Search

Background. The survival of 52 patients with hepatocellular carcinoma (HCC) seen during the last 4 years was analyzed prospectively on the basis of disease stage and nuclear DNA content. Methods. Ploidy was measured by flow cytometry (FCM). Cells for cytologic diagnosis and FCM were collected by ultrasound-guided fine needle aspiration. Results. DNA aneuploidy, which was detected in 62% of the patients, did not correlate with clinicopathologic features, except in the sonographic aspect (P = 0.03). However, ploidy correlated significantly with survival; the survival times for patients with an aneuploid DNA index were significantly shorter than for those with a diploid index (P = 0.02). In a Cox multivariate analysis, DNA content was prognostically significant, as were the grade of cirrhosis severity and the echographic aspect. Conclusions. In addition to the clinicopathologic features observed, FCM DNA analysis of ultrasound-guided fine needle aspirates from HCC is a simple and valid method for estimating a prognosis of these patients. Cancer 1994; 74:599-605.

Background. Flow cytometry (FCM)-measured DNA content may be a predictor in the prognosis of ovarian cancer. Multiple specimens taken from the same ovarian tumor may show a variation in DNA content (i.e., intratumoral DNA heterogeneity). We measured the FCM DNA content of multiple specimens from the same tumor in ovarian cancer, and the relationship among DNA ploidy, intratumoral DNA heterogeneity, retroperitoneal lymph node involvement, and surgical staging was evaluated. Methods. Forty-one patients with primary epithelial ovarian cancer were included in the study. The FCM-measured DNA content of multiple fresh tumor specimens taken from different parts of the same ovarian tumor from each patient was measured. When aneuploidy was observed in at least one specimen from the same tumor, the tumor was defined as an aneuploid tumor. If there were two or more different aneuploid stem lines with a variation of DNA indices (differences of the DNA indices >0.15) from the same tumor, the presence of intratumoral DNA heterogeneity was defined. Results. Diploid tumor was found in 8 (19.5%) of the 41 patients, and aneuploid tumor with intratumoral DNA heterogeneity was found in 20 (48.8/o). None of the eight patients with diploid tumors demonstrated lymph node involvement. In contrast, lymph node involvement was found in 14 (70.0%) of 20 patients with intratumoral DNA heterogeneity. There was a significant different incidence of lymph node involvement between the groups with and without intratumoral DNA heterogeneity (P < 0.01). The incidence of intratumoral DNA heterogeneity significantly correlated with the International Federation of Gynecology and Obstetrics staging (P < 0.01), while that of aneuploid tumor did not. Conclusions. Intratumoral DNA heterogeneity reflected a malignant potential for lymph node involvement and its progression in epithelial ovarian cancer. Cancer 1994; 73:3011-4. Key words: flow cytometry, DNA ploidy, intratumoral DNA heterogeneity, lymph node involvement, ovarian cancer.

Background. DNA ploidy is a significant prognostic factor in patients with prostate cancer. Using DNA/nuclear protein flow cytometry, a subpopulation of tumors with near-diploid DNA is identifiable. The prognostic significance of near-diploidy was examined. Methods. Paraffin-embedded formalin fixed prostate tumor tissue from patients treated at M. D. Anderson Cancer Center with external beam radiation therapy was processed for DNA/nuclear protein flow cytometry. All patients had pretreatment and follow-up serum prostate specific antigen (PSA) levels. Seventy-six specimens were suitable for flow cytometric analysis. Tumors were classified as either diploid (n = 30), near-diploid (n = 24), or nondiploid (n = 22, tetraploid and aneuploid). Median follow-up time was 36 months. Results. Diploid tumors were associated with a significantly better actuarial outcome at 4 years, compared with near-diploid tumors, using either biochemical relapse (rising PSA) or a composite end point of a rising PSA or clinical relapse (16% versus 52% relapse, P < 0.05, log-rank). Moreover, patients who had nondiploid tumors had the worst prognosis (77% relapse, composite end point). No significant difference was observed between diploid and near-diploid neoplasms regarding actuarial local control, freedom from metastasis, freedom from clinical relapse, or overall survival time. A Cox proportional hazards model, using the composite end point of a rising PSA or relapse, was performed with ploidy categorized as diploid, near-diploid, and nondiploid; pretreatment PSA, DNA ploidy, and tumor grade were found to be independent prognostic factors. When ploidy was categorized as diploid or near-diploid (nondiploid tumors excluded), pretreatment serum PSA and DNA ploidy were independent predictors of outcome. Ploidy remained an independent prognostic factor even when nondiploid tumors were excluded. Conclusions. These data show that patients who have near-diploid tumors have an intermediate prognosis between the more favorable diploid tumors and the less favorable nondiploid tumors. Cancer 1994; 73: 1895-903. Key words: prostate cancer, radiation therapy, flow cytometry, ploidy, DNA, nuclear protein.

