Different opinions on classification of DNA histograms produced from paraffin-embedded tissue

Fluorescent labeling of DNA

A well-defined method for staining cellular DNA especially for cell cycle determination is provided. Emphasis is placed on utilizing DNA content and cell sizing measurement to further define cell populations.

Flow cytometry in the diagnosis of peritoneal carcinomatosis

OBJECTIVES

Peritoneal carcinomatosis is the second major cause of ascites. Because of its frequency and poor prognosis, it is important to establish an accurate diagnosis. The aim of this study was to analyze the use of a DNA index, detemined by flow cytometry in the differential diagnosis of ascites, and to compare it to the cytopathological examination.

METHODS

A prospective analysis was carried out on 67 patients (39 female, 28 male; mean age, 53+/-14 yr [range, 5-82]) with ascites of various etiologies. Peritoneal carcinomatosis was detected in 21 patients, whereas in 46 the ascites was of noncarcinomatosis origin.

RESULTS

The sensitivity of the cytopathological examination for the diagnosis of peritoneal carcinomatosis was 42.9%, and the specificity was 100%. The mean DNA index determined by flow cytometry was similar for peritoneal carcinomatosis and noncarcinomatosis patients, being 1.28 versus 1.01, respectively, in the preparations without control lymphocytes and 1.28 versus 1.04, respectively, when control lymphocytes were added. The sensitivity of DNA index cytometry was 57.1% and specificity, 93.5%. The combined use of the DNA index and cytopathological examination did not show an advantage over the use of any of the tests individually, although the DNA index was able to detect half of the cases of peritoneal carcinomatosis in which cytopathological examination was negative. Although the sensitivity was higher when the parameters were associated, the DNA index did not offer a statistically significant advantage over the use of cytopathological examination alone, which in turn had higher specificity.

CONCLUSION

The DNA index presented lower sensitivity for the diagnosis of peritoneal carcinomatosis when used alone, showing no advantage over conventional cytopathological examination. However, the DNA index was able to detect 50.0% of peritoneal carcinomatosis cases whose conventional cytopathological examinations were negative, and could be valuable in these situations.

The role of radical surgery followed by adjuvant therapy for high-risk early-stage cervical carcinoma patients with pelvic lymph node metastasis

OBJECTIVES

To identify a subgroup of high-risk node-positive patients in early-stage cervical cancers and to determine the role of radical hysterectomy followed by adjuvant therapy to these patients.

STUDY DESIGN

We conducted a retrospective review of 482 surgically-treated patients of clinical stage Ib and IIa cervical carcinoma from July 1986 to December 1994 at Kaohsiung Chang Gung Memorial Hospital. Of these, 96 patients had pelvic lymph node metastases. Clinicopathological variables, including the level of pretreatment squamous cell carcinoma antigen (SCC-Ag), DNA flow cytometry analysis, and the use of different adjuvant therapies were studied.

RESULTS

Disease-free survival was significantly worse among patients with S-phase fraction greater than 20% and pretreatment SCC-Ag level above 5 ng/ml. Utilizing these significant variables, we identified two distinct risk groups. Those patients without any of the risk variables were categorized as the low-risk group. Those patients with either one or both risk variables were categorized as the high-risk group. Five-year disease-free survival rates were 74% in the low-risk group and 43% in the high-risk group, (P=0.034). Disease recurred in 30.2% of the low-risk patients and 45. 3% of the high-risk patients. No survival advantages were found by using different adjuvant therapies.

CONCLUSIONS

Radical hysterectomy should not be attempted if either the pretreatment SCC-Ag level is above 5 ng/ml or S-phase fraction of the tumor greater than 20% due to its limited value despite applying aggressive postoperative adjuvant therapy.

Relationship between flow cytometric DNA ploidy and nuclear grade with endocrine dysfunction in adrenal cortical adenomas

OBJECTIVES

To determine the relationship among the DNA ploidy, histopathologic features, and clinical syndrome in adrenal cortical adenomas, because the cells often show variability in nuclear size and configuration.

METHODS

Our study included 44 adenomas associated with primary aldosteronism and 23 adenomas associated with Cushing syndrome. Normal adrenal glands from patients with renal carcinoma served as the controls. Paraffin-embedded tissues were examined for DNA content by flow cytometry. The mean percentage of G(2)/M (4C%) of the control samples was 3.8%. Tetraploid was represented by a histogram with both a 4C peak greater than 9% (mean + 2.4 SD of control samples) and a small 8C peak.

RESULTS

Flow cytometric analysis revealed diploidy in 30, tetraploidy in 27, and aneuploidy in 8 of the 67 adenomas; 2 adenomas could not be classified. All 17 normal adrenal glands showed diploidy. A significant relationship was noted between DNA ploidy and the clinical syndrome (ie, a larger proportion of adenomas with primary aldosteronism had a tetraploid DNA histogram compared with adenomas with Cushing syndrome, P <0. 0001). Adenomas with primary aldosteronism had a significantly higher nuclear grade (III or IV) than did tumors with Cushing syndrome (P = 0.033). A significant relationship was also observed between DNA ploidy and nuclear grade in 57 euploid tumors, with tetraploid tumors often showing the highest nuclear grade (P = 0. 037).

CONCLUSIONS

Our results have demonstrated that adrenal cortical adenomas associated with primary aldosteronism often reveal severe nuclear pleomorphism, indicating that nuclear pleomorphism might be due to a tetraploid stemline.

Value of morphometry, texture analysis, densitometry, and histometry in the differential diagnosis and prognosis of malignant mesothelioma

Malignant mesothelioma is a tumour with increasing incidence due to widespread use of its causative agent, asbestos, in the past decades. The poor survival necessitates a correct differentiation from other lesions at the same site, such as hyperplastic mesothelium and carcinomas metastatic to pleura or peritoneum. Since genetic and immunohistochemical markers are not absolutely differentiating, the diagnosis is based on the histology complemented with (immuno)histochemistry. However, as the tumour presents itself in numerous heterogeneous histological forms, visual evaluation is extremely difficult. In order to evaluate the prognostic and diagnostic performance of syntactic structure analysis (SSA), chromatin texture analysis, densitometry, and morphometry, an automated KNN-classification system has been used to compare Feulgen-stained tissue sections of hyperplastic mesothelium, malignant mesothelioma, and pulmonary adenocarcinoma. In addition, we also studied most discriminative aspects in the differentiation, typing, and prediction of survival. The results indicate that for the diagnosis of malignant mesothelioma, chromatin texture parameters outperform SSA, densitometry, and morphometry (recognition score=96.8 per cent). Most discriminative parameters highlight spatial patterns of the chromatin distribution that are hard to appraise visually and directly show the benefits of a quantitative approach. Typing of the tumour is best described by SSA parameters, relating to the spatial arrangement of the cells in the tissue (recognition score=94.9 per cent). In survival time classifications, chromatin texture yields the highest recognition score (82.9 per cent), although accurate estimations are unreliable due to a large degree of misclassification.

