IUdR polymers for combined continuous low-dose rate and high-dose rate sensitization of experimental human malignant gliomas

Local polymeric delivery enhances IUdR radiosensitization of human malignant gliomas (MG). The combined low-dose rate (LDR) (0.03 Gy/h) and fractionated high-dose rate (HDR) treatments result in cures of experimental MGs. To enhance efficacy, we combined polymeric IUdR delivery, LDR, and HDR for treatments of both subcutaneous and intracranial MGs. In vitro: Cells (U251 MG) were trypsinized and replated in triplicate 1 day prior to LDR irradiation in media either without (control) or with 10 microM IUdR. After 72 hr, LDR irradiation cells were acutely irradiated (1.1 Gy/min) with increasing (0, 1.25, 2.5, 5.0, or 10 Gy) single doses. Implantable IUdR polymers [(poly(bis(p-carboxyphenoxy)-propane) (PCPP): sebaic acid (PCPP:SA), 20:80] (50% loading; 10 mg) were synthesized. In vivo: For flank vs. intracranial tumors, mice had 6 x 10(6) subcutaneous vs. 2 x 10(5) intracranial cells. For intracranial or subcutaneous MGs, mice had intratumoral blank (empty) vs. IUdR polymer treatments. One day after implantation, mice had immediate external LDR (3 cGy/h x 3 days total body irradiation) or HDR (2 Gy BID x 4 days to tumor site) or concurrent treatments. For the in vitro IUdR treatments, LDR resulted in a striking increase in cell-killing when combined with HDR. For the in vivo LDR treatments of flank tumors, the growth delay was greater for the IUdR vs. blank polymer treatments. For the combined LDR and HDR, the IUdR treatments resulted in a dramatic decrease in tumor volumes. On day 60 the log V/V0 were -1.7 +/- 0.22 for combined LDR + HDR + IUdR polymer (P < 0.05 vs. combined LDR + HDR + blank polymer). Survival for the intracranial controls was 22.9 +/- 1.2 days. For the blank polymer + LDR vs. blank polymer + LDR + HDR treatments, survival was 25.3 +/- 1.7 (P = NS) vs. 48.1 +/- 3.5 days (P < 0.05). For IUdR polymer + LDR treatment survival was 27.3 +/- 2.3 days (P = NS). The most striking improvement in survival followed the IUdR polymer + LDR + HDR treatment: 66.0 + 6.4 days (P < 0.05 vs. blank polymer + LDR + HDR). The polymeric IUdR delivery plus combined continuous LDR and HDR treatments results in growth delay and improved survival in animals bearing the MG xenografts. This treatment may hold promise for the treatment of human MGs.     

Synthetic,implantable polymers for local delivery of IUdR to experimental human malignant glioma

Purpose: Recently, polymeric controlled delivery of chemotherapy has been shown to improve survival of patients with malignant glioma. We evaluated whether we could similarly deliver halogenated pyrimidines to experimental intracranial human malignant glioma. To address this issue we studied the in vitro release from polymers and the in vivo drug delivery of IUdR to experimental human U251 glioblastoma xenografts.Methods and Materials: In vitro: To measure release, increasing (10%, 30%, 50%) proportions of IUdR in synthetic [(poly(bis(p-carboxyphenoxy)-propane) (PCPP):sebacic acid (SA) polymer discs were serially incubated in buffered saline and the supernatant fractions were assayed. In vivo: To compare local versus systemic delivery, mice bearing flank xenografts had intratumoral or contralateral flank IUdR polymer (50% loading) treatments. Mice bearing intracranial (i.c.) xenografts had i.c. versus flank IUdR polymer treatments. Four or 8 days after implantation of polymers, mice were sacrificed and the percentage tumor cells that were labeled with IUdR was measured using quantitative microscopic immunohistochemistry.Results: In vitro: Increasing percentage loadings of IUdR resulted in higher percentages of release: 43.7 + 0.1, 70.0 + 0.2, and 90.2 + 0.2 (p < 0.001 ANOVA) for the 10%, 30%, and 50% loadings, respectively. In vivo: For the flank tumors, both the ipsilateral and contralateral IUdR polymers resulted in similarly high percentages labeling of the tumors versus time. For the ipsilateral IUdR polymers, the percentage of tumor cellular labeling after 4 days versus 8 days was 45.8 ± 7.0 versus 40.6 ± 3.9 (p = NS). For the contralateral polymer implants, the percentage of tumor cellular labeling were 43.9 ± 10.1 versus 35.9 ± 5.2 (p = NS) measured 4 days versus 8 days after implantation. For the i.c. tumors treated with extracranial IUdR polymers, the percentage of tumor cellular labeling was low: 13.9 ± 8.8 and 11.2 ± 5.7 measured 4 and 8 days after implantation. For the i.c. tumors having the i.c. IUdR polymers, however, the percentage labeling was comparatively much higher: 34.3 ± 4.9 and 35.3 ± 4.0 on days 4 and 8, respectively. For the i.c. tumors, examination of the percentage cellular labeling versus distance from the implanted IUdR polymer showed that labeling was highest closest to the polymer disc.Conclusion: Synthetic, implantable biodegradable polymers provide the local, controlled release of IUdR and result in the high, local delivery of IUdR to experimental intracranial human malignant glioma. This technique holds promise for the local delivery of IUdR for radiosensitization of human brain tumors.     

