Review articles

1. Application of the immunoperoxidase method for
  histopathological diagnosis of infectious diseases

2. Occupational risk of tuberculosis in pathology
  Textbooks and atlases

3. Recommended textbooks and atlases




Yutaka TSUTSUMI, M.D., Department of Pathology, Fujita Health University School of Medicine, Toyoake, Japan

back next

   The present review describes our own experience in the application of immunoperoxidase staining to routine histo- and cytodiagnosis.

A) Use of commercially available antibodies: 1) Mycobacterial infection (tuberculosis, atypical mycobacterial infection and leprosy) was demonstrated by using a BCG antiserum, with much higher sensitivity than Ziehl-Neelsen's acid-fast method. 2) Chlamydial and bacterial epididymitis was distinguished by immunostaining for C. trachomatis and E. coli. The chlamydial antigens were identified in pap-stained cytologic preparations after bleaching the dyes in acid alcohol, while prostatic malakoplakia was clearly positive for the E. coli antigens.
B) Use of patients' sera: Diluted patients' sera became convenient probes for indirect immunoperoxidase localization of pathogens in paraffin sections, particularly when cellular tissue reaction was evident histologically. Examples included Staphylococcal pyoderma, cat scratch lymphadenitis, cryptococcosis, sporotrichosis, alternariosis, amebic dysentery, acanthoamebic meningoencephalitis, cutaneous and visceral leishmaniasis, schistosomiasis, gnathostomiasis, liver ascariasis, etc. Endogenous human IgG in sections was scarcely detected by the peroxidase labeled secondary antibody. Similarly, sera of animals experimentally infected with Treponema pallidum and Toxoplasma gondii were applicable to human material.
C) Immunostaining and non-isotopic in situ hybridization: Comparison was made in human specimens infected by cytomegalovirus (CMV), human papillomavirus (HPV) and Epstein-Barr virus. In the latter two oncogenic viruses, the viral antigens were less frequently detectable than the viral genomes in cervical severe dysplasia and Hodgkin's disease.
D) Ultrastructural visualization of pathogens in routine material: The antigens of C. trachomatis, E. coli, CMV and HPV were seen directly in paraffin sections by applying pre-embedding immunoelectron microscopy. This approach was useful to confirm the presence of pathogens within the lesions and the specificity of the antibodies. The viral genomes were also identifiable at the ultrastructural level.

 back next

   It is of no doubt that the detection of infectious agents within the lesion is essentially important for the histopathologic diagnosis of infectious diseases. Immunohistochemical demonstration of the pathogen is sensitive enough for this purpose. When one knows the cross-reactivity of the antibodies to be applied, the results will lead the patients directly to the appropriate diagnosis and treatment.
   In the present review, we report variegated applications of immunoperoxidase staining of pathogens to the diagnosis in surgical pathology and clinical cytology. A total of four aspects are demonstrated, including A) representative immunostaining with commercially available antibodies, B) use of diluted patients' sera as the primary antibody, C) comparison of immunostaining with non-isotopic in situ hybridization, and D) ultrastructural demonstration of pathogens in routine material.

 back next

   The paraffin-embedded specimens used in the present series were mainly prepared in the Division of Diagnostic Pathology, Tokai University Hospital, Isehara. The tissues, taken either by biopsy, during surgery or at autopsy, had routinely been fixed in 10% unbuffered formalin for 1 to 14 days. These included 1) granulomatous, fibrous or calcified lesions caused by mycobacterial infection or by sarcoidosis or related conditions for the study using a bacillus Calmette Guerin (BCG) antiserum, 2) acute and chronic epididymitis and prostatic malakoplakia for demonstrating Chlamydia trachomatis and Escherichia coli (E. coli) antigens, 3) histopathology specimens of Staphylococcal pyoderma, cat scratch lymphadenitis, cryptococcal meningitis, cutaneous sporotrichosis, cutaneous alternariosis, amebic dysentery, acanthoamebic meningoencephalitis, cutaneous and visceral leishmaniasis, schistosomiasis, gnathostomiasis, liver abscess caused by old ascariasis, primary and secondary syphilis, and toxoplasma encephalitis for the study using sera of affected patients and experimentally infected animals, 4) cutaneous verruca vulgaris, cervical and penile condyloma acuminatum and cervical severe dysplasia for the study of human papillomavirus (HPV), 5) cytomegalic interstitial pneumonitis for detecting cytomegalovirus (CMV), 6) lymph node lesions of Hodgkin's disease for demonstrating Epstein-Barr virus (EBV), and 8) cytologic cervical pap smears for identifying C. trachomatis and HPV.
   The antigens were visualized by the indirect immunoperoxidase technique in consecutive deparaffinized sections, as described earlier (17, 20). Information of the primary antibodies and patients' and animals' sera is listed in Table 1. The second layer reagents included horseradish peroxidase (HRP)-labeled IgG to rabbit or mouse immunoglobulins or human IgG (Amersham International, Amersham Bucks, England) or to goat immunoglobulins (ICN Immunobiologicals, Irvine, CA, USA) at a 1:50 dilution. Endogenous peroxidase activity was either not blocked or inhibited in methanol containing 0.3% hydrogen peroxide for 20 min. Before antibody incubation, 0.1% trypsin pretreatment for 30 min was required for the detection of EBV-related latent membrane antigen-1 (LMP-1). Incubation time for the antibodies was 30 min. For the HRP coloring reaction, 0.05 M Tris-HCl solution, pH 7.6, containing 20 mg/dl diaminobenzidine tetrahydrochloride (Wako Pure Chemicals, Osaka, Japan), 65 mg/dl sodium azide and 0.003% hydrogen peroxide was employed. The nuclei were counterstained with 5% methyl green buffered by 0.1 M Veronal acetate at pH 4.0 or lightly with hematoxylin. The pap-smeared preparation was immunostained, after taking photomicrographs and bleaching the dyes overnight in acid alcohol (6).
   Commercially available biotinylated probes were used for in situ hybridizing viral genomes in paraffin sections mounted on 3-aminopropyltrimethoxysilane-coated glass slides (19). A double stranded cDNA probe for CMV (Patho-gene II) was purchased from Enzo Biochem, New York, NY, USA, and cDNA probe cocktails for HPV types 6 and 11 (HPV 6/11) and types 16 and 18 (HPV 16/18) were obtained from Life Technologies, Gaithersburg, MD, USA. A biotinylated oligonucleotide probe for small nuclear RNA (EBER-1) of EBV was the product of Dako, Carpinteria, CA, USA. Deparaffinized sections were pretreated with 0.1 mg/ml proteinase K at 37ーC for 20 min. The probes for CMV and HPV and tissue DNA were simultaneously denatured into single stranded by heating the slides on a hot plate at 95ーC for 5 min. The heat denaturation step was skipped for demonstrating the EBV genome. The hybridization was extended overnight at 37ーC in 4x standard sodium citrate (SSC: 0.15 M sodium chloride plus 0.015 M trisodium citrate). Post-hybridization washing was done with 1x SSC at 37ーC for 10 min. HRP-labeled streptavidin (Dako) was used as the second layer reagent.
   For ultrastructural localization of the antigens and genomes by the pre-embedding sequence, the standard inverted beam capsule method directly followed the immunostaining and in situ hybridization for the light microscopic level without any special modification (18, 19). For the cytologic material, the single pap specimen was thus able to be processed finally for electron microscopy (6).
   To demonstrate the specificity of the antibodies, cultured bacterial and fungal colonies suspended in 10% formalin overnight were cytospun onto the glass slides for immunostaining. The diluted BCG antiserum was absorbed for 1 hr with the suspension of fixed Candida albicans. A variety of paraffin sections of infectious lesions were additionally immunostained, in order to show the cross-reactivity of the patients' sera (17). Normal rabbit, mouse, goat or human serum served for negative control staining.

