False-positive squamous cell carcinoma in cervical smears: cytologic-histologic correlation in 19 cases

Cytologic features of squamous intraepithelial lesions (SIL) can mimic those of invasive squamous-cell carcinoma. We compare and correlate the cytological findings of 19 false-positive squamous-cell carcinomas with follow-up cone biopsies or hysterectomy specimens to define which type of dysplasia is more prone to diagnostic errors on cervical Papanicolaou (Pap) smears. Out of 128 patients diagnosed with invasive squamous-cell carcinoma from 1994-2000, 19 (14.8%) with follow-up cone biopsies or hysterectomy specimens were false-positive cases, showing only cervical intraepithelial neoplasia (CIN). We reviewed tissue sections from these 19 cases of CIN for cytologic features of squamous-cell carcinoma, such as markedly pleomorphic and/or dysplastic squamous cells, necrosis, and nucleoli. Twelve of 19 patients (63%) were menopausal. The mean age was 50.5 yr. On review of cervical smears, 18 cases qualified for the cytologic diagnosis of squamous-cell carcinoma, keratinizing type, and one case qualified for squamous-cell carcinoma, nonkeratinizing type. Pleomorphic and/or keratinizing dysplasia was found in 15 out of 19 patients (79%), necrosis within superficial endocervical glands in 9 out of 19 patients (47%), and conspicuous nucleoli in 12 out of 19 patients (63%). One or more of these changes were seen in all but 2 patients (89%). Endocervical gland involvement was present and extensive in 18 of the 19 cases (94%). The mean age was older than expected for SIL (50.5 vs. a reported 40), and matched the mean age found in patients with invasive squamous-cell carcinoma. Pleomorphic and/or keratinizing dysplasia involving endocervical glands may exhibit the cytologic features of squamous-cell carcinoma on cervical Pap smears.     

Atypical squamous epithelium in cytologic specimens from the pancreas: cytological differential diagnosis and clinical implications

Atypical squamous epithelium is an uncommon finding in cytologic specimens obtained from pancreatic lesions. A variety of pathologic conditions can result in the presence of these cells, including primary or metastatic carcinomas, chronic pancreatitis, and squamous metaplasia related to pancreatic or biliary duct stent placement. Primary adenosquamous and squamous-cell carcinomas of the pancreas are rare, representing 3.4% and 1.4 % of pancreatic carcinomas, respectively. Cytologic separation of these malignancies from less ominous metaplasias has immense clinical importance. We reviewed Indiana University Hospital's and Duke University's experiences with atypical squamous epithelium occurring within pancreatic aspirates. Study cases were identified using a computer to search the cytology records of these two institutions. Nine cases with a diagnosis of squamous-cell carcinoma, adenosquamous carcinoma, or atypical squamous epithelium were retrieved from the two institutions' Department of Pathology files. One case of pure squamous-cell carcinoma occurred in a patient with a known pulmonary primary; a single case of adenosquamous carcinoma was diagnosed in a patient with a coexistent endometrial primary; a single sample of adenocarcinoma with squamous differentiation was diagnosed in a patient without other known disease; and four primary squamous-cell carcinomas of the pancreas were detected. In addition, a single case of atypical squamous metaplasia associated with a stent was identified, and one case of atypical squamous epithelium associated with chronic pancreatitis was diagnosed. Despite the reactive atypia present in the examples of metaplastic squamous epithelium, separation of these cases from true squamous-cell carcinoma and adenosquamous carcinoma was achievable by cytologic evaluation. No cytologic criteria aided in separating primary pancreatic carcinomas with squamous differentiation from metastatic lesions. In this study, we report our findings in a series of nine cases where cytology disclosed atypical squamous epithelium in the aspirates derived from pancreatic lesions.     

Atypical glandular cells of undetermined significance (AGUS): clinical considerations and cytohistologic correlation

