Search

Flow cytometry was used to study the incidence of aneuploidy and to determine the significance of multiple sampling from colorectal tumors. DNA ploidy pattern has been proposed as a supplementary prognostic marker, but discrepancies in findings are major. DNA clonal heterogeneity, defined as two or more DNA aneuploid stemlines in the same tumor, is well established. However, most studies have been based on only one biopsy from each tumor. In our study multiple biopsies were taken from 163 patients (88 males and 75 females) electively operated for colorectal cancer. Tumor cells were harvested by fine needle aspiration from fresh frozen biopsies sampled at different sites of each tumor. DNA aneuploidy was detected in tumors from 145 patients (89%), and 18 patients (11%) had a solitary DNA diploid cell population. In a 79 month follow-up period 105 patients had died. Statistical analysis showed that distinction between diploidy and aneuploidy did not predict survival. However, grouping subpopulations into DNA diploid plus near diploid (DNA index (DI) 0. 97-1.15), DNA aneuploid with all aneuploid subpopulations in the interval 1.15-2.06, and DNA aneuploid with subpopulations with DI < 0.97 and/or DI > 2.06, showed a significant difference in survival in a Cox multivariate analysis including Dukes' stage P = 0.049 comparing the second group to the first and P = 0.01 comparing the third group to the first. In 21 (13%) patients only one subpopulation was found, 57 (35%) had two, 44 (27%) had three, and 41 (25%) had four or more different subpopulations. The association of DNA ploidy to survival is shown to be dependent on the number of biopsies analysed., (Copyright 1999 Wiley-Liss, Inc.)

Objective: To form a methodologic basis for DNA analysis of ductal carcinoma in situ (DCIS) and invasive carcinoma (IC) of the breast, including very small lesions, by comparison of flow cytometric (FCM) and image cytometric (ICM) methods for DNA quantitation., Study Design: The material consisted of 41 DCIS lesions and 26 ICs. FCM DNA analysis of unfixed, frozen samples were compared to (1) FCM of formalin-fixed, paraffin-embedded tissue; (2) ICM of imprints; and (3) ICM of paraffin-embedded tissue sections., Results: FCM of unfixed tissue showed higher DNA measurement precision and a higher number of DNA nondiploid clones as compared to the other three methods. For the classification of DNA diploid/nondiploid cases, high concordance rates were found between the methods. Discordant cases were predominantly DNA neardiploid by FCM of unfixed tissue but DNA diploid by the other methods. The reproducibility of the DNA index (DI) was best in the interval 1.2 < DI < or = 2.2; it was 74% for FCM of fixed tissue and 79% for ICM of imprints. Clones with DI > 3 were found almost exclusively by ICM of imprints. For ICM of tissue sections, DI could not be reliably estimated. By ICM, contrary to FCM, a combined DNA diploid and nondiploid pattern was found frequently., Conclusion: Each of the methods has its own advantages and limitations. If possible, FCM should be combined with ICM. FCM of unfixed tissue is superior to the other methods with respect to precise DI estimation. Alternatively, FCM of fixed tissue and ICM of imprints may both give a reliable estimate of DI. ICM of tissue sections can discriminate DNA diploid from nondiploid clones, except for neardiploid subpopulations, and permits the analysis of very small lesions.

Cellular DNA contents of consecutive hydatidiform moles were analysed by flow cytometry of unfixed samples, freed of maternal tissue and treated with detergent and trypsin, using two external controls. DNA-diploidy was found in 62 moles, DNA-triploidy in 42 and DNA-tetraploidy in one. Two non-molar placentas and two invasive moles were found to be DNA-diploid. In 28 of 42 DNA-triploid moles, trophoblastic hyperplasia was not noted, making the discrimination between vesicular abortions with or without trophoblastic hyperplasia unwarranted on genetic grounds. A good fit was obtained to a distribution with one G0-1 peak in 49 cases, whereas 60 cases showed two peaks. Technical artifacts, created by flow cytometry, were excluded. Differences in the nuclear accessibility for the dye and degradation of DNA are unlikely causes of this heterogeneity. The existence, in vivo, of two or more cell populations with different DNA contents is a reasonable explanation. The criteria for persistent trophoblastic disease were standardised. Eight cases of persistent trophoblastic disease were observed among the 62 DNA-diploid moles (13%); no case was observed after DNA-triploid or DNA-tetraploid moles. Combining these data, with those of other studies where ploidy was determined using optimal techniques [10,13], allows the calculation of 95%-confidence-limits for the risk of persistent trophoblastic disease after a triploid mole to 0-2.7%.

