The validation of cervical cytology. Sensitivity, specificity and predictive values

Statistics and methods for the validation of the results of cytologic screening for cervical cancer and its precursors were examined. Many of the methods commonly used, including the calculation of sensitivity and specificity on raw data, contain flaws that undermine their conclusions. Using a large computerized database of 748,871 cytologic screenings of 277,842 women over a ten-year period, the value of screening was examined. Only subsequent histologic examinations within one year were accepted to validate positive initial cytologic diagnoses; only two subsequent cytologic screenings within the next three years were accepted to validate negative initial cytologic diagnoses that had not been followed by a histologic examination. Cases not meeting these criteria were excluded from the initial analysis. From these data, the predictive value of a negative cytologic examination was determined to be 99.8%; the predictive value of a positive cytologic examination was 73.4% for an initial diagnosis of mild-to-moderate dysplasia, 90.6% for a diagnosis of severe dysplasia/carcinoma in situ, 94.5% for a diagnosis of carcinoma in situ or microinvasive carcinoma and 95.5% for an initial diagnosis of invasive carcinoma. Cases with an initial "questionable" cytologic diagnosis had a positive predictive value of only 64.0%. Extrapolation from the validated cases to the entire screened population showed an overall sensitivity of 80% and a specificity of 99.4% for cytologic screening for cervical cancer. The sensitivity was slightly lower for mild and moderate dysplasia (78.1%) and slightly higher for carcinoma in situ and severe dysplasia (81.4%) and invasive carcinoma (82.3%).     

Morphologic scales and relative distances between lesions identified on Papanicolaou smears.

We fit a Gaussian-type curve to a nonmonotonic transform of initial arbitrarily given scales (pseudoscalar transform) of a two-way discrete classification table by maximizing the likelihood (entropy) and computing morphologic scales of clusters of objects identified by visual judgments. The scales give the relative distance between pairs of categories or states. Morphologic scales of histologic lesions (states) identified on Papanicolaou smears were computed from a confusion matrix consisting of 3,545 matched pairs of observations on Papanicolaou smears and colposcopically directed biopsies available from the Gynecologic Cytology Laboratory, University of Minnesota, between 1985 and 1987. The original (uncollapsed) confusion matrix consisted of eight cytologic categories and histologic states: normal plus atypical benign, reactive atypia (including condylomatous changes); mild, moderate and severe dysplasia; squamous carcinoma in situ; invasive squamous cell carcinoma; and other malignancies. The morphologic scales are expressed numerically, which reflects a degree of confusion between two diagnostic categories or states. Except for malignancies other than squamous cell carcinoma, morphologic scales of histologic states seen from cytology retained the order of clinical severity. We detected a high degree of confusion between severe dysplasia and carcinoma in situ. Malignancies other than squamous cell carcinoma fell between moderate and severe dysplasia. Such a transposition of the scales in this group containing adenocarcinoma was likely due to frequent association of adenocarcinoma with cervical intraepithelial neoplasia. Morphologic scales of cytologic categories seen from histology showed high degrees of confusion and transpositions between mild and moderate dysplasia and between severe dysplasia and carcinoma in situ.(ABSTRACT TRUNCATED AT 250 WORDS)     

Early detection of precancerous lesions in dysplasias of the lung by rapid DNA image cytometry

Hematoxylin-and-eosin-stained cytologic smears of sputum from 28 patients with dysplastic and suspicious cell findings were subjected to DNA image cytometry after Feulgen restaining. The nuclear DNA contents were measured with a TV-based image-analysis system, the Leitz TAS plus, combined with an automatic microscope. Computation of DNA data was performed according to an algorithm for the diagnosis and grading of malignancy. Of the 19 cases that were proven to be malignant in the follow-up, either by histologic examination, sputum cytology, fine needle aspiration biopsy or autopsy, the algorithm identified 17 as malignant in a stage (dysplasia) in which cytology was not yet able to present a definite diagnosis of malignancy. Only two cases of bronchial carcinoma were not detected in the state of dysplasia by this procedure. The periods between the DNA diagnosis of malignancy in dysplasia and the morphologic evidence of cancer varied from three days up to six months. Of the 11 cases that had been classified as benign by the algorithm, 9 were confirmed as benign during the clinical follow-up. Rapid DNA image cytometry appears able to separate squamous dysplasias of the lung into precancerous and nonprecancerous lesions.     

