S-100 immunostaining in the Distinction of Borderline Tuberculoid Leprosy from other Cutaneous Granulomas.

BACKGROUND: Histopathologic diagnosis of borderline tuberculoid leprosy (BTL) is fraught with hurdles. It overlaps with other granulomas and documenting nerve involvement is the key to correct diagnosis. This is difficult on H and E sections alone. S-100 immunostaining may help in this regard.
OBJECTIVES: To study the patterns of nerve involvement in BTL and other cutaneous granulomas using S-100 immunostain and compare its sensitivity with that of H and E staining, in both adequate and inadequate biopsies.
MATERIALS AND METHODS: A total of 20 cases of BTL were reviewed. And, 19 biopsies from other cutaneous granulomas were taken as controls. S-100 immunostaining was done on paraffin sections. The pattern of nerve involvement was graded as intact, infiltrated and/or fragmented, intact with perineural inflammation.
RESULTS: Of the 20 cases of BTL, S-100 demonstrated infiltrated and/or fragmented nerves in 15 and absent nerves in 5 cases. H and E stain identified neuritis in eight cases. The sensitivity of S-100 and H and E is 0.78 and 0.41. In the 19 controls, S-100 identified normal nerves in 16 with 7 showing perineural inflammation only and their absence in 2 cases. H and E identified normal nerves in nine cases. The sensitivity of S-100 and H and E is 0.83 and 0.41. In biopsies where subcutis was absent, the sensitivity of S-100 in identifying nerve involvement is 0.66 compared with H and E 0.33.
CONCLUSION: S-100 staining is an efficient ancillary aid in distinguishing BTL from other granulomas and is superior to H and E in identifying nerve involvement, even where subcutis is absent. Infiltration and/or fragmentation of nerves by S-100 is the only reliable marker of BTL.

What was known?

Identifying nerve involvement is the key to diagnosis of leprosy in the tuberculoid spectrum and this is often difficult, owing to sampling errors or complete destruction of nerves.

Introduction

Leprosy is probably the oldest known disease to mankind and is widely prevalent in India. According to the latest data presented by the National Leprosy Eradication Programme NLEP, the annual new case detection rate is 10.48 per 100,000 population, that is, about 1.27 lakh new cases in 2010-11.[1] Prevalence rates have gone up in some endemic areas and leprosy continues to remain a public health burden.

Histopathology is the key to diagnosis of leprosy. Diagnosis usually is not difficult in the lepromatous spectrum because of the abundance of bacilli in the lesions.[2] In tuberculoid and indeterminate forms, diagnosis is more difficult. The granulomas overlap morphologically with other infectious and non-infectious lesions, notably tuberculosis, sarcoidosis and fungal infections, which are also common in our country.[3] The quest for bacilli by Fite-Faraco technique is often unrewarding. It is cutaneous nerve involvement by the inflammatory reaction that permits the differential diagnosis of leprosy from the other cutaneous granulomas. Even this feature is not easily discernable always. The pathologist often faces difficulties in visualizing nerve impairment, especially in inadequate biopsies. In sections stained with H and E, nerve remnants within granulomas are confused with epithelioid cells, fragments of arrector pili muscles, and small vessels.[4]

S-100 is an acidic calcium binding protein, an immunocytochemical marker of Schwann cells of peripheral nerves. The utilization of S-100 protein as a nerve marker in sections of tuberculoid leprosy, sarcoidosis, lupus vulgaris, and fungal infections that share a granulomatous pattern could help in diagnosis and distinction of the above infections. There are older reports that have stressed on its utility but it is still not widely used in practice.[34567] This study seeks to add to the existing data, with particular regard to biopsies deemed “inadequate”.

