9
J Clin Pathol 1981;34:616-624 Interstitial nephritis AJ DIXON,* CG WINEARLS,t MS DUNNILL* From the *Departments of Histopathology and tMedicine, John Radcliffe Hospital, Oxford OX3 9DU SUMMARY The clinical and pathological findings are reviewed in ten cases where renal biopsy showed abnormalities predominantly within the interstitium. In six the nephritis was considered to be drug-induced; in two the aetiology was slightly obscure but the most likely diagnosis was con- sidered to be sarcoidosis. Of the remaining two cases one was chronic pyelonephritis and the other polyarteritis nodosa. The diagnosis and pathogenesis of the renal lesions are discussed and attention is drawn to the importance of distinguishing primary interstitial changes from those found in association with glomerular disease. The value of renal biopsy in diagnosis of glomerular disease is well-established and may be used to predict response to treatment and prognosis. Similarly renal biopsy may provide clear and conclusive histological evidence of acute tubular necrosis, vascular abnor- malities or end-stage renal disease. Yet in our group of patients biopsies have failed to reveal appearances which are characteristic of any of these conditions. The glomeruli appeared essentially normal and although tubular damage was present the most pronounced changes were found in interstitial tissues. Such cases have been labelled interstitial nephritis1 and represent an ill-defined, heterogeneous group, both in clinical manifestation and pathogenesis. We report nine such patients and one in whom profound interstitial changes accompanied a segmental glomer- ulonephritis. Material and methods The indication for renal biopsy was acute oliguric renal failure in five patients, raised blood urea concentrations with a normal urine volume in three patients and haematuria in one patient. In the remaining case a renal biopsy was performed during a renal sympathectomy for recurrent loin pain and fever. The renal tissue was divided into three portions. The largest piece was processed for paraffin embed- ding and serial sections (4 ,um) were stained by haematoxylin and eosin, periodic acid-Schiff, Goldner's trichrome, Martius scarlet blue and Congo red for amyloid. One piece was snap-frozen in liquid nitrogen and sections were stained by the direct immunofluorescent method for IgA, IgG, IgM, IgE, Accepted for publication 12 November 1980 C3 and fibrinogen. The third portion was fixed in 4 % glutaraldehyde and processed to epon embedding for electron microscopy. Results HISTOLOGICAL FINDINGS Major clinical and pathological features of all cases are summarised in Tables 1 and 2. All patients had either normal-sized or slightly enlarged kidneys on x-ray examination except case 9 where they were small and scarred with calyceal dilatation. In the first six patients it is highly probable that the renal condition was drug-induced but, since all were receiving more than one drug, incrimination of a single compound was usually difficult. Ooi et al.2 have drawn attention to this problem of ascribing nephropathy to any one compound because of the extreme complexity of the drug history in many patients. In cases 7 and 8 the most likely diagnosis was sarcoidosis, although neither demonstrated all the typical features of this condition, and in case 9 a diagnosis of chronic pyelonephritis was made, principally on x-ray findings. It is noteworthy that this patient had also received ampicillin. Case 10, although differing from the others in having segmental glomerulonephritis, is included here to illustrate the complexity of the diagnostic problem. This patient underwent two renal biopsies. The first contained seven glomeruli, six of which were normal but one of which had a segmental necrosis containing a few neutrophil polymorpho- nuclear leucocytes. In other respects the pathological findings bore a remarkable resemblance to some other cases (see Tables 1 and 2), notably a marked interstitial infiltrate with numerous eosinophils. Furthermore both IgE and IgG containing plasma 616

Interstitial nephritis

Embed Size (px)

Citation preview

Page 1: Interstitial nephritis

J Clin Pathol 1981;34:616-624

Interstitial nephritisAJ DIXON,* CG WINEARLS,t MS DUNNILL*

From the *Departments of Histopathology and tMedicine, John Radcliffe Hospital, Oxford OX3 9DU

SUMMARY The clinical and pathological findings are reviewed in ten cases where renal biopsyshowed abnormalities predominantly within the interstitium. In six the nephritis was considered tobe drug-induced; in two the aetiology was slightly obscure but the most likely diagnosis was con-sidered to be sarcoidosis. Of the remaining two cases one was chronic pyelonephritis and the otherpolyarteritis nodosa. The diagnosis and pathogenesis of the renal lesions are discussed and attentionis drawn to the importance of distinguishing primary interstitial changes from those found inassociation with glomerular disease.

