\ I. JOISNAI, 01.•^

Volume 67. Number IPrinted ii, the (ISA.

(ISSN 0145-916N)

INTERNATIONALJOU ANAL OF LEPROSY

and Other Mycobacterial Diseases

VOLUME 67, NUMBER 1^

MARCH 1999

Role of S-100 Staining in Differentiating Leprosy fromOther Granulomatous Diseases of the Skinl

Mathew M. Thomas, Mary Jacob, Sushil M. Chandi, Soshamma George,Susanne Pulimood, L. Jeyaseelan, and Charles K. Job''

Leprosy is most easily diagnosed xvhenMycobacterium leprtw are demonstrable indiseased tissues, but this is often difficult inthe indeterminate and in the tuberculoidtypes of leprosy in which M. leprtw arerarely detected. Histological evidence of ac-tive destruction of a cutaneous nerve bygranulomatous inflammation is widely ac-cepted as a feature of tuberculoid leprosy.However, identifyine remnants of the nervein the granuloma may be difficult in someinstances, and morphoiogic similaritieswith granulomatous dermatoses, such as lu-pus vulgaris, secondary syphilis, sarcoido-sis, etc., add to the problem.

S-100 protein is an immunocytochemi-cal marker of Schwann cells of nerves

' Received for puhlication ou 9 July 1998.Accepted for publication in revised tono ou 15December 1998.

M. M. Thomas, M.D. M. Jacob, M.N.A.M.S.; S.George, M.D., Dip. N.B.E.; S. Pulimood, M.D.. De-partment of Dermatology; S. M. Chandi. M.D., Ph.D..F.1.C.(Path.), Department of Pathology; L. Jeyaseelan,M.Sc., Ph.D., Department of Biostatistics, ChristianMedical College and 1lospital. Vellore 632 004, N. A.District, Tam-il Nado, Judia. C. K. Joh, M.D.,F.R.C.Path., St. Thomas Hospital and Leprosy Center,Chettopatto 606 801. T. S. District, Tamil Nado, Ilidia.

Reprint requests to Professor George at the ahoveaddress or FAX 91-416-32035:e-mail: ahraham@ cmc.ernet.in

which is ais° found in normal neural tissue(neurons and glial cens) melanocytes andLangerhans' cells of the skin, interdigitat-ing reticulum cells of the lymphchondrocytes, myoepithelial cells and ductsof the sweat, salivary and mammary glands,serous glands of the lung, fetal neuroblastsand sustentacular cells of the adrenalmedulla ('").

There are three reports in the literaturewhich address the utility of the immunoper-oxidase staining of S-100 protein in skinbiopsies as an aid in the dia,gnosis of tuber-culoid leprosy^".1"). Narayan, et al.have found that S-100 is a sensitive and re-liable marker of nerve damage.

The present study was undertaken to ex-amine with S-100 staining the different pat-terns of cutaneous nerve involvement in thetuberculoid spectrum of leprosy and othergranulomatous conditions of the skin; tocorrelate these with the clinicai and histo-pathological findings and also to determinewhether S-100 staining will demonstratefraLsimented, intiltrated, or the absence ofdermal nerves.

MATERIALS AND METHODSTwenty-five consecutive patients with a

clinicai diagnosis of tuberculoid leprosy(TTHD; N = 9) and borderline tuberculoid

International Jonrnol o/ Lelnos.v^ 1999

^Ftc. I. Fragmented nerves incide a granulo na de-^FIG. 2. Intact cutaneous nerve detected by S-100

^

tccted by S-100 protein (avidin-biotin peroxidase com-^protein (avidin-biotin peroxidase complex x200).plex x200).

Ieprosy (BTHD; N = 16), according to theJob and Chacko classification (`), who at-tended the Dermatology Department ofChristian Medical College Hospital, Vel-lore, India, in 1995 and 1996 were includedin the study. These patients had cutaneousanesthetic hypopigmented or erythematousmacules or plaques and did not have acid-fast bacilli (AFB) in their slit-skin smears;eight of them were in type 1 reaction. Pa-tients who had received prior antileprosytreatment for more than 3 months andwhose skin biopsies were reported as inde-terminate leprosy were excluded from thestudy. Skin biopsies from representative le-sions were examined histologically usinehematoxylin and eosin (H&E) staining anda modified Fite-Fanico technique for M./ehrae. These lesions were reported as ei-ther tuberculoid leprosy (N = 11) or border-line tuberculoid Ieprosy (N = 14). The clin-ica) and histopathological findings, includ-ing the presence of AFB, were noted.

Fifteen consecutive patients with a clini-cal and/or histological diagnosis of uranulo-

matous dermatosis other than leprosy (9cases of tuberculosis of the skin, 1 case ofsarcoidosis and 5 cases of granulomatousinflammation of unknown cause) were in-cluded as controls.