Materials and Methods

A total of 20 skin biopsies with histopathological diagnosis signed out as borderline tuberculoid leprosy (BTL) (as defined by Ridley and Jopling) were included for the study. There were no cases of TT leprosy in the study. A total of 19 biopsies signed out as granulomatous inflammation-lupus vulgaris, tuberculous verrucosa cutis, sarcoidosis, granuloma annulare, granulomatous rosacea, and deep fungal infections were taken as controls. Diagnosis was established based on histopathology, treatment and follow-up data, cultures and polymerase chain reaction for tuberculosis and fungal infections, in some cases.

Formalin fixed, paraffin embedded tissue was cut at 5 microns. Sections were stained by H and E to ensure that the tissue was representative. The original H and E sections, special stains for fungi (periodic acid-Schiff, Gomori methenamine silver) and mycobacteria (Ziehl-Neelsen, modified Fite-Faraco) were reviewed. Immunohistochemistry was done by labeled streptavidin-biotin method using antibody to S-100 (Biogenex clone15E2E2, purified bovine protein) after heat antigen retrieval.

H and E slides and S-100 immunostained slides were examined independently by four different investigators blinded to the original diagnosis. Biopsies were considered adequate when subcutis was present.

Both H and E and S-100 stained slides were studied for the patterns of nerve involvement. They were categorized as: Absent, intact, intact with perineural inflammation, infiltrated and/or fragmented.

Sensitivities of H and E and S-100 stains in identifying nerve involvement were calculated. Chi-square test was used to determine the significance of the various patterns in the two groups, with a P < 0.05 taken as significant.

Results

There were 20 cases of BTL and 19 non-BTL granulomas. Biopsies of 14 BTL and 12 other granulomas were adequate, that is, showed subcutis. Table 1 depicts the comparison of H and E and S-100, including their sensitivity, in identifying nerve involvement.

Table 1

Comparison of H and E staining and S-100 immunostaining in the identification of nerves in both adequate and inadequate biopsies of borderline tuberculoid leprosy(n=20) and other cutaneous granulomas(n=19)

Out of a total of 39 biopsies, H and E identified nerves in 17, while S-100 identified nerves in 31 cases. S-100 proved to be more sensitive than H and E in both adequate and inadequate biopsies with overall sensitivity being 0.78 versus 0.41.

The patterns of nerve involvement in BTL versus other granulomas are shown in Table 2.

Table 2

Patterns of nerve involvement by S-100 staining in borderline tuberculoid leprosy and other cutaneous granulomas

Infiltrated and/or fragmented nerves were seen only in BTL. In BTL, S-100 identified nerve involvement in 15 cases, in comparison to H and E alone, which identified only 8 cases [Figure 1] The presence of normal nerves or nerves with perineural inflammation alone by S-100 favored a diagnosis other than leprosy [Figure 2]. Absence of nerves, even in adequate biopsies, was seen in two cases each of BTL and lupus vulgaris.

Figure 1

(a) Periadnexal granulomas in borderline tuberculoid leprosy with no identifiable nerves (H and E, ×100) (b and c) Nerve fibers infiltrated by inflammatory cells and fragmented (S-100, ×200 and ×400)

Figure 2

(a and b) Perifollicular granulomas in Rosacea with no identifiable nerves (H and E, ×100, ×200) (c and d) Intact nerve fibres with perineural inflammation (S-100, ×200 and ×400). (d) In addition shows an eccrine gland that has stained with S-100

Discussion

Leprosy is a neurotropic disease caused by Mycobacterium leprae.[2] This property is invaluable to a histopathologist making its microscopic diagnosis. In the lepromatous end of the spectrum, a confident diagnosis is possible owing to the abundance of acid fast bacilli, which may be detected even on H and E sections with a bluish hue.[28] TT leprosy, however, is a microscopic challenge due to many reasons. First of all, organisms are not demonstrated by special stains and even if present in low numbers, are elusive except to the most persistent and experienced eyes.[2] Nerve involvement alone is insufficient, as nerve encasement by granulomas may also be encountered in sarcoidosis and lupus vulgaris.[4] In this respect, pathologists seek features like the elongated contour of granulomas as they follow neurovascular bundles of the adnexae, absence of caseation, epidermal changes and according to some, the absence of normal nerves in adequate biopsies as clues to leprosy diagnosis.[6] The last criterion is somewhat debatable, since it is dependent on sampling and the size of tissue. Morphologically, the presence of inflammatory cells or granulomas in the endoneurium are robust indicators. Nerve remnants within granulomas are often confused with epithelioid cells, fragments of smooth muscle, or endothelial cells.[4]

