Cell proliferation in childhood acute leukemia. Comparison of Ki-67 and proliferating cell nuclear antigen immunocytochemical and DNA flow cytometric analysis.

The proliferative activity of bone marrow leukemia cells was determined by DNA flow cytometric (FCM) analysis and labeling index (LI) of Ki-67 monoclonal antibodies and proliferating cell nuclear antigen (PCNA) autoantibodies in 73 children with acute leukemia. LI of Ki-67 varied greatly from patient to patient (range, 0.4% to 42.2%; mean, 18.8%) and differed significantly between acute lymphoblastic leukemia (ALL) and acute nonlymphoblastic leukemia (ANLL). In ALL, the Ki-67 LI showed a positive correlation with the S-phase fraction (SPF) determined by DNA FCM analysis, whereas, in ANLL, there was a discrepancy between the Ki-67 LI and SPF. In contrast, LI of PCNA varied less among the patients (range, 57.2% to 100%; mean, 90.3%), and the value was always higher than that of the Ki-67 LI in individual patients. A significant relationship between PCNA LI and the percentage of blast cells was found in peripheral blood leukocytes from patients with leukemia. These results suggest that the Ki-67 LI reflects differences in the proliferative activity depending on the subtype of the disease and that the PCNA LI is useful as a marker of proliferating cells.     

Percentage of S-phase cells in bone marrow aspirates, biopsy specimens and bone marrow aspirates corrected for blood dilution from patients with acute leukemia

For 14 patients with acute leukemia flow cytometry was used to determine the percentage of cells in S-phase flushed out of a bone marrow biopsy compared with the percentage in a bone marrow aspirate; there was a statistically significant difference (p less than 0.01) between the two. The percentage dilution of the bone marrow aspirate by peripheral blood was then calculated, according to Holdrinet et al. [1], in order to correct the percentage S-phase cells in the aspirate. The data presented show that when the percentage S-phase cells in the aspirate is corrected for blood dilution, it closely approaches the percentage S-phase cells in the biopsy (p greater than 0.10).     

DNA cell cycle distribution and glutathione (GSH) content according to circadian stage in bone marrow of cancer patients.

DNA cell cycle distribution and glutathione (GSH) content in bone marrow were measured both at daytime and midnight over single 24 h periods in 15 cancer patients. Between patients the S-phase demonstrated a difference from lowest to highest value of 700%, whereas the corresponding difference for the G2/M-phase was nearly 900%. The mean GSH content measured in the bone marrow at the two timepoints was 2.24 +/- 0.21 nmol mg-1 protein, range 0.91-4.19 nmol mg-1 protein. A statistically significant higher fraction of cells in S-phase and G2/M-phase was found at daytime as compared to midnight when excluding the four patients with an abnormal circadian variation in cortisol. No significant temporal variation in total bone marrow GSH content was found, although a weak correlation between S-phase and GSH content was demonstrated (r = 0.42; P less than 0.05). This correlation was strengthened when not including the six patients with an abnormal cortisol pattern (4) and bone marrow infiltration (2) (r = 0.66; P = 0.005). Cells in S-phase demonstrated a positive correlation with cells in G2/M-phase (r = 0.64; P less than 0.0001). A negative correlation was found between GSH content and age (r = 0.53; P less than 0.005). Finally, a statistically significant positive correlation was demonstrated between cortisol and both S-phase and G2/M-phase (r = 0.57; P less than 0.001 and r = 0.38; P less than 0.05, respectively). The present study suggests a possibility of optimising cancer therapy and use of hematopoietic growth factors by determining individual average values and circadian stage dependent variation in bone marrow DNA cell cycle distribution. Furthermore, GSH content in bone marrow may predict this tissue's sensitivity to cytotoxic agents.     