Immunologic response to cryoablation of breast cancer

Summary Purpose.With improvements in breast imaging and image-guided interventions, there is interest in ablative techniques for breast cancer. Cryosurgery initiates inflammation and leaves tumor-specific antigens intact, which may induce an anti-tumor immune response. To help define the mechanisms involved in the cryoimmunologic response, we compared cryosurgery to surgery in a murine model of breast cancer. Experimental designBALB/c mice with MT-901 tumors underwent cryoablation or resection. Mice successfully treated were re-challenged with MT-901 or RENCA. Serum cytokine levels were analyzed by ELISA. Tumor draining lymph nodes (TDLN) and spleens were harvested, lymphocytes were activated and assessed for a specific anti-tumor response by both an interferon- (IFN) release assay and ELISPOT. NK cell activity was assessed by cytotoxicity against YAC-1, an NK-susceptible cell line. Results.After re-challenge, tumors developed in 86% of mice treated by surgical excision compared to 16% of mice treated by cryosurgery (p=0.025). Cryoablation of MT-901 had no effect on re-challenge with RENCA. Cryoablation led to significantly higher levels of interleukin (IL)-12 (383.6pg/ml±32.8 versus 251.6±16.5, p=0.025) and IFN- (1564pg/ml±49 versus 1244pg/ml±101, p=0.009), but no changes in IL-4 or IL-10. Tumor-specific T-cell responses were evident after cryosurgery in lymphocytes from TDLN but not from spleen. Cryoablation also increased NK activity compared to surgery (24.5%±17.3 versus 16.5%±5.9, p<0.001). Conclusion.Cryoablation results in the induction of both a tumor-specific T-cell response in the TDLN and increased systemic NK cell activity, which correlates with rejection of tumors upon re-challenge.     

Autologous Adjuvant Linked Fibroblasta Induce Anti-Glioma Immunity: Implications for Development of a Glima Vaccine

Objectives: Adjuvant-linked vaccines have been shown to induce anti-tumor immunity in patients with a variety of solid tumors. In this study we describe an in vitro model of active immunotherapy using autologous fibroblasts as immunogen. Correlative results from glioma patients immunized with autologous fibroblasts are also described. Methods: Peripheral blood lymphocytes (PBLs) from normal subjects were immunized in vitro against autologous skin fibroblasts coupled to the adjuvant muramyl dipeptide. The lymphocytes developed cell-mediated cytotoxicity that was measured with a short-term chromium release assay. Results of in vitro experiments were compared to data derived from glioma patients immunized with subcutaneous injection of an autologous adjuvant-linked fibroblast vaccine. Glioma target cells and fibroblast immunogens were derived from early passage primary tissue culture. Results: A comparison of autologous vs. homologous immunogen indicated that major histocompatibility complex matching was required at the sensitization stage of immunity (17.2 ± 3.4% specific lysis vs. 0.4 ± 3.1%, P < 0.01). Pre-treatment of fibroblast immunogen cells with interferon gamma (IFN-) was found to significantly increase immunity (42.2 ± 10.0%, P < 0.01), as did IFN- pre-treatment of tumor target cells (35.8 ± 9.0%, P < 0.01). The positive effect of IFN- was diminished by treatment of cells with IFN-. These in vitro results correlated well with in vivo data derived from glioma patients immunized with an autologous adjuvant-linked fibroblast vaccine. PBLs from patients developed direct cell-mediated cytotoxicity against autologous tumor cells. Lysis of tumor targets after in vivo immunization increased over a three-week interval (from 1.2 ± 3.0% to 21.0 ± 3.4%, P < 0.01) while lysis of a non-MHC matched control cell line remained essentially unchanged. Conclusions: Specific lysis of glioma targets in vitro was achieved after in vivo sensitization with autologous adjuvant-linked fibroblasts. Collectively, the data indicate that biochemically modified autologous cells can stimulate anti-glioma immunity in humans. The degree of specific immunity seen in our patients compares favorably with other published series using glioma cells as an antigenic source. Accordingly, fibroblasts may represent a practical alternative to glioma cells for vaccine construction.     