 back next

A) Use of commercially available antibodies
1) Detection of mycobacterial antigens:
   The rabbit antiserum to BCG diluted at 1:3,000 was a very sensitive probe for demonstrating mycobacterial antigens in paraffin sections (10, 21). As shown in Table 2, not only exudative (active) tuberculous lesions but also a considerable percentage (19/32 = 59%) of the encapsulated caseous and fibrocalcified lesions showed positivity, while Ziehl-Neelsen's acid-fast staining rarely gave positive findings in such old tuberculous lesions (10). In addition, the judgment whether positive or negative was easily reached in the BCG immunostaining. Hence, the observation time was saved significantly. Foamy positive signals were seen in the cytoplasm of live or dead macrophages: Bacilliform positivity was scarcely obtained. The reason for this should be that the mycobacterial antigens remained in lysosomes of the macrophages even after phagocytosis and digestion of the bacilli. Normal human tissue components and sarcoidosis or related non-caseous granulomas remained unstained. Staining with the normal rabbit serum served as negative controls.
   The BCG antiserum allegedly detecting more than 100 antigenic substances present on the BCG cell body (1) was reactive with atypical mycobacteria and Mycobacterium leprae. In the dermal lesion of lepromatous leprosy, the cytoplasm of foamy macrophages was diffusely and strongly immunoreactive. This antiserum was found to be further cross-reactive with Gram-positive bacteria and fungi. Absorption by formalin-fixed Candida albicans resulted in abolishment of the cross-reaction with the non-mycobacterial species without any change in positive staining in the mycobacterial lesions (10). However, the absorption procedure seemed to be unnecessary for the practical application, since the above bacterial and fungal lesions could easily be distinguished in H&E preparations.
   In the Division of Diagnostic Pathology, Tokai University Hospital, we have applied this diagnostically valuable method, BCG immunostaining, for over years on a routine basis. Recently, an exceptional case was encountered, in which an endoscopic lung biopsy specimen containing small non-caseating granulomas failed to reveal the BCG antigens, but the acid-fast staining demonstrated a few red-stained bacilli. Therefore, we recommend to employ a combination of BCG immunostaining and Ziehl-Neelsen's staining for the diagnosis of mycobacterial diseases. It should be reminded that some antisera give a false positive reaction in mycobacterial lesions (9), since they were raised by immunizing animals with Freund's complete adjuvant containing dead tuberculous bacilli.

2) Detection of chlamydial and bacterial (E. coli) antigens:
   On the basis of immunohistochemistry, active epididymitis was able to be divided into two categories: chlamydia-induced proliferative and non-destructive epididymitis, and bacteria-induced destructive and abscess-forming epididymitis (7). Of 16 surgical lesions of active non-tuberculous epididymitis, six lesions were shown to be chlamydial, and seven bacterial. Chlamydial epididymitis was characterized by the occurrence of cytoplasmic inclusions positively labeled for the chlamydial antigens in some of the proliferative ductal epithelial cells. Periductal inflammation was pronounced, but tissue destructive change was minimal. In contrast, in bacterial epididymitis the cytoplasm of foamy macrophages clustered in abscess cavities or xanthogranulomas showed a diffusely positive reaction for the E. coli antigens. The latter-type lesion was highly destructive, so that it was easy for us to differentiate it from the chlamydial lesion histopathologically, even without aid of immunostaining (7). Appropriate recognition of the histologic features of chlamydial epididymitis seems to be especially important for the prevention of this sexually transmissible infection. Ortho's chlamydial monoclonal antibodies containing two clones were specific for this abortive bacterial species (16). The polyclonal antibody to E. coli was found to be cross-reactive with other Gram-negative bacilli and some Gram-positive cocci, so that the antiserum should be regarded a pan-bacterial reagent (7).
   The chlamydial antigens were also demonstrable in cervical pap smears, after bleaching the dyes in acid alcohol. By photomicrographic comparison of the same cells, it was evident that the cytoplasmic "nebular" inclusion was the site of chlamydial infection (6). The E. coli-associated bacterial antigens were clearly demonstrated in the cytoplasm of foamy macrophages in the lesion of prostatic malacoplakia or related xanthogranulomatous lesions, as has been reported previously (14).