The predictive value of in-vitro techniques in acute non-lymphocytic leukemia

Bone marrow aspirates obtained from 27 patients with acute non-lymphocytic leukemia (ANLL) was cultured at the time of presentation, during remission and at relapse. Growth patterns were assessed throughout the patient's clinical course. The percentage of Ia-positive progenitor cells was assayed by a complement-dependent cytotoxicity assay. The percentage of cells in S phase was measured by a tritiated thymidine suicide index. Growth patterns of leukemic bone marrow samples at presentation showed varied numbers of clusters but only rare colonies. This was not predictive of clinical course. Growth patterns of bone marrow in complete remission from ANLL often had depressed colony numbers. However, some patients in remission had bone marrow growth patterns that approached or reached normal colony numbers, suggesting elimination of residual leukemia. The percentage of cells that expressed Ia antigen at presentation, during remission and at relapse varied widely and was not predictive of long-term remission or early relapse. The percentage of cells in S phase was also highly variable and not predictive of clinical course. At presentation the S-phase percentage correlated with the percentage of cells expressing Ia antigen. However, there was no such correlation during remission.     

Cytokinetic investigation of lung tumors using the anti-bromodeoxyuridine (BUdR) monoclonal antibody method: comparison with DNA flow cytometric data

In order to investigate the cytokinetics of malignant tumors and non-malignant lesions of the lung, tissue samples from 57 patients affected by non-small-cell carcinoma (NSCLC), small-cell carcinoma (SCLC), and benign and inflammatory lesions have been analyzed using the BUdR monoclonal antibody (MAb) method. This method is based on the preparation, at the time of surgery, of viable monocellular suspensions (using collagenase and DNase treatment) and the concomitant administration of BudR. The percentage of BudR-labelled cells was monitored by fluorescent microscopy using an FITC-labelled second antibody. In NSCLC, each histological group showed a wide range of labelling index (LI) values. On the contrary, SCLC exhibited a more homogeneous kinetic behaviour as evidenced by a narrowly distributed, higher LI. Tumors shown to be diploid by flow cytometry did not show a lower LI than aneuploid tumors. Furthermore, differences were constantly observed between the S-phase percent calculated using BUdR and that calculated using the DNA flow cytometric (FC) histogram, the latter always showing higher S-phase values. In an attempt to study the intra-tumor proliferative heterogeneity, multiple-site sampling was performed. Proliferative heterogeneity seemed to be higher inter-tumor than intra-tumor. Finally, a positive correlation (p less than 0.05) was found between LI and the actual doubling time (DT) of the primary tumor mass, evaluated using sequential radiographs. In conclusion, the present BUdR method can be considered a useful source of relevant information on in vivo cell growth, in parallel to other clinical (DT) and biological (DNA content) approaches.     

Evaluation of the malignancy of breast carcinoma using flow cytometry

Flow cytometric DNA analyses of single cell nuclei were performed on nuclear suspensions prepared from tumors of primary breast carcinomas in 40 patients operated on in our hospital. Fluorescence intensities of propidium-iodide stained cell nuclei were analyzed in a SHOWA DENKO SC-20 cell sorter. The DNA index and the percentage of S-phase cells and DNA aneuploidy were calculated from each DNA histogram. There was a tendency for more advanced nodal tumor involvement to have a higher DNA index, increasing DNA aneuploidy proportions and significantly larger percent of S-phase cells.     

DNA and RNA flow cytometric study in multiple myeloma. Clinical correlations

Flow cytometric studies of cellular DNA and RNA content using the acridine-orange technique were conducted in 81 patients with multiple myeloma (MM). All patients were treated with the M-2 protocol and clinical response was evaluated according to the criteria of the Chronic Leukemia-Myeloma Task Force. Aneuploid DNA stemlines were found in 38.2% of untreated patients with a median DNA index (DNA-I) of 1.15 in marrow aspirates and 1.22 in biopsy specimens. The median percentage of cells with abnormal DNA content was 31.5 (aspirates) and 35 (biopsy specimens) and a positive correlation with the percentage of bone marrow plasma cells was observed. Significantly higher proliferation (S-phase) was found in marrow biopsy specimens as compared with marrow aspirates. Significantly higher RNA content (RNA index [RNA-I]) was observed in aneuploid versus diploid patients in biopsy material. There was no difference in response to the Memorial Hospital M-2 protocol between diploid and aneuploid patients. In patients with DNA-I greater than 1.15 remission duration was shorter as compared with DNA-I less than or equal to 1.15. Furthermore, no difference in cellular RNA content was noted between responders and nonresponders. This study demonstrates no correlation between cellular RNA content and response, as previously described for patients treated with vincristine, Adriamycin, and dexamethasone (VAD), but DNA aneuploidy appears to be an adverse prognostic factor in MM patients treated with the M-2 protocol. It also demonstrates that prognostic models for MM are not universal but depend on the chemotherapeutic regimen used.     