Flow cytometric DNA analysis was performed on unfixed frozen samples from 56 breast cancer patients. From each patient, two samples were analyzed. The only difference in the handling of the paired samples was the mechanical disaggregation of one sample by fine-needle aspiration compared to an automated mechanical disaggregation method (Medimachine) of the other sample. With the two methods for tissue disaggregation, the same resolution of the DNA histograms was obtained, indicated by median coefficients of variation (CV) of 1.5% for the DNA diploid G1 peaks. Also, the frequencies of DNA diploid and aneuploid cases as well as the fractions of DNA aneuploid cells were comparable. This indicates that the two methods did not differ in ability to detect DNA aneuploid tumor clones. Automated mechanical disaggregation resulted in DNA histograms with significantly less debris and with lower S-phase fractions. In practice, the procedure of automated mechanical disaggregation was rapid, easy, and safer because of minimal handling of the unfixed tissue compared to the fine-needle aspiration.

Flow cytometric (FCM) DNA analysis of 41 clinical cases of ductal carcinoma in situ (DCIS) of the breast was performed on fine-needle aspirates of unfixed, frozen tissue. Based on analysis of a single tissue sample, abnormal DNA content was found in 35 (85%) of the cases, and 15 (37%) disclosed heterogeneity. If five or more samples were analyzed, then heterogeneity was found in eight of ten cases. No conclusive correlation could be found between DNA index (DI) and histopathology. Our results on DCIS demonstrate a very high concordance with invasive breast cancer with respect to the frequency of DNA nondiploid cases, distribution of DI, occurrence of multiple clones, and S-phase fraction. This indicates that major genetic alterations and DNA heterogeneity are early events in carcinogenesis that are already established at the preinvasive stage.

The prognostic significance of DNA index (DI), S-phase fraction, and heterogeneity determined by flow cytometric DNA analysis was assessed in a prospective study of 249 newly diagnosed transitional cell carcinomas of the bladder. The median observation time was 4.8 years. A total of 456 subpopulations were detected. The S-phases could be estimated in 299 subpopulations. A DI > 1.25 or an S-phase above 9.7% were strongly correlated to invasiveness. One hundred and ten patients were treated with transurethral resection (TUR). Relapse-free survival could not be predicted by the DNA-derived parameters. Univariate analysis of survival showed prognostic significance of diploidy (0.98 < DI < or = 1.02, P = 0.02), hypotetraploidy (1.50 < DI < or = 1.96, P = 0.002), and S-phase size (P = 0.008). Multivariate analysis pointed to the T-classification (RR = 1.64) and hypotetraploidy (RR = 1.57) as prognostic parameters for survival of TUR-treated patients. One hundred and thirty-nine patients received radiotherapy (RT). A significantly better response was found for tumors with a subpopulation with a hypertetraploid DNA content (DI > 2.04, P = 0.05), and a significantly worse response for subpopulations with a maximum S-phase > 24.5% (P = 0.04). T-classification and histological grade had no predictive value. A logistic regression analysis indicated an estimated probability of response to RT of 77% for tumors with a DI > 2.04 and an S-phase < 24.5%, whereas tumors with a DI < 2.04 and an S-phase > 24.5% had only a 28% probability of response. The poor response to RT, predicted by an S-phase > 24.5%, translated into a poor survival, whereas the better treatment response found for patients with a DI > 2.04 did not result in a longer survival. Multivariate analysis pointed to S-phase (RR = 1.70), T-classification (RR = 1.60), and grade (RR = 0.65) as independent prognostic parameters for survival of RT-treated patients.

In 42 tumour samples of human neuroblastoma, histological classification by differentiation (Shimada) was significantly correlated with strong positivity for neuron-specific enolase (NSE) and inversely correlated with rosette formation. Most ganglioneuroblastomas were positive for S-100 protein and reacted strongly with NSE antibody. Histological signs of high proliferative activity included intermediate or high mitosis-karyorrhexis index, necrosis and lack of calcification, which were significantly correlated with each other. Flow cytometric DNA analysis demonstrated that 88% of the tumour samples had DNA aneuploid stem lines. High S phase fraction (greater than or equal to 0.20) was significantly correlated with necrosis and lack of calcification. Univariate analysis of prognosis for 26 patients whose tumour samples were obtained before adjuvant treatment showed that five factors were significantly related to a better outcome: early stage of the disease (stages I, II, IV-S), S phase fraction less than 0.20, favourable Shimada histology, positivity for S-100 protein, and strong positivity for NSE. In multivariate analysis, only S phase fraction or stage of disease remained significantly associated with prognosis. DNA index did not correlate with prognosis in this study.

