Correlation of in vitro bromodeoxyuridine labeling index and DNA aneuploidy with survival or recurrence in brain-tumor patients

Liver hypertrophy and accelerated growth of implanted tumors in nonembolized liver of rabbit after left portal vein embolization

BACKGROUND

Portal vein embolization (PVE) has become a standard preoperative procedure to promote hypertrophy of the future remnant liver to reduce postoperative liver failure. Whether PVE accelerates tumor growth is still controversial. We developed a left PVE procedure and investigated its effect on liver hypertrophy and tumor growth in a rabbit liver tumor model.

MATERIALS AND METHODS

VX2 tumors were implanted in both the external left and right middle lobe (the bilateral group) or in the external left lobe only (the unilateral group) of rabbit liver. Both groups were further divided into a PVE or a sham/control group. Tumor volume and tumor growth rate as volume relative increase were determined by ultrasound. Liver volume-to-body weight index, an index for liver volume, was compared. Serum HGF was measured by ELISA.

RESULTS

In the bilateral PVE group, tumor volume and relative increase value in the nonembolized lobe were significantly (71% and 65%, respectively) greater than those in the control group at 5 d post-PVE. In the unilateral PVE group, liver volume-to-body weight index of the nonembolized lobes was significantly increased by 17%. Increase of serum HGF level after PVE was correlated well with both tumor growth and liver hypertrophy.

CONCLUSIONS

Left PVE promoted both the growth of implanted tumors and liver hypertrophy in the nonembolized liver, in which serum HGF might play an important role.

PCNA and Ki-67 in central nervous system tumors: correlation with the histological type and grade

Determination of proteins in the control of proliferation in normal cells helps a better understanding of cellular transformation and proliferation mechanisms. Measurement of proliferative activity is important in determining the tumor grade, recurrence span and malignancy. Proliferating cell nuclear antigen (PCNA) and Ki-67 are two of the nuclear markers used to demonstrate the proliferative phase of the cell cycle. In the present study, 63 central nervous system (CNS) tumors of various histologic types, diagnosed in Cukurova University Medical Faculty. Department of Pathology and graded according to WHO grading system were examined for PCNA and Ki-67 monoclonal antibodies using immunohistochemistry. Results were analyzed with statistical methods. Distribution of PCNA and Ki-67 LI (labeling index) values were determined for different tumor types. The highest PCNA and Ki-67 LI values were detected in medulloblastoma, malignant meningioma, primitive neuroectodermal tumor (PNET) and glioblastoma (GBM) groups, while pilocytic astrocytoma, meningioma, craniopharyngioma and oligodendroglioma showed the lowest values. In such tumors, the correlation between the increasing grade and PCNA and Ki-67 LI values were statistically significant. A correlation between the clinical outcome and Ki-67 and PCNA LI values was also detected. Conclusively, both markers can be used to evaluate the tumor grade and to asses the possibility of recurrence and malignancy in CNS tumors.

DNA ploidy and cell-cycle analysis in intracranial meningiomas and hemangiopericytomas: a study with high-resolution DNA flow cytometry

Although various DNA flow-cytometric studies have been performed on meningiomas, the role of DNA ploidy and the S-phase fraction (SPF) in predicting biological tumor behavior remains unresolved. Discrepant results in earlier studies might be due to different preparing, staining and measuring techniques; different quality standards; and lack of sophisticated computer software. In this study, high-resolution DNA flow cytometry using the DNA-specific dye DAPI (4', 6'-diamidino-2-phenylindol) was performed on stored frozen tissue from 128 microsurgically resected meningiomas and 7 hemangiopericytomas, including 17 recurrent meningiomas and 4 recurrent hemangiopericytomas. The computer software Multicycle 2.5 was used to determine the ploidy level and to perform cell-cycle analysis. DNA aneuploidy and SPF were significantly higher in atypical, anaplastic and recurrent meningiomas and correlated well with histopathological features such as focal necrosis, infiltration of dura mater and mitotic activity. Among 128 meningiomas, 42 had additional DNA aneuploid stem lines. No association between hypo- and hyperploidy and either histological subtype or clinical outcome was found. In 7 hemangiopericytomas, SPF was significantly higher compared to the benign meningioma group, while only 1 tumor was aneuploid. In all 42 DNA aneuploid tumors, cell-cycle analysis was performed separately for the euploid and aneuploid stem lines. The proliferation parameters (SPF, G2/M phase) were significantly higher in the DNA aneuploid stem lines. DNA ploidy and SPF are thus useful indicators of different biological behavior within identical histological subgroups in meningiomas.