The value of renal biopsy in diagnosis of glomerulardisease is well-established and may be used to predictresponse to treatment and prognosis. Similarly renalbiopsy may provide clear and conclusive histologicalevidence of acute tubular necrosis, vascular abnor-malities or end-stage renal disease. Yet in our groupof patients biopsies have failed to reveal appearanceswhich are characteristic of any of these conditions.The glomeruli appeared essentially normal andalthough tubular damage was present the mostpronounced changes were found in interstitial tissues.Such cases have been labelled interstitial nephritis1and represent an ill-defined, heterogeneous group,both in clinical manifestation and pathogenesis. Wereport nine such patients and one in whom profoundinterstitial changes accompanied a segmental glomer-ulonephritis.

Material and methods

The indication for renal biopsy was acute oliguricrenal failure in five patients, raised blood ureaconcentrations with a normal urine volume in threepatients and haematuria in one patient. In theremaining case a renal biopsy was performed duringa renal sympathectomy for recurrent loin pain andfever.The renal tissue was divided into three portions.

The largest piece was processed for paraffin embed-ding and serial sections (4 ,um) were stained byhaematoxylin and eosin, periodic acid-Schiff,Goldner's trichrome, Martius scarlet blue and Congored for amyloid. One piece was snap-frozen in liquidnitrogen and sections were stained by the directimmunofluorescent method for IgA, IgG, IgM, IgE,

Accepted for publication 12 November 1980

C3 and fibrinogen. The third portion was fixed in 4%glutaraldehyde and processed to epon embeddingfor electron microscopy.

Results

HISTOLOGICAL FINDINGSMajor clinical and pathological features of all casesare summarised in Tables 1 and 2. All patients hadeither normal-sized or slightly enlarged kidneys onx-ray examination except case 9 where they weresmall and scarred with calyceal dilatation. In the firstsix patients it is highly probable that the renalcondition was drug-induced but, since all werereceiving more than one drug, incrimination of asingle compound was usually difficult. Ooi et al.2have drawn attention to this problem of ascribingnephropathy to any one compound because of theextreme complexity of the drug history in manypatients. In cases 7 and 8 the most likely diagnosiswas sarcoidosis, although neither demonstrated allthe typical features of this condition, and in case 9a diagnosis of chronic pyelonephritis was made,principally on x-ray findings. It is noteworthy thatthis patient had also received ampicillin.

Case 10, although differing from the others inhaving segmental glomerulonephritis, is includedhere to illustrate the complexity of the diagnosticproblem. This patient underwent two renal biopsies.The first contained seven glomeruli, six of whichwere normal but one of which had a segmentalnecrosis containing a few neutrophil polymorpho-nuclear leucocytes. In other respects the pathologicalfindings bore a remarkable resemblance to someother cases (see Tables 1 and 2), notably a markedinterstitial infiltrate with numerous eosinophils.Furthermore both IgE and IgG containing plasma

616

Page 2: Interstitial nephritis

Interstitial nephritis

Table 1 Clinical features ofpatients with interstitial nephritis

Case No Age (yr) Sex Antecedent illness Drugs Course Diagnosis

72 F General malaise. Bethanidine Alive but in chronic renal Interstitial nephritis due todermatitis, acuteoliguric renal failure

LorazepamCyclopenthiazideCotrimoxazoleIbuprofenAmitriptylineCarbocisteine"Junipah"-containingphenolphthalein

failure drugs.