Background. The prognostic significance of flow cytometric DNA ploidy and S-phase fraction (SPF) in ovarian cancer has been controversial. In the current study, the authors analyzed tumor heterogeneity in respect to DNA index DI and SPF. Methods. Flow cytometric variation in DI and SPF among representative fresh tumor material from the primary tumor, metastasis, and malignant effusions from the same patient was analyzed. Results. One hundred thirty-two samples from 47 patients were analyzed, and 119 samples from 42 patients were evaluable. Stable DI between different samples was found in 34 patients, whereas heterogeneity was found in 8 patients (19%). The metastases showed stable DNA content. The malignant effusion samples often lacked tumor cells. The representative ones were often DNA diploid. In 21% of the aneuploid samples, the SPF could not be analyzed. In 38% of the aneuploid samples, the stem line constituted less than 15% of measured nuclei. In these samples, a negative correlation between SPF and percentage of aneuploid cells was found, making SPF unreliable. Correct SPF measurement was thus possible in only 41% of the aneuploid samples, and in these tumors, SPF values varied considerably among different samples from the same patient, illustrated by a median SPF difference of 11% (range, 0-28%). Conclusions. Tumor DI heterogeneity existed in 19% of tumors. SPF depended on the amount of aneuploid cells in case of small stem lines and varied considerably, making its use as a prognostic factor doubtful. To ensure that all tumor stem lines are represented, at least two biopsy specimens from any solid tumor should be analyzed. Cancer 1994; 73:1870-7. Key words: ovarian carcinoma, tumor heterogeneity, DNA ploidy, S-phase fraction.

Background. Aneuploidy in DNA flow cytometry (FCM) of musculoskeletal tumors is generally considered to be a sign of malignancy. Previously, giant cell tumor of the bone has been reported to contain aneuploid (near0diploid) DNA stemlines. Otherwise, only sporadic cases have been reported. The authors wanted to study the relationships among DNA FCM, histology, and clinical course of nonmalignant musculoskeletal lesions. Methods. Twenty-eight histologically benign tumors and seven nonnepolastic lesions were subjected to DNA FCM. After tissue preparation mechanically and with ribonuclease and trypsin, the isolated nuclei were stained with propidium iodine using chicken and rainbow trout erythrocytes as controls. In the DNA FCM histograms, ploidy and cell cycle fractions were determined using a computerized mathematical model. The histologic diagnoses were made without knowledge of the DNA FCM results. Results. Aneuploidy was found in eight lesions. A shoulder in the diploid peak, suggesting a diploid and a near-diploid population, was found in DNA histograms of a condensing osteitis of the clavicle (a benign inflammatory process) and of a giant cell tumor of bone. The latter lesion also had a tetraploid population. Six benign tumors--two enchondromas, one osteochondroma, one subcutaneous and one intramuscular lipoma, and a calcifying aponeurotic fibroma--showed clear aneuploidy with separate peaks. The S-phase fraction was less than 10% in all cases. The highest aneuploid population, DNA index = 1.70, in a subcutaneous lipoma, was small, with an undetectable S phase. Despite nonradical operations in seven lesions, no recurrences were observed during a median follow-up of 49 months (range, 28--73 months). Conclusions. Small aneuploid populations with low DNA synthetic activity may be compatible with a benign histologic picture and uneventful clinical course of the musculoskeletal lesion. Cancer 1994; 73:1200--5.

Background. Although flow cytometry (FCM) has become a widely used technique for the measurement of DNA content in solid tumors, the correlation of ploidy analysis by FCM with cytogenetic analysis (CGA) is not well described. The sensitivities of G-banded CGA and FCM were compared to determine the accuracy of the DNA index value (DI) as a measurement of chromosome number. Methods. Tumor specimens from 56 pediatric cases were analyzed for DNA content by both FCM and CGA. Nuclei for FCM were prepared from archival tissue in 53 specimens using a modification of the Hedley technique and from fresh tissue in 3 specimens. Metaphase chromosomes for CGA were prepared from standard solid tumor harvests and Giemsa-trypsin banding procedures. Ploidy status for this study was defined as (1) diploid--DI between 0.97 and 1.03 by FCM or chromosome number [+ or -] 2 from normal by CGA (44-48); and (2) aneuploid--DI < 0.97 or > 1.03 by FCM or total chromosomes < 44 or > 48 by CGA. Results. Forty-nine of the 56 pediatric specimens were evaluable by both techniques. Concordance was observed in 34 cases (69%) between the two techniques in assigning similar ploidy status to a tumor (22 diploid and 12 aneuploid). It also was observed that among the aneuploid concordant cases, the actual DI obtained from archival material could predict total chromosome number with 95% accuracy. The 15 discordant cases showed a distinct aneuploid population by FCM, but were diploid by CGA. Conclusions. A correlation of 69% was obtained between both techniques to assign a similar pleidy status (diploid versus aneuploid) in 56 pediatric solid tumors. These results support the combined use of CGA and FCM to obtain the most complete analysis of DNA content and chromosome abnormalities in pediatric solid tumors. FCM on formalin-fixed, paraffin-embedded tissue can be used to measure total DNA content. Cancer 1993; 72:2033-41.