The 40 skin-biopsy specimens were sub-jected to S-100 immunoperoxidase staining(") in two batches of 20 each. The S-100-stained slides were read without any refer-ence to the clinicai data.

The clinicai diagnoses, histopathologicalfeatures and S-100 staining patterns werethen correlated. Diagnostic efficacy of thetest in terms of specificity and sensitivitywere calculated (by using 2 x 2 tables) us-ing clinico-histopathological features as the"gold standard." Correlation between thepresence of AFB in tissue sections and thepanem of neuritis by S-100 staining and theagreement between the clinica) and histo-logical diagnoses were also analyzed.

RESULTSFour different patterns of nerve involve-

ment were noted by S-100 ininiunoperoxi-dase staining: a) Intiltrated: dark staining,

S-100 stainingpattern ol nerve

involvemcnt^Inflainednerves

Cases (N = 25)^ Controls (N = 15)

Absentnerves

Intactnerves

67, 1^%honras, et al.: S-100 Staining in Dif/erential Dias'nosis^3

TABU. 1. Correlation between imitem of nerve inrolvement hv H&'E staining cmdS-100 staining in cases and controls.

Histopathology (H&E)

Infiltrated^

O^

OFrasntented^

17^

3^

OAbsent^

3^

OIntact^ o

^15

Sensitivity = 100%.Specificity = 100%/.Positive predictive value = 100%.Negative predictive value = 100c/•.

fibrillar structures in a wavy pattern assoei-ated with inflammatory cells; b) Frag-mented: small, dark staining, fibrillar struc-tures, usually multiple, inside a granuloma(Fig. 1); c) Absent: no dark staining fibrillarstructures in or outside the granuloma; andd) Intact: dark staining, large fibrillar struc-tures in a wavy pattern with no inflamma-tory cells inside (Fig. 2). There were seen inand outside the granulomas. Figure 3 showsthe percentage distribution of the four pat-terns. Table 1 shows the correlation be-tween the pattern of nerve involvement byH&E staining and S-100 staining.

By S-100 immunoperoxidase staining,the skin biopsies with tuberculoid leprosy(TTHD) showed only fragmented nerves(100%); whereas those with borderline tu-berculoid leprosy (BTHD) showed infil-trated (7.2%), fragmented (64.2%) and ab-sent nerves (28.6%). Intact nerves were notpresent in there leprosy biopsies. Skin biop-sies from the 15 controls contained intactnerves only. Nine of 14 BTHD and ali ofthe 11 TTHD biopsies showed the frag-mented nerve pattern. Tinis, there was amore significant probability of finding afragmented nerve pattern in TTHD than inBTHD patients (Fischer's exact text, p =0.04).

Of the 8 cases which clinically showedfeatures of type 1 reaction, 5 had frag-mented nerves and 3 showed an absence ofnerves in S-100-stained sections. There wasno significant correlation between the mor-phological type of skin lesions or the sen-sory impairment of the skin lesion and theS-100 staining pattern.

The agreement between the clinicai andhistological diagnoses of leprosy cases was84% (kappa = 0.66) (Table 2).

The presence of AFB in a skin-biopsyspecimen did not significantly affect thepattern of cutaneous nerve destruction.

Of the four cases which did not shownerves by S-100 staining, three (75%) wereclinically in type 1 reaction.

DISCUSSIONThe fragmented pattern of nerve damage

due to derma! neuritis was the most com-mon pattern seen in the 25 leprosy casesand was more often seen in tuberculoid lep-rosy than in borderline tuberculoid leprosy.Intact nerves were not observed in any ofthe leprosy cases. This may be due to theimmunity of the patient mounting a nervedestructive inflammatory response, forminggranulomas. The 15 controls, however,showed intact nerves either inside or out-side the granuloma. This enabled exclusionof leprosy in the controls, including thosewho were reported as having a granuloma-tous inflammation of unknown causes.

TABLE 2. Correlation between clinicaiand histological diagnoses of 25 cases.

Clinica) diagnosisHistopathological diagnosis

13"1'1ID TTHD Total

BTIID'Tf I I I)Total

13

14

38

I l

169

25

11T111) = Borderlinc tuberculoid leprosy."1"I'111) = Tuberculoid leprosy.

4^ International fournal aí Leprosy^ 1999

100%100

90^0%

80'

70

g) 60ooo

^Fragmented Absent^Intact

S100 Staining Pattern

FIG. 3. I listogram showing the distribution ofS-100 staining pattems of cutaneous nerve involve-ment; = leprosy case; i;g1= control.