2 70 F Arthralgia, urticarial rash, Diflunisaleosinophilia, anuria Chlorpheniramine

ProchlorperazineTrifluoperazine

Recovery (steroids) Interstitial nephritis due todiflunisal.

3 23 F Epilepsy, rash, acuteoliguric renal failure

4 56 F Epilepsy, leg veinthrombosis, lobarpneumonia, arthritis,haematuria, eosinophilia

5 66 F Rash, fever, acutenon-oliguric renalfailure

6 22 M Tuberculous adenitis,chronic renal failure

7 16 M Fever, splenomegaly,lymphadenopathy,ascites, thrombo-cytopenia, raised bloodurea concentration

Y 61 F Sixth nerve palsy,hypercalcaemia,splenomegaly, acuteoliguric renal failure

PhenytoinCarbamazepinePhenobarbitoneDiazepamParaldehydeChlorazepam

PhenobarbitoneAspirinNumerous antibioticsAllopurinolAlcoholCyclopenthiazide

Mefanamic acidChlorpheniramine

PASIsoniazidRifampicin

Aspirin

ParacetamolPhenylpropanolamineIndomethacinAmpicillinFlucloxacillin

Recovered from renalfailure but latercommitted suicide

Recovery (steroids)

Recovery (steroids)

Chronic renal failure.(Renal transplantperformed)

Recovery (steroids) butpersistent impairmentof renal function

Recovery (steroids)

Interstitial nephritis due todrugs.

Interstitial nephritisprobably due to drugs.

Interstitial nephritis due tomefenamic acid.

Interstitial nephritispossibly due to drugs.

Possibly sarcoidosis inspite of negative Kveim.Granulomata in spleen.

Possibly sarcoidosis inspite of negative Kveim.

9 36 F Pulmonary tuberculosis, Ampicillinchronic peptic ulcer,recurrent fever, and loinpain

Not known Chronic pyelonephritis.

10 59 M Arthritis, sinusitis,eosinophilia, acuteoliguric renal failure

PhenylbutazoneCo-trimoxazole

Initial recovery withsteroids but died laterin renal failure

Polyarteritis nodosa.

cells were observed on immunofluorescence. Therewas no evidence of immunoglobulins or complementin glomeruli and no electron dense deposits were

found on electron microscopy. A second renalbiopsy specimen from this patient, obtained elevenweeks later, showed a diffuse crescentic nephritiswith an interstitial infiltrate now composed pre-dominantly of mononuclear cells. A diagnosis ofmicroscopic polyarteritis nodosa was made whichwas later confirmed at necropsy.

The pathological findings in the remaining ninecases are summarised below.

Glomerular changes when present were minor. In themajority of cases there was no abnormality on lightor electron microscopy (Fig. 1). In four patients therewas some evidence of nuclear crowding in mesangialstalks but no deposits were seen on electron micro-scopy and there was no fusion of foot processes ofglomerular epithelial cells. There were two instances

617

Page 3: Interstitial nephritis

Table 2 Pathological features in patients with interstitial nephritis

Case No Glomeruli Tubules Interstitium Vessels I,nmunofluorescence

Epithelial Contents Inflammation Cell type Oedema Fibrosis Granulomatadamage

Normal Polymorphs Diffuse Polymorphs +- Present Normal Minor granular

Normal

3 Slight increase in - -

mesangial cells

4 Slight increase in +mesangial cells

5 20 normal2 hyalinised

Red cellsProtein

Protein Diffuse inRed Cells cortex.