Proliferative activity of bone marrow cells in primary dysmyelopoietic (preleukemic) syndromes

The proliferative activity of bone marrow cells was studied in 24 patients with primary dysmyelopoiesis by means of flow cytometry and 3H-TdR autoradiography. Abnormal DNA content was found in two cases with aneuploid karyotypes. DNA content typical of a diploid population was observed in all patients with normal karyotype and in three patients with chromosomal aberrations. The fraction of bone marrow cells in S- and G2-phase was higher in primary acquired sideroblastic anemia and refractory anemia (without excess of blasts) than in refractory anemia with excess of blasts and chronic myelomonocytic leukemia. Regardless to the diagnosis, the patients with low fraction of cells in S- and G2-phase had short survival time and showed high rate of evolution into acute nonlymphoblastic leukemia. The labeling (LI) and mitotic (MI) indexes of both erythroblasts and granulocytic cells were decreased in nearly all patients. The lowest values of LI and MI of the granulocytic compartment were found in the patients who subsequently developed acute leukemia. These data suggest that cytokinetic analysis allows investigators to achieve useful information on the stage of disease in the dysmyelopoietic syndromes.     

The proliferative state of normal hemopoietic stem cells during progression of leukemia. Studies in the BN acute myelocytic leukemia (BNML)

The kinetics of proliferation of normal hemopoietic stem cells (HSC) were investigated during the course of leukemia in a rat model for human acute myelocytic leukemia (BNML). The fraction of stem cells in DNA synthesis phase was determined with the hydroxyurea suicide technique followed by modified spleen colony assays on bone marrow, blood and spleen cell suspensions. It was found that, besides a redistribution and an absolute numerical decrease in the total number of HSC (to 45% of the original number), the percentage of cycling HSC decreased significantly as the leukemia progressed. In the bone marrow the S-phase fraction of HSC decreased from 38% to almost zero. Proliferating stem cells are thought to emigrate from the blood and lodge at extramedullary sites. Proliferation of HSC is best preserved in the spleen (in the terminal stage: 32 to 53% in S-phase). However, normal hemopoiesis is still insufficient and the rats die from the consequences of a lack of mature end cells. A possible mechanism underlying the changes and differences in kinetic behaviour of HSC in the various organs studied is discussed.     

Accuracy of Immunohistochemistry and Flow Cytometry in Estimating the Proportion of Leukemic Cells in S Phase in Chronic Myeloid Leukemia Patients

We compared the proportion of S phase cells in bone marrow and peripheral blood samples obtained from 17 patients with chronic myeloid leukemia (CML). Before sampling all patients received a one hour IV infusion of iododeoxyuridine (IdUrd). The proportion of S phase cells was studied by immunohistochemistry (IHC) in bone marrow biopsies, and by flow cytometry (FCM) in bone marrow aspirates and peripheral blood samples. The IdUrd labelling index (LI) in bone marrow biopsy sections (27.5 ± 1.8%) was significantly higher than the proportion of IdUrd labelled cells in bone marrow aspirate (15.1 ± 2.0%). The percentage of S phase cells in peripheral blood was approximately the same as that in the aspirate (12.4 ± 1.3%) and was correlated with that of bone marrow aspirate indicating a high degree of the aspirate dilution by peripheral blood. It is likely that the differences in % S phase cells in the aspirate and the biopsy result from this dilution. Estimates of the % S phase cells in the peripheral blood study by IHC and FCM were essentially the same. Samples labelled for one hour in vitro resulted in 1.5 fold higher LI than the same samples labelled in vivo. We conclude that estimates of the 8% S phase cells in the bone marrow of patients with CML should be made by infusing patients with IdUrd or BrdUrd with immunohistochemical evaluation of a marrow biopsy. Additionally in vitro labelling is not reflective of the percent S phase cells in vivo in patients.     

