whole. Itmust be emphasised that the preparation obtained fromthese biopsies is not a smear of liver-cells but a section ofliver tissue comprising about 10-20 lobules.

Use of Biopsy Specimens in DiagnosisThe distribution of the case material is shown in table

III. The diagnostic potentialities of the method willbe illustrated by examples from some of the groups.

ACUTE HEPATITIS

Aspiration biopsy in the study of this condition hasbeen discussed elsewhere (Dible et al. 1943). If thebiopsy is performed early in the jaundiced phase, the

acute liver inflammation presents a characteristic picture,and a diagnosis can readily be made from other commoncauses of jaundice. If the biopsy is postponed untilconvalescence, normal liver tissue may be observed, anddiagnosis is then impossible. ,

CIRRHOSIS OF THE LIVER

It is in this group that clinical diagnosis, even withbiochemical assistance, may be very difficult. Liverbiopsy often supplies a definite answer. This is illus-trated by the following case : .

CASE 1.—A soldier, aged 35, had arsenotherapy jaundice inFebruary, 1944. This never fully cleared up, and in July,1944, there was a severe exacerbation of jaundice. Recoverywas slow, and in January, 1945, the patient was invalidedfrom the Army, and in March, 1945, admitted to Hammer-smith Hospital. There was no other history of jaundice, andthe patient had never served overseas. The present com.plaints were some dyspnoea on exertion and discomfort in theleft side of the abdomen. There was no jaundice ; the liveredge was felt 4 cm. below the costal margin and was firm andslightly tender. The spleen was palpable half-way to theumbilicus. There was no ascites or evidence of portal-veinobstruction. The urine contained an excess of urobilin butwas not otherwise abnormal. Laboratory tests were mostlynegative ; plasma bilirubin, phosphatase, cholesterol, anddifferential plasma proteins, the colloidal-gold test (Maclagan

TABLE III-CASES STUDIED BY ASPIRATION LIVER BIOPSY

1944), the intravenous hippuric-acid test (Quick, Ottenstein,and Weltchek 1938), and intravenous galactose test (King andAitken 1940) were all normal. There was 20% retention ofbromsulphthalein 30 min. after the injection of 5 mg. per kg.of body-weight (technique of Helm and Machella 1942).As the trocar and cannula were introduced into the liver for

aspiration biopsy, the organ felt hard and granular. The

histological picture was that of a fully developed portalcirrhosis. There was fatty infiltration of the liver-cells

(fig. 1). In this case it was essential to know the exact stateof the liver. Laboratory tests were equivocal ; aspirationliver biopsy gave a definite diagnosis.

Liver biopsy is useful not only in making a diagnosisof cirrhosis but also in proving a negative. We havestudied cases with histories and physical signs almostidentical with those of case 1, yet with a normal liver onbiopsy. In other cases treatment had been institutedfor a hepatic cirrhosis, diagnosed by clinical or laboratorymeans, yet liver biopsy showed a perfectly normalliver.

’

OBSTRUCTIVE JAUNDICE

Aspiration liver biopsy readily distinguished obstruc-tive icterus of short duration from other types ofjaundice.CASE 2.-A man, aged 60, had had, 3 days before admission

to hospital, a rigor followed by anorexia, flatulence, and afeeling of fullness in the epigastrium. His stools had becomepale, his urine dark. Next day his sclerotics were yellow.The liver was slightly tender and was enlarged 4 cm. below thecostal margin. The gall-bladder and spleen were not palp-able. The urine showed bile pigments and an absence ofurobilin. Laboratory findings were : plasma bilirubin 3-9 mg.per 100 ml., plasma phosphatase 19 units per 100 ml.; galac-tose tolerance was impaired, and hippuric-acid synthesis waslow. All these findings suggested a hepatitis rather than anobstructive jaundice. Liver biopsy, however, showed onlyaccumulations of bile in the canaliculi towards the centres ofthe liver lobules. This is a common finding in obstructivejaundice. There was no hepatitis. The jaundice cleared

400

rapidly, and cholecystograms showed a non-functioning gall-bladder. The probable diagnosis was chronic cholecystitiswith gall-stones.