Focal inmedulla

_L -4-

PolymorphsEpithelialcells

PolymorphsRed cellsProtein

Diffuse

Focal

Epithelial cell Focaldebris 4- +

LymphocytesPlasma cellsEosinophils

Lymphocytes 4- -

Plasma cellsEosinophils

Lymphocytes + tPlasma cellsPolymorphsEosinophils

Lymphocytes 0Plasma cellsPolymorphs

Lymphocytes + +EosinophilsPlasma cells

+

0

deposition of lgGand C3 on sometubular basementmembranes

Present Intimal fibrosis Granular IgG andin arcuate IgM, and linear C3and inter- on tubular basementlobular membranesarteries

None Normal Negative

+ None Normal None available

0 None Intimal fibrosis Negativein inter-lobulararteries

6 Periglomerularfibrosis andintraglomerularfibrosis

7 Normal

Normal

9 Several hyalin-ised. Intra- andperiglomerularfibrosis

ProteinEpithelialcell debris

Focal-!

Epithelial cell Diffusedebris + + +

IId- -'- Protein casts FocalEpithelial +4debris

Protein casts FocalPolymorphs -4-+

Lymphocytes 0

Lymphocytes -rtPlasma cellsPolymorphs

Lymphocytes +Plasma cellsEosinophilsPolymorphs

Lymphocytes 0

+ None Intimal fibrosis None availablein inter-lobularvessels

+ Present Normal A little IgA and C3in mesangium ofglomeruli

+ + + Present Minor degree None availableof intimalfibrosis ofinterlobulararteries

+I-, None Intimal fibrosis None availableand medialhypertrophyof arcuateand inter-lobulararteries

10 6 normal. One + --

with segmentalfibrinoidnecrosis

ProteinOccasionalpolymorphs

Diffuse Lymphocytes + + 0EosinophilsPlasma cells

None Normal IgG and IgEcontaining plasmacells

of periglomerular and intraglomerular fibrosis butthese were probably a reflection of the interstitialdisease rather than of primary glomerular damage.These two cases were those in which clinical detailsindicated chronicity of the renal condition.

Tubules Damage to tubules was variable. Bothproximal and distal portions were affected, althoughif epithelial damage was at all severe, it was very

difficult if not impossible to distinguish between thetwo. There were focal areas where inflammatory

cells penetrated the tubular basement membraneaccompanied by necrosis of lining cells (Fig. 2). Suchareas were not numerous. Many tubules were dilatedand often contained neutrophil polymorphonuclearleucocytes, red blood cells, and epithelial cellulardebris as well as protein casts within their lumen(Fig. 3). Evidence of regeneration with mitoses intubular epithelial cells was seen only in case 5.

Interstitial tissues exhibited the most importantfeatures in biopsy specimens (Fig. 4). Most had some

Dixon, Winearls, Dunnill618

+

Page 4: Interstitial nephritis

Interstitial nephritis

Fig. 1 Case 4: normal glomeruli andoedematous interstitium with cellularinfiltrate. Haematoxylin and eosinx 175.

Fig. 2 Case JO: tubular epithelial cellnecrosis and penetration of tubularbasement membrane by mononuclearinflammatory cells. Haematoxylin andeosin x 175.

4 A 4$

degree of oedema or fibrosis with separation oftubules one from another. The cellular infiltrate wasof variable density and composed primarily ofmononuclear cells. These were mainly lymphocytesbut in several cases plasma cells were prominent andin some there were foci of neutrophil polymor-phonuclear leucocytes. Eosinophils were conspicuousin five cases, four of which were considered to bedrug-induced. It was frequently not possible to assessdistribution of the infiltrate on the material availablebut in those biopsy specimens where the cortico-medually junction was present, there did appear to be

a concentration of inflammatory cells in that region.Cellular infiltrates in the medulla were often lesspronounced than in the cortex. A notable feature infour patients was the presence of small granulomatacomposed of histiocytes, lymphocytes, and plasmacells, but lacking Langhans giant cells (Fig. 5). Therewas no necrosis or caseation. In two of these patientsthe nephritis was thought to be drug-induced, in twothe aetiology was slightly obscure but overall,clinical and pathological findings indicated adiagnosis of sarcoidosis. The granulomata observedwere not, however, typical of this condition.