Background The recurrence or mortality rate of axillary lymph node-negative invasive breast cancer has been associated with the tumor S-phase fraction, which is measured by DNA flow cytometry. Because many of the studies that established this association were performed using frozen, pulverized tumor specimens, this association could not be tested for independence from the established prognostic factors of histologic and nuclear grading. Methods. Histologic, nuclear, and mitotic grades, DNA ploidy, and S-phase fraction (SPF) were determined from paraffin-embedded tumors obtained from 280 women with node-negative invasive ductal carcinomas using standard grading schemes and flow cytometric techniques. These variables were compared with disease-free and cancer-specific survival (CSS) in univariate and multivariate analyses of these patients. Results. Tumor diameter, SPF, histologic grade, and nuclear grade were significant predictors of disease-free survival (DFS); diameter and SPF had significant associations with CSS. Cox analysis showed histologic grade to be the only independent predictor of relapse, whereas diameter and SPF were independent predictors of mortality. The patients with low nuclear or histologic grade tumors had only a 5% risk of recurrence at 5 years. In contrast, 36% of patients in this series with medium-grade or high-grade high SPF tumors had a 30% risk of recurrence over the same interval. Conclusions. Histopathologic grading and flow cytometric determination of SPF appear to provide additive prognostic information for patients with early invasive ductal carcinomas of the breast. Cancer 1993; 72: 1926-32.

Background. The prognostic value of flow cytometric DNA analysis on paraffin-embedded tumor samples has been controversial in esophageal cancer. To clarify its true significance, the authors developed an improved method that excludes the possibility of contamination by lymphocytes in tumor sample. Methods. Single nuclear suspension was prepared from paraffin-embedded samples on 103 patients with squamous cell carcinoma of the esophagus. Both DNA content and nuclear size were simultaneously measured by flow cytometry on 30,000 nuclei, and contaminated lymphocyte nuclei were eliminated from the data by optimal gating. Correlation between DNA ploidy and postoperative survival was examined. Results. Analysis using a flow cytometric cell sorter showed that the frequency of tumor cells in the lymphocyte-reducing gating fraction (LGF) was significantly higher than that in the conventional nongating fraction (NGF). LGF analysis showed aneuploid peaks in 58 patients (56.3%), but NGF analysis showed aneuploid peaks in only 38 patients. LGF analysis revealed that the aneuploid tumors had higher histologic grading (P < 0.05) and worse survival rate (P < 0.01) compared with diploid tumors. However, conventional methods could not detect this difference. Conclusions. Flow cytometric analysis gating by nuclear size may be helpful to detect aneuploid peaks, and for predicting prognosis of patients with squamous cell carcinoma of esophagus. Cancer 1993; 72:1813-8.

Background. DNA ploidy and S-phase fraction measured by DNA flow cytometry were shown to correlate with several clinicopathologic characteristics in several types of tumors. Methods. DNA flow cytometry was performed on 329 samples (164 normal mucosa, 165 tumors) obtained from 165 patients (112 men) with stomach cancer, and the findings were correlated with various clinicopathologic characteristics of the patients in a prospective manner. Results. Seventy-nine of 165 samples (48%) from the tumors gave an aneuploid histogram. None of 164 samples from the normal mucosa showed aneuploidy. There was no significant difference in the frequency of DNA aneuploidy in terms of age, sex, symptom duration, bleeding history, gastric outlet obstruction, weight loss, performance status, serum hemoglobin level, albumin level, creatinine level, tumor size, tumor location in the stomach, and TNM stage. Moderately well differentiated tumors had a significantly higher frequency of aneuploidy compared to well differentiated or undifferentiated tumors. S-phase fraction was obtained in 162 of 164 samples from the normal mucosa, and 123 of 165 samples from the tumors. The overall mean of the S-phase fraction was 4.01% (range, 0.5-23.6%) for the normal mucosa and 13.8% (range, 0-51.9%) for the tumors. Higher S-phase fraction of tumors was correlated with history of weight loss, poorer performance status, and histologically less differentiated tumors. Conclusions. The overall frequency of aneuploidy was 48% in stomach cancer. DNA aneuploidy showed significant correlation with histologic differentiation of tumors. S-phase fraction of the tumors showed significant correlartion with history of weight loss, performance status of the patients, and histologic differentiation of tumors. Cancer 1993; 72:1819-26.