Inflammatory epithelial changes and nonviral cervicovaginal pathogens

To assess the relationship between inflammatory epithelial changes (IECs) and cervicovaginal infection, 249 women with cytologic diagnoses of IEC were studied. The findings showed a Chlamydia infection in 15.7% of the cases, Monilia in 38.1%, Trichomonas in 1.6% and a lateral vaginal wall pH of greater than 5.5 in 29.7%. These results were compared with those for 211 controls drawn from the same population; women with abnormal cytologic reports were excluded from the control group. The controls had a prevalence of yeast and trichomoniasis no different from that of the IEC group. In contrast, the prevalence of chlamydial infection (7.6%) and pH greater than 5.5 (10.7%) were significantly lower in the controls than in the IEC cases. Chlamydia was found in 30.2% of 43 women with an atypical transformation zone, equally distributed between women whose biopsies were normal and those whose biopsies showed cervical intraepithelial neoplasia. The traditional guidelines of treating cases with cytologic diagnoses of IEC prior to obtaining a repeat smear may need reevaluation in the light of these findings.     

Preoperative brush and impression cytology in ocular surface squamous neoplasms

OBJECTIVE: Preoperative cytologic diagnoses of ocular surface squamous neoplasms were evaluated and compared with histologic diagnoses. STUDY DESIGN: Impression cytology (Millipore filter paper) and brush cytology were applied to 32 patients who had conjunctival neoplasms. Papanicolaou-stained cytologic preparations and hematoxylin and eosin-stained histologic sections were examined by light microscopy. RESULTS: The brush technique was used on 27 patients; impression cytology was applied in 5 cases. Cytologic and histologic diagnoses were concordant in 26 cases. Squamous cell carcinoma or carcinoma in situ was diagnosed in 18 and dysplasia in 4 cases. Squamous metaplasia and normal-appearing conjunctival epithelial cells were diagnosed cytologically in four cases; of those histologic diagnoses, one was pterygium and three, conjunctival nevus. Four cases revealed discrepancies between the cytologic and histologic preparations. There was one false positive result, and one case was subconjunctival invasion of basal cell carcinoma of the eyelid. CONCLUSION: Impression and brush cytology are fast, cost-effective, reliable and noninvasive diagnostic tools for ocular surface squamous neoplasms. However, the brush technique has several advantages over impression cytology.     

Inflammatory atypia. An apparent link with subsequent cervical intraepithelial neoplasia explained by cytologic underreading

A review of 5,752 cervical smears done on college students at a medium-sized midwestern university in a 24-month period showed 496 cytologic diagnoses of inflammatory atypia and 132 of dysplasia (cervical intraepithelial neoplasia: CIN). Further retrospective review of a nonselected cohort (178 cases) of the 496 patients with inflammatory atypia showed that their subsequent cytologic smears were more likely to show CIN than could be explained by chance alone. Only ten cases accounted for this difference, and a case-controlled blind review of the original cytologic smears of these ten patients resulted in the reclassification to CIN (mild dysplasia) of seven who had subsequently "progressed" from inflammatory atypia to dysplasia. Only one control smear was reclassified. In this population, underreading of a small number of cervical smears explained a strong statistical apparent correlation between inflammatory atypia and the subsequent development of CIN.     

Aspiration cytology of the conjunctival surface

OBJECTIVE: To evaluate a method of obtaining conjunctival cells that utilizes aspiration of the conjunctival surface. STUDY DESIGN: Eighteen eyes from 16 patients with epibulbar lesions underwent aspiration of the conjunctival surface using a tuberculin syringe without a needle. The aspirated cells were prepared by the cytocentrifuge or Millipore filter technique. The lesions were subsequently biopsied, and the cytologic and histologic diagnoses were compared. RESULTS: Diagnoses of the conjunctival lesions, in decreasing order of frequency, were conjunctival intraepithelial neoplasia (dysplasia, carcinoma in situ), squamous cell carcinoma, squamous metaplasia, nevus, lymphoma and inflammation. In this feasibility study, the cytologic diagnosis accurately correlated with the histologic diagnosis in most cases. There were two cases in which the cytologic diagnosis led to the correct histologic diagnosis. CONCLUSION: Aspiration cytology of the conjunctival surface is useful in obtaining diagnostic material and may be used as a guide to where a biopsy should be performed. Limitations include inadequate samples and lack of observer experience.     