Intracavitary VEGF,bFGF, IL-8, IL-12 levels in primary and recurrent malignant glioma

Intracavitary levels of VEGF, bFGF, IL-8 and IL-12 were evaluated by ELISA in 45 patients, 7 with recurrent anaplastic astrocytoma (rAA), 12 with glioblastoma (GBM) and 26 with recurrent glioblastoma (rGBM). In 25 patients plasma levels of the molecules were also quantitated. Twenty-three healthy controls were also studied for plasma concentrations of the same molecules.Plasma levels of VEGF (mean 33.89 ± 6.71pg/ml) and bFGF (mean 11.1 ± 3.24pg/ml) were higher in patients than in controls (mean 16.78 ± 3.7pg/ml for VEGF, mean 0.21 ± 0.09pg/ml for bFGF) (p = 0.04 and p = 0.001, respectively) while plasma IL-12 levels were lower (mean 45.6 ± 1.5pg/ml in patients, mean 79.7 ± 1.3pg/ml in controls) (p = 0.009).Intracavitary VEGF levels were 5–53.307 fold higher (mean 90,900 ± 24,789pg/ml) than in the corresponding plasma. Also IL-8 concentrations were higher in intracavitary fluid (mean 6,349.76 ± 1,460.93pg/ml) than in plasma (mean 43.44 ± 24.82pg/ml). Maximum VEGF levels were found in tumor fluid of recurrent glioblastoma patients (mean 147,678 ± 39,903pg/ml), intermediate levels in glioblastoma patients (mean 20,322 ± 11,892pg/ml) and lower levels in rAA patients (mean 9,111 ± 5,789pg/ml). The data also suggest that higher intracavitary levels of VEGF and IL-8, and lower IL-12 levels, may be correlated with shorter adjunctive survival times, but more data will need to be collected to establish this correlation clearly.     

In vitro evaluation of a novel chemotherapeutic agent,adozelesin, in gynecologic-cancer cell lines

Summary Adozelesin is a derivative of an extremely cytotoxic compound, CC1065. This entirely new class of drug binds preferentially to DNA and facilitates alkylation reaction. In the present study, we used the adenosine triphosphate (ATP) chemosensitivity assay to compare the cytotoxic potency of Adozelesin with that of common chemotherapeutic agents in ten gynecologic-cancer cell lines. Flow cytometry was also used to study its effects on cell-cycle kinetics. The mean drug concentrations required to produce a 50% reduction in ATP levels as compared with controls [IC₅₀] were: Adriamycin, 0.17±0.06 m; 4OH-Cytoxan, 18±3 m; cisplatin, 17±7 m; 5-fluorouracil, 183±116 m; and Adozelesin, 11.0±5.4pm. Thus, Adozelesin was 10⁴–10⁷ times more potent than Adriamycin, cisplatin, 5-fluorouracil, and Cytoxan. Cell kinetics studies revealed significant S and G2 blocks such as those previously reported for other alkylating agents.Supported in part by an American Cancer Society Clinical Oncology Fellowship and an American Cancer Society/Florida Division Startup Grant (both awarded to H. N. N.)     