Background. The prognostic value of flow cytometric analysis of DNA content is contradictory in renal cell carcinoma (RCC). Methods. Flow cytometric analysis was performed on 72 patients with RCC, and the relationship between the DNA content and pathologic or clinical features was investigated. Results. Aneuploidy was observed in 36 (50%) of these patients. Two tissue samples from primary tumors of 12 patients were analyzed, and heterogeneity was found in 7 (58%) of these patients. The incidence of aneuploidy was significantly higher in high-grade than in low-grade tumors (P = 0.0328). All five tumors with a maximum diameter of 2.5 cm or smaller (T1N0M0) were diploid, and a trend for more aneuploid tumors among high-stage diseases was observed. The survival rate of patients with diploid tumors was not significantly different from that of those with aneuploid tumors in the low-or high-stage group. Conclusions. DNA ploidy assessed with one sample is not a prognostic factor, and heterogeneity between different samples of a given tumor indicates that one sample is not enough for DNA flow cytometry in RCC. Cancer 1993; 72:1319-23.

Background. Although it is important to determine any relationship between tumor DNA ploidy and its biologic behavior, the correlation between DNA-ploidy and the prognosis of patients with ovarian cancer is not conclusive. Accordingly, the authors evaluated the clinical application of DNA ploidy in ovarian cancer. Methods. Flow cytometric measurements were performed in 45 selected patients with well-differentiated serous cystadenocarcinoma of the ovary, Stages Ic-IV. All of them had the same surgical procedure, with retroperitoneal lymphadenectomy including paraaortic nodes, followed by the same postoperative chemotherapeutic regimen. Results. Of the 45 ovarian cancers, 28 were diploid and 17 were aneuploid. The 2-year survival rate and the estimated 5-year survival rate for patients with diploid tumors were significantly greater than those for patients with aneuploid tumors (73.2% versus 46.7% and 29.1% versus 22.4%, respectively). The 2-year survival rate in patients with advanced disease (Stage III or IV) was also significantly higher for those with diploid tumors (53.3% versus 37.4%, respectively), but the estimated 5-year survival rate was similar in both groups (8.9% versus 9.1%, respectively). Patients with advanced disease had aneuploid tumors more frequently than those with early-stage disease. A significantly higher incidence of retroperitoneal lymph node metastasis was observed in aneuploid tumors than in diploid tumors (43.8% in diploid tumors versus 86.7% in aneuploid tumors). The authors found no difference in the response to chemotherapy between diploid and aneuploid tumors. Conclusions. Although tumor DNA ploidy was not as reliable as conventional parameters such as surgical stage in establishing prognosis, it may provide an indicator of lymph node involvement.

Background. The incidence of upper urinary tract tumors is relatively high in southern Taiwan. DNA analysis by means of flow cytometry is not well investigated with regard to tumors of the upper urinary tract and the differences in DNA ploidy between transitional cell carcinomas in endemic and nonendemic areas. Methods. A retrospective nuclear DNA ploidy analysis by flow cytometry comprised 41 formaldehyde solution-fixed, paraffin-embedded tissue specimens of transitional cell carcinoma of the renal pelvis and ureter. The preparation of nuclear suspensions from paraffin-embedded tissue blocks and staining were modified by means of the techniques of Hedley and Vindelov. Results. There was no statistical correlation between DNA ploidy, histologic grade, and pathologic stage; however, 82% of the DNA nondiploid tumors showed tumor progression. In contrast, only 46% of the DNA diploid tumors revealed tumor progression. Among Grade 2 tumors, 85% of the DNA nondiploid tumors showed postoperative tumor progression, whereas only 31% of the DNA diploid tumors showed tumor progression. Seventy-nine percent of the nondiploid patterns were present in patients native to the Pa Chang Valley, where the so-called 'blackfoot disease' and urothelium tumor are endemic, whereas only 22% appeared in patients living. in other areas. Conclusions, DNA flow cytometry can identify a group of patients with poor outcome unpredictable by pathologic examination, and is an important tool in research into the pathogenesis of cancer.