In situ localization of S-phase-specific histone (H3) mRNA in Bowen's disease

PCNA and Ki-67 immunohistochemistry has been used to assess cell proliferation in place of tritiated thymidine or BrdU labeling of S-phase cells. Recently, it has been possible to reliably demonstrate histone H3 mRNA by in situ hybridization in formalin-fixed and paraffin-embedded tissue sections. We have compared this new proliferation marker with Ki-67 and PCNA with regard to distribution of positive cells and labeling indices (LI%) for 22 cases of Bowen's disease. In normal skin, Ki-67-IHC positive cells and histone mRNA positive cells were observed in the basal and suprabasal layers of the epidermis. In Bowen's disease, positive cells with each marker were more frequent in upper neoplastic epidermis than in suprabasal layers, and the average LI%s were markedly elevated with all markers, the scores decreasing in the following order: PCNA-IHC, Ki-67-IHC and H3mRNA-ISH. However, the results of double staining demonstrated that S-phase cells do not necessarily show exactly the same distributions as with PCNA and Ki-67-IHC labeling. H3mRNA-ISH showed three different degrees of reaction with significantly different LI%s, whereas PCNA and Ki-67 LI% did not vary essentially in the same areas. These results strongly suggest that Bowen's disease, which is well known as a low-grade neoplastic state with malignant potential, also demonstrates clear intratumoral heterogeneity of S-phase cells using the H3mRNA-ISH method.

S-phase fraction identifies high-risk subgroups among DNA-diploid breast cancers

The prognostic value of DNA content measured by means of flow cytometry was analyzed in formalin-fixed, paraffin-embedded samples from 231 breast cancer patients treated between 1984 and 1988, with a mean follow-up period of 55 months. We followed the guidelines of a Consensus Meeting held on this issue in Maine, USA, in 1992. DNA-diploid and -aneuploid tumors were evaluated separately for the fraction of cells in S-phase (SPF) contained in them, this being divided into three groups ('high', 'intermediate', and 'low'), defined by the 25th and 75th centile of the SPF-distribution corresponding to either DNA-diploid or DNA-aneuploid tumors. Unequivocally readable histograms were obtained from 174 samples (75.3%). A high SPF in diploid tumors was significantly associated with a higher recurrence rate (p = 0.015), a shorter disease-free survival (p = 0.014), advanced (IIIB) clinical stage (p = 0.034), and almost significantly with total survival (p = 0.055). In a multivariate Cox regression analysis, a high SPF in diploid tumors retained its independent prognostic power, being significantly associated with a shorter disease-free survival (p = 0.00049) and total survival (p = 0.0077). It also allowed to identify a subgroup with an ominous prognosis among patients < or = 50 years of age with early stage tumors. Our results fully validate the recommendations of the 1992 Maine Consensus Meeting.

Interlaboratory reproducibility of semiautomated cell cycle analysis of flow cytometry DNA-histograms obtained from fresh material of 1,295 breast cancer cases

Conflicting prognostic results have been published as to the DNA variables, such as DNA ploidy, DNA index, and % S-phase cells for breast cancer patients. These variables can be obtained by interpreting DNA histograms by cell cycle analysis. Explanations for these conflicting results might be found on the level of the interpretation of the DNA histograms. In a previous study, the semi automated cell cycle analysis computer program MultiCycle (Phoenix Flow Systems, San Diego, CA) showed high intralaboratory reproducibility. However, what types of DNA histograms may cause disagreements was still unclear. The aim of this study was to determine the interlaboratory reproducibility of MultiCycle-based cell cycle analysis of 1,295 flow cytometric DNA histograms derived from fresh frozen breast cancer material and to clarify potential sources of interobserver variation when analyzing DNA histograms. DNA ploidy classification into diploid, hyperdiploid, tetraploid, hypertetraploid, and multiploid showed an interlaboratory agreement of 94% (kappa value = 0.92). The 6% discrepancies (n = 74) were caused by tetraploid peaks, as established in one laboratory, which shifted outside the tetraploid region on reanalysis by the other laboratory (37%), shoulders sometimes interpreted as peaks (24%), small peaks not always recognized as such (24%), fitting failures (10%), and overlooking of tetraploid peaks (5%). Furthermore, the cell cycle analysis variables showed variable reproducibility. The % S-phase cells of the first, second, and third cell cycle showed overall a moderate reproducibility (0.62 < or = R < or = 0.79), but the average % S-phase cells and the average aneuploid % S-phase cells were more reproducible with correlation coefficients of 0.89 and 0.81, respectively. The coefficient of variation of the G0/G1 peak of the first cell cycle, the DNA indices and the % diploid cells were highly reproducible (R > or = 0.94), and the % G2/M-phase cells of the first, second, and third cell cycle were poorly reproducible (0.22 < or = R < or = 0.68). When a cut-point was used at the mean value of 7% for the average % S-phase cells, the number of "threshold discrepancy cases" was 6%. Sources of variation for cell cycle analysis were variations in the debris correction procedures, disagreement about the modes of the aneuploid peaks, disagreement about small peaks, shoulders sometimes interpreted as peaks, and overlooking of tetraploid peaks.

S-phase fraction of 155 soft tissue sarcomas: correlation with clinical outcome

BACKGROUND

Traditionally, grade is considered the most important prognostic factor for soft tissue sarcomas (STS). However, because of the alleged difficulties in reproducibility of grading, new, objectively determined prognostic factors would be of value. The aim of our study was to establish if S-phase fraction (SPF) measured with flow cytometry was of prognostic significance for STS.