Predicting recurrence in axillary-node negative breast cancer patients

Summary This study attempted to identify the risk groups in axillary node negative breast cancer patients using validated first-generation prognostic clinical and pathologic factors. An updated 10-year follow-up in 407 such patients treated by surgery alone at Roswell Park between 1976–1987 showed a 10-year recurrence rate (RR) of 19% (95% confidence interval ±5%). Predictors of outcome were, in order of strength: (1) Tumor size (p= 0.0006); RR at 10 years was 2% ± 4 for tumors 0.5cm, 6% ± 7 for tumors 0.6-1.0cm, 16% ± 9 for 1.1–2cm, 29% ± 12 for 2.1–5cm, and 40% ± 31 over 5cm; (2) Histologic differentiation (p = 0.017); poorly differentiated/anaplastic (P/A) tumors had a greater RR (24% ± 8) than well or moderately differentiated (W/M) tumors (13% ± 8); (3) Age (p = 0.046); patients < 35 showed a RR of 28% ± 20, pts 35–50, 22% ± 10, and pts > 50, 17% ± 7 (p = 0.046). Cox Model analysis showed tumor size (4 groups) significant at < 0.0001, histologic differentiation (2 groups) significant at < 0.0005 after allowing for size, and age (±50) significant at <0.05 after allowing for size and differentiation.Combining these variables into subgroups enables selecting groups at various risks of recurrence. Groups with low risk are: (1) patients with tumor1cm, W/M (0% RR), (2) patients with ductal carcinomain situ with microinvasion (0% RR), and (3) patients with tumors 1cm, P/A (8% RR). In a suggestive finding in this last group, those over age 50 achieved a RR of 3% ± 6, while those age 50 or less had RR 14% ± 15. With the exception of this last group, all should be considered highly curable using loco-regional therapy alone, and might be spared the risks and costs of routine systemic adjuvant therapy. Groups with high risk are: (1) patients with tumors > 2cm (RR 32% ± 12), and (2) patients with tumors 1.1–2cm, P/A (RR 21% ± 14). These should receive adjuvant therapy. Groups with intermediate risk are patients with tumor 1.1-2cm, W/M (RR 12% ± 12). In a suggestive finding, those in this group over age 50 had a RR of 11% ± 12, while those up to 50 had a RR of 17% ± 30. These patients should be considered to be prime candidates for clinical trials.Adding second generation factors such as DNA ploidy or S-phase fraction to first generation factors should provide additional information on which to base therapy decisions, particularly in the gray area of intermediate risk. Our study indicates that node-negative breast cancer patients represent a heterogeneous population in terms of risk and prognosis, and that an individualized approach to adjuvant therapy should be taken.Presented in part at the 14th annual San Antonio Breast Cancer Symposium, December 6-7, 1991, and the American Association for Cancer Research, San Diego, CA, May 20-23, 1992     

Are Childhood and Adult Medulloblastomas Different? A Comparative Study of Clinicopathological Features,Proliferation Index and Apoptotic Index

Childhood medulloblastomas have been suspected to be biologically different from adult tumors, though comparative studies are sparse in the literature. The present study aims to establish any differences or nexus in the biological characteristics between childhood and adult medulloblastomas. A total of 181 medulloblastomas were studied with respect to clinical and histological characteristics, MIB-1 labeling index (MIB-1 LI), apoptotic index (AI), ratio of apoptotic to LI, p53 and Bcl-2 protein expressions. Two-thirds (112) of the 181 medulloblastomas occurred in children (15 years) and 69 in adults (>15 years). Childhood tumors were more commonly of classical histology and midline location while the desmoplastic variant and lateral location occurred more frequently in adults. Adult medulloblastomas were biologically less aggressive, having lower growth rate parameters (mean MIB-1 LI 19.1 ± 15.7; AI 3.73 ± 2.71 and AI:LI 0.207 ± 0.162) as compared to childhood tumors (mean MIB-1 LI 28.3 ± 20.4; AI 2.86 ± 2.14 and AI:LI 0.108 ± 0.111). p53 and Bcl-2 protein expressions were infrequent in all groups of tumors. No difference was noted in any of the parameters when classical and desmoplastic medulloblastomas were compared as a whole. But when compared between the age groups, an interesting observation (hitherto unreported in English literature) was that both classical and desmoplastic variants of childhood medulloblastomas had higher LI, lower AI and lower AI:LI ratio than their counterparts in adults, indicating that differences in growth rates cannot be attributed to differences in the frequency of occurrence of the histological variants in the two age groups. Thus, this study conclusively shows that there is a biological difference between childhood and adult medulloblastomas which is independent of standard histology and appeared to be associated more with age-related factors. This also warrants less-aggressive therapy for adult medulloblastoma.     

Blood vessel invasion as a predictor of long-term survival for Japanese patients with breast cancer