Bone marrow biopsy instead of 'marrow juice' for cell kinetic analysis. Comparison of bone marrow biopsy and aspiration material

Since cell kinetic bone marrow studies have so far exclusively been carried out on aspiration material and have yielded inconsistent or even contradictory results, we investigated the adequacy and reliability of aspirates for cell kinetic analyses in comparison to biopsies. Paired samples of bone marrow (133) were taken simultaneously by aspiration and Jamshidi biopsy from 48 patients with acute leukemias and 67 patients with non-leukemic disorders. Cell kinetic analysis by (1) flow cytometry (FCM) of cellular DNA and RNA content, (2) autoradiography for [3H]TdR pulse labelling indices and (3) liquid scintillation counting of [3H]TdR uptake revealed significantly higher values in biopsies (p less than 0.001) exceeding the corresponding results from aspirates on average by factors of 1.65 for FCM S-phase index, 1.90 for G0/1 cells with high RNA content, 1.82 for [3H]TdR LI and 1.90 for [3H]TdR uptake. In more than 70% of all samples results from biopsies were 1.1-11.4 times higher, indicating that aspirates were equivalent to biopsies in fewer tan 30% of cases. Cell kinetic analysis in vitro blood/biopsy mixtures and measurements of DNA synthesis rate in corresponding aspirates and biopsies revealed that these discrepancies are due to the contamination of aspirates with non-proliferating nucleated blood cells. Biopsy, however, was found to provide representative and reproducible sampling of marrow for cell kinetic studies and should replace the presently used aspirate already characterized as "unreliable marrow juice" by Dameshek et al. in 1937 [18].     

Prognostic value of flow cytometric DNA content analysis in granulosa cell tumor of the ovary

The nuclear DNA content and S-phase fraction of 23 ovarian granulosa cell tumors were measured from paraffin-embedded tissue with flow cytometry. Crude survival of the patients with a euploid tumor (17 diploid, one tetraploid) was more favorable than that of the patients with an aneuploid tumor (n = 5, P = 0.02). The percentage of S-phase cells was a good predictor of survival. If more than 6% S-phase cells were present in the DNA histogram, both crude survival (P = 0.0001) and survival corrected for intercurrent deaths (P = 0.0001) were clearly inferior as compared with tumors with less than 6% S-phase cells. The results indicate that DNA flow cytometric study may provide a rapid and valuable method to predict the biological behavior of granulosa cell tumors of the ovary.     

DNA analysis of breast cancer

Several methods for DNA analysis including DNA histogram and proto-oncogene amplification in human breast cancer, and the results recently reported were reviewed. A large number of DNA histograms obtained by flow cytometry have provided a possible correlation between DNA ploidy pattern and clinical outcome of breast cancer patients. Poor clinicopathologic factors, however, are not always in association with DNA aneuploidy and/or S-phase fraction rate, suggesting that further investigations on DNA ploidy are required. Amplification and/or over expression of proto-oncogenes in human breast cancers have been reported. Among them c-erbB-2 amplification may be one of the candidates for prognostic indicators of breast cancer survival. To determine any reliable biological factors exist further analysis of tumor DNA will be required in breast cancer research.     

Proliferative behaviour of an oestrogen sensitive rat mammary tumour: evidence for a paracrine interaction between tumour and stroma.

An oestrogen-sensitive rat mammary tumour (OES HR1) has been grown in normal female rats and in female and male rats supplemented with oestrone. In some rats, after the tumour was established, both exogenous and endogenous sources of oestrogen were removed--a treatment which inhibited further growth of the tumour. The proliferative characteristics of the tumours were measured by injecting the rats with deoxybromouridine (BrdU) 4 h before removing the tumour. Extracted nuclei were reacted with anti-BrdU and the labelling index and DNA content measured by flow cytometry. A correlation between the number of (diploid) host cells present and the number of (aneuploid) tumour cells in S-phase of the cell cycle was observed. This result suggests that there are paracrine interactions between tumour and host cells. We also observed that, on oestrogen ablation, the labelling index was significantly reduced while the percentage of cells in S-phase changed far less. The demonstration that there are cells in S-phase which are not proliferating highlights a possible problem with the measurement of proliferation in human tumours from a DNA histogram.     