619

Page 5: Interstitial nephritis

Dixon, Winearls, Dunnill...... 4tf *A.

-.: #;.

lctomir

Ultrastructural studies didnotrevealanyelect

Elec trontmicosctopyepeeceo inmtcodi

medmbraes.atheyditd showg focaloarasmiofedamageut

in tubular epithelial cytoplasm in interstitial nephri-

ti5.2 These features were not noticeable in our cases.

tsTm Fig. 3 Case 3: polymorphonuclearleucocyte and epithelial cellular debriswith tubular lumen. Haematoxylin andeosin x 175.

h.

Fig. 4 Case 1: severe interstitialinflammation accompanied by fibrosisand tubular epithelial cell atrophy.

ffi Haematoxylin and eosin x 100.

ImmunohistologyImmunofluorescent preparations showed two caseswith immunoglobulins and complement on tubularbasement membranes. In one this consisted of aminor focal deposition of IgG and C3; the other hada striking linear deposition of C3 in addition togranular deposits of IgG and IgM. In both cases, theevidence pointed strongly to drug-induced disease.

Staining for immunoglobulins and complement inglomeruli was negative in all patients except for one

620

Page 6: Interstitial nephritis

Interstitial nephritis

Fig. 5 Case 2: medullarygranulomata. Haematoxylin and eosinx 300.

Fig. 6 Case 2:lymphocyte between

/ tubular epithelial cells.Electron micrographx 8000.

case of sarcoidosis (case 7) where a little IgA and C3was demonstrable in the mesangium.

Discussion

DIAGNOSISTwo major problems relate to the diagnosis of inter-stitial nephritis on renal biopsy. The first concerns itsdistinction from interstitial changes occurring inassociation with glomerular disease; the second

relates to the many possible aetioiogies of the con-dition and the feasibility of distinguishing betweenthem on biopsy material. Of particular importanceis the diagnosis of interstitial nephritis due to drugs,often presenting with acute renal failure, sincerecovery is usual if the drugs are withdrawn. Manydifferent compounds have been incriminated, someof which are listed in Table 3.3-20The distinction from glomerulonephritis often

presents no difficulty yet it must be noted that minor

621

Page 7: Interstitial nephritis

Dixon, Winearls, Dunnill

Table 3 Some drugs associated with interstitialnephritis3-20

Drug Reference

Sulphonamides Robson et al. (1970)'Penicillin Baldwin et al. (1968),' MWry (1970)'Methicillin Border ct al. (1974),' Mayaud et al. (1975)'Ampicillin Ruley et al. (1974)8Cotrimoxazole Dry et al. (1975)9Cephalosporins Burton et al. (1974),1" Wiles et al. (1979)"Rifampicin Nessi et at. (1976)12Ethambutol Collier et al. (1976)"'Phenindione Sraer et al. (1972)14Diphenylhydantoin Muehrcke and Pirani (1972)"Furosemide Lyons et al. (1973)"Phenylbutazone Russell et al. (1978)1"Allopurinol Gelbart et al. (1977)"1Diflunisal Chan et al. (1980)"Phenobarbitol Faarup and Christenson (1974)2"

glomerular abnormalities are not infrequent ininterstitial nephritis and in several of the patientsreported here there were areas of mesangial nuclearcrowding. In one patient this was associated with alittle IgA and C3 in the mesangium. This was thoughtto be a case of sarcoidosis and it is possible there wassome glomerular involvement2l in addition to pro-nounced interstitial disease. A further problem isillustrated by case 10 where, apart from a segmentalnecrosis in one of seven glomeruli, the findings werevery similar to those in drug-induced nephritisprimarily affecting the interstitium. In particularthere were numerous eosinophils together withIgE-containing plasma cells. The subsequent rapidprogression of this man's disease to diffuse crescenticnephritis emphasises the importance of looking forglomerular involvement in these cases and con-

sidering the diagnosis of polyarteritis nodosa. Thisof course might also be drug-induced.22 Such adiagnosis could easily be missed on a biopsy speci-men containing inadequate numbers of glomeruli.

