Barbara Winkler

Human papillomavirus infection of the esophagus. A clinicopathologic study with demonstration of papillomavirus antigen by the immunoperoxidase technique

Papillomaviruses are known to be oncogenic in animals. In humans they are associated with benign squamous tumors (verruca, condylomata acuminata, and papillomas) in a variety of body sites. Human papillomavirus (HPV) infection of the esophagus, however, has not previously been documented. Recent reports of condylomatous changes in esophageal epithelium adjacent to esophageal carcinoma and the sporadic descriptions of esophageal papillomas in the literature for many years, lend credence to the assumption that HPV may affect the squamous mucous membrane of the esophagus. In the current study 75 cases, including 2 papillomas and 73 focal epithelial hyperplasia of the esophagus, were examined for histologic evidence of HPV infection as characterized by the presence of koilocytosis, giant and multinucleated cells, dyskeratosis, hyperkeratosis, acanthosis, papillomatosis, and anisonucleosis. Thirteen of the cases—the 2 papillomas and 11 of the focal epithelial hyperplasias—contained distinctive histologic evidence of HPV infection. The presence of HPV antigens was demonstrated by immunoperoxidase (IMPO) in the 4 of the 13 cases (31%). In the remaining cases the IMPO was equivocal in two and negative in seven.

Endometrial polyps in postmenopausal patients receiving tamoxifen.

The histologic features of an endometrial polyp include irregular, often dilated glands, thick-walled blood vessels, and a fibrotic stroma. Such polyps may be responsive to some chemotherapeutic drugs that can exert hormonal effects. We report on endometrial polyps detected in three postmenopausal patients who were receiving tamoxifen for treatment of metastatic breast carcinoma. The clinical presentation in all cases was vaginal bleeding and all had documented uterine enlargement suggestive of an intrauterine malignancy. The polyps were large, measuring up to 9 cm in largest diameter. On histologic examination each polyp had extensive cystic glandular hyperplasia. In one case foci of atypical epithelial proliferation and predecidualization were noted. The atypical proliferation suggested a borderline neoplastic process and was strongly positive for carcinoembryonic antigen. These findings underscore the marked proliferative changes that can be induced in endometrial polyps in postmenopausal women receiving hormonally active chemotherapeutic agents.

Koilocytotic lesions of the cervix. The relationship of mitotic abnormalities to the presence of papillomavirus antigens and nuclear DNA content.

It has been reported that abnormal mitotic figures (AMFs) occur principally in aneuploid lesions and that aneuploidy is a diagnostic feature of non‐endocrine‐dependent epithelial cancer precursors and cancers. Recently, AMFs have been reported in cervical lesions interpreted as flat condylomata, and it has been suggested by several authors that AMFs may not be diagnostic or aneuploidy or neoplasia, particularly in human papillomavirus‐(HPV)‐induced lesions. Although it is conceivable that AMFs may be a regular feature of HPV infection, their association with cytologic atypia and their presence in higher grades of cervical intraepithelial neoplasia (CIN) suggests that AMFs may herald the presence of a different lesion than the pure flat condyloma. In the current study, koilocytotic cervical lesions thought to be HPV‐induced were examined microscopically for the presence of AMFs, and the findings were correlated with the presence of HPV as determined by immunoperoxidase and nuclear DNA distribution patterns as measured by Feulgen microspectrophotometry. In unselected lesions originally diagnosed as flat cervical condylomata, AMFs were surprisingly common (22.6%), and did not correlate with the extent of koilocytosis. Immunoperoxidase (IMPO) stains were performed in 35 cases with AMFs, and were negative for HPV in 74.3% and positive in 22.8%. However, among the cases evaluated by IMPO, there was an inverse relationship between the presence of mitotic abnormalities and the expression of HPV antigen. Nine of 11 (81.8%) lesions containing AMFs were aneuploid, and 2 of 11 (18.2%) were polyploid. Abnormal mitotic figures have a range of morphology and frequency in koilocytotic cervical lesions. Although the biology of these lesions is not well‐defined, the presence of AMFs may identify a subgroup of HPV‐induced cervical atypias which represent a transition between flat cervical conylomata and CIN. Cancer 53:1081‐1087, 1984.

Cervical Intraepithelial Neoplasia and Condyloma

The nomenclature and classification of intraepithelial squamous lesions of the cervix have a controversial history,17 now further complicated by the introduction of new terminology for the classification of cervical human papillomavirus (HPV) infection.39 The classic terminology separates noninvasive cervical lesions into two groups: dysplasia and carcinoma in situ (CIS), implying a biologic distinction between these two entities that can be reproducibly distinguished histologically.21,151 In the cytologic nomenclature using this system, dysplasia is considered a benign to possibly malignant squamous epithelial atypia, whereas CIS is designated as positive for malignant cells. In 1961, at the First International Congress on Exfoliative Cytology, the Committee on Histological Terminology for Lesions of the Uterus Cervix defined CIS as follows231: “Only those cases should be classified as CIS which, in the absence of invasion, show a surface lining epithelium in which, throughout its whole thickness, no differentiation takes place. The process may involve the lining of the cervical glands.” It is recognized that the cells of uppermost layers may show some slight flattening. The very rare case of an otherwise characteristic CIS that shows a greater degree of differentiation belongs to the exceptions for which no classification can provide. Dysplasia of the cervix was defined as 231: “... all other (than CIS) disturbances of differentiation of the squamous epithelial lining of surface and glands.... They may be characterized as of high or low degree, terms which are preferable to suspicious and non-suspicious, as the proposed terms describe the histological appearance and do not express an opinion.” The narrow and arbitrary definition of CIS and the broad, all encompassing definition of dysplasia led to diagnostic uncertainties and to a classification based solely on histology, divorced from clinical and therapeutic considerations. 21,102,173

