Reproducibility and efficiency testing of two sampling methods for image analysis in cytology.

OBJECTIVE: Quantitative nuclear DNA analyses and nuclear morphometry have become a widely used tool in pathology. Nevertheless, studies done by different observers utilizing the same methods very often have shown different results. One of the reasons could be that different sampling methods were used. In this study two selection methods for image analyses (the consecutive selection method and the random selection method) were tested for reproducibility and efficiency. STUDY DESIGN: Using an interactive imaging analysis system, 10 cases (5 breast adenocarcinomas, 3 prostatic cancers, 1 sarcoma, 2 prostatic hyperplasias) were measured twice by using each sampling method. Median nuclear area and entropy (i.e., a statistical descriptor of DNA histograms) were the parameters calculated. RESULTS: The results showed that using the consecutive selection method, the number of individual nuclei in a cytologic sample that should be measured to obtain reproducible results was 150 for nuclear area and 110 for entropy. The corresponding figures using random systematic selection method were 80 for assessment of nuclear area and 100 for assessment of entropy. CONCLUSION: Both methods showed high reproducibility and a good correlation with each other. The mathematical idea of "plateau values," which was used to estimate the number of measurements needed, was considered a good contribution to increasing the efficiency of both methods.

p53, Ki-ras, and DNA ploidy in human pancreatic ductal adenocarcinomas.

Ki-ras mutations and DNA aneuploidy are common findings in human pancreatic ductal adenocarcinomas. An altered p53 tumor-suppressor gene has been suggested to cooperate with activated Ki-ras in malignant cellular transformation and could enhance genomic instability. We have investigated a panel of well-documented pancreatic carcinomas with defined ploidy and Ki-ras mutations for the presence and pattern of genetic alterations of the p53 gene, their coincidence with Ki-ras point mutations, and their correlation with DNA ploidy, tumor pathology, and clinical course. DNA was isolated from formalin-fixed and paraffin-embedded tumor tissue and polymerase-chain-reaction-amplified fragments of the p53 gene exons 5 to 9 were screened by the single-strand conformation polymorphism method. The positive cases were further examined for mutations by direct sequencing. Twenty-nine of seventy-one (41%) tumors showed mutations of the p53 gene, however, five tumors carried two mutations resulting in a total of 34/71 (48%) genetic alterations of the p53 gene. The majority were missense point mutations and distributed primarily within the evolutionary conserved domains (62%). Ten of Thirty-four (29%) affected the hotspot codons 248, 273, and 282, respectively, and 21/34 (62%) of the p53 gene mutations clustered on exons 7 and 8. Transitions (71%) predominated over transversions (15%), deletions were identified in 7/34 (21%) tumors. One third of the carcinomas showed both Ki-ras codon 12 and p53 gene mutations. p53 mutations correlated with distant metastasis (p < 0.05) and survival (p < 0.05). DNA triploidy was associated with a mutated Ki-ras gene (p < 0.05) as well as with double mutations of c-Ki-ras and p53 (p < 0.05). Unlike most other malignant tumors pancreatic ductal adenocarcinomas exhibit a significantly higher incidence of c-Ki-ras than p53 gene mutations. However, like other neoplasms p53 gene mutations seem to be associated with a metastatic phenotype possibly acquired during tumor progression.

Morphometric characteristics in adenocarcinoma of the pancreas and chronic pancreatitis.

BACKGROUND: Morphometric analysis whereby size and form of cellular nuclei are transformed into quantities has previously been shown to be a valuable adjunct to the histopathologic differential diagnosis between chronic pancreatitis and pancreatic carcinoma. The present study aims to assess the clinical value of morphometry performed on cytologic material from benign and malignant pancreatic lesions. METHODS: Cytologic specimens from 100 patients with the diagnosis of pancreatic carcinoma and 15 patients with chronic pancreatitis were evaluated by interactive morphometry using a digital image analyzer system. RESULTS: There were significant differences (p < 0.001) for all morphometric variables between the malignant and benign groups (mean area p 50, 135.41 microns 2 versus 69.66 microns 2; anisokaryosis, 0.74 versus 0.41; and polymorphism, 0.13 versus 0.09). CONCLUSIONS: Morphometry may be used as a complementary tool in the cytologic diagnosis of pancreatic carcinoma.

A morphometric filter improves the diagnostic value of morphometric analyses of frozen histopathological sections from mammary tumours.