Interstitial nephritis without glomerular diseasemay be found in association with a number ofconditions.' Its aetiology may be difficult if notimpossible to determine on biopsy alone. Relianceon clinical and radiological evidence is often required,particularly in diagnosing conditions such as pyelone-phritis, renal papillary necrosis or urinary tractobstruction, all of which may result in substantialinflammatory cell infiltration of the renal inter-stitium. This is illustrated by case 9.

Features in favour of a drug-induced interstitialnephritis include the presence of numerous eosino-phils in the infiltrate, IgE-containing plasma cellsand deposition of immunoglobulins and complementon tubular basement membranes. Unfortunatelynone of these features is specific for drug-induceddisease and none is invariably present even in thoseinstances where there is a clear indication of drug

hypersensitivity. In the majority of instances a focallymphocytic and plasma cell infiltrate predominates.Neutrophil polymorphonuclear leucocytes are notunusual and when numerous, raise the possibility ofan acute infection and clearly this must be excluded.Unfortunately many of the drugs producing inter-stitial nephritis are antibiotics and will have beenprescribed originally in order to treat an infection.Withdrawal of such drugs may thus present a thera-peutic dilemma.A further diagnostic problem is posed by the

presence of granulomata. These are sometimesobserved in cases of drug-induced renal disease, butmay also be associated with sarcoidosis, and tuber-culosis should of course be excluded.

Tubular changes are seen in most cases of inter-stitial nephritis, their nature depending more on theseverity and duration of the condition than on theaetiology. Tubular atrophy with interstitial fibrosistypify chronic disease; more characteristic of theacute stages are focal areas of epithelial necrosis,often associated with interstitial oedema. Intra-luminal inflammatory cells and proteinaceous castsare frequently present and it is notable that poly-morphs within tubules are not found exclusively inpyelonephritis. Red cells are sometimes seen in thetubular lumen and haematuria may be a presentingfeature. Their origin is a matter for speculation. It isunlikely to be glomerular and in all probability thered cells leak from peritubular capillaries in areas ofacute inflammation.

PATHOGEN ESISThe underlying mechanisms producing the lesionsare not understood. It seems that a variety of initialinsults may give rise to similar histological appear-ances. Possible aetiological factors include: ischae-mia; an immunological reaction; direct toxic actionon renal tubular epithelium; infection; obstructionto urine outflow. That chronic ischaemic injury mayresult in cellular infiltration of the renal interstitiumseems clear. The role of shock as a principal cause ofacute interstitial nephritis is more doubtful. There isseldom a definite record of a hypotensive episode.The similarity of tubular lesions to those in acutetubular necrosis does not imply an identical cause.Furthermore the interstitial cellular reaction isabsent or scanty, certainly in early stages of acutetubular necrosis, whereas it is a prime and earlyfeature of interstitial nephritis. However, it must beborne in mind that the two conditions may coexist,particularly if antibiotics have been administered in asevere case of bacteraemic shock.Tubular damage can be associated with deposition

of immunoglobulins and complement componentson the tubular basement membrane. This has been

622

Page 8: Interstitial nephritis

Interstitial nephritis

reported in systemic lupus erythematosus23 and inrapidly progressive glomerulonephritis where it isassociated with circulating antiglomerular and anti-tubular basement membrane antibodies.24 Anti-tubular basement membrane antibodies may also beresponsible for tubulointerstitial damage occurringin some renal allografts. They have been reported tooccur in association with tubular damage due tomethicillin6 but have not been found in many otherinstances of drug-induced disease. There was someevidence of a similar mechanism in two of our cases.Although these had a focal rather than a lineardeposition of immunoglobulin, one had a strikinglinear deposition of C3. Electron dense deposits werenot observed in the tubular basement membranes