Association of DNA index and S-phase fraction with prognosis of nodes positive early breast cancer

DNA index and S-phase fraction of 490 primary breast carcinomas were measured by flow cytometry, using paraffin-embedded tissue as a starting point. All patients had involvement of axillary lymph nodes and all were randomized onto one of the Ludwig Breast Cancer Trials I-IV between July 1978 and August 1981. The presence of an abnormal DNA stemline correlated with involvement of four or more axillary lymph nodes (P = 0.09), postmenopausal status (P = 0.008), estrogen receptor negativity (P = 0.03), and high tumor grade (P = 0.0005). Disease-free and overall survival for these patients was worse than for those with tumors of normal DNA content, but DNA aneuploidy did not have independent prognostic significance when allowance was made for its correlation with other prognostic features. The percentage of cells in S phase, which is an index of cell proliferation, was also analyzed in 285 DNA histograms considered to be capable of giving a valid estimate. Disease-free survival for 154 patients with S phase less than or equal to 10% was significantly longer than for those with S phase greater than 10% (P = 0.0008). S-phase greater than 10% was strongly correlated with high tumor grade (P less than 0.00001) and abnormal DNA index (P less than 0.00001) but only weakly correlated with nodal, hormone receptor, and menopausal status. Multivariate analysis using a Cox proportional hazard model suggested that the prognostic significance of the percentage of S phase was related to the association with tumor grade. Because it gives rapid, objectively determined information about biological aggressiveness DNA flow cytometry may well establish a role in routine clinical practice, but further technical refinements are required before it can be used for treatment decision making in nodes positive early breast cancer patients.     

Immunohistochemical approach to reveal the growth potential of uterine cancers using anti-BrdU antibody and Ki-67

To reveal the growth potential of uterine cancer, the population of S-phase and proliferating cells were examined with immunohistochemical technique using anti-BrdU antibody and Ki-67. BrdU is a thymidine analogue that is also incorporated into nuclear DNA and S-phase cells are recognized with anti-BrdU antibody. Ki-67 reacts a nuclear antigen present in proliferating cells (late G1, S, M and G2 phase). The percentage of BrdU-labeled cells (labeling index: LI) and that of cells recognized with Ki-67 (growth fraction: GF) were calculated. LI and GF were examined in 12 cervical cancers (LI: 16.0 +/- 6.0; GF: 32.2 +/- 11.2, mean +/- SD), 18 normal ectocervical portions (LI: 6.9 +/- 3.1; GF: 11.4 +/- 5.0), 13 endometrial cancers (LI: 15.9 +/- 5.0; GF: 26.2 +/- 9.0) and 11 normal endometrial tissues (LI: 12.2 +/- 6.1; GF 20.3 +/- 6.8). Indices of GF were always higher than LI in any cases. Both indices of LI and GF in malignant cases were higher than normal cases, therefore, high growth potential of uterine cancer was demonstrated. Both LI and GF of squamous cell carcinoma were higher than adenocarcinoma. In general, the values of GF were parallel to those of LI, and a regression line: Y = 1.44 X + 3.93 (r = 0.80) was obtained. Therefore, S-phase cells occupied about 60-70% of all proliferating cells. These results showed that these two parameters were useful to evaluate growth potential of uterine cancer, but calculation of GF using Ki-67 might be superior to LI using anti-BrdU antibody in terms of simplicity and rapidity.(ABSTRACT TRUNCATED AT 250 WORDS)     

Expression of nucleolar antigen p145 in bone marrow cells of patients with myeloid leukemias

