Alpha-lactalbumin production in human mammary carcinoma.

Alpha-Lactalbumin was isolated from mature human milk and utilized as an immunogen in rabbits. A radioimmunoassay was developed that was capable of detecting nanogram quantities of the antigen. Alpha-Lactalbumin synthesis was detected in human breast cancer cells (MCF-7) cultivated as a continuous cell line in vitro. Other human carcinoma epithelial cell lines (throat and cervix) failed to react in this assay. The ability to synthesize alpha-lactalbumin by breast carcinoma cells appeared to be independent of the addition of prolactin to the culture medium.

Placental infection with human papillomavirus is associated with spontaneous preterm delivery.

BACKGROUND: We sought to determine if human papillomavirus (HPV) infection of extravillous trophoblast cells reduces cell invasion and if placental infection is associated with adverse reproductive outcomes attributed to placental dysfunction. METHODS: We conducted apoptosis and invasion assays using extravillous trophoblast (HTR-8/SVneo) cells that were transfected with a plasmid (pAT-HPV-16) containing the entire HPV-16 genome. In order to associate HPV infection with reproductive outcomes, we conducted a case-control study to detect HPV DNA in the extravillous trophoblast region of placentas from cases of spontaneous preterm delivery, severe pre-eclampsia requiring delivery at <37 weeks and controls who delivered at term. RESULTS: Rates of apoptosis were 3- to 6-fold greater in transfected cells than in non-transfected cells or cells transfected with an empty plasmid. Invasion of transfected cells through extracellular matrices was 25-58% lower than that of the controls. HPV was detected more frequently in placentas from spontaneous preterm deliveries than in placentas from controls (P = 0.03). Identification of HPV in placentas from cases of pre-eclampsia was not significantly different to controls. CONCLUSIONS: HPV infection of extravillous trophoblast induces cell death and may reduce placental invasion into the uterine wall. Thus, HPV infection may cause placental dysfunction and is associated with adverse pregnancy outcomes, including spontaneous preterm delivery.

Adverse reproductive outcomes in urban women with adeno-associated virus-2 infections in early pregnancy.

BACKGROUND: We demonstrated recently that adeno-associated virus-2 (AAV-2) DNA was detected significantly more frequently in placental trophoblast cells from cases of severe pre-eclampsia than from normal term deliveries. Here, we sought to determine if maternal AAV-2 infection early in pregnancy preceded adverse outcomes resulting from placental dysfunction. METHODS: We collected first trimester maternal serum samples and compared anti-AAV-2 IgM antibody levels (indicating primary infection or reactivation of latent AAV-2) between controls delivered at term (n = 106) and three groups of cases: spontaneous abortions (n = 34), spontaneous preterm deliveries (n = 24) and women with at least one outcome usually attributed to placental dysfunction, including pre-eclampsia, intrauterine growth restriction (IUGR) or stillbirth (n = 20). The seroprevalence of immunoglobulin G (IgG) antibodies against AAV-2 and IgM antibodies against viruses that promote AAV-2 replication [adenovirus and cytomegalovirus (CMV)] were also determined. RESULTS: First trimester maternal IgM seropositivity was 5.6 times more prevalent among pre-eclampsia/IUGR/stillbirth cases (P = 0.0004) and 7.6 times more prevalent among preterm deliveries (P < 0.0001) than among controls. CMV and adenovirus IgM antibodies and chronic AAV-2 infections (IgG seropositivity) were not associated with adverse pregnancy outcomes. CONCLUSIONS: Primary or reactivated AAV-2 infection (maternal IgM seropositivity) early in pregnancy was associated with adverse reproductive outcomes associated with placental dysfunction, including pre-eclampsia, stillbirth and spontaneous preterm delivery.

Differential response of malignant BALB/c mammary epithelial cells to the multiplication-stimulating activity of insulin.