S-phase fraction, 5-bromo-2'-deoxy-uridine labelling index, duration of S-phase, potential doubling time, and DNA index in benign and malignant brain tumors

Seventy-one histologically malignant brain tumors, 52 histologically benign brain tumors, and 14 cerebral metastases were characterized according to DNA content and proliferative capacity. DNA ploidy, DNA index (DI), S-phase fraction (SPF), 5-bromo-2'-deoxy-uridine (BrdUrd) labelling index (LI), duration of S-phase (Ts), and potential doubling time (Tpot) were assessed by flow cytometry (FCM). In histologically benign tumors, a high percentage of DNA diploid tumors and a low proliferative capacity in DNA diploid tumors were found. Histologically malignant tumors and cerebral metastases were both found to be characterized by a low percentage of DNA diploid tumors and a high proliferative capacity in DNA diploid tumors. The proliferative capacity of DNA aneuploid benign tumors and that of DNA aneuploid malignant tumors, however, appeared not to differ significantly. The number of DNA aneuploid tumors was small. Duration of S-phase was short (range 3.9-4.7 hr) and appeared not to differ between the three groups. From this, the observed differences in Tpot values should be accredited mainly to differences in LI. High-grade as well as low-grade gliomas both appeared to be characterized by malignant (FCM) features, i.e., 1) a high percentage DNA aneuploidy, 2) a high mean DI (for DI > 1), and 3) a high proliferative capacity.

In vivo determination of 5-bromo-2'-deoxyuridine incorporation into DNA tumor tissue by a new 32P-postlabelling thin-layer chromatographic method

The halopyrimidine 5-bromo-2'-deoxyuridine (BUDR) can serve as one of many indicators of tumor malignity, complementary to histologic grade. We have developed a thin-layer chromatographic (TLC) technique that can assess tumor DNA base composition and analogue (BUDR) incorporation which vies with immunochemistry for BUDR. This requires post-labeling DNA by nick-translation and radioactive 5'-phosphorylation of representative 32P-alpha-dNMPs (deoxynucleotide monophosphates). Subsequent 3'-monophosphate digest exchanges a radioactive 32PO4 for the neighboring cold nucleotide. Separation in two dimensional PEI-cellulose TLC is carried out in acetic acid, (NH4)2SO4, and (NH4)HSO4. TLC of dNMPs was applied to control HeLa DNA, and HeLa cells receiving BUDR. BUDR is detected in 10(6) HeLa cells after 12-72 h incubations. Findings in HeLa DNA demonstrate normal TLC retention factors for all 32P-dNMPs. Two dimensional R(F) (x,y axes in cm) demonstrate: dAMP=1.4, 9.4; dCMP=10.0, 13.5; dGMP=4.6, 4.4; dTMP=9.0, 7.4; and BUDRMP 6.4, 6.6. This technique quantifies BUDR--which parallels tumor S phase, and serves as an indicator of labelling index (LI).

DNA ploidy and S-phase in recurrent astrocytomas: a retrospective study by flow cytometry of deparaffinized specimens

Twenty-two patients with recurrent astrocytic tumors were treated surgically two or even three times. At the time of the first surgery 6 tumors were fibrillary astrocytomas (grade II), 9 anaplastic astrocytomas (grade III) and 7 glioblastomas (grade IV). Histopathological specimens from second surgery demonstrated in 12 cases signs of higher grades of malignancy. Flow cytometry (FCM) did not reveal any significant changes of S-phase fraction (p = 0.55). This study supports the theory that, given enough time, the histopathology of brain tumors change significantly from more benign forms to more malignant ones. The flow cytometry (FCM) could trace a weak tendency to higher S-phase fractions at the time of the second surgery. No apparent change of ploidy pattern was observed. In spite of the unequivocal histopathological changes of the recurrent astrocytomas the flow cytometry failed to indicate similar changes in terms of ploidy and S-phase fraction parameters.