Cytologic features of premalignant glandular lesions in the upper gastrointestinal tract

The cytologic findings in 13 endoscopic brushing specimens from biopsy-proven premalignant glandular lesions (PGLs) of the upper gastrointestinal tract were reviewed retrospectively. The specimens were from ten patients: three with dysplasia in Barrett's esophagus, four with gastric adenomas and three with duodenal adenomas. One dysplasia in Barrett's esophagus and four adenomas (two gastric and two duodenal) had coexisting adenocarcinomas. Most pure PGLs were characterized cytologically by cohesive three-dimensional clusters of cells with more-or-less uniformly enlarged nuclei and an increased nuclear/cytoplasmic ratio. Crowding and molding were present within these clusters; however, the cells were arranged in a somewhat orderly or palisading fashion, instead of entirely haphazardly. In cases of carcinoma coexisting with adenoma or dysplasia, the atypical cells tended to be more pleomorphic and dyshesive. In one specimen from an adenocarcinoma arising in an adenoma, the adenomatous and carcinomatous components could be distinguished cytologically.     

Membrane connectivity estimated by digital image analysis of HER2 immunohistochemistry is concordant with visual scoring and fluorescence in situ hybridization results: algorithm evaluation on breast cancer tissue microarrays

Because metaplastic carcinoma of the breast encompasses a great variety of histopathology, diagnostic challenges abound, especially within the realm of cytology. The authors compiled and studied an eight-case series comprised of metaplastic breast carcinomas and lesions initially suspicious cytologically for metaplastic carcinoma in order to assess the degree of cytologic-histologic correlation and to identify recurring problematic themes surrounding the cytology-based diagnosis of this neoplasm. The cytologic and histologic slides from eight cases suspicious for metaplastic breast carcinoma diagnosed by fine needle aspiration (FNA) were collected and analyzed through a seven-year retrospective search of case files at our institution. Based on cytologic characteristics, the cases were separated into three groups. Group 1 consisted of three cases presenting with poorly differentiated adenocarcinoma and squamoid components on FNA. Group 2 was composed of two cases that featured a monophasic, malignant ductal cell population on cytology, while the cytologic specimens for the third group of cases presented with a mesenchymal component with or without a malignant glandular constituent. Cytologic-histologic correlation was present in two of three cases demonstrating a mesenchymal component, and there was 100% sensitivity in the cytologic detection of those mesenchymal elements. However, in only one of three cases was there an accurate cytologic diagnosis of metaplastic carcinoma when squamoid changes were present on FNA. Both cases demonstrating only malignant glandular elements on cytologic specimens revealed an additional component of malignant squamous differentiation upon the examination of mastectomy-derived tissue. These results indicate that squamous-like changes identified on FNA should be interpreted with caution and that sampling error remains a problematic recurrence in cytology. Regardless, there appears to be promise concerning the accurate cytologic diagnosis of metaplastic carcinoma when the lesion is characterized by a mesenchymal component. A study implementing a larger case number is essential in determining the significance of these findings.     

Interobserver variability in cytologic subclassification of squamous intraepithelial lesions--the Bethesda System vs. World Health Organization classification

The aim of the study was to compare interobserver variability for The Bethesda System (TBS) and World Health Organization (WHO) classification of cervical squamous intraepithelial lesions. A total of 1,000 conventional Papanicolaou smears (156 positive and 884 negative) were examined "blindly" by three cytologists and one cytotechnician. The degree of observer agreement was expressed by kappa statistics using a program for the calculation of interobserver variation and association "Agree" (Svanholm and Jergensen, 1989). Kappa (kappa) was determined for each cytologic diagnosis within a particular classification and total for either classification. The association with and separation from other diagnoses was determined for each cytologic diagnosis in the form of conditional probability (P(j)). In WHO classification, the diagnoses of dysplasia media and dysplasia gravis showed poor reproducibility (kappa = 0.114 and kappa = 0.259, respectively), the diagnosis of dysplasia levis good reproducibility (kappa = 0.639), and the diagnosis of carcinoma in situ excellent reproducibility (kappa = 0.762). WHO classification yielded pool kappa of 0.741. In TBS classification, the diagnosis of LSIL showed good, and HSIL excellent reproducibility (kappa = 0.542 and kappa = 0.763, respectively). TBS classification yielded pool kappa of 0.699. Dysplasia media (P(j) = 0.121) and dysplasia gravis (P(j) = 0.274) were found to be morphologically poorly defined, and carcinoma in situ (P(j) = 0.777) and dysplasia levis (P(j) = 0.651) well defined diagnoses. LSIL was morphologically moderately defined (P(j) = 0.587) and HSIL well defined (P(j) = 0.789) diagnosis. Accordingly, TBS does not substantially improve diagnostic reproducibility of the cytologic diagnoses of squamous intraepithelial lesions, while providing considerably less information to the clinician than the four-grade dysplasia/CIS terminology, thus eliminating the opportunity of choosing a different procedure for the diagnosis of dysplasia media, which is of utmost importance in the population of young nulliparae.     