It is possible that cellular immunity may play apart in causation. The evidence for this is largelyconjectural and based on the observation that inmany instances the cells causing disruption of thetubules are lymphocytes.2 The presence of eosino-phils and of IgE-containing plasma cells in theinterstitial tissue20 indicates that a type I allergicreaction may sometimes be important. In drug-induced interstitial nephritis, evidence for a hyper-sensitivity reaction rather than a direct toxic reactionto the offending drugs has been cogently summarisedby Mery and Morel-Maroger (1976).25 They notedthat it occurred in only a small proportion of treatedpatients, was not dose-related, and was often accom-panied by other evidence of hypersensitivity such asfever, skin rash, and eosinophilia. Furthermore, theyclaimed that circulating antibodies reacting with theoffending drug could frequently be detected, thattests for cell-mediated immunity were sometimespositive, and that similar renal reactions occurred iffurther exposure to the drug or chemically-relatedcompounds took place.

Interstitial nephritis was more frequently referredto in early literature before the antibiotic era, usuallyin association with infectious fevers.26 These patientsdiffered from most of the recently reported cases inthat functional disturbance was minimal. Yet manypatients with interstitial nephritis are suffering frominfections either in the urinary tract or elsewhere inthe body. Treatment of these patients with anti-biotics must leave some doubt as to the part playedby the underlying infection and the role of antibioticsthemselves in the aetiology of the renal disease.

Primary renal papillary damage from any cause-for example, urinary tract obstruction, reflux,diabetes mellitus, sickle cell disease or analgesicabuse may be associated with inflammatory cellinfiltrates in the interstitium.27 To what extent this isrelated to mechanical damage produced by obstruc-tion to urinary outflow and to what extent otherfactors are involved is not clear.

Pathogenesis of renal fJilureThe clinical presentation of interstitial nephritisvaries from minor disorders of renal function toacute oliguric renal failure. The underlying physio-pathology of the latter has received little attention.The lack of a hypotensive episode and the presence ofnormal glomeruli with patent capillaries excludeslack of glomerular filtration as a primary cause.Oliguria may occur in the absence of apparentobstruction to urine flow by medullary lesionsalthough inevitably restricted biopsy material cannotprovide conclusive evidence on this point and it israrely possible to study these cases at necropsy.Tubular damage associated with back filtration ofglomerular filtrate through focal areas in the necrotictubules is another possible explanation. The presenceof interstitial oedema which is a common finding inthis condition would be in accord with such ahypothesis. This is an area where the relation ofpathological findings to function is somewhatobscure.

References

Heptinstall RH. Interstitial nephritis. Am J Pathol 1976;83:214-36.

2 Ooi BS, Wellington J, Jao W, First MR, Mancilla R,Pollak VE. Acute interstitial nephritis. A clinical andpathologic study based on renal biopsies. AmJMed 1975;59:614-29.

3 Robson M, Levi J, Dolberg L, Rosenfeld JB. Acutetubulo-interstitial nephritis following sulfadiazinetherapy. Isr J Med Sci 1970;6:561-6.

4 Baldwin DS, Levin BB, McCluskey RT, Gallo G. Renalfailure and interstitial nephritis due to penicillin andmethicillin. N Engl J Med 1968;279:1245-52.

Mery J-Ph. Les nephropathies interstitielles aigues pro-voqu6es par les p6nicillines. Ann Med Interne (Paris)1970;121 :811-6.

6 Border WA, Lehman DH, Egan JD, Sass HJ, Glode JE,Wilson CB. Antitubular basement membrane antibodiesin methicillin nephritis. NEnglJ Med 1974;291 :381-4.

7 Mayaud C, Kanfer A, Kourilsky 0, Sraer JD. Interstitialnephritis after methicillin. N Engl J Med 1975;292:1132-3.

8 Ruley EJ, Lisi LM. Interstitial nephritis and renal failuredue to ampicillin. J Pediatr 1974;84:873-7.

9 Dry J, Leynadier F, Herman D, Pradalier A. L'associationsulfamethoxazole trimethoprime (cotrimoxazole). Reac-tion immuno-allergique inhabituelle. Nouv Presse Med1975 ;4:36.