Mucoepidermoid carcinoma arising in an intraparotid lymph node

A well‐differentiated mucoepidermoid carcinoma that was confined to, and apparently arose within, an intraparotid lymph node is reported. Salivary gland ducts and acini often are found within intraparotid lymph nodes, and occasionally within extraparotid nodes. Salivary gland tumors, both benign and malignant, can develop within this ectopic salivary tissue. When a malignant salivary‐gland‐type neoplasm is found within an intraparotid or periparotid lymph node, the possibility exists that the tumor has arisen within the node and does not necessarily represent a metastasis from some other occult site.

Carcinoma and Metastatic Tumors of the Cervix

The concept of microinvasive carcinoma of the cervix was first introduced in 1947 by Mestwerdt.125 Microinvasive carcinoma (MICA) is considered a preclinical stage in the progressive spectrum of cervical intraepithelial neoplasia (CIN) and frank clinical invasive carcinoma of the cervix uteri. It is classified as clinical stage Ia according to the 1985 FIGO (International Federation of Gynecologists and Obstetricians) staging of carcinoma of the cervix (Table 8.1). Because the lesion cannot be visualized on gross inspection, the diagnosis is based on histologic examination of cervical tissue.

Benign Diseases of the Cervix

Cervicitis can be divided into noninfectious and infectious etiologies. Whatever the etiology, the tissue response of the cervix to injury is limited and reflects the basic mechanisms of inflammation and repair. Two types of morphologic changes, however, that are often encountered in association with a variety of inflammatory diseases deserve specific attention. These are atypia of repair and hyperkeratosis and parakeratosis.

Chlamydial endometritis: a histological and immunohistochemical analysis

A series of 90 endometrial biopsies and curettings originally diagnosed as chronic endometritis were reviewed and histological findings of plasma cells, lymphoid infiltrate, stromal necrosis, acute inflammation, lymphoid follicles, and epithelial atypia were correlated with the demonstration of chlamydial antigens by the immunoperoxidase technique. Chlamydial antigens were localized within endometrial epithelial cells in four cases. Although these four cases represented only 4% of the total number, chlamydial immunoperoxidase positivity was best discriminated by the severity of the inflammation and the presence of an acute inflammatory infiltrate. Among cases of severe endometritis 22% were chlamydia-positive, and in those cases with an associated acute inflammatory infiltrate, 57% were positive.A high index of suspicion of chlamydial infection should exist when severe endometritis is diagnosed in patients with clinical histories of post-abortal state, pelvic inflammatory disease, secondary infertility or menometrorrhagia, and chronic pelvic pain.

Localizing chlamydial infection in cervical biopsies with the immunoperoxidase technique.

SummaryOne hundred and two cervical biopsy specimens containing varying degress of chronic inflammation were stained for chlamydial antigens with the immunoperoxidase technique. Seven cases (6.9%) were positive. Histologically, six (84%) of the Chlamydia-positive cases contained severe chronic inflammation, all contained reparative atypia, and two (28%), follicular cervicitis. When evaluated separately, 22% (six of 27) of the specimens with severe inflammation were positive in contrast to 0% (0 of 45) of cases with mild inflammation. Positively staining cells were located primarily in columnar epithelium and reparative atypia and occasionally in areas of immature squamous metaplasia. The cytological finding which correlated with positive staining was cytoplasmic vacuolation; however, cytoplasmic vacuoles were common in cells which did not stain positively, and it was impossible to predict on histological grounds which cells/specimens would stain positively by immunoperoxidase. Because of these findings, the presence of chlamydial infection should be strongly suspected whenever the cervical biopsy specimen contains severe inflammation and repair. Although tissue staining may not be as sensitive as culture for diagnostic purposes, it can be performed rapidly and simply and may be a useful special stain in cases where the diagnosis of chlamydial infection is not suspected clinically or cultures are not immediately available.

Chlamydial cervicitis and cervical intraepithelial neoplasia: An immunohistochemical analysis

Studies using serological and culture techniques indicate that chlamydial infection is frequently associated with cervical intraepithelial neoplasia (CIN). This relationship was investigated by examining a series of biopsies containing normal epithelium, non-neoplastic condylomatous epithelium, and neoplastic epithelium (CIN). In each case the degree of inflammation and the presence of reparative atypia were recorded from an examination of the hematoxylin and eosin stained sections and serial sections were stained for chlamydial antigens using a polyclonal antichlamydial antibody and the immunoperoxidase technique. Overall, staining for chlamydia was positive in 0, 2, and 16% of biopsies with mild, moderate, and severe inflammation, respectively. In cases of severe inflammation positive staining was present in 20, 25, and 8% of biopsies containing non-neoplastic, condylomatous, and neoplastic epithelium, respectively. In all positive cases the staining was most prevalent in areas of the most intense inflammation. A spectrum of squamous epithelial changes was found in the infected biopsies and their distinction from CIN is discussed. This study suggests that tissue staining for chlamydia is related more to the extent of the coexisting inflammation than the presence or absence of CIN. Chlamydial infection, however, was frequently associated with inflammation-related squamous atypia in the transformation zone, which may be confused histologically with CIN.