Comparison of mathematical formulas used for estimation of DNA synthesis time of bromodeoxyuridine-labelled cell populations with different proliferative characteristics

Growth kinetic data of human tumours, obtained by flow cytometric analysis of cells labelled with bromodeoxyuridine (BrdUrd) might provide prognostic information and allow prediction of response to radio- and chemotherapy. However, the theoretical models applied for calculation of growth kinetic data are not fully evaluated. The purpose of this study was to investigate the dependence of the estimation of DNA synthesis time (Ts) on sampling time after BrdUrd labelling, using four different mathematical formulas (Begg et al., White & Meistrich, White et al. and Johansson et al.) which have been developed for the evaluation of flow cytometry-derived data of BrdUrd-labelled cells. In addition, we have investigated the influence of the growth kinetic properties of the cell populations using two cultured cell lines (one slow and one fast growing), and two hetero-transplanted human tumours. The dependence of the estimation of Ts on sampling time was more or less pronounced, depending on the cell population examined and on the formula used. In the fast growing cell line, the estimates of Ts did not vary significantly with sampling time when using the formulas by White et al., whereas in the slow growing cell line, the estimates of Ts did not show any significant dependence on sampling time when using the formula by Johansson et al. In the tumours, the estimation of Ts depended on sampling time with all formulas used, although to different degrees. In one of the tumours, this was mainly caused by the influence of mouse cells, as we demonstrate. Our results indicate that the proliferative characteristics of a cell population should be taken into consideration when choosing a mathematical formula in order to attain Ts values that are independent of sampling time.

A correlative study of in vivo and in vitro labeling index using bromodeoxyuridine in human brain tumors

The labeling index (LI) of 216 cases of human brain tumors was determined by the immunohistochemical technique with monoclonal antibody to bromodeoxyuridine (BrdU). The proliferative potential of 110 cases was estimated using the intra-operative intravenous infusion of BrdU at a dose of 200 mg/sq m. In another 106 cases, the in vitro technique of incubating freshly resected tumor tissue fragments with 100 microM bromodeoxyuridine was used. The BrdU LI in these tumors was then correlated with the histological types and the data as determined by both the in vivo and in vitro BrdU incorporation were compared. The results indicate that although in vivo and in vitro techniques could possibly provide equivalent data in some histologic types, a clear statistically valid proof however is not apparent from this study.

Prognostic factors in gliomas. A multivariate analysis of clinical, pathologic, flow cytometric, cytogenetic, and molecular markers

BACKGROUND

The ability to divide subsets of patients with glial neoplasms into prognostic groups currently is limited because only a few clinical and pathologic variables are available. The goal of this investigation was to identify biologic factors of potential prognostic value in patients with cerebral gliomas.

METHODS

This prospective investigation used clinical, pathologic, flow cytometric, cytogenetic, and molecular genetic variables as potential prognostic factors in 207 patients with newly diagnosed gliomas (153 astrocytic tumors of the fibrillary type, 31 oligodendrogliomas, and 23 pilocytic astrocytomas). Classification and regression tree (CART) analysis was performed as part of the multivariate statistical analysis.

RESULTS

The age of the patient and the grade of the tumor were confirmed as strong prognostic factors. Cytogenetic or molecular genetic abnormalities of chromosomes 7 and 10 were associated with poor survival in univariate analysis (P < 0.0001). CART multivariate analysis identified several subsets of patients with different prognoses. In the subset of patients younger than 66.5 years with Grade 4 tumors, the survival time was longer for those with aneuploid tumors than for those with nonaneuploid tumors. In the subset of patients with Grades 1-3 tumors, the survival time was longer for those whose tumors had a %G2M of less than 6.9 than for those whose tumors had a %G2M of 6.9 or greater.

CONCLUSION

This investigation provides further evidence that flow cytometry, cytogenetic, and molecular genetic factors that may have prognostic value in patients with gliomas can be identified.