A wide range of frequencies has been reported for blood vessel invasion (BVI) among patients with breast cancer, however, the prognostic significance of BVI remains controversial. Three hundred ninety-eight Japanese patients with breast cancer, operated on during the period between 1971 and 1987, were studied. We investigated five factors, including BVI, lymph-node status (n), clinical tumor size (T), histological grade (HG), and tumor necrosis (TN), followed for a median of 10years. BVI was detected by hematoxylin and eosin (HE) staining and both factor VIII-related antigen and elastica van Gieson staining. BVI detected by HE staining alone was defined as BVIh. The subtypes of BVI were classified as follows: BVI e, BVI detected only by elastica van Gieson staining; BVI f, BVI detected only by factor VIII-related antigen staining; and BVI e/f, BVI detected by both factor VIII-related antigen and elastica van Gieson staining. BVI-positive tumors were defined as lesions showing BVI e, BVI f, or BVI e/f. BVI and BVIh were presented in 27.4%, 6.5% of all cases, respectively. The mean diameters of the calibers of BVI e, BVI f, and BVI e/f were 141.9±80.5m, 61.0±37.4m, 136.0±102.0m, respectively (P < 0.0001). Seventy-three patients (18.3%) had recurrence and 60 patients (15.1%) died of breast cancer. Univariate analysis showed that BVIh (P<0.0001), BVI (P<0.0001), n, T, and HG were significantly predictive of 20-year RFS and OS. Multivariate analysis showed that BVI (P<0.0001, P=0.0088, respectively), n, T, and HG were all significant and independent prognostic factors for RFS and OS. On the other hand, BVIh was an independent factor for RFS (P= 0.0475), but of borderline significance for OS (P= 0.0506). When stratified by BVI, BVI e, and BVI e/f were significantly predictive of 20-year RFS or OS (P>0.0001). We can confirm BVI, especially BVI e and BVI e/f, are significant independent prognostic factors associated with long-term survival in Japanese breast cancer patients.     

Bioavailability and pharmacokinetics of the investigational anticancer agent XK469 (NSC 698215) in rats following oral and intravenous administration

Purpose To determine the oral bioavailability of R-XK469, a water-soluble investigational anticancer agent undergoing phase I clinical trials as an intravenous product.Methods R-XK469 was administered to two groups of catheterized Sprague-Dawley rats via the oral and IV routes at a dose of 10 mg/kg and blood samples were collected at predetermined times. XK469 in plasma samples was quantified using a HPLC method. The pharmacokinetic parameters were computed using WinNonlin 4.0.1 software.Results The pharmacokinetic parameters of XK469 following oral and IV administrations, respectively, were (mean±SD): C_max 138±64 and 404±355 g/ml; AUC_0– 2381±773 and 2854±1924 g h/ml; and elimination half-life (T_1/2) 12.9±5.8 and 13.5±7.8 h T_max was 2.92±1.92 h following oral dosing. Oral R-XK469 was 83% bioavailable.Conclusion Together with the antitumor efficacy of oral XK469 shown in preclinical models and its schedule dependency, these results indicate the promise of developing an oral dosage form of R-XK469 for clinical development.     

Clinical pharmacology of aminoglutethimide in patients with metastatic breast cancer

Summary The pharmacology of aminoglutethimide (AG) was studied in two subsequent trials without hydrocortisone supplementation. A total of 79 patients with metastatic breast cancer entered the study, and their plasma and urine samples were analyzed by high-performance liquid chromatography (HPLC). Thirty evaluable patients with a median age of 57 years (range, 37–79) were treated with the standard dose of 1000 mg/day, and 37 evaluable patients with a median age of 59 years (range, 35–79) received 500 mg/day. The median follow-up in the two groups was 5 months (range, 1–16) and 4 months (range, 1–21), respectively. After the first oral dose of 500 mg, peak plasma concentrations of AG were observed 1–4 h after administration in 15 patients. The elimination half-life was 10.1±1.7h (mean ±SD) after initial dosage; it decreased significantly to 6.9±1.2 h after 8 weeks of treatment. The area under the curve of AG concentrations was 92.5±14.2 g/ml x h. The total clearance rate was 5.5±0.91/h and the volume of distribution was 80±11 l. About 23% of the drug was excreted unchanged in the urine. The major metabolite, N-acetyl-AG (AAG), had the same half-life as AG. A comparison on day 7 of treatment revealed that doses of 1000 and 500 mg yielded AG plasma concentrations of 9.0±1.2 and 4.5±0.5 g/ml, respectively. After 1 month of treatment, however, AG plasma levels of 6–7 and 4–5 g/ml were observed, respectively. A 50% reduction of dose, therefore, resulted in only 30% lower AG levels during continuous treatment. Apparently, the induction of metabolism is of greater importance in standard-dose than in lower dose treatment. The plasma concentrations of AG did not bear a relationship to the clinical response.Abbreviations used AG aminoglutethimide - AAG N-acetylaminoglutethimide - G glutethimide - HPLC high-performance liquid chromatography - SD standard deviation - SE standard error of the mean - AUC area under the concentration versus time curve     

Measurement of in vivo glucose transport from blood to tissue of experimentally-induced glioma in rat brain