B) Use of patients' sera and experimentally infected animals' sera
   Patients' sera are expected to contain high-titered IgG-type antibodies against the pathogen, particularly when cellular tissue reactions such as abscesses and granulomas are evident histologically. We have successfully identified pathogens at the site of infection in paraffin sections, by employing patients' sera diluted at 1:10 to 1:1,000 as the primary antibody and HRP-labeled anti-human IgG as the second layer reagent. Examples included Staphylococcal pyoderma, cat scratch lymphadenitis caused by infection of rickettsia-like small bacteria (2), cutaneous sporotrichosis, cryptococcal meningitis, cutaneous alternariosis, amebic dysentery, acanthoamebic meningoencephalitis, cutaneous and visceral leishmaniasis, schistosomiasis, gnathostomiasis, and ascariasis-induced liver abscess, as described earlier (17, 20). In formalin-fixed, paraffin-embedded sections, endogenous human IgG was hardly demonstrated by the HRP-labeled second antibody. For this purpose, the indirect immunoperoxidase method should thus be chosen: Sequences showing a higher sensitivity, such as the streptavidin-biotinylated peroxidase complex method and labeled streptavidin-biotin method, would result in high background staining due to detection of endogenous IgG in tissue fluid.
   It should be noted that IgG in the patients' sera showed cross-reactivity to related pathogens wider than expected (17). In bacterial and fungal infection, the sera served as pan-bacterial and pan-fungal probes, respectively. In protozoal and helminthic infection, the specificity was much narrower with limited cross-reactions. In spite of such broad specificity in the former, this convenient procedure is excellent in selectively identifying the pathogen within the lesion in question. What one should do is, instead of ordering to purchase an expensive antibody of unknown quality, to make a phone call to clinicians or laboratory technicians to ask to save a small aliquot of patients' sera, soon after the microscopic confirmation of host response in histopathology specimens. This is particularly true when specific antibodies are not listed in the commercial catalog.
Similarly, sera of animals experimentally infected with pathogens became a powerful weapon (8). The serum of a rabbit with experimental orchitis provoked by the local injection of Treponema pallidum was useful in demonstrating the spiral microbes in biopsy specimens of primary syphilis in the penis and in those of secondary syphilitic gastritis. The serum of a mouse with experimentally induced toxoplasma encephalitis was specifically reactive to cell bodies of Toxoplasma gondii in the human brain and subarachnoid space.

C) Comparison of immunostaining with non-isotopic in situ hybridization
1) CMV identification:
   In cytomegalic pneumonitis, the same cells with intranuclear inclusions of varying sizes were labeled comparably by both of the histochemical methods. The CMV antigens detected with Polysciences' antiserum were localized in the cytoplasm of the infected pneumocytes, while the CMV-DNA signals were observed mainly in their nuclei (11).

2) HPV identification:
   Details were reported earlier (18, 19). Among the HPV-induced squamous lesions, skin wart (verruca vulgaris), known to be caused by infection of HPV types 1, 2 or 4, was positive by Dako's antiserum against the 5.4 kDa HPV common capsid antigen. The positive signals, typically seen in the pyknotic nuclei with perinuclear haloes, were distributed mainly in the granular layer and hyperkeratotic cornified layer. HPV 6/11 and HPV 16/18 genomes were not identified. In condyloma acuminatum with koilocytic change, more cells were labeled for HPV 6/11 DNA than for the viral antigen. HPV 16/18 DNA was negative. In severe dysplasia of the uterine cervix, a few dyskaryotic cells located in the uppermost part of the dysplastic mucosa showed dot-like nuclear positivity of oncogenic HPV 16/18 DNA. The viral antigen and HPV 6/11 DNA were undetectable. The oncogenic HPV DNA was also detectable in the nuclei of severely dyskaryotic cells in the cervical smear preparation, when silane-coated glass slides were utilized for smearing to avoid sloughing the cells off.

3) EBV identification:
   In the lymph node involved by Hodgkin's disease, both the EBV-related LMP-1 antigen and EBV-related small nuclear RNA (EBER-1) were demonstrated predominantly in Hodgkin's cells and Reed-Sternberg's giant cells. LMP-1 was seen in the cytoplasm, while EBER-1 was localized in the nuclei. In general, more cells were labeled for the viral genome than for the viral antigen (13). However, some cases revealed antigenic predominance. Two nodular sclerosis-type lesions were negative for EBV-related expression. At present, we recommend the combined application of immunostaining and in situ hybridization for detecting the oncogenic EBV virus in such neoplastic lesions as Hodgkin's disease (3), non-Hodgkin's lymphomas of Burkitt or T-cell type (4), nasopharyngeal carcinoma (23) and blue cell carcinoma of the stomach (15).

D) Ultrastructural demonstration of pathogens in routine material
1) Chlamydial and bacterial (E. coli) antigens:
   The immunostained sections as described in A-2) were further processed for pre-embedding immunoelectron microscopy. For example, chlamydial bodies in duct epithelial cells and rod-shaped bacilliform bodies in histiocytes can be demonstrated ultrastructurally in paraffin-embedded epididymitis. The cell walls were positively labeled by the diaminobenzidine-osmium tetroxide complex. Although the morphologic preservation was not good enough, infection of pathogenic microbes was clearly demonstrated in the archival material at the ultrastructural level (7). Similarly, intracellular chlamydial infection was confirmed ultrastructurally in metaplastic cells in cervical smear preparations (6). Not only the chlamydial antigens but also the chlamydial genome were able to be demonstrated in acetone-fixed cultured McCoy cells at the ultrastructural level (5).
   Ultrastructural localization of infectious agents in routine specimens is valuable for demonstrating the intracellular site of infection, particularly when the retrospective study is inevitable or the number of infected cells is limited. In addition, this approach makes a useful tool for confirming the specificity of the antibody used.