Later in the course of obstructive jaundice, diagnosismay not be so simple. The secondary changes in theliver produce a picture of biliary cirrhosis which can beexceedingly difficult to differentiate from cirrhosisdeveloping in the absence of disease of the biliary tract.

MALIGNANT DISEASE OF THE LIVER, WITHOUT JAUNDICE

The diagnosis of localised lesions by puncture of theliver with a narrow-bored instrument must be a matter ofchance, but surprisingly often a portion of a localisedlesion is obtained for section, and it may indicate the siteof the primary disease.

CASE 3.-A carpenter, aged 65, complained of anorexia,epigastric discomfort, constipation, and loss of weight.Apart from wasting, nothing abnormal was found on physicalexamination. A fractional test-meal showed a histamine-fast achlorhydria. Radiograms showed no abnormality inchest, oesophagus, or stomach. Stools were examined onfive occasions, but no occult blood was found. A definitediagnosis could not be made.

’

A fortnight later a small firm nodule was noticed in theregion of the liver. Liver biopsy was performed, and thespecimen obtained proved on section to be a squamous-celledcarcinoma (fig. 2). Gall-bladder, renal tract, bronchus, andoesophagus were all considered as the site of the primarylesion. This latter site was finally incriminated by repeatingthe barium swallow, when a considerable pressure deformityand irregularity of the oesophagus were seen. The diagnosisof oesophageal squamous-celled carcinoma with hepaticmetastases was later confirmed by autopsy.A second case in which aspiration hepatic biopsy showed

squamous-celled carcinoma, proved to be a symptomlessbronchial carcinoma. Melanocarcinoma (primary inthe eye), sarcoma of the liver (primary in the uterus), andadenocarcinoma (primary in the upper rectum) have alsobeen identified by aspiration liver biopsy. This sort ofpositive information is of value if surgical intervention isbeing considered. However, if aspiration liver biopsydoes not demonstrate malignant disease, its absencecannot be assumed.

DISEASES OF THE BLOOD

The use of aspiration liver biopsy in the diagnosis ofobscure haematological disorders is illustrated by thefollowing case : ’

CASE 4.-A grocer, aged 54, complained of exhaustion,progressive pallor, and breathlessness. The skin was lemon-yellow ; the tongue was smooth ; the spleen was just palp-able ; there was no hepatomegaly or lymphadenopathy ; theurine constantly showed excess of urobilin; fractionaltest-meal revealed histamine -fast’achlorhydria. Examinationof the peripheral blood showed erythrocytes 2,000,000 perc.mm., Hb (Haden) 38%, colour index 1’18, mean corpuscularvolume 107.5 cµ, reticulocytes 0-8%, and leucocytes 3000 perc.mm., of which 1500 were lymphocytes. Smears of sternalmarrow revealed a preponderance of late erythroblasts andbasophil normoblasts ; there were small numbers of megalo-blasts.The patient was treated with iron, various liver prepara-

tions, including protolysed liver by mouth, and with ascorbicacid. There was no improvement. He was sustained withblood-transfusions. The provisional diagnosis was megalo-cytic anaemia (refractory type). Six weeks later the liver was

palpable and the spleen was larger. Aspiration liver biopsyshowed increased cells in the sinusoids and portal tracts (fig. 3).The cells were primitive and gave the impression of beingformed from the sinusoidal epithelium. There was no evi-dence of phagocytic activity. There was little differentiationof the cells ; iron stains showed diffuse siderosis. The picturemost closely resembled the histiocytic medullary reticulosisdescribed by Bodley Scott and Robb-Smith (1939). The

patient eventually died with terminal purpura and jaundice.Necropsy confirmed the diagnosis.

This case illustrates the part liver biopsy may have toplay in the elucidation of " refractory " anaemias. Wehave used the method in other blood disorders. Bloodformation in liver has been demonstrated in two casesof leucoerythroblastic anaemia. A case of infectiousmononucleosis showed infiltration of the sinusoids with

immature members of the white-cell series. TypicalGauclier cells were demonstrated in a case of Gaucher’sdisease. A suspected instance of Banti’s syndrome wasfound to have a normal liver on biopsy. Diffuse haemo-siderosis has been demonstrated and used as confirmatoryevidence for a diagnosis of acholuric jaundice in a patientwho had no increase in red-cell fragility and subsequentlyunderwent splenectomy with a satisfactory result.