By statistical analysis, it was found thatS-100 staining was as good as H&E stainingfor the diagnosis of tubercuioid forms ofleprosy. II was, however, superior to H&Estaining in identifying the destruction ofderma! nerves in leprosy. Fieury and Bacchi(5) in their study found that 8 out of 9 biop-sies with a clinicai diagnosis of tuberculoidleprosy, but with no histological or bacterio-logical evidence of the same, showed cuta-neous nerve alteration by S-100 staining.They concluded that the use of S-100 stain-ing to visualize remnants of peripheralnerves represents an efficient auxiliary aidin the diagnosis of leprosy. In a study byJob, et al. (7) out of 20 skin biopsies withclinicai suspicion of tuberculoid leprosy,histologically reported as granulomatous in-fiammation, 14 showed nerve fragments inthe granulomas, enabling a definite diagno-sis of tuberculoid leprosy. Singh, et al. (')concluded that the complete absence ofnerve twigs, either inside or outside thegranuioma, is the most reliable criterion forthe diagnosis of tuberculoid leprosy. In con-trast, in our series of 25 cases 80% showedfragmented nerves, while only 16% did nothave demonstrabie nerves. The presentstudy clearly confirms the observation ofearlier workers that the disorganization anddestruction of dermal nerves is a far moreconsistent and frequent finding (84%) thanthe absence of nerves (16%) in the diagnosisof tuberculoid disease.

While agreement of clinicai and histo-logical diagnoses in earlier studies showed

an average agreement of 50% (3.15) in thepresent study it was 84%. Disagreementcentered upon a) the absence of AFB and b)nerve-related histopathology.

In this study, 36% (9/25) of the skin-smear-negative leprosy cases showed AFBin tissue sections in the dermal nerves in 7and in the granulomas in 2. This was con-sistent with the finding of Fieury andAranda (4)• However, in a recent study byPonnighaus, et al. (") there were AFB inthe skin sections in 3.3% and in the nervebiopsies in 36.3% of the patients. The pres-ence of AFB in the derma! nerves or granu-lomas did not have a significam effect onthe pattern of nerve destruction.

With the World Health Organization'sgoal of reducing the prevalence of leprosyto 1 per 10,000 population by the year 2000(17), as the incidence of leprosy declines, theexpertise to diagnose leprosy may ais° beexpected to decrease and S-100 immu-noperoxidase staining could prove to be aneffective aid in the diagnosis and study ofpatterns of cutaneous nerve involvement inthe tuberculoid forms of leprosy.

SUMMARYSi ice Mycobacterium leprae are rarely

demonstrable in the tuberculoid spectrum ofleprosy, a confirmatory diagnosis of leprosycan be made on the basis of finding activedestruction of cutaneous nerves by granulo-matous infiammation in a skin biopsy. Im-munoperoxidase staining for 5-100 protein,which is a marker for Schwann cells, wasused to delineate nerves in lesionai skinbiopsies of 25 patients with tuberculoid andborderline tuberculoid leprosy as well as 15controls with nonieprous granuiomatous in-fiammation. Four different patterns of nervedamage were observed: infiltrated, frag-mented, absent, and intact. Ali of the non-leprous granulomatous dermatoses showedonly intact nerves, either inside or outsidethe granuloma, and so S-100 staining canbe used to rule out leprosy.

RESUMENDebido a que M■robacterium hprae es raramente

demostrable en Ia lepra tuberculoide, cl diagnósticocontirmatorio de la enfermedad puede hacerse si se ex-amina Ia destrucción activa de los nervios cutáneos porla inflamación granulomatosa en una biopsia de piel.En el presente estudio, la tinción inununoquímica COO

50'

40

301

201

10, 4%urn

infiltrated

16%

'I

67, 1^Thomas, et al.: S-100 Stainit ig in Differential Diagnosis^5

anticuerpos marcados con peroxidasa para Ia proteínaS-100, la cual es un marcador de las células de Shwann,fue usada para delinear los nervios eu las biopsias depiei lesionai de 25 pacientes con lepra tuberculoide y tu-berculoide subpolar, y en 15 controles con inflamacióngranulomatosa no leprosa. Se observaron 4 patronesdiferentes de dafio a nervios; infiltrados, fragmentados,ausentes e intactos. Todas Ias dermatosis granulo-matosas no leprosas mostraron sói° nervios intactostanto dentro como ibera dei granuloma. La tinción paraS-100 peude usarse para descartar la lepra.