METHODS

In this study, we included all 193 adult STS patients with superficial trunk or limb tumors who were treated by the Helsinki University Central Hospital (HUCH) STS group between January 1987 and May 1993. One hundred and seventy-two formalin fixed paraffin embedded tumor samples were available. SPF measurement was successful in 155 cases.

RESULTS

Eighty-six cases were diploid. Ploidy was found to have no effect on overall survival. The median SPF was 6.8% (diploid tumors, 4% and nondiploid tumors, 12.9%). A high SPF predicted a shorter survival in patients with diploid tumors (P=0.003). The prognostic value was even stronger when we studied disease specific survival and excluded from analysis samples that contained less than 50% tumor cells (P=0.011). However, no prognostic value could be detected in nondiploid tumors or in the material as a whole.

CONCLUSIONS

Our results suggest that high SPF is an adverse prognostic factor for survival of patients with diploid STS. However, further studies are needed to confirm these results.

Use of light scatter when recording a DNA histogram from paraffin-embedded tissue

The quality of a DNA histogram as recorded in a flow cytometer can often be improved by gating on forward and orthogonal light scatter. This may be helpful when measuring DNA histograms from formalin-fixed, paraffin-embedded material. We report that nuclei from many human tumours scatter more light orthogonally than those from normal stroma. Gating on light scatter enabled the DNA histogram from the tumour to be recorded with reduced contamination from normal cells. In studies of human mammary carcinomas, we have found that this method has the potential to improve the estimation of S phase and the measurement of DNA index (DI) when the DI is close to 1 (near diploid).

Reproducibility in DNA flow cytometric analysis of breast cancer: comparison of 12 laboratories' results for 67 sample homogenates

Flow cytometric (FCM) DNA analysis yields information on ploidy status and the S-phase fraction (SPF), variables of prognostic importance in breast cancer. The clinical value of the SPF is currently being evaluated in prospective randomized trials. The widespread use of FCM DNA analysis emphasizes the importance of reproducibility (both intra- and interlaboratory). In this study, 67 nuclear suspensions of breast cancer samples were analyzed by 12 laboratories routinely performing FCM DNA analysis in breast cancer. No general guidelines were imposed; each laboratory used its own standard protocols. For DNA ploidy status (diploid vs. non-diploid), agreement was complete for 79% (53/67) of the samples, compared with 64% (43/67) of samples when tetraploidy was considered [i.e., euploid (diploid+tetraploid) vs. aneuploid (the remaining non-diploid)]. For the SPF, pairwise comparison of the results of all 12 laboratories yielded a mean Spearman's rank correlation of 0.78 (range: 0.54-0.93). For those 39 samples being categorized in low or high SPF by all laboratories, all agreed in 14 samples (36%). Similar patterns were obtained with kappa measures, agreement being good for ploidy status (diploid vs. non-diploid; overall kappa = 0.87 and 0.74 for euploid vs. aneuploid), but moderate for the SPF [overall kappa = 0.47 (for low SPF vs. high SPF vs. "no SPF reported")]. Discrepancies were chiefly attributable to differences in the categorization of the S-phase values, rather than in FCM procedures, other critical differences being in the detection and interpretation of near-diploid and small non-diploid cell populations, the definition of tetraploidy, and the choice and execution of the method used for S-phase estimation. Based on the observations of this study, detailed guidelines for FCM analysis and interpretation of data are proposed in the Appendix. Some issues remain, however, e.g., to standardize a method for S-phase calculation and tetraploid definition.

DNA ploidy as a predictor of cervical metastasis in advanced squamous carcinoma of the tongue

Flow cytometric analysis of nuclear DNA content was performed in 34 squamous cell carcinomas of oral cavity and base of tongue tumours using archived paraffin-embedded tissues. The DNA content was correlated with the presence of cervical metastasis. Ten and 24 patients were classified as early (I & II) and advanced (III & IV) clinical stages, respectively. The DNA index (DI) was grouped into diploid (DI 0.85-1.15) and non-diploid. Seven (70%) tumours were non-diploid in clinical stages I & II. Four out of 7 (57%) developed initial and late cervical lymph node metastasis (p > 0.05). There were 15 (62.5%) non-diploid tumours in clinical stages III & IV. Thirteen out of 15 (86.7%) had cervical lymph node metastases (p < 0.05). However, the survival period and the incidence of recurrent disease for the whole group did not show any association with DNA ploidy. Our results suggest that DNA content may be useful as a reliable predictor of regional metastasis in advanced stage carcinoma of the tongue.

Role of the pathologist in biomarker studies

Cancer chemoprevention is defined as intervention by chemical agents prior to invasion to inhibit or slow the carcinogenic process. Using surrogate endpoint biomarkers in chemoprevention studies may reduce the size, length and cost of clinical prospective randomized trials in high-risk populations. Intermediate biomarkers are measurable alterations in the tissues at risk and include differentiation, genetic composition, biochemical expression, and proliferation. Assessment is possible because invasive epithelial neoplasms are known to begin as intraepithelial proliferations with a spectrum of cellular abnormalities extending to carcinoma in situ. Genetic heterogeneity begins in the intraepithelial phase; a stochastic accumulation of genetic errors characterizes the progression of clonal evolution within the tumor through the process of invasion and metastasis. Pathologic features associated with this process include tumor classification as well as whether it is intraepithelial or invasive. If the process is intraepithelial, the grade and extent of the intraepithelial lesion are reported. If the neoplasm is invasive, tumor size, extent, degree of differentiation (histologic and nuclear grade), mitotic rate, vascular invasion, and lymph node involvement are evaluated. In assessing biomarkers relevant chemoprevention, and without complete regression of the neoplasm with the chemopreventive agent or agents, measurable parameters along with histopathologic features are applicable. Three methods readily applicable for this purpose that can be applied to paraffin-embedded, formalin-fixed tissue include quantitative pathology, immunohistochemistry, and molecular biologic applications. These methods require some consistency in handling and processing the tissues under study; results may deteriorate due to a number of processing variables, including time to fixation, time in fixative, and fixative type. Quantitative pathology, including static image analysis and flow cytometry, can determine total DNA content. Using static image analysis, very small tumors can be studied. In addition, adjacent intraepithelial and invasive components of a tumor may be studied from a single slide. Steroid receptors, oncogenes, and other proteins detectable through immunohistochemical or molecular biologic methods can be quantitated by this technique as well. Cell cycle synthetic function is assayable by both methods. Flow cytometry can calculate the total percentage of cells in S-phase, or the tumor cell S-phase fraction based on the percentage of cells detected between the G0, G1 peak and the G2 + M peak. A similar approach is generally not applicable with current image analysis equipment; however, cell cycle related proteins such as MIB-1 (Ki-67 associated) can be quantified. Immunohistochemical methods can employ a wide variety of monoclonal antibodies to detect oncogene related proteins, including HER-2/neu (c-erbB-2) and p53. Molecular biologic methods, including in situ hybridization, polymerase chain reaction, and in situ PCR, can have many applications when applied to paraffin-embedded tissues, including detection of viral DNA, identification and measurement of apoptosis, and defining gene deletions.