AMYLOID DISEASE

This condition may be extremely difficult to diagnose.Liver may be affected in the absence of renal changes.Infiltration of the liver may be associated with normalblood biochemistry and liver-function tests. TheCongo-red test is often normal or equivocal (Stemmer-man and Auerbach 1944).

CASE 5.-A clerk, aged 29, was in 1941 found to have anintestinal ileosigmoid fistula of unknown aetiology. Therewas hepatomegaly, the Congo-red test was positive, and theurine showed albumin and casts. In November, 1941, anileocaecostomy was performed, but it proved impossible to closethe fistula. A liver biopsy made at operation showed heavyinfiltration with amyloid.

In Novembei, 1943, the patient was readmitted to hospital.The diarrhoea had persisted, there being 4-5 motions a day.Otherwise he felt well and had gained weight. The liver wasjust palpable. The urine was free of albumin. Urea clear.ance was 83% of normal. The results of the hippuric-acid,galactose-tolerance, and bromsulphthalein tests were allnormal. The Congo-red test showed 50% retention of the dyein the plasma 60 min. after injection. This figure is just withinnormal limits. It was desired to know whether there was stillamyloid disease in the liver. Aspiration liver biopsy (fig. 4)showed that there was still infiltration with amyloid, althoughthis was considerably less coarse than in 1941.The other two cases of amyloid disease studied were

associated with pulmonary tuberculosis ; a positivediagnosis was made by liver biopsy.

KALA- AZAR

In the tropics splenic and liver puncture are commonlyused in the diagnosis of kala-azar. The specimensobtained are usually smears of blood from the liver orspleen which may or may not contain the parasite.Aspiration liver biopsy can show not only the Leishman-Donovan bodies but arlso the characteristic liver histology,which may itself be almost pathognomonic.

CASE 6.-A soldier, aged 28, had served in the NorthAfrican and Sicilian campaigns. On return to England hecomplained of rigors, malaise, nausea, and vomiting. Theliver and spleen were enlarged. There were enlarged axillaryand inguinal lymph-nodes. There was an irregular pyrexiaof 100-102°. The patient was anaemic, the erythrocytes being3,300,000 per c.mm. and the haemoglobin 57% (Haldane).The leucocytes were 2850 per c.mm., with a relative lympho-cytosis and 15% immature polymorphs. No parasites wereobtained by sternal or splenic puncture. The fever did notrespond to sulphathiazole or mepacrine. and a provisionaldiagnosis of kala-azar was made. The patient received teninjections of stilbamidine (total 1-37 g.). He was referred toHammersmith Hospital for aspiratibn liver biopsy. Thebiopsy specimen (figs. 5 and 6) showed well-marked portalzonal infiltrations as well as scattered islands of endothelialproliferation in the lobules. These were very numerous.Leishman-stained sections showed the characteristic Leish-man-Donovan bodies in the reticulo-endothelial proliferations.It is interesting to note that parasites could still be demon-strated after the course of stilbamidine.

The other two instances of kala-azar showed the same ,general liver histology as case 6. In one the parasitescould not be demonstrated. In the other they probablyexisted in a degenerate form. In both these cases smearsof liver blood would probably have been classed as nega-tive. The characteristic histological picture shown byliver biopsy gives useful evidence in distinguishing thecondition from other causes of hepatomegaly andsplenomegaly.

MISCELLANEOUS CASESThis group comprised a wide variety of diseases, That

there were 34 cases in,this group which were suspectedon clinical and laboratory grounds, to have a liver lesion,most often cirrhosis, but showed normal liver histology,

401

indicates the frequency with which liver disease is mis-diagnosed. Moreover, " hepatic insufficiency " is oftenpostulated in general diseases, especially those, such asrheumatoid arthritis, which are of obscure aetiology.

,

Therapeutic agents--e.g., mepacrine-have, largely onthe basis of massive dosage in animals, been held tocause liver lesions. Aspiration liver biopsy has been ofvalue in showing the absence of liver damage in theseand other conditions.