2) CMV antigens and genome:
   The CMV antigens were ultrastructurally observed in the viral particles distributed in the cytoplasm of the infected cell, whereas the viral genome was localized in both the cytoplasmic viral particles and the matrix of the nuclear inclusion (11). Particulate signals were also sparsely seen in the nucleus. The diffuse (non-particulate) nuclear labeling by in situ hybridization indicated active intranuclear replication of viral DNA prior to the formation of viral particles in the cytoplasm. In our condition employing routine paraffin sections, the nuclear labeling was much clearer and denser than the finding of in situ hybridization at the electron microscopic level reported by Wolber et al. (22), who performed the pre-embedding sequence using a streptavidin-labeled colloidal gold reagent in Triton X-100 treated cultured cells.

3) HPV antigens and genome:
   HPV-related antigens and genome can be demonstrated at the ultrastructural level in verrucous and precancerous squamous lesions (18, 19). In verruca vulgaris, HPV antigen-immunoreactive viral particles measuring 50-70 nm were ultrastructurally identified in the nuclei of superficial keratinocytes with a perinuclear haloe and keratohyaline granules. In koilocytic cells of condyloma acuminatum, the nuclear matrix was filled with fine particles, 15-20 nm in size, positively labeled for HPV 6/11 DNA. This pattern is indicative of intranuclear distribution of episomes, consisting of naked DNA without capsid formation. In severe dysplasia of the uterine cervix, viral DNA signals of HPV 16/18 type corresponded to part of nuclear chromatin in a localized islet-like fashion. The features were strongly suggestive of integration of the oncogenic viral DNA into host genomic information in this precancerous lesion (19). Similar intranuclear localization of HPV 16 DNA was recently described by Lin, et al. in CaSki cell line (12).

 back next

   A wide variety of infectious agents have so far been detected immunohistochemically in routine histo- and cytopathologic specimens (20). The present article covers limited aspects of this approach, but what the author wanted to demonstrate is the unexpectedly powerful potentiality of routinely processed paraffin sections. On occasion, as in the case of immunostaining using antisera against BCG and E. coli and patients' sera, the sensitivity of detection is more important than the specificity. The reasons for this are as follows: 1) Basically, the histologic pattern of host reaction is the key aspect for presuming the causative agent. 2) Determination whether chlamydial, bacterial, fungal or viral should be a tentative goal of report in histologic examination. 3) Whether the lesion is caused by E. coli, Pseudomonas aeruginosa or Mycobacterium tuberculosis, for example, should primarily be determined by non-histologic methodology, such as microbial culture, serologic test and DNA analysis. It must be emphasized again that the appropriate histologic and cytologic diagnosis, not rarely aided by the appropriate immunohistochemical staining, is essentially beneficial for patients to lead to the appropriate treatment.

 back next


1. Closs, O., Harboe, M., Axelsen, N.H., Bunch-Christensen, K. and Magnusson, M.: The antigen of Mycobacterium bovis, strain BCG, studied by crossed immunoelectrophoresis: A reference system. Scand. J. Immunol. 12: 249-263, 1980.
2. Dolan, M.J., Wong, M.T., Regnery, R.L., Jorgensen, J.H., Garcia, M., Peters, J. and Drehner, D.: Syndrome of R. henselae adenitis suggesting cat scratch disease. Ann. Intern. Med. 118: 331-336, 1993.
3. Herbst, H., Niedobitek, G., Kneba, M., Hummel, M., Finn, T., Anagnostopoulos, I., Bergholz, M., Krieger, G. and Stein, H.: High incidence of Epstein-Barr virus genomes in Hodgkin's disease. Am. J. Pathol. 137: 13-18, 1990.
4. Ho., F.C., Srivastrava, G., Loke, S.L., Fu, K.H., Leung, B.P., Liang, R. and Choy, D.: Presence of Epstein-Barr virus DNA in nasal lymphomas of B and T cell type. Hematol. Oncol. 8: 271-281, 1990.
5. Hori, S., Kawai, K., Tsutsumi, Y. and Osamura, R.Y.: Ultrastructural demonstration of Chlamydia trachomatis DNA by in situ hybridization using a biotinylated DNA probe, in comparison with immunoelectron microscopy. Med. Sci. Res. 19: 429-430, 1991.
6. Hori, S., Itoh, H., Tsutsumi, Y. and Osamura, R.Y.: Immunoelectron microscopic detection of chlamydial antigens in Papanicolaou-stained vaginal smears. Acta Cytol 38, 1994 (in press).
7. Hori, S. and Tsutsumi, Y.: Histologic differentiation between chlamydial and bacterial epididymitis: Non-destructive and proliferative versus destructive and abscess-forming. Immunohistochemical and clinicopathologic findings. Hum. Pathol. 25, 1994 (in press).
8. Hunter, E.F., Greer, P.W., Swisher B.L., Simons, A.R., Farshy, C.E., Crawford, J.A. and Sulzer, K.R.: Immunofluorescent staining of Treponema in tissues fixed with formalin. Arch. Pathol. Lab. Med. 108: 878-880, 1984.
9. Kahn, H.J. and Thorner, P.S.: "False immunohistochemical positivity" associated with mycobacterial infection in acquired immune deficiency syndrome. Am. J. Surg. Pathol. 16: 1126, 1992.
10. Kawai, K. and Tsutsumi, Y.: Detection of acid-fast bacilli by the immunoperoxidase method. Comparison with the conventional acid-fast staining. Byori-to-Rinsho 2: 862-867, 1984 (in Japanese).
11. Kawai, K., Hori, S., Yamazaki, H. and Osamura, R.Y.: Detection of cytomegalovirus in paraffin sections by in situ hybridization. Byori Gijutsu 39: 13-18, 1989 (in Japanese).
12. Lin, C.-T., Chen, C.-C., How, S.-W., Huang W.-M. and Peck, K.: Localization of HPV-16 DNA sequence in CaSki cells by electron microscopic hybridocytochemistry. J. Histochem. Cytochem. 40: 467-473, 1992.
13. Nakui, M. and Tsutsumi, Y.: Histochemical studies of EB virus-infected cells in Hodgkin's disease. Tr. Soc. Pathol. Jpn. 83: 308, 1994 (Abstr. in Japanese).
14. Qualman, S.J., Gupta, P.K. and Mendelsohn, G.: Intracellular Escherichia coli in urinary malacoplakia: A reservoir of infection and its therapeutic implications. Am. J. Clin. Pathol. 81: 35-42, 1984.
15. Shibata, D., Tokunaga, M., Uemura, Y., Sato, E. and Tanaka, S.: Association of Epstein-Barr virus with undifferentiated gastric carcinomas with intense lymphoid infiltration. Lymphoepithelioma-like carcinoma. Am. J. Pathol. 139: 469-474, 1991.
16. Stamm, W.E., Tam, M., Koester, M. and Cles, L.: Detection of Chlamydia trachomatis inclusions in McCoy cell cultures with fluorescein-conjugated monoclonal antibodies. J. Clin. Microbiol. 17: 666-668, 1983.
17. Tsutsumi, Y., Kawai, K. and Nagakura K.: Use of patients' sera or immunoperoxidase demonstration of infectious agents in paraffin sections. Acta Pathol. Jpn. 41: 673-679, 1991.
18. Tsutsumi, Y., Kawai, K., Hori, S. and Osamura, R.Y.: Ultrastructural visualization of human papillomavirus DNA in verrucous and precancerous squamous lesions. Acta Pathol. Jpn. 41: 757-762, 1991.
19. Tsutsumi, Y.: Electron microscopic localization of pathogenic genomes by in situ hybridization. Byori-to-Rinsho 10: 1405-1409, 1992 (in Japanese).
20. Tsutsumi, Y.: Immunohistochemistry in infectious diseases. Byori-to-Rinsho 11: 320-327, 1993 (in Japanese).
21. Wiley, E.L., Mulhollan, T.J., Beck, B., Tyndall, J.A. and Freeman, R.G.: Polyclonal antibodies raised against Bacillus Calmette-Guerin, Mycobacterium duvalii, and Mycobacterium paratuberculosis used to detect mycobacteria in tissue with the use of immunohistochemical techniques. Am. J. Clin. Pathol. 94: 307-312, 1990.
22. Wolber, R.A., Beals, T.F., Lloyd, R.V. and Maassab, H.F.: Ultrastructural localization of viral nucleic acid by in situ hybridization. Lab. Invest. 59: 144-151.
23. 23. Wu, T.-C., Mann, R.B., Epstein, J.I., MacMahon, E., Lee, W.A., Charache, P. Hayward, S.D., Kurman, R.J., Hayward, G.S. and Ambinder, R.F.: Abundant expression of EBER-1 small nuclear RNA in nasopharyngeal carcinoma. A morphologically distinctive target for detection of Epstein-Barr virus in formalin-fixed paraffin-embedded carcinoma specimens. Am. J. Pathol. 138: 1461-1469, 1991.