Relationship of DNA ploidy to histology and prognosis in rhabdomyosarcoma. Comparison of flow cytometry and image analysis

BACKGROUND

Although DNA ploidy correlates with prognosis in certain childhood cancers, e.g., neuroblastoma, its significance in rhabdomyosarcoma (RMS) is unclear and controversial.

METHODS

Ploidy by flow cytometry (FCM) and image analysis (IA) in 26 of 27 children with RMS (17 embryonal, 3 mixed embryonal/alveolar, 5 alveolar, 1 anaplastic, 1 ectomesenchymoma) and 4 adults with pleomorphic RMS were evaluated. Statistical comparisons were analyzed between DNA content and gender, age, localization, Intergroup Rhabdomyosarcoma Study (IRS) group, and histopathologic subtype. Survival analyses were performed by the Kaplan-Meier test using the approximate chi-square statistic for the log rank test.

RESULTS

The concordance rate between FCM and IA was 26 of 30 (87%); FCM was not performed in one tumor. Image analysis was more sensitive than FCM in detecting aneuploidy. Furthermore, DNA content was associated significantly with histologic subtype (P = 0.031); embryonal histology commonly was hyperdiploid (mean, 1.44; median, 1.27), whereas alveolar histology usually was near-tetraploid (mean, 1.83; median, 1.95). All four adult patients with pleomorphic RMS were aneuploid, with one showing multiple DNA peaks. No correlation between DNA content and survival was observed in the children with RMS. However, IRS group (P = 0.011) and patient age (P = 0.036) were independent prognostic indicators significantly related to survival. All adult patients died of their disease.

CONCLUSIONS

Although ploidy correlates with histologic subtype, DNA content is not significantly predictive of prognosis in patients with RMS. Age at diagnosis and IRS group are independent predictors of clinical outcome in children with RMS.

Near-diploid transitional cell carcinoma: a preliminary report

DNA ploidy analysis has been accepted as an important prognostic factor for patients with transitional cell carcinoma (TCC). However, there was few information dealing with the clinical relevance of slightly aberrant DNA content by flow cytometry (FCM). Here we present five cases of near-diploid (ND) tumours, with DNA index (DI) varying from 0.92 to 1.14, obtained from a prospective study of fifty-one cases (9.8%). The frequency of ND tumours showed a tendency to decrease with increasing tumour stage. Higher fraction of tumour proliferation, defined by Ki-67 index, showed a steady increment from 3.4 to 23.5% with occurrence of gross chromosomal changes. In contrast, the expression of epidermal growth factor receptor (EGFR) decreased from 48.3 to 35.3% for diploid (n = 29) through aneuploid (n = 17) tumours. All three ND bladder cancers had recurrence of one to three times with median follow-up of 36 months. The incidences of tumour recurrence (60%) and cancer death (20%) in ND tumours were intermediate between the aneuploid and diploid TCCs. But, flow DNA analysis of paraffin-embedded ND tumours revealed wide and symmetrical G0/G1 peak with DI varying from 5.6 to 13.0. Our limited experience suggests the necessity of special treatment for G0/G1 peaks having CV values greater than 5.5% from paraffin-embedded urothelial carcinomas.

DNA-ploidy, G2M-fractions and prognosis of stages B and C prostate carcinoma

Paraffin embedded tissue of 49 stage C and 27 stage B prostate adenocarcinomas was investigated by flow cytometry. All patients were treated by radical prostatectomy with pelvic lymphadenectomy and followed up for 5-10 years. The tumour was separated from the benign tissue to increase the proportion of tumour cells. Ten stage C and seven stage B carcinomas had to be excluded because of poor fixation. Six of the 39 (15%) stage C and 1/20 (5%) stage B carcinomas were aneuploid. Cell cycle analysis was done with correction for sliced nuclei and background subtraction. The threshold between carcinomas with low and with increased ("tetraploid") G2M-fraction was determined by comparing carcinomas with and without tumour progression. Sixty-seven percent of the patients with non-euploid stage C carcinomas and 11% of those with euploid carcinomas suffered from tumour progression (P < 0.01). The respective values for the stage B carcinomas were 67% and 6% (P < 0.01). These results demonstrate the strong prognostic impact of DNA-ploidy and G2M-fractions for each individual patient.

Automated peak detection and cell cycle analysis of flow cytometric DNA histograms

We describe an algorithm for fully automated flow cytometric DNA histogram classification and analysis that provides rapid, reproducible determination of DNA index and S-phase fraction (SPF). Automated classification agreed with subjective assessment of DNA ploidy in 96-98% of DNA histograms. Automated and conventional analyses of DNA index (r = 0.95) and SPF (r = 0.89) were also highly correlated with one another. In a series of 86 node-negative breast carcinomas, SPF calculated with the fully automated method was a significant predictor of 10 year survival (p = 0.009). Automation greatly increased the speed of DNA histogram analysis, allowing evaluation of the same set of histograms with different methods. In a preliminary study exploring the optimization of DNA histogram analysis, the best association between SPF and prognosis of breast cancer patients was achieved using sliced nuclei debris modeling, reporting only the aneuploid SPF (in aneuploid histograms), while excluding small aneuploid clones (< 15% of total cell count) from evaluation. In conclusion, automated DNA histogram analysis does not replace the need for close human supervision but provides a useful guideline for less experienced users, facilitates interlaboratory comparisons, and makes possible extensive reanalyses of large data sets.