The diagnoses of atypical glandular cells of undetermined significance (AGUS) made upon evaluation of cervical/vaginal (Pap) smears is examined to ascertain salient clinical and cytologic features that may lead to better characterization of the true nature of these lesions. Prior history of squamous dysplasia, age of the patient, and the occurrence of abnormal microbiopsy tissue fragments are investigated to determine their value in the proper evaluation of AGUS specimens. Of the 86,234 Pap smears submitted to our laboratory during a period of 2 yr, 187 (0.2%) were diagnosed as AGUS. Available follow-up in 128 (69%) cases revealed 54 (42%) significant tissue proven abnormalities, the majority (55%, 30 patients) of which were diagnosed as squamous intraepithelial lesions (SIL). Squamous dysplasia is significantly more common in women younger than 40 (15/18, 83%) and in patients with prior history of SIL (29/30, 97%). In addition, all nine patients diagnosed with endometrial lesions on subsequent histology were older than 40. Age, however, was not a discriminating factor in women proven to have endocervical glandular lesions. Additionally, certain tissue fragment cytomorphologic features were significantly more often observed on follow-up in specific histologic diagnostic categories. The Pap smears of patients diagnosed with SIL were noted to contain tissue fragments composed of both dysplastic squamous and benign glandular cells in 29 of 30 (97%). The presence of two distinct populations of glandular tissue fragments (typical and atypical) was found in the Pap smears of all nine women with endometrial abnormalities and in the smears of most women subsequently diagnosed with endocervical glandular lesions (87%, 13/15). These observations suggest that a more specific and clinically useful Pap smear interpretation other than AGUS is often possible by consideration of the patient's age and prior history along with the correct identification of the type of atypical cells observed in abnormal tissue fragments.     

Can glandular lesions be diagnosed in pap smear cytology?

Many reports have been published on the accuracy of the cervical vaginal smear for squamous lesions, and the literature contains fewer reports on the accuracy of the cervical vaginal smear for glandular lesions. The sensitivity of glandular lesion diagnosis depends on the subtype of lesion. The diagnostic sensitivity is highest for invasive endocervical adenocarcinoma and lowest for endometrial adenocarcinoma. The ability of some of the Bethesda system categories for glandular lesions to describe what they purport to describe is questionable. The Bethesda system categories of adenocarcinoma accurately classify adenocarcinomas. The Bethesda System category of atypical glandular cells of undetermined significance (AGUS) is a misnomer. Although many cases of adenocarcinoma in-situ are placed in this category, follow-up of patients with AGUS show that the majority of patients with clinically significant lesions have squamous dysplasias. Other categories of AGUS, such as AGUS favor endometrial origin, are more appropriately named and encompass endometrial lesions which are either neoplastic or non-neoplastic.     

Microinvasive adenocarcinoma of the cervix: A cytohistopathologic study of 40 cases

Cervical smears obtained from 40 women with a histologic diagnosis of microinvasive adenocarcinoma (MIA) were reexamined for features of invasion. In our study MIA was defined as stromal invasion by adenocarcinoma cells to 5 mm or less beyond a surface epithelium and without lymphovascular involvement. In 24 cervices, squamous carcinoma in situ was a coincidental histologic finding. All 40 smears contained atypical glandular cells (AGC) forming pseudosyncytial clusters, and 12 showed additional features suggestive of invasion; pleomorphic nuclei, coarse irregular chromatin, karyorrhectic nuclei, and cell detritus. The invasive features tended to occur together and were found more than twice as frequently in this group than in the remaining 28 smears. There was little difference between the two groups in the frequency of super-crowded cell clusters, acini, cell strips, isolated cells, nuclear hyperchromasia, macronucleoli, or normal cervical glandular cells. Fourteen smears from the 24 cervices with squamous carcinoma in situ contained cells from a high-grade squamous intraepithelial lesion. It is concluded that cervical cytologic examination has a sensitivity of 30% (12/40) for the identification of stromal microinvasion in adenocarcinoma in situ. Cell detritus, present in 5/12 smears, is considered highly specific for invasion but lacks sensitivity. Due to the more proximal location and ease of sampling, cytologic examination has a sensitivity of 58% (14/24) for histologically confirmed coexisting squamous lesions. Diagn. Cytopathol. 16:430–436, 1997. © 1997 Wiley-Liss, Inc.     