Association of human papillomavirus and Chlamydia infections with incidence cervical neoplasia

Incidence cervical neoplasia is defined as disease that becomes manifest during a given period of observation. Association with preceding genital infections having characteristic cytologic findings would seem to be more likely for incidence than for prevalence cases since the usual long latency period of carcinoma in situ (CIS) could allow resolution of infectious processes. For this reason, it was elected to examine the preceding Papanicolaou smears from patients with tissue-confirmed incidence CIS or invasive epidermoid carcinoma. There were 67 women with biopsy-proven CIS or invasive carcinoma of the uterine cervix identified in the files of the University of New Mexico Cytopathology Laboratory from 1966 to 1982 who had two initial negative smears as well as smears at intervals of three years or less. All cytologic smears prior to tissue diagnosis were rescreened for confirmation of cytologic atypia or its absence as well as for morphologic evidence of human papillomavirus (HPV) or chlamydial infections. Control cases matched for age, gravidity, ethnicity and number of smears were reviewed in an identical manner. Koilocytes indicative of HPV infection were found in 17 index cases (25%) and 5 controls (7%) (p = 0.005). Chlamydial infections were identified in 18 index cases (27%) and in 4 controls (6%) (p = 0.001). The times required for conversion from smear negativity to malignancy were determined for each incidence case. The results showed great variability but suggest that the progression to malignancy is not hastened in women with antecedent HPV or chlamydial infections. Our results indicate that the presence of koilocytes and/or chlamydial inclusions in cervical smears serves to identify a group of women with a significantly increased risk of developing cervical carcinoma, even in the absence of concurrent dysplasia.     

Contribution of exfoliative cytology to the diagnosis of laryngeal lesions

OBJECTIVE: To estimate the diagnostic accuracy and reliability of exfoliative laryngeal cytology. STUDY DESIGN: Over three years (1996-1999) cytologic smears were obtained from clinically suspicious laryngeal lesions during laryngoscopy in a total of 31 selected patients (28 males and 3 females with an age range from 28-90 years). The cytologic diagnoses were analyzed and correlated with the histologic and final clinical diagnoses in 17 and 14 cases, respectively. Cytologic identification of the exact histologic type of the lesion was evaluated in 17 patients from whom both cytologic smears and biopsy material were obtained. RESULTS: The overall specificity was 100%, with no false positive diagnoses. The overall sensitivity was 93.3%, with one false negative cytologic diagnosis, in a case of non-Hodgkin's lymphoma. Cytohistologic correlation showed complete agreement between cytologic and histologic diagnoses in five of six benign lesions, in four cases of dysplasia and in six cases of squamous cell carcinoma. The overall diagnostic accuracy of cytology was 96.7% CONCLUSION: Exfoliative cytology by the smear technique is a reliable and accurate method in clinically suspected laryngeal lesions. Moreover, exfoliative cytology may be applied as the only alternative diagnostic method, especially in elderly patients with coexistent cardiorespiratory problems, when biopsy is not advisable or indicated.     