Summary In experimentally-induced F98 glioma of rat brain, regional blood flow and glucose transfer were assessed by means of double tracer autoradiography to measure Michaelis-Menten constants for the determination of unidirectional glucose transport across the blood-tumor and blood-brain barrier. In brain regions opposite the tumor hemisphere, the maximal glucose transport rate constant, T_m, ranged from 1.41 ± 0.12 to 3.22 ± 0.29 mol/g/min and the half saturation transport constant of glucose, K_t, varied from 2.78 ± 0.83 to 5.6 ± 1.94 mol/ml (estimate ± standard error of the estimate) yielding a normoglycemic unidirectional glucose inward transport which ranged from 1.24 ± 0.24 to 1.97 ± 0.13 mol/g/min (mean ± standard deviation). In the tumor periphery, the T_m and the Kt values were 3.64 ± 0.56 mol/g/ml and 7.32 ± 2.12 mol/min, and in the tumor center, 1.77 ± 0.25 mol/ml and 2.76 ± 1.13 mol/min, respectively. The unidirectional glucose influx of tumor periphery and center in normoglycemia was 1.98 ± 0.22 and 1.34 ± 0.16 mol/g/min, respectively. Despite comparable unidirectional glucose influxes, however, glucose metabolism of tumor tissue located in the periphery (0.83 ± 0.12 mol/g/min) and the center (0.41 ± 0.10 mol/g/min) of the tumor mass exceeded that of normal gray matter by about 68% and 100% which indicates uncoupling between glucose transport and phosphorylation in experimentally-induced F98 glioma of rat brain.     

Extensive and saturable accumulation of paclitaxel by the human platelet

Little is known about the cellular distribution of paclitaxel in humans. In the present study we examined the distribution of [³H]-paclitaxel in human blood. When 1 M paclitaxel was incubated with fresh blood at 37°C, the platelet/plasma concentration ratio was 240±17 (mean±SEM), whereas the red blood cell (RBC)/plasma concentration ratio was only 0.59±0.05. In kinetics experiments using platelet-rich plasma, we observed that the platelet accumulation of paclitaxel was highly temperature- and concentration-dependent. Scatchard analysis of the 37°C uptake data demonstrated a dissociation constant (K _app) of 0.80±0.10 M and a maximal binding capacity of 672±102 pmol/10⁹ platelets. It is proposed that the platelet accumulation of paclitaxel reflects binding to microtubules and may serve as a useful model for binding to less accessible cellular sites.     

Evaluation of malignancy using Ki-67 labeling index for gastric stromal tumor

Background. Assessment of malignant potential in gastrointestinal stromal tumors (GISTs) is still problematic. The maximum tumor diameter and the mitotic index are generally used as an index of malignancy of GISTs. The Ki-67 labeling index has recently been used as an index of cell growth, and the prognosis of GISTs was reported to be significantly poor when the value of this index was 10% or higher. Methods. Clinicopathological and immunohistological factors were analyzed in 15 patients who underwent surgical resection of gastric stromal tumors at our department between April 1997 and July 2002. The patients were divided into metastasis/recurrence and benign groups. Also, the relationship of changes in the Ki-67 labeling index to the degree of malignancy in recurrent lesions was assessed in an 84-year-old woman who underwent five reoperations because of recurrences in the peritoneum. Results. Significant differences were noted between the metastasis/recurrence and benign groups in relation to the mean maximum tumor diameter (186.7 ± 80.8mm vs 41.3 ± 22.9mm), mitotic index (88.3 ± 5.0/50 high-power fields [HPF] vs 3.0 ± 2.9/50 HPF), and the Ki-67 labeling index (11.4 ± 2.5% vs 0.01 ± 0.51%). In the patient who had metastasis to the liver 3.5 years after initial operation and underwent five reoperations before death, the intervals until detection of recurrence tended to be shortened gradually. The Ki-67 labeling index varied with each operation, and tended to be higher at the time of reoperations than at the initial operation. Conclusion. The maximum tumor diameter, mitotic index, and Ki-67 labeling index were useful as an index of malignancy for gastric stromal tumor. The efficacy of surgical resection alone may be insufficient in patients with disseminated metastasis to the peritoneum.     