Atypical glandular cells of undetermined significance. Outcome predictions based on human papillomavirus testing

Cases of atypical glandular cells (AGC) diagnosed on liquid-based preparations were culled from a 3-year period. When available, residual cellular material was analyzed for human papillomavirus (HPV) by polymerase chain reaction and correlated with cytologic and histologic (biopsy) outcome. Of 178,994 cytologic cases, 187 (0.1045%) contained AGC compared with 8,740 (4.8828%) atypical squamous cells (ASC) for an AGC/ASC ratio of 0.021. HPV results and follow-up were available for 108 specimens from 106 patients. Depending on the end-point (histologic/cyto-logic), the sensitivity range of HPV testing for significant cervical disease (high-grade squamous intraepithelial lesion [SIL], adenocarcinoma in situ [ACIS], invasive carcinoma) was 83% with a specificity range of 78% to 82%, a positive predictive value of 57% to 61%, and a negative predictive value of 91% to 95%. Fifteen false-positive results included concurrent ASC or low-grade SIL, ASC on follow-up cytology, and previous ACIS with a negative follow-up cone biopsy result. Noncervical glandular neoplasia (including atypical endometrial hyperplasia) was confirmed in 13 cases (1 recurrent), only 2 of which scored positive for HPV. HPV-positive AGC has a substantially higher positive predictive value for significant disease than ASC (61% vs historic 20%) and merits consideration in the triage of patients with atypical endocervical cells not otherwise specified. However, noncervical or other HPV-negative glandular neoplasia must be considered in all patients with AGC, particularly older patients.     

Atypical mitoses in dysplasias of the Barrett's mucosa

The frequency of mitotic figures, the proportion of atypical mitoses, and the spatial position (vertical or horizontal) of metaphasal plates were studied in resected specimens from 18 patients having Barrett's mucosa with dysplasia, and/or invasive adenocarcinoma. The number of dividing glandular cells increased (by comparison with the non-dysplastic Barrett's mucosa) in areas with dysplasia and with invasive adenocarcinoma, the largest percentage was found in the latter. Atypical mitoses were often found in areas with glandular dysplasia and with invasive adenocarcinoma. The highest frequency was found in adenocarcinoma. While 90.8% of the metaphases in Barrett's mucosa without dysplasia had a vertical position, only 39.5% of the mitoses in areas with dysplasia had the same position. It is conceivable, that the number of mitoses, the number of atypical mitoses as well as the spatial position of mitosis should be registered in the Barrett's mucosa with dysplasia in attempts to learn more about the biological behaviour of these lesions.     

Aspiration cytology of cystic carcinoma of the breast

Cystic carcinomas of the breast are rarely encountered in fine-needle aspiration (FNA) biopsies. The most common entities comprise intracystic papillary adenocarcinoma, ductal adenocarcinoma with cystic degeneration including comedo forms of ductal adenocarcinoma in situ, medullary carcinoma, squamous carcinoma, and cystic hypersecretory ductal adenocarcinoma. The cytologic diagnosis is often hampered by sparse cellularity, abundant obscuring blood, necrotic debris, and degenerative changes in diagnostic cells. We report on the cytologic features of 10 cases of cystic carcinoma, including 12 FNA biopsies with radiologic and surgical correlation. The original cytologic diagnoses for these cases were: benign (2 cases), atypical (2 cases), suspicious (3 cases), and positive for malignant cells (3 cases). On repeat FNA, one benign case and one atypical case were reclassified, respectively, as atypical and suspicious for carcinoma. The follow-up diagnoses were 5 intracystic papillary adenocarcinomas and 5 cystic ductal adenocarcinomas. Despite 2 false-negative cases, all cases were adequately managed. Correlation with clinical and radiologic findings and direct sampling of any solid component of these cystic neoplasms are crucial in diagnosis and management.     

Atypical endocervical glandular cells: Accuracy of cytologic diagnosis

Atypical cells thought to be of endocervical glandular origin often cause diagnostic uncertainty in cervicovaginal smears. For this reason consecutive cases of endocervical glandular atypia diagnosed in smears were correlated with subsequent biopsy diagnoses and then retrospectively reviewed. Smears were originally diagnosed as “mild glandular atypia, probably reactive” or “severe glandular atypia, suggestive of adenocarcinoma in situ” (AIS). Biopsy follow-up was obtained on 34 of 58 patients diagnosed with severe endocervical glandular atypia. Nine patients (26%) had AIS, three with concomitant high-grade squamous intraepithelial lesions (HSIL) and two with invasive adenocarcinoma. Eighteen patients (53%) had HSIL only. Seven had benign changes. Of 152 patients diagnosed with mild glandular atypia, biopsy follow-up was obtained on 40. One patient had AIS; 14 (35%) had HSIL; one had low-grade SIL (LSIL); and 24 (60%) had benign changes. Blinded review of these smears yielded results similar to those in the biopsy follow-up, that is, the prediction of AIS on smears included most cases of AIS, some invasive adenocarcinomas, a significant number of HSIL cases and a few benign lesions. A review diagnosis of “reactive glandular cells” proved to be HSIL in 31% of cases and AIS in one case. We conclude that patients with a diagnosis of severe glandular atypia in smears may prove to have AIS or invasive adenocarcinoma, but often have HSIL without concomitant AIS. In addition, although “reactive” glandular atypia in smears usually reflects a benign condition, a significant minority of such patients prove to have HSIL. © 1995 Wiley-Liss, Inc.     