Investigations were done to distinguish normal from malignant BALB/c mammary epithelial cells by growth differences. In a high concentration (15%) of serum, normal epithelial cells (from mice at midterm of first pregnancy) and malignant BALB/c mammary epithelial cells in primary culture divided at the same rate (doubling time = 24 hr) and exhibited an extremely short longevity (three doublings). Insulin did not affect growth rate or longevity. Growth rate of normal and malignant cellss was serum-dependent. At serum concentrations limiting for normal cell multiplication, malignant cells showed no proliferative advantage. Insulin complemented but did not replace limiting serum for multiplication of both normal and malignant cells. When allowed to become confluent before three doublings, normal and malignant cell exhibited a contact-mediated (or density-mediated) growth inhibition. Both normal and malignant cells were arrested in the G1 phase of the cell cycle at the same density. Near physiologic doses of insulin stimulated DNA synthesis and mitosis in malignant cells but not in normal cells. Approximately 6% of malignant cells were stimulated to divide in the presence of insulin. Maximum DNA synthesis occurred at 24 hours and maximum mitotic activity at 30 hours after addition of insulin. The data suggested that malignant mammary epithelial cells are more sensitive to insulin as an overgrowth-stimulating factor than are normal cells.

Presence of a mouse mammary tumor virus-related antigen in human breast carcinoma cells and its absence from normal mammary epithelial cells.

Antigen(s) related to the major external glycoprotein (gp52) of mouse mammary tumor virus was detected in the human breast cancer cell line MCF-7. No such antigenic determinants were detectable in normal human mammary epithelial cells.

Induction of endogenous mammary tumor virus expression during hormonal induction of mammary adenoacanthomas and carcinomas of BALB/c female mice.

Treatment of BALB/c female mice with pituitary isografts, under conditions that established mammary hyperplasia in an anovulatory condition, enhanced mammary tumor development with prior onset of dysplastic foci in lobular parenchyma. Tumor onset began at 8 months (mean onset time, 18 mo); the 25-month incidence was 100%. Adenoacanthomatous tumors appeared first. Adenocarcinomas appeared only after more than 14 months of continuous hormone stimulation. Dysplastic and neoplastic changes occurred while blood levels of the three major mammotropic hormones were physiologic. Murine mammary tumor virus (MuMTV) p28 was detected in all tumors tested, independent of time of tumor appearance or tumor type, although keratinizing cells in adenoacanthomatous tumors did not contribute to MuMTV antigen expression. MuMTV gp52 was detected in only a small fraction of cells in a few tumors. MuMTV RNA contents were above normal in all tumors tested. Neither MuMTV structural antigens nor RNA was induced in normal glands by the same hormone treatment that ultimately resulted in dysplasia and tumor formation and elevated levels of these viral markers in neoplasms. The MuMTV RNA in all hormone-induced tumors was readily distinguishable in base sequence from standard MuMTV RNA but indistinguishable from MuMTV RNA recovered from lactating mammary glands of BALB/c females carrying only endogenous MuMTV proviruses, suggesting that endogenous MuMTV RNA sequences were induced in hormone-induced neoplasms. RNA indistinguishable from MuMTV sequences present in hormone-induced primary tumors was also detected in multiple genomic equivalents in two independently derived hormone-induced premalignant alveolar hyperplasias. MuMTV p28 was detected, but gp52 was not. The same hormone stimulus that generated 100% tumors in normal gland greatly accelerated tumor development in premalignant hyperplasias but did not amplify MuMTV RNA or antigens in either hyperplasias or the tumors derived from them. B-type virions were not detected in these tissues, in either hypophyseal implant-stimulated or virgin hosts. Cell-free virions were not detected in culture. These data suggest that the replication of MuMTV induced in hormone-induced neoplasms is defective. Details of its expression suggest that if involved in events leading to tumors, its involvement is insufficient cause for those tumors.

Augmentation of the response of normal mammary epithelial cells to estradiol by mammary stroma.

A two-dimensional monolayer culture system is described in which mammary stromal cells and colonies of normal epithelium are allowed to confront each other en bloc. Epithelial (and stromal) cell growth was inhibited by confrontation. Epithelial cell growth was reinitiated by estradiol (10(-8) M) but not by dexamethasone. Reinitiated growth was inhibited by tamoxifen (10(-6) M). Contact between stromal cells and epithelium was critical for the response to estradiol, and photomicrographic evidence was obtained that estradiol stimulated invasion of the stroma-epithelia interface. These observations are organized into a model for mitogenic action of estradiol that seems to reconcile observed disparities in the action of the hormone in vivo and in vitro.

Hormonal induction of mammary tumor viruses and its implications for carcinogenesis.