Prognostic significance of bromodeoxyuridine labeling in primary and recurrent glioblastoma multiforme

To evaluate the prognostic significance of bromodeoxyuridine (BUdR) labeling index (LI) and to estimate tumor proliferative potential, BUdR LI was examined in 98 patients having a primary diagnosis of glioblastoma multiforme (GBM); 55 underwent infusion of 200 mg/m2 of BUdR at the time of the primary resection and 49 underwent infusion at the time of the second resection. The tumors of six patients were labeled at both operations. The tumor specimens were stained with hematoxylin and eosin for histopathology and by immunohistochemical methods to determine the ratio of labeled to unlabeled cells, i.e., BUdR LI. The median BUdR LIs for the primary and recurrent GBM were significantly different at 6.8 and 2.6%, respectively (P < 0.0001). However, there was no significant association between BUdR LI at the first or second operation and survival (log rank, P = 0.12; Cox regression analysis, P = 0.91; log rank, P = 0.55; Cox regression analysis, P = 0.17, respectively). Patients who underwent a second operation within 10 months of the first operation had a lower BUdR LI than did patients with a longer interval between procedures (P = 0.078; Spearman rank correlation, 0.26). The aggressive biological behavior of GBM is dependent on complex cellular kinetics, not simply on the number of cells within the S phase of the cell cycle. Caution should be used when determining prognosis with BUdR LI in GBM.

Cell kinetic analysis in pediatric brain and spinal tumors: a study of 117 cases with Ki-67 quantitation and flow cytometry

We present cell kinetic data including Ki-67 quantitation and flow cytometry on 117 pediatric brain/spinal cord tumors and review the literature. Although, in general, these proliferation indices are in agreement with the histologic grade, they are useful in prognostication in some instances when the histological features of malignancy are equivocal. Specific examples in which flow cytometry may prove particularly useful in this context are childhood ependymomas, which do not show frank anaplasia but have cellular foci with focal increase in mitoses, and choroid plexus neoplasms, where elevated S phase fractions have been associated with an adverse outcome. Thus Ki-67 quantitation and flow cytometry not only serve as useful adjuncts to conventional histologic grading but also in specific instances may provide new information on tumor prognosis.

Proliferative potential and prognostic evaluation of low-grade astrocytomas

Despite their histological similarity, low-grade astrocytomas vary widely in their clinical behavior. To elucidate this variable behavior, we measured the proliferative potential of 69 primary and 18 recurrent low-grade astrocytomas and correlated the results with the clinical characteristics and outcome. Each patient received an intravenous infusion of bromodeoxyuridine (BUdR); BUdR-labeled nuclei in excised tumor specimens were identified by immunoperoxidase staining. The BUdR labeling index (LI), or S-phase fraction, ranged from < 1 to 9.3%; the LI was < 1% in 64 (74%) patients and > or = 1% in 23 patients (26%). The LI did not appear to be associated with age, sex, tumor location, or whether the tumor was primary or recurrent. A Cox proportional-hazards analysis of the influence of the LI and other variables (age, sex, tumor location, extent of surgery, primary versus recurrent tumor) on survival showed that the LI and extent of surgery (total resection, subtotal resection, biopsy) were significant in predicting both survival and progression-free survival for all patients and for patients with primary tumors. The LI was also significant in predicting progression-free survival for patients with recurrent tumors. The correlation between the BUdR LI and both survival and time to recurrence suggests that the outcome of low-grade astrocytomas varies according to the proliferative potential. The growth rate of these histologically similar tumors should be assessed individually in order to select specific treatment.