The expression of a cell cycle-related nucleolar protein (p145) antigen was examined in the bone marrow aspirates of 45 individuals, three of whom had no malignant disease; 30 had a diagnosis of acute myeloid leukemia (AML), and 12 suffered from chronic myeloid leukemia (CML). While no evidence of p145 expression was found in the three normal bone marrow samples, it was noted to be the highest in patients with active leukemia, be they AML or blastic crisis of CML. There was a direct correlation between the percentage of blasts and the percentage of p145-positive cells in all patients. Double labeling with tritiated thymidine and p145 in AML patients with active leukemia showed that the majority of S-phase cells contained p145. Myeloblasts in both chronic phase and blastic crisis of CML expressed p145. Nine of 12 AML patients studied during remission had less than 5% p145-positive cells, but three showed 11%, 16%, and 33% positive cells. Since functionally/morphologically, these marrows were normal, the appearance of p145 may indicate a proliferative abnormality preceding maturation arrest and development of relapse. Thus we conclude that p145 is more commonly associated with immature cells and may serve as an early indicator of relapse in AML, but requires further study with larger numbers of patients.     

Indirubin in the treatment of chronic myelocytic leukemia (CML)--estimation of labelling index of bone marrow cells by 3H-TdR

Using 3H-TdR labelling radioautography in vitro, labelling index (LI) of the bone marrow cells from patients with CML was estimated before and after treatment by indirubin, and compared with myleran. The results showed that LI of the bone marrow cells was prone to decline after treatment by indirubin or myleran, and it was most marked on the myelocyte and polychromatophilic erythroblast. The LI reduction was less by indirubin than myleran. It suggests that the inhibition of indirubin on the bone marrow cells be weaker than that of myleran. However, long-term administration of indirubin would inhibit not only the granulocytic series but also the red cell series to a certain degree.     

Comparative analysis of radiometric and autoradiographic methods in the investigation of DNA synthetic activity of tumor cells in vitro

DNA synthesis in cell cultures of Ehrlich ascites carcinoma (EAC) was studied by means of autoradiography and liquid scintillation radiometry. The following parameters were compared for estimation of DNA synthesis: a) total radioactivity of incorporated [H-3]thymidine registered by a liquid scintillation counter, b) labeling index (LI) reflecting the relative number of cells which synthesize DNA, and c) mean grain count (MGC) indicating average intensity of DNA synthesis in the S-phase of the cell cycle. Kinetics of total radioactivity of incorporated [H-3]thymidine was determined to correlate better with LI. Changes in DNA synthesis in vitro were shown to be monitored with a higher level of significance by the radiometric technique, but autoradiography was more sensitive in some cases. Priority and informative value of the two methods were discussed. Autoradiography and liquid scintillation radiometry were concluded to be mutually complementary for investigation of DNA synthesis in tumor cells in vitro. Combination of the two methods was recommended for investigation of influence of cell proliferation inhibitors on DNA synthesis.     

The reproducibility of flow cytometric analyses in human tumors. Methodological aspects

Various methodological aspects of flow cytometry were studied in material from endometrial carcinoma, ovarian tumors and bladder carcinoma. Measurements of identical samples on two different occasions gave an excellent correlation of the obtained DNA values, r = 0.997 and a good reproducibility of the S-phase rate, r = 0.87. In large tumors different DNA values were found in 5/36 cases when central and surface biopsies were compared (indicating tumor heterogeneity) which stresses the importance of multiple biopsies. The S-phase rate of the surface biopsies was generally higher. Comparing staining with ethidium bromide and DAPI, a good correspondence of ploidy determinations in human bladder tumors was found, provided that diploid cells from the normal tissue component were used as internal reference. When ethidium bromide staining was used, there was a good agreement between the values of tumor ploidy obtained by external standardization using lymphocytes and internal standardization using diploid tumor cells, respectively. In DAPI staining with lymphocytes as an external standard the tumor ploidy was systematically overestimated by about 10% due to suboptimal staining of lymphocytes after 3 hours. This difference decreased after 18 hours of staining. Determination of S-phase fraction showed a good correlation between DAPI and ethidium bromide stained bladder tumor samples (r = 0.86). Human lymphocytes as an external standard showed good reproducibility. In conclusion, flow cytometric measurements of the ploidy level and S-phase rate are highly reproducible provided that tumor heterogeneity is taken into account and proper preparation and standardization methods are used.