SummaryThe technique of aspiration liver biopsy is described.

Difficulties and risks are discussed. _

An analysis of 264- biopsies is presented. Representa-tive cases have been selected, and described to demon-strate the diagnostic use of the method.

I am indebted to the Medical Research Council for an

expenses grant for technical assistance; to Major-General A. G.Biggam, Lieut.-Colonels W. R. M. Drew and W. H. Hargreaves,and Major James Marshall for many of the cases studied ; toMr. J. R. Baker and Mr. J. C. Griffin for the histological pre-parations ; to Mr. E. V. Willmott for the photomicrographs;and especially to Prof. J. H. Dible and Dr. J. McMichael fortheir assistance and criticism.

REFERENCES

Baron, E. (1939) Arch. intern. Med. 63, 276.Beek, C. van, Haex, A. J. C. (1943) Acta. med. scand. 113, 125.Bergstrand, H. (1930) Ueber die akute und chronische gelbe Leber-

atrophie, Leipzig.Bingel, A. (1923) Verh. dtsch. Ges. inn. Med. 35, 210.Dible, J. H., McMichael, J., Sherlock, S. P. V. (1943) Lancet, ii, 402.Hatieganu, I., Sparchez, T., Radu, P., Macavei, I. (1943) Wien. klin.

Wschr, 56, 21.Helm, J. D., Machella, T. E. (1942) Amer. J. digest. Dis. 9, 141.Himsworth, H. P., Glynn, L. E. (1944) Lancet, i, 457.Hoffbauer, F. W. (1945) J. lab. clin. med. 30, 381.Huard, P., May, J. M., Joyeux, B. (1935) Ann. Anat. path. 12, 1118.Iversen, P., Roholm, K. (1939) Acta. med. scand. 102, 1.King, E. J., Aitken, R. S. (1940) Lancet, ii, 543.Lucké, B. (1944) Amer. J. Path. 20, 471.Maclagan, N. F. (1944) Brit. J. exp. Path. 25, 15.Miller, J., Rutherford, A. (1923) Quart. J. med. 17, 81. Olivet, J. (1926) Med. Klinik, 22, 1440.Quick, A. J., Ottenstein, H. N., Weltchek, H. (1938) Proc. Soc. exp.

Biol., N.Y. 38, 77.Scott, R. B., Robb-Smith, A. H. T. (1939) Lancet, ii, 194.Stemmerman, M. G., Auerbach, O. (1944) Amer. J. med. Sci. 208,

305.Stewart, M. J. (1917) Proc. R. Soc. Med. 10, 10.Tripoli, C. J., Fader, D. E. (1941) Amer. J. clin. Path. 11, 516.Watson, C. J. (1944) Ibid, 14, 129.

RETROPULSED INTERVERTEBRAL DISK

PRODUCING FROIN’S SYNDROME

REPORT OF A CASE

R. R. HUGHES, M D LPOOL, M R C PMAJOR RAMC, MEDICAL SPECIALIST AND GRADED NEUROLOGIST

WHILE a minor degree of spinal block is commonlyassociated with a retropulsed intervertebral disk, thiscase is of interest in that the block was apparentlycomplete.

. A man, aged 33, apart from two previous attacks of pleurisy,had been quite well until the onset of his present trouble.The condition began in November, 1943, with pains in bothbuttocks ; during the next few weeks the pains gradually spread down the back of both legs and calves to the aukles.They were aching in type and were associated with a stabbingpain of the same distribution on eoughing.’ About the timeof onset frequency and urgency of micturition lasted for abouttwo weeks and then completely cleared. In January, 1944,the patient noticed a numb " dead " sensation over the leftbuttock, the back of the calf, and the outer side of the leftankle and foot. At the same time the legs became weak, andhe could only walk with difficulty, being unable to move thetoes of his left foot.On admission to hospital in February, 1944, his most trouble-

some symptoms were the numbness and weakness of the legs ;pain was still present but not severe. He could walk only withdifficulty by holding on to the foot of the bed. On examina-tion there was loss of the normal curve of the lumbar spineand gross limitation of movement in all directions ; therewas no scoliosis. There was also generalised tenderness ondeep palpation in the lumbosacral area but no tenderness onpressure over either sciatic nerve. The loss of muscular