Table 1. Primary antibodies and patients' and animals' sera used in the present study


Species Source _@Dilution

C. trachomatis
E. coli

Treponema pallidum
Toxoplasma gondii

rabbit 1
mouse* 2
rabbit 1

goat 3
rabbit 1
rabbit 4
mouse 5


common mycobacterial Ag
detection kit
cross-reactive with
  other bacterial species
cytoplasmic antigens
5.4 kDa capsid antigen
experimental orchitis
experimental encephalitis

Sera of patients suffering from:
Staphylococcal pyoderma
cat scratch lymphadenitis
cutaneous sporotrichosis
cryptococcal meningitis
cutaneous alternariosis
amebic dysentery
acanthamebic meningoencephalitis
cutaneous leishmaniasis
visceral leishmaniasis
schistosomiasis haematobium
liver ascariasis


MRSA cultured
history of scratch by a cat
Sporotrichin reaction positive
opportunistic infection
clinically diagnosed
high immunofluorescent titer
high immunofluorescent titer
high immunofluorescent titer
high immunofluorescent titer
high immunofluorescent titer
gel diffusion test positive
gel diffusion test positive

*: monoclonal antibody (two-clone cocktail: A21.65 and K14.67)
1: DakoCytomation, Carpinteria, CA, USA
2: Ortho Diagnostic Systems, Rantan, NJ, USA
3: Polysciences, Warrington, PA, USA
4: Tuberculosis Research Institute, Kiyose, Japan
5: Department of Parasitology, Tokai University School of Medicine, Isehara, Japan


Table 2. Summary of BCG immunostaining and Ziehl-Neelsen's acid fast staining in mycobacterial and non-mycobacterial granulomatous lesions

Type of lesion

Z-N staining BCG immunostaining

Exudative lesion
Caseous granuloma
Non-caseous granuloma
Fibrocalcified lesion

5/6 (83%)
3/11 (27%)
0/6 ( 0%)
0/21 ( 0%)

6/6 (100%)
5/11 (45%)
1/6 (17%)
14/21 (67%)

Atypical mycobacterium infection
  Lepromatous leprosy
  Tuberculoid leprosy
Sarcoid-type granuloma
  Sarcoid-like reaction

8/44 (18%)

2/2 (100%)

2/2 (100%)*
0/3 ( 0%)

0/7 ( 0%)
0/3 ( 0%)
26/44 (59%)

2/2 (100%)

2/2 (100%)
0/3 ( 0%)

0/7 ( 0%)
0/3 ( 0%)

* Focally positive after Fite's modification


 back next

Occupational Risk of Tuberculosis in Pathology

Yutaka Tsutsumi, M.D.
Department of Pathology
Fujita Health University School of Medicine, Toyoake, Japan