A patient is described with chronic myelogenous leukemia in blastic crisis, in whom numerous circulating platelet fragments and megakaryocytic nuclei were present, with 50% blasts and 50% micromegakaryocytes in the marrow. The blasts expressed myeloid-associated antigens CD34, CD33, and CD13, whereas the micromegakaryocytes were positive for CD41, CD42b, and CD61. These findings suggested a myeloblastic transformation with a possible megakaryoblastic component. Cytogenetic analysis showed rearrangement of 3q26 in the form of t(2;3) (p13; q26), in addition to t(9; 22) (q34; q11). Dual-color flow cytometric analysis of DNA content of CD42b-positive cells showed that the micromegakaryocytes were predominantly 2N, indicating a maturation block before nuclear endoreplication and polyploidization. These findings confirmed a combined myeloblastic and megakaryoblastic transformation. It is concluded that dual-color flow cytometric DNA analysis is a useful method for the investigation of abnormal megakaryocytopoiesis in hematologic malignancies.

Background. The prognostic significance of tumor DNA ploidy in patients with cancer of the pancreas has not been defined because conflicting results have been reported. Methods. DNA content was measured in 56 ductal adenocarcinomas of the pancreas. DNA ploidy status was evaluated by flow cytometry in nuclei isolated from paraffin-embedded tumor tissues. Results. An abnormal DNA stemline was observed in 27 (48%) patients. The percentage of aneuploid tumors was significantly increased in tumors classified as Stage III/IV (53%) compared with those classified as Stage I (22%). A borderline significant association existed between DNA ploidy and radicality of surgery (P = 0.08). The median survival of patients with diploid carcinomas was 6.9 months (standard error, [+ or -] 0.9) in comparison to 4.5 [+ or -] 1.2 months for patients with aneuploid tumors (P = 0.013 by generalized Wilcoxon test; P = 0.023 by generalized Savage test). Although a selection bias cannot be excluded, survival of patients with a radical resection was longer than that of patients with a nonradical resection (P = 0.0008 and P = 0.0085, respectively). In addition, presence of distant metastasis (P = 0.0006 [Wilcoxon test] and P = 0.033 [Savage test]) could be identified as a prognostic factor. In a Cox regression model, results of surgery and DNA ploidy were independent prognostic variables. Conclusions. Because DNA ploidy has a significant impact on prognosis in pancreatic cancer, it should be used as a variable for stratified randomization of patients in therapeutic trials.

Background. The proliferative characteristics of acute nonlymphoblastic leukemia (ANLL) were studied in vivo, and data were correlated with response to chemotherapy and survival. Methods. Sixty-five patients with untreated ANLL and 15 patients with solid tumors and normal bone marrow (BM) received 250 mg/[m.sup.2] of bromodeoxyuridine (BUdR); bivariate flow cytometric (FCM) analysis then was used to measure cell BUdR incorporation and DNA content to obtain a complete set of kinetic parameters (i.e., BM BUdR-labeling index, DNA-synthesis time, potential doubling time [Tpot], and cell production rate). The percentage of blasts with positive results for proliferating cell nuclear antigen (PCNA) also was obtained by FCM analysis on the same BM samples, and these kinetic parameters were derived specifically for the ANLL proliferating compartment (growth fraction). Induction therapy, consisting of vincristine, arabinosylcytosine, and daunomycin, was administered subsequently to the patients with ANLL. Results. Overall ANLL proliferative activity was lower than normal myelopoiesis, and a short Tpot was found to be a favorable factor for achieving complete remission (CR), the duration of CR, and survival. When the growth fraction was considered, however, ANLL proliferative activity was higher and more like that of normal BM. The kinetic differences detected in the PCNA-positive cells of patients with CR and no response and those with CR and survival durations above and below the median values for the entire series were highly significant in univariate analysis and retained a strong independent prognostic value when multivariate analysis was performed. Conclusions. These data show the clinical feasibility of a detailed study of cell kinetics by means of new FCM-based techniques and reinforce the clinical value of pretreatment proliferative activity in ANLL. Cancer 1993; 71:2739-45.

Background. DNA ploidy and cell kinetics determined by flow-cytometric (FCM) analysis have provided relevant information on the natural history of breast cancer. These highly feasible and evaluable assays are particularly appropriate for small samples from fine needle aspirates (FNA) to complement cytologic information for diagnosis of breast lesions. Methods. DNA content and the percentage of cells in the cell cycle phases were evaluated on FCM DNA histograms from 281 consecutive FNA of benign and malignant breast lesions. Results. FCM analysis was possible in 93% of FNA. Qualitative concordance in DNA content between FNA and surgical specimens was observed in 82% of the patients. Of the 145 aneuploid FNA, 144 (99%) were histologically diagnosed as carcinomas. The presence of only diploid clones was observed in 115 FNA: rapidly proliferating (S+[G.sub.2]M > 12%) diploid lesions proved to be histologically malignant in 67% of patients, and slowly proliferating diploid lesions were histologically benign in 79% of patients. False-positive results of ploidy and cell kinetics were observed in 1 and 10 patients, respectively. Of the 163 cancers evaluable for cytology and FCM, 158 were positive at cytology. The presence of aneuploid or of diploid rapidly proliferating clones in cytologically false-negative aspirates resulted in the detection of four additional cancers. Conclusions. Flow cytometric analysis of ploidy and cell kinetics on FNA could be a valuable tool complementary to cytology for breast cancer diagnosis. Cancer 1993; 71:2522-7.