Cytohistological correlation in patients with atypical glandular cells on Papanicolaou test in a Chilean population

The standard screening test for detecting cervical lesions and cancers is a Papanicolaou (Pap) smear. While squamous cell abnormalities remain the most common positive Pap test result, cytologic findings of glandular cell abnormalities have become more frequent in recent decades. The 2014 Bethesda System for reporting cervical cytology includes the classification “atypical glandular cells” (AGC). AGC have morphological abnormalities that fall outside the range of reactive changes, but are insufficient for a diagnosis of invasive adenocarcinoma. In several histologic follow‐up studies, most AGC cases were found to represent a benign condition. In the current study, we evaluate the significance of AGC cytology findings by analyzing the histologic follow‐up results of a large number of patients with AGC. Most patients with AGC in this study were found to have a significant lesion on follow‐up (63.9%), with negative histologic results in only 36.1% of patients. Among patients with significant lesions, the most common result was low‐grade squamous intraepithelial lesion (26.6%), followed by high‐grade squamous intraepithelial lesion (23.2%). This provides further evidence to support the Chilean Clinical Guidelines for Cervical Cancer, which recommends diagnostic follow‐up studies in all women with AGC to minimize the chance of undetected serious cervical disease.     

Women with atypical glandular cells: a long-term follow-up study in a high-risk population

To determine the incidence of clinically significant lesions in high-risk patients with atypical glandular cells (AGCs) after 4 to 6 years of follow-up, we reviewed repeated Papanicolaou (Pap) test and surgical pathology results for a 3-year period for 337 patients; 62 (18.4%) had only repeated Pap smears; 84 had Pap smear and histologic evaluations. In a range of repeated Pap smears from 1 to 11 (mean, 4.2), 9 patients had persistent AGCs/atypical squamous cells; remaining Pap smears were judged normal. Histologic follow-up revealed a clinically significant lesion in 110 (40.1%) of 274 patients--low-grade squamous intraepithelial lesion (LSIL), 46; high-grade squamous intraepithelial lesion (HSIL), 47; endocervical adenocarcinoma in situ (AIS), 3; endometrial hyperplasia, 4; endocervical or endometrial adenocarcinoma, 10. Among patients with histologic follow-up, 14 lesions (12.7% of patients with clinically significant lesions) were diagnosed after a mean of 37 months (range, 21-59 months): LSIL, 7; HSIL, 4; AIS, 1; endometrial adenocarcinoma, 2. Seven patients had negative cytologic and/or histologic evaluations between the initial cytologic AGC diagnosis and the final histologic diagnosis. Patients with AGCs are at risk of harboring clinically significant uterine lesions and should be followed up for a substantial period despite initial negative findings.     

Malignant glandular lesions and glandular differentiation in invasive/noninvasive urothelial carcinoma of the urinary bladder

Although the lumen of the urinary bladder is covered with only urothelial epithelium, malign glandular lesions (eg, nonurachal adenocarcinoma) and benign lesions (eg, cystitis cystica and cystitis glandularis) can also rarely occur in this site due to its characteristic embryologic development. Glandular differentiation is uncommon in urothelial carcinomas and is even less common in noninvasive urothelial cancers. In addition, in situ urothelial carcinomas are more likely to progress in the presence of glandular differentiation toward high-grade urothelial carcinomas and/or aggressive urothelial carcinomas. Pure nonurachal adenocarcinomas and mixed carcinomas (urothelial carcinoma and adenocarcinoma) are very rare, and their pathogenesis is not clear. Most of the nonurachal adenocarcinomas are thought to arise on the grounds of cystitis glandularus with intestinal metaplasia. Here, I present 2 cases with noninvasive urothelial carcinoma with substantial glandular differentiation showing progression to signet ring cell carcinoma and invasive urothelial carcinoma, one case with mixed carcinoma (urothelial carcinoma and adenocarcinoma) and another case with pure adenocarcinoma developing from cystitis glandularis with intestinal metaplasia, and discuss malign glandular lesions in the bladder and invasive/noninvasive urothelial carcinomas with glandular differentiation.     