   Infectious diseases of healthcare workers contracted through healthcare practices are called occupational infections. Tuberculosis is transmitted by inhaling contaminated air; i.e. "airborne transmission" caused by air-dispersed particulates (droplet nuclei) smaller than 5 mm in size. It is important to distinguish this from droplet transmission.1) Viral hepatitis caused by needlestick or a cut by sharps is another serious infection for healthcare workers. Treponema pallidum, human T-cell leukemia virus, human immunodeficiency virus and Creutzfeldt-Jakob's disease prion could also comprise a potential biohazard. Of no doubt is that the autopsy accompanies a direct exposure to various pathogens, including unexpected ones, and thus carries the highest risk of biohazard in all the medical practices available.2-5)

 back next

Risk of Tuberculosis Associated with Autopsy Services
   It is regrettable to say that at present, the percentage of correct clinical diagnosis of tuberculosis is much less than 50% in Japan.6) Namely, it is not a rare occasion to firstly see tuberculous lesions during post-mortem examination which were clinically undetected prior to the patient's death. In Tokai University Hospital, Isehara, Japan, 169 (6.2%) of 2705 autopsied cadavers for 13 years possessed infectious tuberculosis lesions in the lung, such as caseous foci and exudative lesions.7) In order to prevent biohazard of tuberculosis during autopsy, the importance of the appropriate recognition of the gross appearance of tuberculous lesions should be emphasized. Mycobacterium tuberculosis is one of the pathogens showing the highest infectivity. No less than 50% of those who inhale ten to the first power level of tubercle bacilli will contract tuberculosis, if they lack immunity. When the autopsy is undertaken, it is imperative for pathologists and autopsy assistants to have full knowledge on the infectivity of M. tuberculosis, such as the relative resistance against disinfectants and the route of infection (airborne transmission). Tuberculosis should not be classified into the orphan disease category. The biohazard is particularly high in case of infection of multidrug-resistant M. tuberculosis.8) In 1997, the Japanese newspapers described full of stories about an obituary notice for a nurse in a city hospital who had taken care of her tuberculosis patient. This tells that the multidrug-resistant M. tuberculosis has made the first advent in Japan. Regrettably to say, we still not rarely see Japanese newspaper articles describing endemics of tuberculosis in schools and hospital wards.
   Let's take a look at the Japan's current anti-infection status against tubercle bacilli in the autopsy facility. In regret, we must admit that we in the pathology field have been working in an excessively defenseless environment. This will continue in the future, unless we break with the false tradition, in which veteran pathologists ironically express that a pathologist becomes full-fledged only when he or she falls ill with tuberculosis.
   In 1988, we conducted a questionnaire-based statistical study on the incidence rate of tuberculosis among the pathology-related personnel (pathologist and pathology technicians) (n = 2,388) across the country. The data were based upon the history of treatment of tuberculosis after being engaged in pathology divisions (Table 1).9) The control group (n = 414) consists of staff and technologists in the departments of public health and preventive medicine. The annual incidence rate of tuberculosis among the pathology workers was proven to be 639.5 per 100,000 population. A total of 129 (5.4%) pathology workers were actually treated for occupation-related pulmonary tuberculosis. The odds ratio (relative risk) against the control group (incidence: 94.2/100,000 population) was 19.0. This is absolutely abnormal when compared to the incidence rates of 53.9 for the entire Japanese population and 30.0 for the Japanese healthy worker population (excluding children and the aged).
   The incidence rate of tuberculosis among the Japanese pathology workers stayed at 559.3 without sign of improvement over the past 10 years up until 1988, with even a greater rate at 673.8 seen for the younger generation with the age of 40 or less. It should be noted that these figures are even greater than the data (547.0) among the British pathologists in the 1950's.10) In the 1980's, the annual incidence of tuberculosis among the British healthcare workers, except for mortuary attendants, became lower than that of the general population.11) Also noteworthy was the fact that the incidence rate for the Japanese technicians who had assisted in post-mortem examination (n = 753) was as significantly high as 823.8/100,000 population. The odds ratio to the technician group not involved in the autopsy assistance (n = 422; the incidence 125.1/100,000) was 6.65. It was concluded that the extremely high incidence rate among the Japanese pathology workers was attributed to the fact that they were prone to be exposed to air-floating tubercle bacilli within the autopsy facility. The average ages at onset were 32.3±7.0 for pathologists and 29.6±8.0 for pathology technicians. A similar tendency was recorded in the questionnaire-based study in California in1979:3) The ages at onset ranged from 20 to 35 years in 75% of physicians suffering from tuberculosis.
   It was disclosed that the majority of Japanese healthcare workers suffering from tuberculosis showed a positive tuberculin skin test according to BCG vaccination. This means that the effect of BCG vaccination in the childhood has its own limitation in immunity to tuberculosis in adulthood,12,13) especially when there is a significantly high risk of exposure to a large amount of tubercle bacilli during medical practices, particularly the autopsy.

 back next

Risk of Tuberculosis Associated with Intraoperative Frozen Diagnosis
   Biohazard of tuberculosis also results from frozen tissue sectioning during the intraoperative frozen diagnosis. Tuberculous bacilli can survive inside the cryostat set around -20℃. The probability of mycobacterial airborne transmission greatly increases, especially when the frozen specimen is blown with a cryospray. Making a gross diagnosis of tuberculosis is very important in this situation.14) Namely, no frozen sections should be prepared, if tuberculosis is grossly suspected on a coin lesion of the lung. In the pathology laboratory, at least two cryostats should be furnished in order to have one machine ready to use while another is being decontaminated. Preferably, we should have biosafety-equipped machines.
   In order to know how biohazardous the cryosectioning procedure is, we did a model experiment using frozen colonies of Penicillin-resistant Staphylococcus saprophyticus.15) After cryosectioning of the bacterial colonies embedded in the OCT compound, 50 L air was vacuum-sampled at the height of the nose by a rotary compressor at five different sites in our "open-styled" cutting room, where a cryostat was placed in the corner. As a result, the bacteria were cultured on Penicillin-containing agar plates from the sampled air at all the sites examined, including the point 10 m apart from the machine, particularly after the use of the cryospray. The viability of the bacteria after freezing was expectedly low around 0.1%-1%. This experiment thus indicated a high biohazard during cryosectioning of tissue sections infected with cryotolerant M. tuberculosis.