Background. In ovarian carcinoma, DNA ploidy measured by flow cytometric (FCM) analysis is an independent prognostic factor. However, limited sampling may underestimate the extent of ploidy variation (i.e., heterogeneity). Uncovering these hidden populations may explain poor outcomes in patients with ostensibly favorable ploidy patterns. The authors examined ploidy in a mean of 6.4 tumor samples per patient to better assess the occurrence of heterogeneity. Methods. FCM analysis was performed on multiple samples from 19 cases of advanced, serous ovarian carcinoma. Tumors were considered as heterogeneous by two definitions: (1) the presence of more than one ploidy pattern (e.g., diploidy and tetraploidy); and (2) the existence of DNA indices of sufficient variation so as to characterize two distinct populations of neoplastic cells. Results. With the first definition, 47% (9 of 19) of the tumors were heterogeneous, 37% (7 of 19) homogeneous-aneuploid, and 16% (3 of 19) homogeneous-diploid. These three groups showed no significant differences in histologic type, grade, patient age, stage, or survival. With the second definition, 27% (4 of 15) of the nondiploid cases were heterogeneous and 73% (11 of 15) were homogeneous. When these two groups were compared as to type, grade, patient age, and stage, no significant differences were demonstrated. However, the median survival of the patients with heterogeneous tumors was significantly longer (P < 0.05) than the survival of patients with homogeneous tumors. Conclusions. Ovarian carcinoma heterogeneity is high when multiple sites are assayed. This suggests conservative interpretation of ploidy when only a single sample is analyzed and examination of multiple samples when practicable. Cancer 1993; 71:2261-7.

Background. The clinical behavior of soft tissue vascular tumors is difficult to predict on histologic grounds alone. Methods. To assess the usefulness of DNA flow cytometry in predicting the biologic behavior of these tumors, the authors studied 51 soft tissue vascular tumors by DNA flow cytometry of paraffin-embedded tissue. Results. All 20 capillary hemangiomas, one epithelioid hemangioma, two spindle cell hemangioendotheliomas, and two benign hemangiopericytomas had a diploid DNA content. Of the 20 patients with angiosarcomas, their ages ranged from 9-80 years (mean, 44.0 years), and the mean follow-up period was 16.5 months. Ten tumors (including two postmastectomy angiosarcomas) were cutaneous, six were deep soft tissues, and four were from other sites. These tumors were classified histologically into three types: sinusoidal (n = 10), capillary (n = 7), and mixed type (n = 3). Thirteen of the angiosarcomas (65%) were diploid, and seven (35%) were aneuploid. Three patients with angiosarcoma had lung metastases at initial diagnosis, and seven had them later in the course. Nine patients died of their disease within a period of 18 months after the initial diagnosis. All five patients with malignant hemangiopericytoma died of their disease. Three of these lesions had a diploid DNA content, and two had an aneuploid content. Although it was found that all benign and intermediate-grade malignant tumors were diploid, there were no significant correlations among histologic type, DNA ploidy, and clinical outcome in angiosarcomas and malignant hemangiopericytomas. Conclusions. Based on this study, DNA analysis has limited value in predicting the biologic behavior of malignant vascular tumors of the soft tissue. Cancer 1993; 71:2233-41.

A large group of patients with node-positive breast cancer was divided into a training set (n = 851) and a validation set (n = 432) to demonstrate techniques for integrating steroid hormone receptor status, DNA flow cytometric findings, and other prognostic factors to predict patient survival. Multivariate analyses showed that estrogen receptor status, the number of involved axillary lymph nodes, patient age, S-phase fraction, progesterone receptor status, and tumor size were significant predictors of survival in patients with node-positive breast cancer. Techniques for optimizing and validating a cut point for a new prognostic factor and for examining alternative representations of prognostic factors were demonstrated. Prognostic indexes were created that could be used to identify patients with very good or very poor prognoses. Cancer 1993; 71:2157-62.