Cell kinetics in human breast cancer: comparison between the prognostic value of the cytofluorimetric S-phase fraction and that of the antibodies to Ki-67 and PCNA antigens detected by immunocytochemistry

The determination of cell proliferation is one of the more widely used tools for assessing prognosis. However, additional research in this field is warranted because today there are several methodological procedures available for monitoring cell kinetics and it has still not been established which is the most reliable marker of proliferation and which possesses the greatest prognostic value. We performed this study in a series of primary invasive breast cancers to compare the prognostic value of S-phase fraction (SPF) by flow cytometry, the most widely used method for detecting proliferation at present, with that of antibodies to Ki-67 and PC-10 to proliferating-cell nuclear antigen (PCNA) detected by immunocytochemical methods. A significant linear relationship was observed only between SPF and Ki-67. In univariate analysis SPF and Ki-67 values, nodal status, histological grading and peritumoral lymphatic-vessel invasion were significant predictors of relapse-free survival (RFS). As far as overall survival (OS) is concerned, only SPF, Ki-67 and nodal status were significantly associated with the risk of death. PCNA had no prognostic value for either RFS or OS. In multivariate analysis only SPF and nodal status retained a significant and independent prognostic value. Neither the cell-kinetics parameters assessed by immunocytochemistry (i.e. Ki-67 and PCNA) nor histological grading were independent prognosticators. In conclusion, our results provide evidence that the determination of SPF by flow cytometry was the strongest cell-kinetics marker used to assess prognosis in this series of breast cancers. However, different and novel markers of cell kinetics need to be compared in larger series in order to identify the best one.

The uses and limitations of DNA flow cytometry in stage IB or II cervical carcinoma

BACKGROUND

The role of flow cytometry in predicting prognosis for cervical carcinoma remains unclear.

METHODS

Flow cytometric analysis was performed on tissues, fixed in formaldehyde solution and embedded in paraffin, from 411 patients with Stage IB or II cervical carcinoma who had been treated with radical abdominal hysterectomy and bilateral pelvic lymphadenectomy.

RESULTS

DNA aneuploid-multiploid tumors were found in 37.5%, tetraploid in 4.6%, and diploid-peridiploid in 57.9%. Five-year recurrence-free survival rates of the three groups were 74.3%, 77.8%, and 76.4%, respectively (P > 0.05). DNA aneuploidy and DNA index (DI) of greater than 1.3 were highly correlated to parametria extension. In univariate analysis, pelvic lymph node metastases, stage, parametrial extension, depth of cervical stromal invasion, tumor size, and DI (1.3, 1.4, 1.5 as breakpoint) were significant prognostic factors. DNA ploidy, S-phase fraction, and S-G2M fraction were not significant. In multivariate analysis, DI of greater than 1.3, pelvic node metastases, clinical Stage II, and depth of stromal invasion greater than two-thirds of full cervical thickness were independent and significant variables. The prognostic index (PI), defined by the model, was able to categorize the patients into three distinct risk groups. The 5-year recurrence free survival rates of the low-, intermediate-, and high-risk groups were 89.5%, 73.0%, and 58.9%, respectively (P < 0.0001).

CONCLUSIONS

The prognostic value of the DI as a single variable is promising and warrants additional investigation to establish its appropriate use.

Prognostic importance of DNA flow cytometric variables in rhabdomyosarcomas

AIM

To determine whether DNA ploidy patterns and S phase fraction offer prognostic information in patients with rhabdomyosarcoma (RMS).

METHODS

DNA flow cytometry was performed on formalin fixed, paraffin wax embedded samples from primary tumours, and metastatic deposits or recurrences in 70 patients. DNA histogram analysis was done using a semi-automated cell cycle analysis program.

RESULTS

Of the 70 primary tumours, 23 were DNA diploid, 32 DNA aneuploid, eight DNA multiploid, and seven DNA tetraploid. The prognosis for DNA aneuploid patterns was favourable, intermediate within the group of DNA tetraploid tumours and poor among patients with DNA diploid and DNA multiploid tumours (p = 0.009). In multivariate analysis (Cox regression model) DNA ploidy was an important independent prognostic factor, along with TNM stage, localisation, and histopathological classification. Ten out of 32 patients with a high S phase fraction (> 15%) with primary RMS achieved long term survival in contrast to 20 out of 29 patients with a low S phase fraction (< or = 15%) (p = 0.008). In 24 cases the DNA ploidy of cases of relapse was analysed. Of the 15 cases, in which stem line changes had occurred, 13 died of disease. No stem line changes were noted in nine cases and in this group four patients died of disease (p = 0.02).

CONCLUSIONS

Assessment of DNA ploidy and S phase fraction in primary RMS and evaluation of stem line changes in cases of relapse are important variables in predicting prognosis.

Prognostic value of DNA ploidy in squamous cell carcinoma of esophagus. Analyzed with improved flow cytometric measurement

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.

DNA aneuploidy in ectopic pregnancy and spontaneous abortions

The nuclear DNA content of 55 ectopic tubal pregnancies was studied by flow cytometry from paraffin embedded tissue blocks. An abnormal amount of DNA content was found in 24% of the cases. This was a significantly higher percentage than encountered in 92 spontaneous abortions studied by the same method in the same population (8%, P = 0.01). The result indicates that, in addition to the maternal factors, abnormal embryogenesis with grave chromosomal aberrations may play a major role in the etiology of ectopic pregnancy.

DNA and S-phase fraction analysis by flow cytometry in prostate cancer. Clinicopathologic implications

BACKGROUND

Available prognostic factors for prostate cancer have not been proven consistently useful. The authors evaluated the prognostic value of DNA analysis by flow cytometry (FCM) in prostate cancer.

METHODS

Paraffin-embedded tumor specimens taken at tru-cut biopsy or transurethral resection (TURP) from 81 patients with prostate cancer were analyzed for DNA content and S-phase fraction (SPF) by FCM according to the method of Hedley et al (1983).