Frozen sections of 202 consecutive breast tumour cases were analyzed by morphometric quantitation of nuclear features. Nuclei were selected at random. Conventional light microscope examination of the paraffin-embedded specimens revealed 144 cases of cancer and 56 benign tumours. Using multivariate discriminant analysis of morphometric features, all but two of the benign cases and 79% of the malignant tumours were correctly classified. When a morphometrically based dynamic filter set to exclude 'non-diagnostic' nuclei was used, the correctly classified malignant cases rose to 86% Morphometry is a fast, reproducible and efficient method that can be used in conjunction with the histomorphological diagnosis of mammary frozen sections. The combination of systematic sampling and an objective dynamic filter may be a powerful approach to quantitative analyses of tumours from other sites. However, it is also likely that efficiency can be improved by combining nuclear morphometric features with structural, histochemical and molecular biological data. The combination of traditional histomorphological examination with quantitative information may well increase the diagnostic accuracy in individual patients.

Morphometry and prognosis in cancer of the pancreatic head.

The histological slides of 39 cases of cancer of the pancreatic head were analysed using an interactive image analyser system. Some 14 cases were classified as periampullary, 25 as ductal pancreatic cancer. All cases had undergone radical tumor resection according to Whipple's procedure. Morphometric data, tumor size and metastatic nodal involvement were correlated to prognosis. Univariate statistical analysis showed that the classical differentiation between ductal and periampullary cancer was a weaker prognosticator than morphometric variables. In fact, multivariant statistical analysis showed that the morphometric variable irregularity was the best prognosticator (p = 0.0001). No other variable added significant prognostic information. Irregularity is a newly developed variable describing the nuclear shape corrected for roundness. We conclude that morphometry can be of essential prognostic information for the clinician in cancer of the pancreatic head.

Discrimination of pancreatic adenocarcinomas from chronic pancreatitis by morphometric analysis.

Chronic pancreatitis and pancreatic ductal adenocarcinoma show similar gross and microscopic anatomical features. Morphological examination alone is not always sufficient in diagnostic practice to make the clinically important discrimination between these two entities. Cases of pancreatic tumors were analysed in a morphometric study to evaluate the discriminatory value of nuclear and nucleolar features. Histologic sections of pancreas from 18 cases of chronic pancreatitis and 33 cases of ductal adenocarcinoma were included either into a learning or a test set. A multivariable discriminatory rule was derived from the learning set of 23 cases including nuclear polymorphism and nucleolar density. When applied to the test set, all 28 cases of adenocarcinomas and chronic pancreatitis were correctly classified. Distributional features describing nucleolar density and variation in nuclear size and shape were the most efficient discriminatory variables. Morphometry is shown to be a simple and fast cell analytical method which can support clinical judgement in distinguishing between chronic pancreatitis and pancreatic ductal adenocarcinoma.

Surface immunoglobulin density in the differential diagnosis of B-cell chronic lymphocytic leukemia and leukemic immunocytoma.

Using flow cytometry peripheral blood samples of 37 consecutive patients with B-cell chronic lymphocytic leukemia (B-CLL) and 17 consecutive patients with leukemic immunocytoma (IC) were studied in order to determine quantitative differences in the surface immunoglobulin (slg) density. In 8/37 (21.6%) cases of B-CLL and 1/17 (5.9%) cases of IC slg staining remained in the control level. Analysis of slg-positive cases demonstrated a close association between the amount of slg and diagnosis: per case the mean calculated fluorescence intensity for IC lymphocytes was 209.7 arbitrary linear intensity units (IU) (median: 156.4, standard error of the mean (SEM): 53.7) and for B-CLL lymphocytes 10.8 IU (median: 7.3, SEM: 1.1; p less than 0.0001). Altogether, 94.6% of all B-CLL patients and 76.5% of all IC patients were correctly classified when a cut-off point was fixed at a mean fluorescence intensity value of 20.0 IU. The percentage of leukemic cells as characterized by CD19 and HLA-DR reactivity was significantly lower in cases of IC (p less than 0.03 and p less than 0.01, respectively). In both entities disease progression occurred more frequently in advanced stages (II-IV) according to the Rai classification (p less than 0.01). In progressive disease rather than in stable disease circulating T lymphocytes were shown to express decreased amounts of surface CD3 antigen (p less than 0.02). We conclude that the quantitative assessment of surface antigens in addition to their qualitative characterization provides accurate information. In particular, the diagnostic discrimination between B-CLL and IC may be improved by determining the lymphocytes' slg amount.

Quantitative nuclear DNA content in fine needle aspirates of pancreatic cancer.

The quantitative nuclear DNA content of nuclei from fine needle aspirations from 70 patients with pancreatic cancer was measured using an image analyser system. Retrospective analysis of patients indicated that cases with tumour stemlines in the diploid region had the best chance for radical surgery (p less than 0.006) and the best probability of survival (p less than 0.0004). The prognosis for patients with tetraploid stemlines was intermediate and was poorest for patients with no stemlines in the diploid-tetraploid region. From those parameters assessed postoperatively, only the tumour stage added appreciable information on prognosis to the preoperative quantitative DNA content obtainable. Therefore, in patients with pancreatic cancer the quantitative DNA content should be taken into account in planning treatment and assessing prognosis. Furthermore, the quantitative DNA content may have a major role in stratification for further treatment trials.