Preneoplastic lesions of pulmonary carcinoma

The World Health Organization (WHO) 2004 classification includes 3 categories of pulmonary preneoplastic lesions, including squamous dysplasia and carcinoma in situ (CIS) for squamous cell carcinoma, atypical adenomatous hyperplasia (AAH) for the majority of adenocarcinomas and diffuse idiopathic pulmonary neuroendocrine cell hyperplasia (DIPNECH) for carcinoids. The distinction of the 3 grades of squamous dysplasia and CIS is mainly based on the degree by which the basal cell zone is expanded, the degree of cellular atypia and the level of mitoses. The category AAH consists of a proliferation of atypical epithelial cells with Clara cells or type 2 pneumocyte features. They grow along the alveolar septae in a lepidic fashion, sometimes reaching into the terminal bronchioles. In contrast to the newly described adenocarcinoma in situ (AIS), AAH is smaller (≤ 5 mm), has a lower cell density and a lower degree of cellular atypia. The putative cancer stem cells of peripheral adenocarcinomas reside in the bronchioloalveolar duct junction, while those of central squamous cell carcinomas are located in the basal cell compartment of the bronchi. This review provides an overview of the current knowledge on preneoplastic lesions of the lungs and their clinical impact.     

In-situ and microinvasive adenocarcinoma of the uterine cervix. A clinical, cytologic and histologic study of 14 cases.

In-situ and microinvasive adenocarcinoma of the uterine cervix. A clinical, cytologic and histologic study of 14 cases. Am J Clin Pathol 64: 155-70, 1975. The clinicopathologic features of 14 cases of in-situ and microinvasive adenocarcinoma of the cervix, encountered during the period 1961-1974, form the basis of this report. Seven cases represent examples of in-situ and seven, microinvasive adenocarcinoma. Severe dysplasia and carcinoma in-situ of the squamous epithelium were coincidental finds in 12 patiients. Papanicolaou smears were abnormal in all 14 patients. In nine, malignant glandular cells were identified. Five smears contained both malignant squamous and glandular cells. The mean age of the 14 patients was 38.4 years. Thirteen patients had borne one to six children. Two were pregnant at the time of diagnosis, and four were on birth control medication. Thirteen of the patients underwent conization biopsy. Hysterectomy in one form or another was curative in all. The relatively infrequent occurrence of in-situ adenocarcinoma is probably due to the low incidence of cervical adenocarcinoma and to the fact that the lesion is focal and easily overlooked. Five of the cases in the present series were a chance finding in a review of 200 consecutive cone biopsies of the cervix. The resemblance of the in-situ lesions to carcinoma and the presence of morphologically similar glandular tissue adjacent to foci of microinvasion suggests that the in-situ lesion is the precursor of invasive adenocarcinoma of the cerivs.     

Human papillomavirus genotyping by oligonucleotide microarray and p16 expression in uterine cervical intraepithelial neoplasm and in invasive carcinoma in Korean women

For evaluating the diagnostic significance of p16(INK4A) over-expression in the uterine cervical intraepithelial neoplasm and in invasive carcinoma, human papillomavirus (HPV) was detected and genotyped by oligonucleotide microarray in archival tissues of 117 cervical specimens, including 47 invasive squamous cell carcinomas (SCC), 30 cases of cervical intraepithelial neoplasia (CIN), 20 adenocarcinomas, and 20 cases of non-neoplastic cervix. The expression of p16(INK4A) protein was immunohistochemically studied in these cases and in five HPV-positive and one HPV-negative cervical cancer cell lines. HPV was detected in 50% of CIN, 61.7% of SCC, and 45.5% of adenocarcinomas. p16(INK4A) expression was seen in all 20 cases of adenocarcinoma, 78.7% (37/47) of SCC, and 96.7% (29/30) of CIN, but not in any cases of the non-neoplastic cervix. There was no difference in p16(INK4A) expression between the HPV-positive and HPV-negative cervical lesions. All HPV-positive and -negative cervical cancer cell lines expressed p16(INK4A) protein. In conclusion, the presence of p16(INK4A) expression in cervical squamous and glandular epithelium indicates the existence of dysplasia or malignancy in the uterine cervix, regardless of HPV infection.     