RESULTS

Thirty-five of the 63 (55.5%) evaluable DNA histograms had a diploid pattern, 18 (28.5%) a distinct aneuploid peak, and 10 (16%) a tetraploid pattern. An association was established between DNA ploidy abnormalities and Gleason score (P < 0.04) or presence of metastases at diagnosis (P < 0.0002). At Kaplan-Meier analysis, overall survival was significantly longer (P < 0.0002) in patients with diploid than in those with nondiploid tumors. Among patients with different risk categories, i.e. tumor size, Gleason score, and metastases at presentation, ploidy improved the detection of patients with poorer survival, with the exception of those with T1-T2 tumors. Cox regression analysis showed that ploidy was significantly related to survival. Bivariate models containing ploidy and SPF or Gleason score had a predictive value similar to that including all variables.

CONCLUSION

The study data show that DNA ploidy provides additional prognostic information in patients with locally advanced or metastatic prostatic cancer. The role of SPF remains to be established.

Flow cytometric analysis of DNA and nuclear protein in paraffin-embedded tissue

Previously we described the simultaneous quantification of DNA and nuclear protein in unfixed tissue from solid tumors. The resultant 2 parameter flow cytometric analysis has several advantages over that of DNA alone. In this report, we describe a modification of the technique for the analysis of formalin-fixed paraffin-embedded tissue. Paraffin-embedded material was prepared by hydrating sections, incubating in 0.5% pepsin solution, washing, and resuspending in buffer containing nonionic detergent. The nuclei were then stained with fluorescein isothiocyanate and propidium iodide in the presence of ribonuclease. Several solid tumor tissue types have been analyzed, including breast, colon, kidney, and thymus. The best results were obtained when the initial pepsin treatment was for 1.5 h, instead of 0.5 h. Pepsin treatment for 1.5 h improved the CVs of both the DNA and nuclear protein parameters, and did not appear to reduce nuclear protein levels or to cause significant disintegration of nuclei. The DNA/nuclear protein histograms of unfixed and fixed, paraffin-embedded tissue were similar. Since tumor nuclei typically have higher protein levels than DNA-diploid nuclei, the technique reduces population overlapping and permits less subjective identification of DNA aneuploidy.

Effect of DNA ploidy classification on prognosis in breast cancer

A series of 327 breast cancers was analyzed for DNA ploidy by flow cytometry from paraffin-embedded tissue, and the resulting DNA histograms were classified independently by 6 researchers in the field as DNA diploid (Di), aneuploid (An), tetraploid (Te), multiploid (Mu), or technically uninterpretable. The frequency of diploid, aneuploid, tetraploid and multiploid cancers varied from 28 to 41%, 33 to 49%, 8 to 21% and 2 to 6%. According to the scale Di-An-Te-Mu, DNA ploidy was not significantly associated with breast-cancer mortality by 2 classifiers, but if DNA euploid cancers (Di+Te) were tested against non-euploid, or diploid cancers against non-diploid, all classifiers found DNA euploid and diploid cancers to have better prognosis. Mortality associated with diploid or tetraploid cancers decreased with improving histogram quality and increasing uniformity of classification, whereas that associated with aneuploid cancers remained unaltered. Among the cases where all classifiers agreed on ploidy, tetraploid, diploid and aneuploid cancers were associated with 100%, 88% and 68% 5-year survival rates. In this sub-set the S-phase fraction and possibly DNA ploidy were independent prognostic factors, together with histological grade, axillary node status, and primary tumor size.

Flow-cytometric DNA analysis of stages IB and IIA cervical carcinoma

The prognostic significance of flow-cytometric DNA analysis was assessed in 375 stages IB and IIA squamous cell carcinoma patients treated with radical hysterectomy and lymphadenectomy at the Mayo Clinic between 1956 and 1985. Paraffin-embedded samples containing at least 20% tumor were dewaxed, rehydrated, stained with propidium iodide, and analyzed. Among 344 assessable samples, 136 (40%) were diploid and 208 (60%) were nondiploid (26 tetraploid, 158 aneuploid, and 24 polyploid). Diploid cases were further subclassified: 25 high proliferative phase (HPP) (S+G2M greater than 20%) and 111 low proliferative phase. No significant correlation was noted between DNA diploid patterns and stage, tumor size, grade, or histotype, but HPP diploid tumors had a significantly higher risk of nodal metastasis. With a mean follow-up period of 150 months, 62 patients died of disease. No significant difference was observed in survival rates (SR) between diploid and nondiploid tumors, but the subset of HPP diploid tumors had a prognosis significantly worse than that of any other group (P less than 0.01). Other significant variables included nodal metastases, parametrial extension, age, and clinical stage. While ploidy patterns did not assign additional risk to node-positive lesions, HPP diploid tumors in node-negative patients were associated with a significantly lower SR. Multivariate analyses in node-negative patients demonstrated that stage, histologic subtype, and HPP diploid patterns retained prognostic independence.

The study of endocrine tumors by flow and image cytometry

Both DNA flow and image cytometry are methods that can be used for the quantitative determination of cellular DNA content. Objective, quantitative analysis of cellular morphology can also be obtained using image cytometry. Data thus generated have been shown to be of diagnostic and prognostic use in the study of many solid tumors and would be of particular value in the evaluation of endocrine tumors that show a poor correlation between their histology and biological behavior. The main application of these techniques to endocrine tumors has been in the analysis of thyroid nodules, although a limited number of studies of parathyroid, pituitary, adrenal, and pancreatic neuroendocrine tumors and tumors of the dispersed neuroendocrine system have been reported. Review of these studies shows that in the endocrine organs DNA and morphometric measurements have a very limited role in the diagnosis of individual cases, but are important as prognostic variables. The high incidence of abnormal DNA content in histologically benign lesions of the endocrine glands has important biological implications. Further investigation of this phenomenon may help to elucidate the process of endocrine tumorigenesis.