Washless methods for double staining of nuclear antigen and DNA in unfixed nuclei were compared with established methods for staining of fixed cells. The methods were tested on phytohemagglutinin (PHA)-stimulated normal human blood lymphocytes for the double staining of 1) Ki-67 antigen and DNA and 2) bromodeoxyuridine (BrdUrd) and DNA, in continuously BrdUrd-labeled cells. With respect to the discrimination between antigen-positive and -negative subpopulations, there was no statistically significant differences between the results from direct (Ki-67) or indirect (Ki-67 or BrdUrd) washless staining of unfixed nuclei and the results from staining of fixed cells. Washless staining of unfixed nuclei was found to be rapid and simple and resulted in greater precision of the DNA analysis and in less aggregation and loss of cells.

Establishing flow cytometric DNA analysis as a clinical routine procedure requires adequate and proven guidelines, by which the data can be obtained and interpreted to directly influence management of the individual patient with a specific neoplasm. The present paper is intended as a contribution to such guidelines, of which only fragments are available today. We have previously described a system of methods, designed for routine flow cytometric DNA analysis. In the present status report our experience, based on approximately 18,000 samples (clinical and experimental) is summarised. Sample acquisition with fine-needle aspiration, storage at -80 degrees C, internal standardization by chicken (CRBC) and trout red blood cells (TRBC), staining with propidium iodide (PI), and analysis in the flow cytometer is recapitulated, with emphasis on previously unpublished aspects. The method of statistical analysis which has an integrating role is described in some detail. A lack of linearity between channel number and DNA content was determined experimentally, and the coefficient of variation (CV) was found to decrease with increasing channel number. The corrections in the algorithm of deconvolution made necessary by these findings are fundamental for estimating the end results. The zero point adjustment and procedures for changing from one batch of standards to another are described. A systematic approach to interpretation of DNA histograms is attempted and illustrated by data from clinical specimens of malignant lymphoma, breast cancer, small cell lung cancer, cancer of the oral cavity, and bladder cancer. Some problems are still unsolved and visual inspection is required to determine if the quality of the individual histogram is satisfactory. Inspection of the fluorescence/light scatter dot-plot provides additional information for the recognition of artifacts. The results stress that good quality DNA histograms with as small CVs as possible are important for interpretation of the data. It is essential that statistical methods are employed to extract the key end-point results. These are the number of subpopulations and their relative representation, and for each subpopulation the DNA index (DI) and the fractions of cells in the cell cycle phases. For the DNA data to have any rationally based impact on clinical decision making, it must be demonstrated that they have an independent prognostic value. Strategies for final evaluation are discussed. Multicenter trials on fresh material, to accrue quickly the number of patients necessary for firm conclusions, are suggested.

Twelve methods for analysing FCM-histograms were compared using the same set of data. Some of the histograms that were analysed were simulated by computer and some were taken from experiments. Simulated data were generated assuming synchronously growing cell populations and (i) measurement coefficients of variation (CV) from 2 to 16%; (ii) constant measurement CV or VC's increasing from G1 to G2 phase, and (iii) varying fractions of cells in each phase. Simulated data were also generated assuming synchronous cell populations in which a block in early S phase was applied and released. DNA histograms were measured for L-929 cells at various times after mitotic selection. Labelling indices were also measured for these cells at the same time. The fractions of cells in the G1, S, and (G2 + M) phases were calculated by each analytical method and compared with the actual fractions used for simulation, or in case of experimental data, with autoradiographic results. Generally, all methods yielded reasonably accurate fractions of cells in each phase with relative errors in the range of 10-20%. However, most methods tended to overestimate G1 fractions and underestimate S fractions. In addition, variations in the shape of the S phase distribution often caused considerable errors. Phase fractions were also calculated for histograms of kinetically perturbed populations, simulated as well as experimental. The errors were only slightly larger than for histograms from asynchronously growing cell populations.

A simple procedure for long-term storage of cells for flow cytometric DNA analysis was developed and tested. The cells were stored as single cells or fine-needle aspirates suspended in a citrate buffer with dimethylsulfoxide (DMSO), or as small blocks of tissue from solid tumors. The cells were stored for up to one year by freezing at -80 degrees C. Statistical analysis of the results showed no change in the fractions of cells in the cell cycle phases as determined by deconvolution of the DNA-histograms. It was found that in addition to the intrinsic sample variation from the parameter estimation by deconvolution, there was significant intraday and interday variation. Hence the most accurate results are obtained if different aliquots of a sample are measured on different days rather than on the same day. Use of the storage method thus has the potential of increasing the accuracy of the analysis. The storage method makes sample collection independent of immediate subsequent analysis. This has enabled us to perform large internally controlled experiments, involving more samples than can be analyzed in one day, to examine tumor samples from different hospitals and to utilize fully the capacity of our flow cytometer. The method was a prerequisite for developing an accurate standardization procedure for DNA content determination.