Evaluation of plasma 5-fluorouracil nucleoside levels in patients with metastatic breast cancer: relationships with toxicities

This paper describes the relationship between 5-fluorouracil (FUra)-derived toxicities and plasma levels of the FUra anabolites 5-fluorouridine (FUrd) and 5-fluoro-2-deoxyuridine (FdUrd) monitored in patients receiving continuous infusions of FUra (1000 mg/m² per 24 h) over 5 days preceded by the administration of cisplatin (100 mg/m²). A total of 63 courses of this treatment were given as second-line chemotherapy to 17 patients with metastatic breast cancer. The active FUra anabolites FUrd and FdUrd were monitored twice daily in the plasma by highperformance liquid chromatography. Data were analyzed using multiple analysis of variance (ANOVA). Only a low proportion of patients exhibited measurable plasmatic levels of FUrd (43%) and FdUrd (70%). The areas under the plasma concentration-time curves (AUC) determined over 120 h for FUrd (AUC_FUrd) and for FdUrd (AUC_FdUrd) were found to be statistically significantly different for chemotherapy cycles with and those without myelosuppression. Chemotherapy cycles without neutropenia were associated with low AUC_FUrd values (mean±SEM, 2.9±0.7 g ml^–1 h) and high AUC_FdUrd values (14.1±2.7 g ml^–1 h), respectively, whereas courses with myelosuppression (WHO grades 2–4) showed inverse profiles with high AUC_FUrd values (16.3±2.3 g ml^–1 h) and low AUC_FdUrd values (3.1±1.0 g ml^–1 h), respectively. A statistically significant difference in AUC_FdUrd values was also observed between cycles with and those without mucositis (P=0.0027), with AUC_FdUrd values being 22.6±5.6 and 7.8±1.9 g ml^–1 h, respectively. Whereas hematotoxicity could be correlated with both AUC_FUrd and AUC_FdUrd values, mucositis was associated with high AUC_FdUrd levels. Moreover, a negative correlation was found between the AUCs determined for FUrd and FdUrd (P=0.002), indicating that activation of FUra via FUrd or via FdUrd may involve competitive processes. Therefore, to follow the development of the major FUra-derived toxicities, measurement of FUrd and FdUrd plasma levels appeared very attractive.     

In vitro stability,plasma protein binding and blood cell partitioning of 14C-carboplatin

Summary Radiochemically pure ¹⁴C-labeled carboplatin, cis-diammine [1,1-cyclobutane (1-¹⁴C) dicarboxylato (2-)-0,0'] platinum (II), was added to fresh human, dog and rat plasma, at concentrations ranging from 1 to 100 g ¹⁴C-carboplatin/ml. After 10 min incubation at ambient temperature, the plasma was ultrafiltered in Amicon Centrifree micropartition units to generate protein-free plasma ultrafiltrate (PU). Total radioactivity was determined by liquid scintillation counting. A mean (±SD) of 102%±2.0%, 99.5%±1.9%, and 99.0%±1.0% of the ¹⁴C-carboplatin added to fresh human, dog and rat plasma respectibely was recovered in the PU. ¹⁴C-carboplatin was incubated at 37°C with fresh plasma (60g/ml) and urine (200 g/ml) from humans and dogs for 120 h, and samples were removed at appropriate times for analysis of carboplatin, 1,1-cyclobutane dicarboxylic acid and cyclobutane carboxylic acid. The latter were separated by HPLC on a C-18 column with a mobile phase of H₂O/CH₃CN/0.3 M tetrabutylammonium phosphate (880:50:20 v/v/v), and the column eluants at the retention time of each compound were collected and counted for total radioactivity. Carboplatin degraded in each of the matrices with a corresponding release of 1,1-cyclobutane dicarboxylic acid. ¹⁴C-carboplatin (50 g/ml) was incubated at 37°C with fresh human, dog and rat blood and the distribution of radioactivity into the cellular fraction was determined. Radioactivity did not distribute into the blood cells of humans or dogs, but after 5 h, 44% of the radioactivity in rat blood was associated with the cellular fraction. These results show that carboplatin, at physiological concentrations, does not bind instantaneously and reversibly to the plasma proteins of rat, dog or human, and that the molecule slowly degrades in plasma and urine in vitro with the release of 1,1-cyclobutane dicarboxylic acid. The remaining diammine platinum (II) portion of the molecule therefore accounts for the essentially irreversible protein binding of the platinum from carboplatin.     