The purpose of this study was to evaluate the possibility that hormones and mouse mammary tumor virus (MuMTV) are cocarcinogenic for mammary epithelium. Results of our investigations in vivo and in vitro suggest: (a) Hormones promote mammary carcinogenesis in BALB/c females whether MuMTV is germinally (BALB/c) or horizontally (BALB/cfC3H) transmitted; the rate of carcinogenesis in BALB/cfC3H females is substantially faster than it is in BALB/c, but the final mammary carcinoma incidence is approximately the same. The rate-limiting step in malignant transformation in BALB/c, which infectious MuMTV overcomes, is in premalignant transformation from normal. (b) One characteristic of horizontal MuMTV transmission in BALB/c that is not observed in germinal transmission is integration of new MuMTV sequences in mammary cell DNA. Integration is mammary cell specific and constant at 2 to 3 copies/cell from tumor to tumor. (c) MuMTV expression is changed in mammary epithelial cells during hormonal carcinogenesis. The nature of the change is qualitatively similar in both BALB/c and BALB/cfC3H. Expression of envelope glycopeptides (glycoprotein with a molecular weight of 52,000) is induced, which correlates with amplification of MuMTV RNA sequence content. Quantitative differences exist in induced levels in BALB/c and BALB/cfC3H. (d) MuMTV RNA and a glycoprotein with a molecular weight of 52,000 were not inducible with dexamethasone in normal mammary epithelial cells in culture. These structural components were induced in both premalignant (BALB/cfC3H) and malignant (BALB/c and BALB/cfC3H) cells. MuMTV RNA was induced by dexamethasone in normal cells pretreated with 5-iodo-2'-deoxyuridine. (e) Both premalignant and malignant cells have altered (vis-à-vis normal) surfaces, discernible by differences in reactivity with concanavalin A in hemadsorption assays. Indirect evidence suggests that the alteration includes membrane incorporation of MuMTV-related determinants of a glycoprotein with a molecular weight of 52,000. (f) Malignant cells exhibit enhanced sensitivity to insulin for reinitiation of DNA synthesis and mitosis in contact-inhibited homotypical monolayers. These findings have been organized into a hormonal cocarcinogenesis hypothesis in which expression of germinally transmitted MuMTV genes is the proximal cause of neoplastic transformation.

Evidence for a direct growth-stimulating effect of estradiol on human MCF-7 cells in vivo.

The MCF-7 continuous line of human breast cancer cells requires that athymic nude mice receive supplemental estrogen so that inocula can produce progressively growing tumors. Although these cells contain a typical estrogen receptor complex, the lack of consistent growth stimulation induced by estrogens added to in vitro culture systems has raised the question as to whether this class of hormones acts directly upon the cells or induces a second message produced in other tissues. The present experiments were designed to test the effect of estradiol on the growth of these cells in vivo by exposing them directly to the hormone prior to its absorption into the hepatic portal circulation and subsequent metabolic inactivation. Tumor fragments that were placed next to an estradiol-containing pellet in the spleen grew to produce grossly evident tumor masses, whereas those in the subcutis of the same animals did not, although some minute residua did remain. In the splenic tumors, the mitotic index of the MCF-7 cells immediately adjacent to the estrogen pellets was 2.4 times that of cells on the other side of the same tumor and 3.5 times that of those in the minute s.c. residua. We interpret these data as indicating that in vivo estradiol is acting directly upon the MCF-7 cells to increase their rate of proliferation rather than to initiate the production of a second message to be released into the circulation. Additionally, it was found that s.c. tumors that were decreasing in volume subsequent to withdrawal of systemic estrogen still contained dividing neoplastic cells but with a lower frequency than that seen in progressively growting tumors stimulated with estradiol. This finding indicates that MCF-7 cells can proliferate in vivo in the absence of a substantial amount of estrogen but only at a rate insufficient to sustain progressive tumor growth.

Site-selective growth of a hormone-responsive human breast carcinoma in athymic mice.

MCF-7 is a human breast cancer line which requires estradiol supplementation for growth in s.c. tissues of athymic mice. In order to evaluate the influence of host site on hormone-dependent tumorigenicity and growth, MCF-7 cells were inoculated into tissues varying in ability to concentrate exogenous estradiol. Tumorigenicity was defined in terms of latency, threshold inoculum size, and tumor growth and progression. We observed that sites such as lung rarely supported MCF-7 tumors. However, moderately estrophilic sites such as mammary fat and adjacent subcutaneum and dermis supported the growth of small MCF-7 tumors from large tumor cell inocula, but only in estrogenized mice. In contrast, the highly estrophilic sites, brain and periuterine tissues, produced rapidly growing tumors from small tumor cell inocula. Only in periuterine tissues did tumors develop in the absence of exogenous estradiol. These studies demonstrate that tumorigenicity and growth rates of estrogen-dependent MCF-7 tumors vary as a function of tissue implantation site.