10 Burton JR, Lichtenstein NS, Colin RB, Hyslop NE Jr.Acute renal failure during cephalothin therapy. JAMA1974;229 :679-82.

Wiles CM, Assem ESK, Cohen SL, Fisher C. Cephradine-induced interstitial nephritis. Clin Exp Immnol 1979;36:342-6.

12 Nessi R, Bonoldi GL, Redaelli B, Di Filippo G. Acuterenal failure after rifampicin: a case report and survey ofthe literature. Nephron 1976;16:148-59.

13 Collier J, Joekes AM, Philalithis PE, Thompson FD. Twocases of ethambutol nephrotoxicity. Br Med J 1976;ii:1105-6.

14 Sraer JD, Beaufils P, Morel-Maroger L, Richet G. Nephrite

623

Page 9: Interstitial nephritis

Dixon, Winearls, Dunnill

interstitielle chronique due a la phenylindanedione faisantsuite a une insufficience renale aigue. Noiov Presse Med1972;1 :193-6.

5 Muehrcke RC, Pirani CL. Renal biopsy: an adjunct in thestudy of kidney disease. In Black DAK, ed. Renaldisease. 3rd ed. Oxford, London, Edinburgh: BlackwellScientific Publications, 1972:110-53.

16 Lyons H, Pinn VW, Cortell S, Cohen JJ, Harrington JT.Allergic interstitial nephritis causing reversible renalfailure in four patients with idiopathic nephroticsyndrome. N EngIJ Med 1973;288:124-8.

17 Russell GI, Bing RF, Walls J, Pettigrew NM. Interstitialnephritis in a case of phenylbutazone hypersensitivity.Br Med J 1978;i: 1322.

8 Gelbart DR, Weinstein AB, Fajardo LF. Allopurinol-induced interstitial nephritis. Ann Intern Med 1977;86:196-8.

19 Chan LK, Winearls CG, Oliver DO, Dunnill MS. Acuteinterstitial nephritis and erythroderma associated withdifiunisal. Br MedJ 1980;i:84-5.

20 Faarup P, Christenson E. IgE containing plasma cells inacute tubulointerstitial nephropathy. Lancet 1974 ;ii :718.

2 Waldek S, Agius-Ferrante AM, Lawler W. Renal failuredue to glomerulonephritis in sarcoidosis. Br MedJ 1978;i: 1110-1.

22 Heptinstall RH. Pathologs' of the kidnei. 2nd ed. Boston:Little, Brown and Co, 1974.

23 Andres GA, McCluskey RT. Tubular and interstitial renaldisease due to immunologic mechanisms. Kidnej Int1975 ;7 :271-89.

24 Andres G, Brentjens J, Kohli R, et al. Histology of humantubulo-interstitial nephritis associated with antibodies torenal basement membranes. Kidne i Int 1978;13:480-91.

25 Mry J-Ph, Morel-Maroger L. Acute interstitial nephritis.A hypersensitivity reaction to drugs. In: Giovannetti S,Bonomini V, D'Amico G, eds. Proceeedings of the 6thinternational congress on nephrologv, Florenc e 1975.Basel: Karger, 1976:524-9.

t Councilman WT. Acute interstitial nephritis. J E.xp Med1898 ;3 :393-420.

27 Freedman LR. Interstitial nephritides. In: Giovanetti S,Bonomini V, D'Amico G, eds. Proc-eedings of the 6thinternational congress on nephrologv, Florence 1975.Basel: Karger, 1976:542-51.

Requests for reprints to: Dr MS Dunnill, Departmentof Histopathology, John Radcliffe Hospital, Headington,Oxford OX3 9DU, England.

624