 back next

Preventive Procedures against Infection
   The Center for Disease Control and Prevention, Atlanta, USA, has published general guidelines for preventing transmission of M. tuberculosis in healthcare facilities.16) A pathology-oriented guideline is provided from the Japanese Society of Pathology, concerning the designing of an biosafety-equipped autopsy facility and the safety and preventive procedures against infection.17) The former is important, but it involves a tremendous amount of costs. Listed below is a set of preventive procedures that are feasible to have them applied to every post-mortem examination practice from the viewpoint of "universal precautions" or "standard precautions".18)

1. No patient's record nor X-ray films are allowed to bring into the autopsy room.
2. Use a surgical mask equipped with activated carbon particles. Preferably, the mask with an eye shield should be worn to protect the eyes of the operators. Wear also a disposable cap and an arm/elbow protector.
3. The operators should be suitably attired, and if possible, use a disposable and water-proof autopsy garb and apron. However, this is contradictory to the viewpoint of the cost effectiveness and the volume reduction of medical wastes.
4. Observers or witnesses including clinicians should similarly and suitably be attired as much as possible.
5. Blood and other body fluids should be vacuumed by using an electric suction pump device.
6. After the organs and tissues were properly examined, put them back into the cadaver's body cavity leaving the necessary portions for gross and microscopic examinations.
7. The entire autopsy operation should be performed on a dissecting table as much as possible. The prosector must avoid staining the floor by blood or body fluids. Care should be taken lest body fluids and rinse water should be splashed around.
8. When the autopsy record form is stained with blood or body fluids, the record should be transcribed in a new form. If this is not feasible, the stained portion(s) should be marked for subsequent disinfection using hypochlorous acid solution.
9. After autopsy, all the equipment used, the boots, the aprons, the dissecting table, the sectioning table and the floor should thoroughly be disinfected with diluted hypochlorous acid solution.
10. After completion of the examination, a cadaver transportation stretcher is carried into the autopsy room. This must be preceded by cleaning and disinfection of the floor of the room.
11. After the post-mortem examination, the attendees must take a shower and wash their hair.
12. The used mask, gloves, elbow covers, etc. should be disposed into a special container, sealed properly and incinerated as a matter of routine for infectious medical wastes.
13. If practical, a bone-cutting device exclusively provided for infectious disease contamination should be used.
14. The autopsy results should be reported as prompt as possible, so that the autopsy findings are appropriately utilized in the hospital infection control activity.

 back next

Additional Precautions for Post-mortem Examinations for Patient Cadaver with Active Tuberculosis

1. Those who tuberculin skin test is negative should be excluded from the attendees list. When active tuberculous lesions are unexpectedly identified during autopsy, the skin test-negative attendees should immediately be replaced by the one(s) showing a positive skin reaction.
2. No observers are allowed to be admitted, or let them walk out of the autopsy room immediately.
3. Wear a respirator (specialized face-fit protector) with a disposable anti- particulate filter, type N95, available from 3M Healthcare Inc., which is recommended by the Center for Disease Control and Prevention, Atlanta, USA.16) For the perfect protection, use an infection-protective helmet (Steri-Shield Turbo II, developed by M&M Inc).
4. Preferably, formalin solution should be injected into the bronchial lumen to fix the removed lung tissue.
5. Incision or slicing of the involved organs and tissues should be done as minimal as possible, in order to avoid to yield infectious droplets and particles.
6. Frozen sections must not be prepared from the involved tissues. In case where frozen sectioning is needed, this should be preceded by thorough fixation of the tissues in the fixative such as a paraformaldehyde solution.
7. When bone tissue is sampled in cases of vertebral caries or miliary tuberculosis, use a chisel rather than the electric bone cutter lest bone powders should scatter around. When the electric bone cutter is used, protective procedures (e.g. covering by a plastic sheet or bag) should be applied.
8. Photograph taking of organs and tissues should be preceded by sufficient fixation in formalin solution. Do not take photographs when they are unfixed.
9. Used tools and utensils should be decontaminated with an appropriate disinfectant such as hypochlorous acid solution. To achieve complete disinfection, they should be soaked in a solution of glutaraldehyde, a divalent aldehyde molecule with a high microbicidal activity, or be autoclaved. Disposable items should be used when necessary.
10. Enough ventilation of the autopsy facility is needed after completion of the autopsy. A HEPA filter should be set at the ventilation hole.
11. After several weeks, the attendees should receive additional health check-ups such as a tuberculin skin test and chest X-ray examination.
12. It is important to note that every effort be made on a continuing basis to sharpen the ability to make a gross diagnosis of tuberculosis.

 back next

   The concept of the AIDS prevention slogan "Vaccination, the Name of the Intelligence" is also applicable to the prevention of the occupation-related infections. This should not merely be borne in mind because of the threat of tuberculosis or liver cirrhosis, but because there is a need to take effective procedures against every possible biohazard. The idea of "universal precautions" or "standard precautions" is very important here.18) It makes us realize keenly that there is a vital need to create a safe and favorable work environment and provide all the concerned with appropriate opportunities of reeducation and discussion on the biohazard and biosafety.5,15)