Small colonic microsatellite unstable adenocarcinomas and high-grade epithelial dysplasias in sessile serrated adenoma polypectomy specimens: a study of eight cases

Eight sessile serrated adenoma (SSA), right colon polypectomies with focal invasive adenocarcinoma or high-grade dysplasia were studied to identify features indicating a high risk of transformation and characterize the morphologic features of serrated dysplasia; 6 cases had invasive adenocarcinoma; 2 were high-grade dysplasia. All 8 were microsatellite unstable-high and had absent hMLH1 nuclear immunoreactivity. The mean patient age at polypectomy was 69.5 years (range, 57.1-83.9 years). Mean polyp maximum dimension was 8.5 mm (range, 6-12 mm). The majority of each polyp was nonmalignant SSA. All 8 cases had an abrupt transition from benign to high-grade in situ or invasive malignancy. In the 6 invasive adenocarcinomas, the neoplasm extended directly down into the submucosa without lateral intramucosal spread. The mean maximum dimension of the invasive adenocarcinoma was 2.9 mm (range, 2-4 mm). All 8 cases had high-grade serrated-type dysplasia. The nonmalignant SSAs had marked expansion of the proliferative zone. Crypts adjacent to malignancy had moderately enlarged nuclei, irregular nuclear membranes, and overly prominent nucleoli. SSA crypts were lined by a variety of gastric-type cells; no cell type predominated. Foci of adjacent crypts had similar cytologic features. Small proximal SSAs can transform into adenocarcinoma without a component of adenomatous dysplasia.     

Diagnostic value and cost-effectiveness of on-site evaluation of fine-needle aspiration specimens: review of 5,688 cases

Despite the increasing utilization of the ThinPrep Pap Test (TP), limited data exist regarding the cytomorphologic features and patterns of invasive squamous-cell carcinoma in TP specimens. We analyzed a series of TP specimens from patients with histologically confirmed invasive squamous carcinomas of the cervix. Patients with biopsy-proven invasive squamous-cell carcinoma of the cervix with a TP cervical cytologic specimen within the previous 2 mo were identified. The TP slides were analyzed for overall cellularity (percent circle coverage by epithelial cells), tumor cellularity, tumor diathesis, inflammation, coexistent dysplasia, and keratinization. Tumor cellularity was defined as <5%, 5-50%, and >50% of slide cellularity. In all 13 cases that were identified, a cytologic diagnosis of either invasive squamous-cell carcinoma or suspicious for invasive squamous-cell carcinoma was made. In 7/13 cases (54%), epithelial cells covered <40% of the slide circle. Epithelial cells covered >40% of the slide circle in 6/13 cases (46%). Tumor cellularity covered <5% of the slide circle in 4/13 cases (31%), 5-50% in 7/13 cases (54%), and >50% in 2/13 cases (15%). A tumor diathesis was present in 12/13 cases (92%). Inflammation was absent in 1/13 cases (8%), mild in 8/13 cases (62%), moderate in 2/13 cases (15%), and severe in 1/13 cases (8%). Coexistent dysplasia was identified in 12/13 cases (92%). Keratinization was identified in 9/13 specimens (69%). In the vast majority of patients, a diagnosis of squamous-cell carcinoma was rendered on the TP cervical specimen, despite a pattern of decreased cell coverage. It could be hypothesized that tumor diathesis and inflammation may be the etiology for decreased cellularity by blocking filter coverage by epithelial cells. This cellular pattern with diathesis in the ThinPrep smear may be a useful clue to look carefully for diagnostic cells of squamous-cell carcinoma.     

Synchronous squamous and glandular neoplasia of the anal canal.

A 48 year old man presented with invasive adenocarcinoma in the wall of a non-healing anal fistula. The subsequent abdomino-perineal resection specimen showed residual invasive carcinoma coexisting with in situ carcinoma of anal glands as well as in situ squamous carcinoma of the anal canal. The epithelium of the anal canal had koilocytotic features. DNA hybridisation studies by the dot blot technique showed weak positivity for human papillomavirus (HPV) subtypes 16, 18. This case illustrates a number of important points--namely, anal fistulas, particularly non-healing fistulas should be biopsied to exclude malignancy; some adenocarcinomas of the anal arise in anal glands; the coexistence of glandular and squamous carcinoma with evidence of HPV infection is highly reminiscent of similar synchronous lesions of the uterine cervix and suggests that HPV may have an aetiological role in both squamous and glandular carcinomas of the anal canal.