Clinical application of flow cytometry for DNA analysis of solid tumors

Recent developments of flow cytometry (FCM) technology which make multiple correlative biological measurements on normal and neoplastic cells is affecting areas of diagnostic pathology as well as research fields, and a general understanding of FCM techniques is essential for pathologists. Today, FCM DNA measurements of tumors also becomes routine in the clinical and/or pathological laboratory for aid in cancer diagnosis and cancer treatment. It can also contribute to diagnosis of tumors as a supplemental method to conventional histopathology, and DNA ploidy and the percentage of S-phase fraction are considered as complementary prognostic parameters independent of the stage of disease. This article reviews clinical applications of flow cytometry focusing on the DNA measurements of solid tumors, and related practical issues, such as the methodology for nuclear DNA measurement, interpretation of DNA histograms and the relationship of DNA ploidy and S-phase fraction to clinical and pathological features of human solid tumors.

DNA ploidy in cell nuclei from paraffin-embedded material--comparison of results from two laboratories

In 49 pairs of contiguous sections from paraffin-embedded prostatic cancer tissue, the DNA indices (DIs) were determined by flow cytometry (FCM) at 2 different laboratories. In 3 of 45 pairs of evaluable nuclear suspensions, DIs of 1.1 (DNA aneuploid) were found at Laboratory 1, whereas all 3 tumours were classified as DNA diploid at Laboratory 2. In the remaining 42 specimens, the correlation between the DIs was excellent, though the application of strictly defined DNA ploidy ranges led to different DNA ploidy allocation in 3 cases. It is concluded that in 85-90% of the cases, reliable DIs can be obtained by FCM done in paraffin-embedded material at different laboratories. Slight technical variations and interpretation differences may lead to different ploidy allocation in 10-15% of the cases.

Abnormal DNA content as a biomarker of large bowel cancer risk and prognosis

Aneuploid cell populations can be defined as those that contain an abnormal number of chromosomes or an abnormal amount of DNA. Aneuploidy can be reliably detected by flow cytometric analysis of DNA content. This technique not only identifies aneuploid cell populations but can also quantify the percent of cells in various phases of the cell cycle, thus giving an indication of the proliferative activity of a tissue. Aneuploidy occurs in approximately 60% of established colorectal cancers, and many studies have demonstrated that patients with aneuploid tumors have a poorer prognosis than patients with diploid colon cancers. Some studies have suggested that the proliferative rate of tumors, as assessed by the percent of cells in S phase, also has prognostic significance. Until recently, aneuploidy was thought to occur only in malignant tissues, but it has been clearly shown that aneuploid cell populations can be identified in benign adenomatous polyps as well as in non-neoplastic-appearing mucosa of patients with chronic ulcerative colitis and Barrett's esophagus. In chronic ulcerative colitis, aneuploidy occurs more frequently in patients with dysplasia or cancer than in those with no evidence of neoplasia. Similarly, dysplastic and malignant biopsies are more commonly aneuploid than non-neoplastic biopsies. Patients who have undergone colectomy for cancer or dysplasia in the setting of chronic ulcerative colitis frequently have multiple areas of aneuploidy throughout the remainder of their colon. Whether aneuploidy can be useful as a marker of cancer risk in patients with chronic ulcerative colitis deserves further investigation.

Node negative breast cancer: the prognostic value of DNA ploidy for long-term survival

The DNA content of breast tumours from 170 patients who presented between 1978 and 1980 was measured by flow cytometry. The relationship between tumour ploidy and disease outcome was assessed and its association with other prognostic factors evaluated. Compared with those with diploid tumours, patients with aneuploid tumours had significantly earlier relapse and shorter survival (P less than 0.0001). Tumour ploidy was strongly related to grade (P less than 0.001), but there was no significant association between DNA ploidy and c-erb-B-2 expression, lymph node status or tumour size. In lymph node negative and c-erb-B-2 negative patients, aneuploid tumours were associated with a poorer prognosis (P less than 0.001) than diploid tumours. Multivariate analysis showed that tumour ploidy gave independent information on disease free and overall survival. Tumour ploidy may be used as an independent prognostic variable in patients with breast cancer and it may be helpful in defining patients within the node negative or c-erb-B-2 negative groups likely to have a poor outcome who might benefit from adjuvant treatment.

The prognostic value of deoxyribonucleic acid flow cytometric analysis in stage D2 prostatic carcinoma

This study was designed to compare the prognostic potential of tumor grade and ploidy status in patients with stage D2 prostate cancer. Two outcome groups were selected on the basis of survival after orchiectomy: a bad outcome group consisting of 66 patients who died of the disease within 12 months and a good outcome group comprising 37 patients who survived beyond 5 years. Tumors were classified histologically as well (17%), moderately (17%) or poorly (66%) differentiated. Tumor grade was a significant predictor of outcome, with 76% of poorly differentiated tumors in the bad outcome group and 65% of well differentiated tumors in the good outcome group (p less than 0.005). Deoxyribonucleic acid (DNA) ploidy analysis was performed on formalin fixed, paraffin embedded samples of the primary tumor to yield 97 final tracings that were classified using set criteria for DNA ploidy status. Over-all, 54% of the tumors were nondiploid (33% aneuploid and 21% tetraploid) and the remaining 46% were diploid. DNA ploidy status was a significant indicator of outcome (p less than 0.001), with 64% of diploid tumors in the good outcome group and 88% of the nondiploid tumors in the poor outcome group. Tetraploid tumors behaved no differently from other nondiploid tumors. We conclude that DNA ploidy status and tumor grading are significant independent predictors of outcome after orchiectomy and when combined yield important additional prognostic information.

An improved Hedley method for preparation of paraffin-embedded tissues for flow cytometric analysis of ploidy and S-phase

A modification of the Hedley-method for flow cytometric DNA analysis of paraffin-embedded tissues is presented. Dewaxed and dehydrated tissue from paraffin blocks was incubated with subtilisin Carlsberg (pronase, Sigma protease XXIV) and then stained directly without washing and centrifugation. The loss of material was minimized, which was advantageous, for example, for the analysis of core-biopsies, and all measured samples showed extremely low frequencies of clumped cell nuclei. This made is easier to detect polyploid nuclei and even rare nuclei of high ploidy could be identified. S-phase analyses were more precise, since the background originating from clumped debris particles was very low. The improved method was applied to the estimation of frequencies of high-polyploid nuclei found in various diploid, tetraploid, and aneuploid human myosarcomas of the uterus.