Correlation of proto-oncogene expression and proliferation and meningiomas

Proliferation and proto-oncogene expression in 19 meningiomas of typical and atypical histology were analyzed in an attempt to understand the mechanism of growth that characterizes the neoplastic process in these tumors. Proliferation was estimated as the proliferative index by the enumeration of S-phase cells in imprints of tumor tissue exposed to bromodeoxyuridine in vitro, and the gene expression of c-myc, c-fos, c-src, c-H-ras, N-myc, acidic and basic fibroblast growth factor, insulin-like growth factors I and II, platelet-derived growth factor-alpha, and epidermal growth factor was quantified by messenger ribonucleic acid dot-blot hybridization assay. Atypical and malignant tumors had significantly higher proliferative indexes than did their nonmalignant counterparts. Levels of c-myc and c-fos messenger ribonucleic acid were elevated more than fivefold in 72 and 78% of the tumors, respectively, relative to the lowest levels detected in the series. Levels of growth factor messenger ribonucleic acid were sporadically elevated; 37 to 44% of tumors had more than fivefold enhanced levels of acidic and basic fibroblast growth factor. Positive correlations between proliferation and proto-oncogene/growth factor expression were found for c-myc in atypical/malignant tumors and for epidermal growth factor in fibroblastic meningiomas. Deregulated expression of c-myc and c-fos common to both typical and atypical tumors suggests that these are early events in the meningioma tumor process that may disturb the control of cell differentiation and together with fibroblast growth factors are likely to endow the transformed cell with a selective growth advantage by reducing the requirement for exogenous mitogens and by providing a niche for the growth of the tumor clone. Positive correlation of c-myc levels with proliferation in atypical/malignant meningiomas implies that this is a feature of malignancy and indicates continued disruption of the negative regulation of proto-oncogene expression, perhaps by tumor suppressor gene losses, during the course of tumor progression.

Detection of proliferating cell nuclear antigen in gliomas and adjacent resection margins

We describe a technique for estimating the number of proliferating cells in gliomas and adjacent resection cavities after tumor removal. Proliferating cell nuclear antigen (PCNA) is a nuclear protein associated with the cell cycle. Anti-PCNA antibody staining provides a semiquantitative estimate of the number of proliferating cells found in fixed tissue embedded in paraffin. The extent of the staining of tumor cells of glioblastoma multiforme, anaplastic astrocytomas, low-grade gliomas, and other lesions of the brain with anti-PCNA antibody is correlated with the histological diagnosis. In addition, the labeling of the margins after resection of gliomas and other lesions with anti-PCNA antibody is also associated with the histological diagnosis of the lesion. This technique may be useful in estimating the "biological" extent of resection and in predicting the recurrence patterns of gliomas.

Clinical evaluation of DNA index in human brain tumors

We evaluated the DNA index values in 28 brain-tumor patients showing DNA aneuploidy on the DNA histogram obtained by flow cytometry. In 17 cases among them, the DNA index values were compared to BUdR labeling indices (BUdR-LI). Average DNA indices of individual tumor types were 1.89 +/- 0.10 (6 glioblastomas multiforme), 1.93 +/- 0.05 (4 anaplastic astrocytomas), 1.68 +/- 0.09 (3 malignant meningiomas), 1.24 +/- 0.03 (7 meningiomas), 1.19 +/- 0.05 (2 pituitary adenomas) and 1.86 +/- 0.24 (6 metastatic brain tumors). Slow-growing tumors such as meningioma and pituitary adenoma, had lower values than malignant tumors, except a case of a metastatic brain tumor. All malignant meningiomas had higher DNA index values than the other meningiomas. No difference was observed between the DNA index values of glioblastomas multiforme and anaplastic astrocytomas. DNA index values were correlated with BUdR-LI values (p < 0.05), and seven cases with low DNA index values (1.13-1.31) had low BUdR-LI values of less than 1%. Others with high DNA index values of 1.44 or more had high BUdR-LI values of 3.5% or more. These results suggest that high DNA index values is related to tumor malignancy, and despite the presence of some disparities, many slow-growing brain tumor has low DNA index values.