Phenotypic variance among cells isolated from spontaneous mouse mammary tumors in primary suspension culture.

A method is given for selecting epithelial cells directly from primary mammary tumors in Methocel suspension culture. The frequency of colony-forming units in primary tumors was approximately 10(-4). Colonies grew by cell division; formation and growth of colonies was cell density dependent. Five Methocel isolates were established in monolayer culture and characterized. Two were epithelial, evidenced by functional occluding junctions. The other three were not typed in vitro, although they formed carcinomas in vivo. All were subtetraploid by passage 10. There were variations in ability of the five Methocel isolates to reclone in suspension that appeared to be due to the evolution of anchorage-dependent variants during their growth in monolayer culture. These variants could be purified by limiting dilution plating on solid substrates. The five Methocel isolates and their derivative variants were used to determine correlations between transformation markers and tumorigenicity. Only three Methocel-derived sublines of nine tested, including two recloned in Methocel, were tumorigenic at all when inoculated in two sites of three syngeneic hosts, one athymic. The other six were nontumorigenic. The tumorigenic sublines were less tumorigenic than uncultured cells of parent tumors or parent tumor cells grown in primary monolayer culture. Thus, anchorage-independent growth is not a reliable marker for the tumorigenic mammary phenotype. No correlation was found between two other "contact-related" transformation markers, rapid growth rate and monolayer overgrowth, and tumorigenicity. The three transformation markers were expressed independently. Both tumorigenic and nontumorigenic sublines expressed mammary tumor virus antigens M.W. 28,000 protein and M.W. 52,000 glycoprotein, although only a minor fraction of cells contained the M.W. 52,000 glycoprotein. These data emphasize the heterogeneity of phenotypes in mammary tumors as well as differences between fibroblasts and mammary epithelium in models of neoplastic transformation.

Concanavalin A-mediated hemadsorption by normal and malignant human mammary epithelial cells.

The concanavalin A (Con A) reactivity of malignant and normal human mammary epithelial cells in culture was determined with a hemadsorption assay. Human erythrocytes were treated with various concentrations of Con A, and these indicator Red Blood Cells were incubated with the test cells in situ in culture dishes. The Con A concentration at which approximately 50% of the test cells adsorbed erythrocytes ([Con a] 1/2 max) was determined. Five malignant epithelial cell lines and the primary cultures derived from 3 pleural effusions and 20 solid tumors were tested. Primary cultures of normal epithelial cells were established from human milk samples obtained from 3 separate donors. The average [Con A] 1/2 max value for the 5 cell lines and the pleural effusion cultures was 6 and 5 microgram/ml, respectively. The average [Con A] 1/2 max value for the 20 solid breast tumors was 20 microgram/ml. In contrast to the malignant cells, normal mammary epithelial cells did not adsorb erythrocytes coated with as much as 100 microgram Con A per ml. These results show that Con A reactivity distinguishes normal from malignant human mammary epithelial cells.

Isolation and characterization of a spontaneously immortalized human breast epithelial cell line, MCF-10.

Two sublines of a breast epithelial cell culture, MCF-10, derived from human fibrocystic mammary tissue exhibit immortality after extended cultivation in low calcium concentrations (0.03-0.06 mM) and floating transfers in low calcium (MCF-10F), or by trypsin-Versene passages in the customary (normal) calcium levels, 1.05 mM (MCF-10A). Both sublines have been maintained as separate entities after 2.3 years (849 days) in vitro and at present have been in culture for longer than 4 years. MCF-10 has the characteristics of normal breast epithelium by the following criteria: (a) lack of tumorigenicity in nude mice; (b) three-dimensional growth in collagen; (c) growth in culture that is controlled by hormones and growth factors; (d) lack of anchorage-independent growth; and (e) dome formation in confluent cultures. Cytogenetic analysis prior to immortalization showed normal diploid cells; although later passages showed minimal rearrangement and near-diploidy, the immortal cells were not karyotypically normal. The emergence of an immortal culture in normal calcium media was not an inherent characteristic of the original tissue from which MCF-10 was derived since reactivated cryo-preserved cells from cultures grown for 0.3 and 1.2 years in low calcium were incapable of sustained growth in normal calcium.

Cytometrically coherent transfer of receptor proteins on microporous membranes.