Improved prognostic impact of S-phase values from paraffin-embedded breast and prostate carcinomas after correcting for nuclear slicing

Nuclear debris may significantly interfere with the analysis of S-phase fraction (SPF) from paraffin-embedded tumors. We used a background subtraction algorithm to compensate for the effects of slicing of tumor cell nuclei during preparation of paraffin-embedded specimens. DNA histograms were analyzed from 88 node-negative breast and from 78 prostatic carcinomas. Median SPFs corrected for nuclear slicing were lower than uncorrected ones in both breast cancer (7.6% vs. 5.7%) and prostate cancer (6.7% vs. 4.2%). The median SPF value in each group was used as a cut-off point in survival studies. As compared with the uncorrected SPFs, corrected SPF levels resulted in a more significant survival difference between breast cancer patients with above and below median SPF (p = 0.0014 vs. p = 0.014) and in a higher relative risk (RR) of death (4.5 vs. 3.1). The same was true for prostate cancer survival (p less than 0.0001 vs. p = 0.002) and RR (5.3 vs. 3.1). Compared with the exponential background subtraction method, the sliced nuclei correction was more reproducible and could be applied in all evaluable histograms without the risk of overcompensation. In conclusion, our results support the use of background correction with the sliced nuclei model in DNA flow cytometric studies of archival tissues.

Inter-laboratory comparison of DNA flow cytometric results from paraffin-embedded breast carcinomas

Consecutive sections from 33 paraffin-embedded human breast carcinomas without intratumor heterogeneity were sent for flow cytometric (FCM) DNA analysis in two experienced laboratories. FCM instruments, run conditions, and tumor disaggregation procedures were different in the two laboratories. In four cases (12%) the laboratories reported a different DNA ploidy and DNA index (DI). These variations were due to analytical reasons, differences in the detection rates of near-diploid and tetraploid DIs, not due to interpretation of data or the criteria used for aneuploidy. There was a significant correlation between S-phase fractions (SPF) obtained in the two laboratories (r = 0.90, p less than 0.0001) if only cases with concordant DI were included. Discordant DI usually led to very different SPF values.

DNA index and S-phase fraction and their combination as prognostic factors in operable ductal breast carcinoma

The prognostic significance of DNA ploidy, DNA index (DI), and S-phase fraction (SPF) and their various combinations were studied together with 16 other clinicopathologic factors in 222 patients with operable invasive ductal breast carcinoma. The patients have been followed for a minimum of 22 years after the diagnosis or until death. Nuclear DNA content was determined by flow cytometry from paraffin-embedded tissue. Patients with DNA diploid cancer (n = 57, 26%) had better survival rate corrected for intercurrent deaths than patients with nondiploid cancer (P = 0.002), and also, a small SPF (less than or equal to 14%, calculated in 134 cases) was associated with a favorable outcome in a univariate analysis (P = 0.01). The prognostic value of the DI and SPF was increased if they were combined. The most effective combination was obtained if diploid cancers were grouped together with DNA aneuploid cancers with a DI less than 2.1 and an SPF less than 14%. This combination had considerable prognostic value in a univariate analysis (P = 0.0002) and had independent prognostic value (P = 0.04) in Cox's multivariate analysis together with the primary tumor size (P less than 0.001) in axillary node negative patients but not in axillary node positive patients. In the whole series the presence of axillary nodal metastases (P less than 0.001), high mitotic count (P less than 0.001), a large primary tumor size (P = 0.001), poorly circumscribed tumor margin (P = 0.005), and slight or absent tubule formation (P = 0.05) were the only independent prognostic factors in a multivariate analysis.

Clinical significance of nuclear DNA content in pancreatic carcinoma

The nuclear DNA content of 62 pancreatic adenocarcinomas was analysed by flow cytometry from paraffin-embedded material. Radical surgery could be performed in 12 of the 24 cases with diploid carcinoma, but only in 3 of the 38 cases with a non-diploid tumour (P = 0.0002); the radically resected carcinomas also had a lower fraction of cells in the S-phase (P = 0.009). Non-diploid nuclear DNA content (38 cases, 61 per cent) was associated with advanced stage (P = 0.002), poor histological differentiation (grade II or III, P = 0.004), and primary tumour site in the body or the tail as compared with the head (P = 0.01). The median survival time of the patients with diploid carcinoma was 13 +/- 3 (SE) months, and that of the patients with non-diploid carcinoma 3 +/- 1 months (P = 0.0001). The DNA index with the cutoff value 1.4 was a slightly more powerful prognostic factor than DNA ploidy, and it was the most important independent prognostic factor in Cox's multivariate analysis (P less than 0.001) followed by histological grade (P less than 0.03). We conclude that diploid pancreatic carcinomas are associated with a longer survival than the non-diploid ones, and that radically operable carcinomas form a special subgroup with frequent diploidy and less aggressive biological behaviour.

Evidence for false aneuploid peaks in flow cytometric analysis of paraffin-embedded tissue

It has recently been shown that bimodal histograms with false aneuploid peaks may be obtained by DNA flow cytometry from histologically normal tissue allowed to autolyze. To investigate if such peaks can be generated from surgically excised archival tissue, 198 paraffin blocks from 179 patients containing histologically normal spleen (n = 65), liver (n = 26), thyroid (n = 32), pancreas (n = 19), salivary gland (n = 49), or lymph node tissue (n = 7), obtained from the archives of two university pathology departments, were analyzed for nuclear DNA content. The great majority (n = 160, 83.8%) of the 191 interpretable histograms had a single symmetrical G1 peak; and 8 histograms, all produced from liver tissue had a tetraploid pattern. A slight or a prominent repeatable deviation in the G1 peak outline was present in 14 (7.3%) cases. A peak resembling an aneuploid G1 peak with a DNA index (DI) ranging from 1.14 to 1.38 was repeatedly produced from 9 (4.7%) blocks containing histologically normal or inflamed splenic (n = 3), pancreatic (n = 3), liver (n = 1), thyroid (n = 1), or lymph node (n = 1) tissue. The three abnormal peaks produced from pancreatic tissue were rounded in shape and resembled closely the ones that can be obtained from autolytic pancreatic tissue, and the six remaining extra peaks were all fused with the "diploid" peak. In conclusion, false peaks, probably caused by degradation of the nuclear contents during formalin fixation or before it, may rarely be obtained from surgical paraffin-embedded samples.