Determination of nuclear DNA content by flow cytometry requires comparison with a reference standard. The use of external standards such as lymphocytes or granulocytes is time-consuming and inaccurate. Chicken red blood cells (CRBC) have a DNA content of 35% of the human diploid value and have been widely used as internal standard. The ratio calculated on the basis of the peak channel numbers of the standard and the sample and used to indicate the DNA content (DNA ratio) is, however, very sensitive to changes in the zero level adjustment of the flow cytometer. If two internal standards are used the DNA ratio becomes independent of the zero level. Rainbow trout red blood cells (TRBC) have a DNA content of 80% of human diploid cells. A mixture of CRBC and TRBC was prepared and stored in small aliquots at -80 degrees C. This mixture was added to the sample before staining. The day-to-day variation of the DNA ratio obtained by use of the two standards was smaller than that obtained by CRBC alone. The possibility of sex related differences in DNA content of CRBC and TRBC was examined. The results indicated that a new batch of standards should be tested against the old batch to avoid the introduction of a systematic error.

A new modification of our detergent technique for the preparation of nuclei for flow cytometric DNA analysis is described. The attainment of low coefficients of variation of the peaks and of quantitative staining of nuclei from different tissues was a problem with the original method. This was solved in the new modification by trypsinization of the unfixed nuclei. The nuclei were stabilized by spermine. A simple procedure for long-term storage of samples at -80 degrees C was integrated into the method. The fluorescence of the nuclei was stable for at least 3 hours after staining. Light exposure protection of the samples was essential. No cell loss was caused by storage or staining. The method was successfully applied on samples including: (a) Normal tissues--human lymphocytes, granulocytes and spleen. Mouse lymphocytes, bone marrow, spleen, liver, kidney and thymus. (b) Human neoplasms--lung cancer, breast cancer, lymphoma, leukemia, bladder cancer and cancer of the oral cavity. (c) Human tumors in nude mice--breast cancer, lung cancer, melanoma and colon cancer. (d) Mouse ascites tumors--JB-1, L 1210, Ehrlich and P 383. It therefore seems well suited as a routine clinical procedure.

A new method is described for flow cytometric cell cycle analysis of normal and psoriatic human epidermis, based on non-enzymatic tissue disaggregation. The epidermis was isolated by treatment with acetic acid and stored by freezing. After thawing, the epidermis was disintegrated into a nuclear suspension by 3 steps: incubation with dithiotreitol, whirling in a buffer (pH 7.4) with the non-ionic detergent Nonidet P40, EGTA, RNase and spermine, and whirling after addition of citric acid to a final concentration of 1% (pH 2.4). The suspension was stained with propidium iodide and filtered before flow cytometry. The yield of suspended nuclei was approximately 70% of the original number of cells in the tissue. The detergent/citric acid method was found to be preferable to an ultrasonication method previously used on human epidermis. All cell cycle and cell maturation stages were represented in the detergent/citric acid suspension, in contrast to the selection of immature G1, S and G2 stages with enzymatic methods. In the analysis of psoriatic epidermis inadequately matured (parakeratotic) cells were present in the suspension and had to be discriminated by gating on light scattering intensity, as they were not susceptible to lysis and did not stain properly. The fraction of S phase nuclei was on average 1.9% in normal and 7.7% in psoriatic epidermis, thus confirming the results of other investigators using enzymes. The presence of mitotic figures in the suspension was demonstrated by flow sorting. In this way the mitotic fraction was estimated to 0.06% in normal and 0.22% in psoriatic epidermis, confirming histological data of other investigators.

Single-cell DNA measurements obtained using flow cytometry have previously been used as a diagnostic tool for various malignant diseases. We used this technique to characterize the dermal infiltrates of 14 patients with psoriatic skin lesions and 22 patients with various benign skin disorders. The DNA histograms of all of the patients exhibited an increased number of cells showing propidium-iodide fluorescence in the hyperdiploid region as compared to those seen in control DNA histograms of normal blood mononuclear cells. The calculated DNA indices revealed one hyperdiploid (G/G) peak in 17 cases and two hyperdiploid (G/G) peaks in 13 cases. The results suggest that the dermal lymphocytic infiltrate consists of (1) lymphocytes in the cell-division cycle, (2) non-cycling lymphocytes (G) and (3) activated lymphocytes that may either remain in a non-cycling state or enter the cell cycle. [ABSTRACT FROM AUTHOR]