S-100 is an immunohistochemical marker for Schwann cells, melanocytes, Langerhans cells, chondrocytes, and myoepithelial cells.[3] It is a useful tool to define nerves, especially those within or in the vicinity of granulomas. Previous studies have demonstrated the superiority of S-100 over routine H and E staining.[3467] Fleury and Bacchi[4] were the first to use this technique in biopsies of clinically suspected BTL, but with no histologic or bacteriologic proof. They affirmed its utility by demonstrating both nerve dissociation and presence of nerve twigs within granulomas. Perineural infiltration by inflammatory cells was also noted in this study. Thomas et al.,[3] used S-100 staining on 25 samples from BTL. They noted that nerve fragmentation was the commonest pattern seen in 80% of cases, followed by total absence of nerves 16%. Singh et al., found that complete absence of nerve twigs in an adequate biopsy is the most reliable criterion for leprosy.

The present study reaffirms the superiority of S-100 over H and E in identifying nerves, whether normal or diseased. Even in biopsies considered inadequate owing to lack of subcutis, S-100 is twice as sensitive as routine staining. None of the previous reports have commented on this aspect. This is a very important and useful finding particularly for lesions on the face, where a deep biopsy may not be feasible.

In this study, infiltration and/or fragmentation of nerves by inflammatory cells or granulomas by S-100 emerged as the only significant decisive factor for leprosy. We noted the absence of nerves in adequate biopsies, touted by earlier studies as virtually diagnostic of leprosy, in both BTL and other granulomas and it did not help us in making the distinction.[6] Lack of nerves could imply sampling error and is not conclusive of leprosy. In such a situation, good clinicopathologic correlation is obligatory.

The findings of normal nerves alone or perineural inflammation alone on S-100 staining were exclusively seen in other granulomas. In leprosy, these findings were seen focally, but always in association with nerve infiltration/fragmentation. On H and E staining, the presence of normal nerves in cutaneous granulomas is a snare leading to a false negative diagnosis. S-100 is very useful in this regard, since it demonstrates both involved and normal nerves. The previous studies have not emphasized this feature.[346]

One has to also be aware of certain pitfalls when interpreting S-100 staining. Myoepithelial cells, Langerhans cells, and smooth muscle also stain positively.[3] Particularly, Langerhans cells are seen within granulomas and are difficult to tell apart from completely fragmented nerves. In this study, we have not considered isolated S-100 positive cells as nerves. They had to be present in small clusters or in the contour of nerve twigs. There are reports of positive staining of mycetoma granules and M. leprae by S-100.[9]

To summarize, S-100 is superior to H and E in identifying nerves, whether intact or damaged, even in biopsies lacking subcutis. Infiltration with or without fragmentation of nerves is the only reliable indicator of leprosy. Absence of nerves in adequate biopsies is not exclusive to leprosy and should not be regarded as decisive. The finding of normal nerves on H and E is a potential source for misdiagnosis as both fragmented and intact nerves are seen in leprosy. We therefore root for performing S-100 immunostain routinely in all biopsies with a clinical suspicion of BTL. It is a simple, practical, and extremely valuable adjunct that facilitates early diagnosis and treatment of leprosy.

What is new?

This study reaffirms the role of S-100 immunostaining in identifying nerve involvement in BT leprosy.

Fragmented and/or infiltrated nerves are specific for leprosy.

Absence of nerves cannot be taken as conclusive proof for leprosy.

Both intact and infiltrated nerves may be seen in a single biopsy.