Role of biliary brush cytology in primary sclerosing cholangitis

OBJECTIVE: To assess the role of brush cytology in the routine evaluation of patients with primary sclerosing cholangitis (PSC). STUDY DESIGN: From January 1995 to June 2000, 64 brush cytology specimens were obtained from 21 patients who had at least one cytologic sample obtained during endoscopic retrograde cholangiography. All patients had a diagnosis of primary sclerosing cholangitis. Cases were classified as benign, atypical or malignant according to major cytologic criteria (nuclear contour and chromatin irregularities) and minor cytologic criteria (polarity, cellularity, nuclear enlargement, mitosis, increased nuclear/cytoplasmic ratio) used by us to diagnose biliary brush cytology. Follow-up was available in all cases. RESULTS: Diagnoses were benign (13), atypical (5) and malignant (3) on cytology. Follow-up of the 13 benign cases showed bile duct stones (2), gallbladder adenocarcinoma at cholecystectomy (1), ascending cholangitis (1) and clinically/cytologically by benign follow-up (9). Five of 13 benign cases had subsequent liver transplantation for liver failure, with explants showing changes of primary sclerosing cholangitis. Of the 3 malignant cases, 1 had carcinoma in situ on biopsy, with the explanted liver showing high grade dysplasia; the second patient had cholangiocarcinoma on explant; and the third had hepatocellular carcinoma on liver five needle aspiration. The 5 patients with atypical cytology were reclassified on review as reactive (3) and atypical not otherwise specified (2). Follow-up showed benign disease in 3 of 3 atypical cases reclassified as reactive; 2 of 2 reclassified as atypical not otherwise specified showed low grade dysplasia in the explant. CONCLUSION: The overall incidence of malignancy was low (3 of 21) in patients with PSC. Bile duct brushing is a sensitive method of detecting neoplasia in the setting of PSC when well-defined cytologic criteria are applied.     

Stereotaxic fine needle aspiration cytology of clinically occult malignant and premalignant breast lesions

Stereotaxic fine needle aspiration (FNA) cytology was used to study clinically occult (nonpalpable) breast lesions in 114 consecutive patients with mammographically suspicious findings prior to excisional biopsy. The aspirate contained insufficient material for cytologic evaluation in 15 cases (13.2%), which were histologically diagnosed as benign (7 cases), atypical hyperplasia (7 cases) or carcinoma in situ (1 case). The cytologic findings indicated a benign lesion in 77 cases (67.5%), which were histologically diagnosed as benign (71 cases) or atypical ductal hyperplasia (6 cases). The cytologic sample showed atypia in eight cases (7.0%), which were histologically diagnosed as severe atypical ductal hyperplasia (three cases), carcinoma in situ (one case) or proliferative fibrocystic disease (four cases). In the eight cases (7.0%) cytologically interpreted as probably malignant, histology confirmed six invasive carcinomas, one carcinoma in situ and one fibrocystic disease. Of six cases (4.4%) cytologically reported as malignant, five were histologically diagnosed as invasive carcinoma and one as carcinoma in situ. Overall, stereotaxic FNA cytology reported as malignant or probably malignant 14 of the 15 cases with a histologic confirmation of malignancy, for a sensitivity of 93.3%. Cytology correctly identified 78 of the 83 histologically negative cases, for a specificity of 94.0%. The 16 cases histologically diagnosed as ductal hyperplasia, which carries a high risk for subsequent malignancy, were studied in detail in an effort to define histologic and cytologic criteria for this entity. Using selected histologic criteria, 11 of these cases were graded as showing mild-to-moderate atypical hyperplasia and 5 as showing severe atypical hyperplasia. Three of the latter cases were similarly identified by an analogous cytologic grading; the other two cases had insufficient cytologic samples. The total results in this series of 114 cases support the use of stereotaxic FNA cytology in the diagnosis of these nonpalpable breast lesions, examples of which are illustrated. In particular, it may help to raise the low specificity yielded by mammography alone, which would represent a significant advance for the patient in terms of the accuracy, expediency and reduced cost of diagnosing these lesions.     

Are oral contraceptive-associated liver cell adenomas premalignant?

From January 1977 to June 1990, 1,670 patients with a liver mass or masses underwent transabdominal fine needle aspiration biopsy of the liver. Of those cases, 99 were diagnosed cytologically as "hepatocellular carcinoma" and 9 as "consistent with liver cell adenoma." Among the 99 patients with hepatocellular carcinoma, 3 were users of oral contraceptives. The nine patients with liver cell adenoma were all users of oral contraceptives. Of them, two developed foci or areas of liver cell dysplasia within the adenomas. The mean periods of oral contraceptive use among these three groups of patients were different. It appeared that patients started to develop liver cell adenoma after five years of oral contraceptive use (mean, 6.3 years). Foci or areas of liver cell dysplasia began to arise within liver cell adenomas after 8 years of oral contraceptive use (mean, 9 years), and the patients started to develop hepatocellular carcinoma after 10 years of contraceptive use (mean, 11 years). The cytologic features of liver cell dysplasia strikingly mimicked those of hepatocellular carcinoma. From this study, the foci or areas of liver cell dysplasia appear to be the missing link responsible for the transformation of liver cell adenoma to carcinoma. It is believed that liver cell adenomas are not premalignant and may undergo reversible change after withdrawal of causative agents, whereas foci or areas of liver cell dysplasia within the adenomas are irreversible, premalignant changes and may transform into hepatocellular carcinoma.     