Background. The prognosis of patients with early ovarian cancer is good as compared with that of patients with advanced disease. However, there are no methods for predicting prognosis of early ovarian cancer, on which treatment decisions can be based. Methods. The prognostic significance of DNA flow cytometric and morphometric analysis was evaluated in 64 surgically treated patients with well-differentiated early-stage (International Federation of Gynecology and Obstetrics [FIGO] Stage IA, IB, IC, and IIA) epithelial ovarian cancer. The extent of the well-defined staging procedure was assessed strictly in every patient. Results. Only five patients died of recurrent ovarian cancer; all of these patients belonged to the inaccurately staged group. No significant relationship was found between clinicopathologic characteristics, such as menopausal status, FIGO stage, histologic cell type, and 5-year disease-free survival rate. Forty-two of the tumors had a mitotic activity index (MAI) of less than 30, and 43 of the tumors showed a volume percentage epithelium (VPE) of less than 65. Neither as a single parameter nor in combination did MAT and VPE correlate significantly with disease-free survival. Thirty-two tumors (50%) were DNA diploid, 15 were considered wide-CV-diploid, and 17 were aneuploid. Nearly 90% of the tumors showed DNA indices (DI) of 1.40 or less. The 5-year disease-free survival rate was significantly lower (61%) for patients with DI greater than 1.40 than for those with DI of 1.40 or less (96%) (P < 0.005). From the total number of five patients who died of their disease, three had tumors with DI in the tetraploid range. Conclusions. In conclusion, DNA flow cytometric analysis may indicate a group of high-risk patients; morphometric analysis does not add supplementary prognostic information. Cancer 1993; 71:787-95.

Background. Although the biologic characteristics of papillary cystic tumors (PCT) generally indicate a good prognosis, a malignant form has been reported. Methods. Twenty-two examples of PCT were examined to assess their malignant potential by histologic, flow cytometric, and immunohistochemical studies. Results. Three had hepatic, peritoneal, and/or lymph node metastases (metastasizing tumors [MT]); the other 19 PCT had no such metastatic features (nonmetastasizing tumors [NMT]). Venous invasion was detected in all three MT and two NMT. Compared with the 19 NMT, the MT had a higher nuclear grade and more prominent necrobiotic nests characterized by aggregates of cells with pyknotic nuclei and eosinophilic cytoplasm. In the flow cytometric analysis of cellular DNA content, one MT was aneuploid, and eight NMT, were diploid. Immunohistochemically, there was no difference between the MT and NMT. Conclusions. These results support the assumption that venous invasion, nuclear grade, and prominent necrobiotic nests are useful as histologic parameters to detect the malignant potential of PCT. Cancer 1993; 71: 82-92.

Background. Bladder washing specimens containing inflammatory or squamous cells have been difficult to accurately analyze with single-parameter DNA flow cytometric (FCM) methods. Methods. The anticytokeratin 18 antibody, CK5, was used in a multiparameter assay of 275 bladder washing and voided urine specimens to immunoselect only the bladder transitional cells for DNA analysis. Results. Flow cytometric detection of transitional cell carcinoma was increased by immunoselection of CK5-positive cells in specimens from patients with disease. Unfortunately, a similar increase in hyperdiploid cells in pathologically benign specimens was observed, which resulted in a false-positive rate of 45%. In some instances, multiparameter FCM assays with CK5 could detect aneuploid cell populations not clearly evident by single-parameter analysis. Conclusions. However, the results from this study of the hyperdiploid cell fraction showed that the increased sensitivity resulting from the use of CK5 was not clinically useful because of the decrease in specificity. Cancer 1992; 70:2879-84.

DNA ploidy of paraffin-embedded tissue from 11 patients with epithelioid sarcoma and 20 control specimens (granuloma annulare and sarcoid) was studied. DNA aneuploidy was found in 64% of the epithelioid sarcomas and in none of the control samples. Aneuploidy did not correlate with necrosis, vascular invasion, or mitotic rate. There was also no association of aneuploidy with adverse outcome. Cancer 1992; 70:2823-6.