 back next


1) Garner JS. Guideline for isolation precautions in hospitals. The Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 17: 53-80, 1996.
2) Craven RB, Wenzel RP, Atuk NO. Minimizing tuberculosis risk to hospital personnel and students exposed to unsuspected disease. Ann Intern Med 82: 628-632, 1975.
3) Barrett-Connor E. The epidemiology of tuberculosis in physicians. JAMA 241: 33-38, 1979.
4) Kantor HS, Poblete R, Pusateri SL. Nosocomial transmission of tuberculosis from unsuspected disease. Am J Med 84: 833-838, 1988.
5) Tsutsumi Y. Waste disposal and biohazard in pathology services: The current status and improvements. Iryo Haikibutsu Kenkyu (J Med Waste Disposal) 7: 35-44, 1995 (in Japanese).
6) Murata T, Ihara S, Kawamura M, et al. The present status of tuberculosis in autopsy cases of Suzuka Central Hospital. J Suzuka Cent Hosp 3: 25-27, 1996 (in Japanese).
7) Sugita M, Tsutsumi Y, Suchi M, et al. High incidence of pulmonary tuberculosis in pathologists of Tokai University Hospital: An epidemiological study. Tokai J Exp Clin Med 14: 55-59, 1989.
8) Beck-Sagu C, Dooley SW, Hutton MD, et al. Hospital outbreak of multidrug-resistant Mycobacterium tuberculosis infections. Factors in transmission to staff and HIV-infected patients. JAMA 268: 1280-1286, 1992.
9) Sugita M, Tsutsumi Y, Suchi M, et al. Pulmonary tuberculosis. An occupational hazard for pathologists and pathology technicians in Japan. Acta Pathol Jpn 40: 116-127, 1990.
10) Reid DD. Incidence of tuberculosis among workers in medical laboratories. Br Med J 2: 10-14, 1957.
11) Lunn JA, Mayho V. Incidence of pulmonary tuberculosis by occupation of hospital employees in the National Health Service in England and Wales 1980-84. J Soc Occup Med 39: 30-33, 1989.
12) Editorial. BCG: Bad news from India. Lancet I (8159): 73-74, 1980.
13) Roche PE, Triccas JA, Winter N. BCG vaccination against tuberculosis: Past disappointments and future hopes. Trends Microbiol 3: 397-401, 1995.
14) Tsutsumi Y. Biohazard associated with intraoperative frozen diagnosis and its prevention. Byori-to-Rinsho (Pathol Clin Med) 9: 430-431, 1991 (in Japanese).
15) Tsutsumi Y. Tuberculosis. Occupation-acquired infection in pathology field. Byori-to-Rinsho (Pathol Clin Med) 13: 1705-1708, 1995 (in Japanese).
16) The Center for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in healthcare facilities, 1994. MMWR 43: RR-13, 1994.
17) The Committee for Diagnostic Pathology Practices of the Japan Society of Pathology (ed). Manual on the preventive procedures for infectious diseases and waste disposal in the field of pathology. Transact Jpn Soc Pathol 84 (Suppl), 1995 (in Japanese).
18) Gammon J. Infection Control. A British Perspective. Bridge & District NHS Trust, 1995.


Table 1: Annual Incidence of Pulmonary Tuberculosis among Pathology Workers in Japan

  Pathologists + pathology technicians (n = 2,388)
Pathologists (n = 1,201)
Pathologists, female (n = 88)
Pathology technicians (n = 1,187)
Pathology technicians (assisting at autopsy) (n = 753)
Pathology Technicians (not assisting at autopsy) (n = 422)
Pathologists + pathology technicians (1978~1988)
Pathologists + pathology technicians (1978~1988)
   (40 years of age or younger)
Pathology staff (other than pathologists and technicians) (n = 207)
Public health/preventive medicine staff (n = 732)
Public health/preventive medicine staff
   (medical doctors + technicians) (n = 414)
Japan Railway (JR) employees (n = 500,000) (1982)
Nippon Telephone and Telegram (NTT) employees (n = 42,000) (1982)
Japanese population (1982)
United Kingdom population (1982)
United Kingdom pathologists (1953~1955)
United Kingdom morbid anatomy-related personnel (1971)



Note 1:  


The data are based upon the questionnaire-based statistical survey on the
   treatment history against tuberculosis (data/100,000 population/year)
The year 1982 corresponds to the median value of the work period (6 years)
   of the Japanese pathology workers
Summarized from the article (Ref. 9) published in Acta Pathol Jpn 40: 116-
   127, 1990 by Sugita M, et al.



Textbooks and Atlases
   As well-written textbooks and atlases of infectious disease pathology, the followings are generally recommended.

1) Chandler FW, Kaplan W, Ajello L. A Colour Atlas and Textbook of the Histopathology of Mycotic Diseases. Wolfe Medical Publications, London, 1980.
2) Farrar WE, Wood MJ, Innes JA, Tubbs H. Infectious Diseases. Text and Color Atlas, 2nd Ed, Gower Medical Publishing, London, 1992.
3) Blaser MJ, Smith PD, Ravdin JI, Greenberg HB, Guerrant RL. Infections of the Gastrointestinal Tract. Raven Press, New York, 1995.
4) Orihel TC, Ash LR. Parasites in Human Tissues. ASCP Press, Chicago, 1995.
5) Connor DH, Chandler FW, Schwartz DA, Manz HJ, Lack EE. Pathology of Infectious Diseases. Appleton & Lange, Stamford, 1997.
6) Horsburgh CR Jr, Nelson AM. Pathology of Emerging Infections. ASM Press, Washington, DC, 1997.
7) Nelson AM, Horsburgh CR Jr. Pathology of Emerging Infections-2. ASM Press, Washington, DC, 1998.
8) Gardiner CH, Fayer R, Dubey JP. An Atlas of Protozoan Parasites in Animal Tissues, 2nd Ed, AFIP, American Registry of Pathology, Washington, DC, 1998.
9) Collier L, Balows A, Sussman M. Topley & Wilson's Microbiology and Microbial Infections, vols 1-6, 9th Ed, Arnold, London, 1998.
10) Ishikura H, Aihara M, Itakura H, Kikuchi K. Host Response to International Parasitic Zoonoses. Springer-Verlag, Tokyo, 1998.
11) Fernando RL, Fernando SSE, Leong AS-Y. Tropical Infectious Diseases. Epidemiology, Investigation, Diagnosis and Management. Blackwell Science Asia, Carlton South, 2000.
12) Tsutsumi Y. Atlas of Infectious Disease Pathology. Bunkodo, Tokyo, 2000 (in Japanese).


close window