Nuclear DNA distribution pattern of the parenchymal cells in adenocarcinomas of the pancreas and in chronic pancreatitis. A study of archival specimens using both image and flow cytometry.

The nuclear DNA distribution pattern of the neoplastic parenchymal cells of 100 conventionally formalin-fixed and paraffin-embedded specimens from pancreatic adenocarcinomas and from 8 specimens of chronic pancreatitis was assessed by means of image cytometry. All material originated from pancreatic restrictions. Evaluable DNA histograms could be obtained for 77 carcinomas, and clinical data were available for 71 of these. In these 71 specimens, the nuclear DNA ploidy pattern was also investigated by means of flow cytometry. In 76 of the 77 cases, the image-cytometric DNA ploidy pattern obtained showed a "nondiploid" distribution with modal values as high as 8.5 c. In 21 cases, the neoplastic cells showed modal values in the "triploid" region. The analogous 71 flow-cytometric DNA histograms could only be evaluated in 50 cases because of excessively high amounts of background and/or excessively broad peaks. In 47 cases, the nuclear DNA histogram was nondiploid according to both techniques. The patients with carcinomas whose cell nuclei showed a triploid DNA distribution showed a significantly shorter survival time than those with tumor cell populations of nontriploid DNA distribution patterns. In the 8 specimens of chronic pancreatitis, the parenchymal cells were all equipped with nuclei showing diploid DNA distribution patterns.

Prognostic significance of tumor cell morphometry, histopathology, and clinical parameters in advanced ovarian carcinoma.

The prognostic value of clinicopathologic data and tumor cell morphometry was assessed in 49 consecutive patients with advanced epithelial ovarian carcinoma (Stage III-IV as defined by the International Federation of Gynaecologists and Obstetricians [FIGO]). Single-variant analysis proved that nuclear area at the 10th, 50th, and 90th percentiles and nuclear density were the strongest prognostic factors. To assess difference in prognostic influence between morphometric and clinicopathologic factors, Cox regression analysis was carried out and revealed nuclear density and age as prognostically significant (p less than 0.0003 and p less than 0.0004, respectively). Choice of chemotherapy, FIGO stage, histologic grade, and mitotic activity index (MAI) were not of additional prognostic importance. Objective measurement of morphometric factors is simple, highly reproducible, and adequate for routine work and promises to be of clinical value in therapeutic decision-making in advanced ovarian carcinoma.

Methodological aspects of DNA image cytometry in formalin-fixed paraffin-embedded material from pancreatic adenocarcinoma.

Deparaffinized and disintegrated material from conventionally formalin-fixed and paraffin-embedded surgical specimens of 100 cases of ductal adenocarcinoma of the pancreas was Feulgen-stained, and the cytochemical DNA distribution patterns of at least 100 single tumour cells and 50 "control" cells (fibrocytes) were assessed by means of image cytometry (ICM). In 77 cases a sufficient number of neoplastic cells could be obtained for these DNA assessments. The fairly high number (23) of cases that had to be excluded due to too small amounts of disintegrated cells or cell nuclei may be explained by the high content of connective tissue stroma in these pancreatic adenocarcinomas. The tumour cell nuclei in 76 of these 77 cases showed cytochemically a clear-cut "non-diploid" DNA distribution pattern. This observation reflects the well-known highly malignant growth potential of this carcinoma. Despite the fact that about 1/4 of the tumours had to be excluded, the main result of our methodological study is, after all that conventionally formalin-fixed paraffin-embedded specimens of most pancreatic adenocarcinomas can be successfully used for the deparaffinization-disintegration procedure preceding the nuclear DNA assessments by means of ICM. Additional studies are, however, required to obtain the diagnostic and prognostic impact of the results of such cytochemical analyses of the DNA distribution pattern in adenocarcinomas of the pancreas.

The value of morphometry to predict chemotherapy response in advanced ovarian cancer.

We evaluated the correlation of morphometric parameters to chemosensitivity. 63 patients with palliatively operated advanced epithelial ovarian cancer were investigated concerning their response to chemotherapy. Using multiple linear stepwise discriminant analysis of five morphometrical parameters 13 out of 17 responders and 37 out of 46 non-responders were correctly classified (76.5% sensitivity, 80.4% specificity, 79.4% efficiency). The five parameters were: nuclear area at the 10th percentile, standard deviation of the nuclear area, median value of the nuclear ovality, number of cells per area and mitotic activity index. To assess the performance of the discrimination formula when applied to new cases, the "leave one out" method was used. For our data the following corrected classification rates were obtained: 58.8% responders (10/17), 76.1% non-responders (35/46) (efficiency 71.4%). Morphometry is a fast and reproducible method to objectively record a tumor's morphology. Our results indicate that there is a correlation between morphometrical features, response to chemotherapy and survival, which should be tested in further studies.