power was due largely to weakness of the flexors and extensorsof the ankle-joints and toes on both sides, but no muscle was

completely paralysed. The ankle-jerks were both absent,knee-jerks were normal, and plantar responses were flexor.Sensation of pinprick and light touch was slightly impairedover the outer aspect of the left calf and the dorsum andouter part of the sole of the left foot. Straight leg raisingwas limited to 45° on both sides. Apart from the abovefindings his nervous and other systems were normal.Radiography of the lumbar spine and pelvis revealed no

abnormality. On insertion of a needle into the subarachnoidspace between the 4th and 5th lumbar spines the cerebrospinal-fluid pressure was found to be so low that it could not bemeasured with a manometer. On abdominal compressionthe rate of drip of fluid from the needle slightly increased, *buton jugular compression the rate did not change. Only about1 c.cm. of fluid could be obtained, on which the pathologistreported as follows : 2750 red cells and 10 lymphocytes perc.mm., protein 280 mg. per 100 c.cm., fluid slightly xantho-chromic and clots spontaneously. A needle was then insertedbetween the 2nd and 3rd lumbar spines ; here the cere-

brospinal-fluid pressure was. 50 mm. and increased to 80 mm.on jugular compression. The Wassermann reaction was

negative in both blood and CSF.The patient was transferred to an EMS Neurosurgical

Unit for further investigation and treatment. Radio-

graphy after intrathecal injection of’ Lipiodol ’ demonstratedthe upper level of the block to be about the centre of the bodyof the 4th lumbar vertebra. Laminectomy was performedon March 8, 1944, and on palpation of the dural sac a largemass could be felt, mainly on the right side, at the level ofthe disk between the 4th and 5th lumbar vertebrae. Onfurther exploration this proved to be sequestrated diskmaterial, which was easily removed. It measured 3-5 X 1-4

sq. cm. and weighed 1-4 g.Convalescence was uneventful and was accompanied by

rapid improvement in the motor and sensory changes.I wish to thank Colonel W. H. O’Riordan, RAMC, for

permission to publish notes on this case, and Mr. E. B. C.Hughes, first assistant to an EMS Neurosurgical Unit,for kindly supplying details of the operative procedureand findings.

INTERDIGITAL RINGWORM

TREATED WITH SOLUTION OF SULPHURATED LIME

THEODORE JAMES, M B CAPETOWN

SQUADRON-LEADER RAF

A PATIENT who had " tried everything " for an intract-able tinea cruris, which he had had more than sixmonths, was treated with Vleminckx’s solution (liquorcalcis sulphuratse, BPC). After four days’ treatmentthe condition was cured and the patient so pleased that,on his own initiative, he tried the solution on a long-standing ringworm of his toes, which also cleared upin less than a week. This unexpected result led me totry this solution for all cases of interdigital ringwormwhich subsequently came under my care. Until thenI had used various medicaments according to theiravailability ; but only when Vleminckx’s solution wasused were the results uniformly gratifying. ,

In warm climates interdigital ringworm is especiallyprevalent and, although only occasionally almostcrippling, always a nuisance because of persistentirritation about and between the tees. I have treatedsuccessfully, at a conservative estimate, 100 cases, rangingfrom slight, of a few days’ duration, to those that hadbeen neglected for months and on which a secondaryinfection had been imposed. In a’number of instancesthe sole, the dorsum, or both surfaces of the foot hadbecome involved, and the patients were kept off theirfeet until cured. In only one case did dysidrosis of thehands, an unusual complication of ringworm of the feet,develop, and this appeared during treatment of the feet.This complication became aggravated after the ring-worm had been cleared. The shortest treatment forthe slight condition was three days, and the longestfor the severer ten days. There were no facilities formicroscopical identification of the infecting agent.The formula and directions for dispensing liquor calcis

sulphuratae (Vleminckx’s solution) are given in the BritishPharmaceutical Codex (1934), but under Service conditions,the following method proved effective. Quicklime, 25 g.,was slaked with an equal quantity of water, 50 g. of sublimed