A new method (Freeze-Transfer) is described for performing high-resolution immunocytochemistry for soluble cell proteins on frozen sections of biological tissues that involves thaw-mounting frozen tissue sections directly onto the surface of nitrocellulose thin films instead of directly onto glass slides. This technically straight-forward change in methodology resulted in chromogenic immunocytochemical assays for Her-2 and EGF receptors that were 1-2 orders of magnitude more sensitive while still fully utilizing the diagnostic resolving power of light microscopy. The effects of membrane pore size and surface chemistry on the resolution and intensity of Her-2 signal suggest that the enhanced sensitivity of Freeze-Transfer was caused by the cytologically coherent transfer of target molecules normally lost from cut surfaces of cells mounted on nonporous glass during assay.

Estrogen responsive proliferation of clonal human breast carcinoma cells in athymic mice.

A clone of MCF-7 (MCF-7ED), a cell in continuous in vitro cultivation derived from an estrogen-responsive human breast carcinoma, requires estrogen supplementation for progressive exponential (double time: 50--85 h) tumor growth in the mammary fat of athymic mice. The plasma concentration of estradiol which stimulated exponential growth of MCF-7ED corresponded to physiologic premenopausal levels in women. The tumors were carcinomas with murine and MCF-7ED cells intermixed. MCF-7ED cells could be repurified in subcultures of mixed tumors. Comparative studies of breast and non-breast cell lines showed concordance between the presence of estradiol receptor, sensitivity to the anti-estrogen tamoxifen for growth in vitro, and estradiol dependence for tumorigenic growth in athymic mice. Progesterone alone did not stimulate MCF-7ED growth, but acted synergistically with estrogen. Progesterone's action was to decrease tumor latent period, not to increase final tumor incidence or growth rate. Under estrogen-deficient conditions, condsitions approximating postmenopausal status in women, (10(-10) M in plasma), a dormant state was established between MCF-7ED cells and murine mammary stroma which could be maintained several months. The dormant state could be broken by introduction of estradiol, but not progesterone. This system should be useful for defining host and cancer cell determinants in estrogen-responsive breast cancer growth.

Estrogen induced growth of human breast cancer cells (MCF-7) in athymic nude mice is enhanced by secretions from a transplantable pituitary tumor.

MCF-7, a human breast carcinoma cell line, was maintained s.c. in female athymic nude mice for a period of 5-6 weeks. Administration of estrogen (s.c. pellet of 17 beta-estradiol and estrone in drinking water, 0.5 mg/l) to these mice resulted in sustained (P less than 0.001) growth of MCF-7 tumors. Grafting of a prolactin and growth hormone secreting rat pituitary tumor to the estrogen-treated mice resulted in an increased (P less than 0.05) rate of MCF-7 tumor growth. MCF-7 did not grow in athymic nude mice grafted with rat pituitary tumor alone or in mice without hormone treatment (controls). Thus, secretions of pituitary hormones alone are not capable of promoting in vivo growth of MCF-7 although such secretions significantly enhance estrogen-induced growth of this cell line. A synergism between pituitary hormones and estrogen for in vivo growth of a human breast carcinoma has been demonstrated in this study.

Lobectomy with tangential pulmonary artery resection without regard to pulmonary function.

BACKGROUND: Non-small cell carcinoma of the lung invading the pulmonary artery (PA) has traditionally been treated by pneumonectomy. Although PA resection and reconstruction (PAR) has begun to gain acceptance, previous series of PAR by the simplest technique of tangential excision and primary repair have been unfavorable. We have maintained a policy of performing PAR preferentially whenever anatomically feasible, and usually this has been possible by tangential excision and primary repair. This study sought to determine if this approach is sound. METHODS: Retrospective clinical and pathologic review. RESULTS: Thirty-three PARs were performed from 1992 to 1999. The patients, followed 6 to 65 months (mean 25), were aged 36 to 80 years (mean 61), and their tumors were pathologic stage IB (n = 7), IIB (n = 13), IIIA (n = 9), and IIIB (n = 4). The mean preoperative forced expiratory volume in 1 second was 70% predicted. The procedures included 14 bronchial sleeve lobectomies with PAR and 19 simple lobectomies with PAR. The PARs were performed without heparinization and included 19 tangential excisions with primary closure, 11 larger tangential excisions with pericardial patch closure, and 3 sleeve resections. There were no operative deaths and 2 (6.1%) early major complications, all unrelated to the PAR. Thirteen patients (39%) had early minor complications. Four-year Kaplan-Meier survival was 48.3% for stages I/II and 45% for stage III. Ipsilateral, central, intrathoracic recurrence occurred in 3 patients (9.1%). CONCLUSIONS: These data are not dramatically different from those reported for standard resections. Although the numbers are small, the results suggest that lobectomy with PAR by tangential excision is an acceptable alternative to pneumonectomy whenever anatomically possible.