RÉSUMÉComme Mycobacterium lepra(' est rarement mise

eu évidence dans les formes tuberculoïdes da la lepre,un diagnostic de certitude de la lepre peut etre effectuéen se basant sur la trouvaille, dans une biopsie cutanée,de nerfs cutanés détruits par l' inflammation granulo-mateuse. Le marquage de la protéine S-100 par im-munopéroxydase, qui est spécifique des cellules deSchwann, a été utilisée sur de biopsies cutanées poursouligner les nerfs dans les lesions de 25 patients at-leints de lèpre tuberculoide et borderline tuberculoïdeainsi que 15 contrôles présentant des intlammationsgranulomateuse non-lépreuses. Quatre aspects dif-férents d'atteinte nerveuse ont été observes: infiltre,fragmente, absent et intact. Toldes les lesions granulo-mateuses non-lépreuses ne montraient que des nerfsintacts, que ce soit à Eintérieur ou à l'extérieur dugrantilôme, et done l'immuno-marquage par Ia pro-téine S-100 peut être utilisé pour écarter la lèpre.

REFERENCESI. BONI, R., BURG, G., Do(iNoGifu, A., 1LG, E. CO,

SCHAFER, B. W., MULLER, B. and HEIZNIANN, C. W.

Immunohistochemical localization of the Ca2+binding S-100 proteins in normal human skin andmelanocytic lesion. Br. J. Dermatol. 137 (1997)39-43.CALissANo, P., MOORE, B. W. and FR1ESEN, A. Ef-fect of calei um on S-100, a protein of the nervoussystem. Biochemistry 8(1969) 4318-4326.

3. FINE, P. E. M., Jon, C. K., Lucxx, S. B., MEYERS,

W. M., PONNIGHAUS, J. M. and STERNE, J. A. C.Extent, origin and implication of interobservervariation in the histoptithological diagnosis of sus-pected leprosy. Int. J. Lepr. 61 (1993) 270-282.

4. Faniet:, R. N. and AlIANDA, C. M. Detection ofAFB in tuberculoid biopsies. (Letter) Int. J. Lepr.63 (1995) 103.

5. FI.EuRr, R. N. and BACCEII, C. E. S-100 proteinand immunoperoxidase technique as an aid in thehistoptithologic diagnosis of leprosy. Int. J. Lepr.55 (1987) 338-344.

6. JAcon, M., CtiAck(), C. J. G., ARUNTHATI11, S. andCIIANDI, S. M. Signiticance and interpretation ofnerve changes in skin biopsies when present in as-sociation with common dermatological condi-tions. (Abstract) Int. J. Lepr. 52 Suppl. (1984) 755.

7. 1013, C. K. Nerve damage in leprosy. (XIII Lep-rosy Congress State of the Art Lecture) Int. J.Lepr. 57 (1989) 532-539.

8. Jou, C. K. and CtiAcK(), C. J. G. A simplified sixgroup classification of leprosy. Lepr. lndia 54(1981) 14-16.

9. J013, C. K., DRAIN, V., DENIING, A. T., HAST1NGS, R.C. and GER13142, M. A. Role of S-100 protein as amarker for Schwann cells in the diagnosis of tu-berculoid leprosy. (Letter) Int. J. Lepr. 58 (1990)392-393.

10. KIIAN, H. J., MARES, A., Timm, H. R. T. and BAU-

MAI, R. Role of antibody to S-100 protein in diag-nostic pathology. Ana J. Clin. Pathol. 79 (1983)341-347.

11. MookE, B. W. A soluble protein, characteristic ofthe nervous system. Biochem. Biophys. Res. Com-mun. 19 (1965) 739-744.

12. NARAYAN, R., MAM:SM:ARI, P. K. and DESIKAN, K.V. Detection of S-100 protein and anticeramideantibodies in leprosy patients by ELISA. Lepr.Rev. 68 (1997) 117-124.

13. PONNIGHAUS, J. M., LEINHARDT, C., LUCAS, S.,

FINE, P. E. M. and STERNE, J. A. C. Comparison ofbacillary indexes in slit skin smears, skin andnerve biopsies; a study from Malawi. Int. J. Lepr.65 (1997) 211-216.

14. SCHAFER, B. W. and HEIZNIANN, C. W. The 5-100family of EF-hand calcium binding proteins func-tions and pathology. Trends Biochem. Sem. 21(1996) 134-140.

15. SIIIIGAL, V. N., REGE, V. L. and RIA', M. Correla-tion between clinicai and histopathological classi-fication of leprosy. Int. J. Lepr. 45 (1977)278-280.

16. SINGif, N., AR0RA, V. K., RANIAN1, M., TICKOO, S.

K. and BluvriA, A. An evaluation of the S-100stain in the histological diagnosis of tuberculoidleprosy and other granulomatous dermatoses. Int.J. Lepr. 62 (1994) 263-267.

17. WoRLD flum:rit ORGANIZATION. Action Pro-gramme for the Elimination of Leprosy. Status Re-port. Geneva: World Health Organization, 1996.