Silvano Bosari, M.D.,*,[unkeyable] Arthur K. C. Lee, M.B.B.S.,* Steven R. Tahan, M.D.,[unkeyable] Mary Ann T. Figoni, B.S., M.T. (ASCP),[unkeyable] Brian D. Wiley, B.S.,* Gerald J. Heatley, M.S., Section and Mark L. Silverman, M.D.* Methods. The prognostic significance of flow cytometric analysis in patients with node-negative invasive breast carcinoma was evaluated in a retrospective series of 158 patients with a minimum follow-up study of 9 years. Results. The ploidy status could be assessed in 147 specimens (93%), and the proliferative phase or S-phase fraction (SPF) could be assessed in 136 tumors (86%); 70 tumors (48%) were diploid, 49 tumors (33%) were aneuploid, and 28 tumors (19%) were tetraploid. Ploidy status and SPF were correlated significantly with tumor size, histologic grade, nuclear grade, and mitotic rate. By itself, ploidy was not a statistically significant prognostic factor, although all of the patients with multiploid and hypertetraploid tumors had recurrence of disease. The SPF was related significantly to recurrence of disease (P = 0.04). However, when multivariate analysis of various histopathologic variables was performed, SPF ceased to be a significant prognostic determinant, whereas peritumoral lymphovascular invasion was the most important variable. The combination of tumor size and flow cytometric parameters permitted stratification into three groups with different prognoses at the 9-year follow-up review (P < 0.001). In the low-risk group (diploid tumors [is less than or equal to] 2 cm in diameter with a low SPF or small tetraploid tumors), the recurrence rate was 12%. In the intermediate-risk group (diploid tumors > 2 cm in diameter with a low SPF or aneuploid tumors with a low SPF), the recurrence rate was 21%. In the high- risk group (diploid or aneuploid tumors with a high SPF or large tetraploid tumors), the recurrence rate was 49%. The high-risk group status remained a significant variable in the Cox proportional hazards multivariate analysis model. Conclusions. These results indicate that flow cytometry in breast carcinoma contributes useful but limited prognostic information and stress the importance of using multiple prognostic factors to improve prognostication and optimize patient management. Cancer 1992; 70:1943-1950.

Masaharu Fukunaga, M.D.,* Takashi Nikaido, M.D.,* Tadakazu Shimoda, M.D.,* Shinichiro Ushigome, M.D.,* and Kazuhito Nakamori, M.D.[unkeyable] Background and Methods. Flow cytometric DNA analysis was performed on 30 cases of giant cell tumor (GCT) of bone with the use of paraffin-embedded sections. Results. According to the criteria of Huvos, they were classified histologically into three groups: Grade 1, 26 cases; Grade 2, 4 cases; and Grade 3, no cases. Among the Grade 1 cases, 21 were diploid and 5 were aneuploid. Of the four Grade 2 cases, three were diploid and one was aneuploid. Nine patients had local relapses. Among four patients with complications by lung metastases, two have remained well at 18 and 157 months with the metastases. The other two patients, who had Grade 1 DNA diploid GCT of the 11th thoracic spine, had malignant transformation (osteosarcoma) resulting from radiation therapy. In one patient, the primary lesion exposed to radation and the lung lesions were diploid, but in the other patient both were found to be aneuploid at autopsy. No significant differences of S-phase fraction were observed between two different grade groups. There was no significant correlation among DNA ploidy, histologic grade, and the presence or absence of lung metastases. Conclusions. Based on this study, the DNA analysis has a limitation in predicting the biologic behavior of GCT. Cancer 1992; 70:1886-1894.

Background. This study was designed to evaluate the efficacy of nuclear DNA content analysis in determining the prognosis of carcinoma of the pancreas. Methods. Resected and paraffin-embedded specimens from 72 patients with duct cell carcinoma of the pancreas were examined, and flow cytometry was used to explore the relationship between DNA ploidy and TNM classification or histologic grade. Results. DNA aneuploidy was found histologically in 42.9%, 56.8%, and 71.4% of Grade 1, 2, and 3 tumors, respectively. DNA ploidy showed a statistically significant correlation with T category and retroperitoneal invasion. The cumulative survival rate of patients with retroperitoneal invasion was shorter than that of those without retroperitoneal invasion. Among the patients with retroperitoneal invasion, those with DNA aneuploidy had a significantly shorter survival time than did those with DNA diploidy. Conclusions. The DNA ploidy pattern, in combination with the presence or absence of retroperitoneal invasion, appears to be useful in predicting the prognosis for duct cell adenocarcinoma of the pancreas. Cancer 1992; 70:1069-1074.

Clinical and pathologic data of 36 patients with transitional cell carcinoma of the bladder were investigated to determine the significance on patient survival of these factors: pathologic grade and stage; the immunohistochemistry of eight cell and tumor markers; nuclear DNA flow cytometric parameters; and patient smoking status. The bivariate and multivariate statistical analysis significantly correlated patient survival rates with the immunohistochemical expression of blood group, isoantigens A (P < 0.05), O(H) (P = 0.001), the oncogene-related protein ORP-p21 (P < 0.05), the pathologic grade and stage (P = 0.002), and the tumor DNA ploidy (P < 0.05). Smoking status correlated aneuploidy (P < 0.05) and tumor expression of ORP-p21 (P < 0.05) with the patient survival rate. Despite the relatively small number of patients in this study, the results suggest that the clinicopathologic variables are significant factors in survival of bladder cancer. Cancer 1992; 70:799-807.