Follow-up of morphologically reactive lymphoid proliferations in fine-needle aspirates of elderly patients.

Fine-needle aspiration (FNA) is an effective tool in evaluating the cause of lymphadenopathy. While the morphologic diagnosis of a reactive lymphoid proliferation is common in younger patients, this diagnosis should be made carefully in older patients (those over the age of 50 yr) in light of the facts that such purely reactive conditions occur much less frequently in this population, and that follow-up of these patients reveals a malignancy (usually lymphoma) in a significant number of cases. In this series, we identified 40 patients with a morphologic diagnosis of reactive lymphoid proliferation on FNA and obtained their follow-up information. Of 19 patients under the age of 50 yr, 5 underwent subsequent biopsies and only one revealed a definitive malignancy (5%). In contrast, 7 of 21 patients over the age of 50 yr underwent a subsequent biopsy, and 6 were found to have a malignancy (5 malignant lymphomas, 1 metastatic melanoma). The higher rate of positive follow-up (29%) in this age group supports previous suggestions that morphologically reactive (mixed) lymphoid proliferations be viewed with increased suspicion in the elderly patient, and that additional studies, such as flow cytometry, be performed when material is available.

Duodenal carcinoid tumor: report of a case diagnosed by endoscopic ultrasound-guided fine-needle aspiration biopsy with immunocytochemical correlation.

Fine-needle aspiration biopsy is a reliable and accurate method for the endoscopic diagnosis of gastrointestinal malignancies and it is particularly well suited for evaluation of submucosal lesions. We report the cytopathologic findings of a case of malignant carcinoid tumor of a 44-year-old male who presented with melena and a nonhealing duodenal ulcer. Endoscopic ultrasound examination revealed a submucosal lesion in the pyloric region. Fine-needle aspiration revealed abundant cellularity with tumor cells arranged in sheets and loose groups and dispersed single cells in a clean background. Papillary fragments, capillaries cuffed by tumor cells, and rosette formation were also noted. The cells were moderate in size, round to oval, with a small subpopulation of spindle-shaped cells. The nuclei were uniform, round to oval, with smooth nuclear borders. The chromatin pattern was finely granular with a salt-and-pepper appearance. The cytoplasm of the cells was small to moderate in amount, pale, and showed fine granularity. The differential diagnosis included a neuroendocrine neoplasm vs. an epithelioid gastrointestinal stromal tumor. The tumor cells were focally positive for chromogranin and negative for CD34, supporting the diagnosis of a neuroendocrine neoplasm. The differential diagnosis of primary gastrointestinal carcinoid tumors from gastrointestinal stromal tumors can be very difficult in cytologic material. In cases when diagnostic material is scant, or only present on one smear, the use of smear division and cell transfer in order to perform immunocytochemical stains may be of considerable value to confirm the neuroendocrine nature of the neoplasms.

Evaluation of mild-to-moderate dysplasia on cervical-endocervical (Pap) smear: a subgroup of patients who bridge LSIL and HSIL.

The Bethesda System recommends Pap smear diagnosis to be based on the most abnormal cells present, regardless of number. Our reporting system includes a diagnostic category of mild-to-moderate dysplasia (D1-2), defined as cases with only a few moderately dysplastic cells. We evaluated the validity of a D1-2 diagnostic category by reevaluation of 58 cases with subsequent follow-up (up to 24 months). On biopsy and/or Pap smear follow-up, 24 cases (41%) showed LSIL and 34 cases (59%) showed HSIL. Index smears from these cases were examined by two cytopathologists blinded to patient follow-up for the following morphologic features: volumes (scale 1-4) of LSIL, HSIL, and dyskeratosis, HSIL as single cells or syncytial fragments, and acute inflammation. None of the morphologic features evaluated were significantly different between the LSIL and HSIL follow-up groups based on univariate and multivariate logistic regression analysis. The diagnosis of D1-2 on Pap smear is a valid diagnostic category that defines a subgroup of patients with both LSIL and HSIL follow-up, which cannot be reliably predicted based on morphology alone.