Herpes genitalis in women attending Planned Parenthood of New York City

A search for cytologic evidence of herpes genitalis in women attending centers of Planned Parenthood of New York City was undertaken to determine whether the choice of contraceptives could be playing a meaningful role in the transmission or manifestation of this disease. The study covered 2 1/2 years with correlated clinical data about contraceptives used. Cervical and vagina; cytology specimons were collected routinely at least once yearly. The affected cells are readily recognized in Papanicolaou stained specimens and have a high degree of diagnostic specificity. The disease is transmitted by direct conjugal contact. In some women it its affected by the menstrual cycle suggesting that hormonal factors have an influences. About 67,000 cytology specimens from 43,331 women were studied. Among these were 37 women with cytologic evidence of herpes (.09%). Those who were most likely to have the disease were young, had few or no children and a low family income. Trichomonas vaginalis was also found in 25 of the women with herpes and monilia in 2. The bacterial flora was mixed without Doderlein bacilli in most of the cases of herpes. Almost all the cases of herpes were discovered in the spring or early summer. Little difference was found between the women using different forms of contraceptive. The women using no contraceptives had a much higher rate of herpes (.3%) than any other group. There was no example of herpes in women with cervical cancer or carcinoma in situ. 2 had dysplasia before herpes was found but both had been treated and had no evidence of the previous dysplasia. It seems clear that the use of contraceptives does not add to the risk of herpes genitalis and may in some way lessen the risk. These data are within the range of reported cases from comparable clinics elsewhere.     

Pemphigus vulgaris and microinvasive squamous-cell carcinoma of the uterine cervix

Although patients with disseminated pemphigus vulgaris may have involvement of the uterine cervix, such involvement is often detected only after vaginal discharge or bleeding. When a cervical smear is obtained, distinctive cytologic abnormalities may be observed; these may be attributed to the changes of pemphigus or to an associated reparative/inflammatory reaction. This study documents the first two cases of microinvasive squamous-cell carcinoma of the uterine cervix developing in association with uterine cervical pemphigus. The gross pathologic, cytologic and histologic features of these lesions are illustrated. The cytologic criteria that may be helpful in distinguishing between cells derived from microinvasive squamous-cell carcinoma and pemphigus of the uterine cervix are described.     

Human papillomavirus infection of the cervix: the atypical condyloma

We report on 162 cases of human papillomavirus (HPV) infection of the cervix seen in a two-year period in which the cell sample showed such marked atypia that errors of interpretation could easily have been made. These atypical condylomata are difficult to diagnose cytologically as well as histologically because they mimic dysplasia or carcinoma in situ and, on smears, even invasive squamous carcinoma. HPV particles associated with fibrillar material were found within nuclei of these lesions; their nature was further proved by the immunoperoxidase test. This new form of HPV infection of the cervix showed a 9.1% rate of progression to more advanced cervical lesions. The cytologic finding of atypical condylomata is an indication for colposcopy, confirmative biopsy and appropriate treatment.     

Prediction of cervical neoplasia diagnosis groups. Discriminant analysis on digitized cell images

The purpose of this study was to develop discriminant analysis models for predicting cervical dysplasia/neoplasia case diagnoses using cytometric features derived from the digital image analysis of cell monolayers. The data base consisted of 925 cells from 27 cases diagnosed either as moderate dysplasia (n = 10), severe dysplasia (n = 5), carcinoma in situ (n = 8) or invasive carcinoma (n = 4) on both tissue biopsy and monolayer preparations. Cell features examined were cell diameter, nuclear diameter, nuclear mean optical density (OD), nuclear integrated OD (IOD), nuclear OD standard deviation, normalized IOD, nuclear texture and nuclear-cytoplasmic ratio. Features derived from cells visually classified as moderate dysplasia correctly predicted the case diagnosis of moderate dysplasia versus more severe disease for 85% of the cells. Prediction models using summary measures (mean and variance) derived from all visually classified abnormal cells within each case correctly separated all cases into their respective diagnostic categories. These findings suggest that dysplastic cells in a cytologic sample have features that collectively reflect the tissue diagnosis, regardless of the visual differences among the cells. Such information has potential use for diagnosis and possibly for prognosis.