Use of [18F]fluorodeoxyglucose positron emission tomography in patients with primary malignant brain tumors

In patients with malignant gliomas, [18F]fluorodeoxyglucose positron emission tomography (FDG-PET) may discriminate tumor progression from radionecrosis. We evaluated data from 50 patients undergoing FDG-PET for suspicion of tumor progression. Forty-nine were treated with surgery, 48 with radiotherapy, and 37 with chemotherapy. Twenty-one had intensive radiotherapy with either three daily treatments in two 5-day periods and intravenous carboplatin (17) or interstitial brachytherapy or stereotactic radiotherapy. Twenty underwent surgery after magnetic resonance imaging/FDG-PET; 9 demonstrated increased uptake of FDG and evidence of tumor, whereas 6 had decreased uptake and no evidence of tumor. In 5 patients, there was no correlation (all had intensive radiotherapy). In 17 patients who received bromodeoxyuridine intravenously just before surgery, the bromodeoxyuridine labeling index corresponded to the histological appearance in all but 2 patients (both had received intensive radiotherapy). In 30 patients without surgery, decreased uptake of FDG suggested prolonged survival; increased uptake of FDG did not predict survival. Eight of 10 with intensive radiotherapy had decreased label uptake. We conclude FDG-PET for evaluation of patients with possible recurrent tumors requires more study. In patients with intensive radiotherapy, FDG-PET results cannot be correlated accurately with tumor progression.

Interphase cytogenetics: a new tool for the study of genetic changes in brain tumors

Interphase cytogenetics is the application of nonradioactive in situ hybridization with chromosome-specific DNA probes to interphase nuclei. In this study, interphase cytogenetics was used to investigate 66 primary brain tumors (33 gliomas, 30 meningiomas, and three medulloblastomas) for numerical chromosomal aberrations of chromosomes 1, 6, 7, 10, 11, 17, 18, X, and Y. Of the 33 gliomas (17 astrocytomas grades II, III, and IV, five oligoastrocytomas, seven oligodendrogliomas, and four ependymal tumors), 22 were near diploid, while the remaining 11 showed a significant triploid or tetraploid component. The predominant specific aberrations in gliomas were an over-representation of chromosome 7 (13 cases) and an under-representation of chromosome 10 (16 cases). These changes were observed in grade III and grade IV astrocytomas, as well as in oligodendrogliomas. Other frequent numerical changes were a gain of chromosome 17 (six cases) and a loss of chromosome 18 (seven cases). This loss of chromosome 18 seemed relatively specific for gliomas with an oligodendroglial component (six cases). Only two of 33 gliomas displayed no genetic abnormality with the probes used. Seven patients with astrocytomas died of their brain tumor during the clinical follow-up period. Their astrocytomas did not show a different chromosomal constitution compared to the other gliomas. For the meningiomas, the probe panel was extended with a probe specific for chromosome 22. Loss of chromosome 22 was obvious in 21 of the 30 meningiomas, and was the sole abnormality in 11 meningiomas; in the other 10, this loss was associated with other chromosomal changes. Five of these tumors with additional aberrations were recurrent or atypical meningiomas. It is suggested that interphase cytogenetics can contribute to a better understanding of the biological behavior of these tumors and possibly result in better insights into prognosis and strategies for therapy.

Biology and treatment of gliomas

This review discusses some of the recent advances in glioma research and treatment. Our understanding of the characteristics of these tumors has been strengthened by the application of molecular biologic and genetic techniques to pathologic grading and therapy outcome. Newer attempts to correlate imaging modalities to pathologic grading are also discussed. It is anticipated that these developments will strengthen our ability to design improved treatment strategies, an essential goal inasmuch as current treatment schemes have limited benefit. More work needs to be done to understand the biology of these tumors especially the complex interactions of their cytokine expression, multiplicity of genetic abnormalities, and their local environment. Only then will be able to develop improved therapeutic interventions.

Pituitary adenoma proliferative indices and risk of recurrence

Twenty unselected pituitary adenomas have been examined for proliferative indices (PIs), and anterior pituitary hormone expression. All but two of the tumours were non-functional with proliferative indices from less than 0.1 to 0.5%. Two tumours were null cell adenomas with PIs less than 0.1 and 0.2%. The PIs of the three recurrent tumours was less than 1.0%. Gonadotrophin and TSH immunoreactivity was heterogenous and was found in 12/20 (60%) of the tumours. There was no significant relationship between PI, hormone expression or any other measured parameter. The biological behaviour of pituitary adenomas with a PI of less than 0.1% is uncertain, but those with a PI of greater than 0.1% are more likely to recur although longer follow-up is needed to confirm this.