Combination treatment of cis- and carboplatin in cancers restricted to the peritoneal cavity in the rat

In the present study, cisplatin (cDDP) and carboplatin (CBDCA) were combined in different in vitro and in vivo assays to determine whether combined cDDP and CBDCA treatment would eventually lead to a better antitumor response. Co-incubation of CC531 cells with cDDP and CBDCA led to higher intracellular Pt concentrations (30.5±3.4 ng Pt/10⁶ cells) than did cDDP (16.9±9.4 ng Pt/10⁶ cells) or CBDCA (1.28±0.72 ng Pt/10⁶ cells) incubation alone. In survival assays an additive cell kill was seen after combined treatment with cDDP and CBDCA. DNA binding experiments using isolated salmon-sperm DNA exposed to the drugs separately or in combination were in agreement with the survival studies (for cDDP a binding of 12.42 g Pt/mg DNA; for CBDCA, 0.49 g Pt/mg DNA at 76 h). Toxicity studies in rats treated with cDDP plus CBDCA required a dose reduction for cDDP amounting to 20% of the MTD, whereas the CBDCA dose could be maintained. Pharmacokinetics studies showed higher AUCs andt _1/2 in plasma as well as the peritoneal cavity after combined treatment with cDDP and CBDCA (both given i.p.) or following cDDP given i.p. and CBDCA given i.v. Pt concentrations in peritoneal tumors corresponded with these observations, with higher Pt concentrations following combined treatment than after single-agent injection. In addition, combined adminstration of cDDP i.p. and CBDCA i.v. led to higher Pt concentrations in peritoneal tumors than did administration of both drugs i.p. (3.93±0.9 vs 2.76±0.2 mg Pt/g tissue). The higher Pt concentrations in the peritoneal tumors after combined treatment was associated with a significantly better antitumor response in comparison with that observed after single-agent treatment (a growth delay of 30.2±5.6 days for cDDP i.p. plus CBDCA i.v. vs 16.1±5.4 days for cDDP alone and 10.8±4.2 days for CBDCA alone).     

Prolactin binding by human mammary carcinoma: Relationship to estrogen receptor protein concentration and patient age

Summary Biopsy specimens of 55 human mammary carcinomas (38 primary and 17 metastatic) were assayed for prolactin receptors (PrlR). Prolactin bound specifically to 32 (58%) of the tumor biopsy specimens. The apparent K _d for PrlR in individual tumors ranged from 15 pM to 2. 3 nM (mean 600 pM,n = 5) and the concentration of PrlR ranged from 0 to 44.5 fmoles/mg protein. Estrogen receptors (ERP) were also detected in 28 of the 32 tumors which had PrlR. Overall, there was no correlation between PrlR and ERP. However, the mean concentration of PrlR was significantly higher (p <0.01) in tumors with 6–100 fmoles/mg protein ERP ( 13 fmoles PrlR) than in tumors with either <6 or >250 fmoles ERP (4.0 ± 0.4 and 6.5 ± 1.8 respectively fmoles PrlR). Analysis of PrlR concentration as a function of patient age also showed no overall correlation, but the mean PrlR in tumors from women aged 60–70 was significantly higher (p <0.01) than in those from either younger or older women. A higher concentration of PrlR was observed in tumors which were classified histologically as medium or well differentiated (6.1 ± 1.2 and 11.1 ± 2.1, respectively) than in those classified as poorly differentiated (3.3 ± 1.2) (p <0.03). There was a negative correlation between PrlR concentration and membrane yield from the tumors (r = 0.43,p <0.002). The membrane yield correlated with the ratio of tumor cells to stroma (histologically) (r = 0.63,p <0.001). In tumors from 12 patients with metastatic disease on whom follow up after endocrine-related therapy was available, the mean PrlR concentration was significantly higher in the non-responding group (8.2 ± 3.0) than in the responding group (3.4 ± 4.2,p = 0.05). Address for reprints: Dr. Barbara Rae-Venter, Department of Surgery, University of Texas Medical Branch, Galveston, TX 77550.     

BCNU-sensitivity in parental cells and clones from four freshly resected near-diploid human gliomas: An astrocytoma,an anaplastic astrocytoma and two glioblastomas multiforme

Summary We compared the BCNU sensitivity of 4 freshly resected tumors (astrocytoma WM, and malignant gliomas MK, MB, and AM) and their clones to their karyology. The majority of primary cells in all 4 tumors had near-diploid chromosome numbers (2n±) and all were resistant to concentrations of BCNU exceeding 10g/ ml. Followingin vitro cultivation, the cells from tumors WM and MB retained their 2n ± modal chromosome number with little change in the complexity of the karyotype. In contrast, tumors MK and AM demonstrated a more unstable genome. The modal chromosome number of MK shifted from 45 to 86 and that of tumor AM from 45 to 90. Karyotyping demonstrated additional ploidy changes and new marker chromosomes in both tumors. The colony forming assay (CFA) performed on thein vitro cultivated cells demonstrated little change in the sensitivity to BCNU in tumors WM and MB, while tumors MK and AM exhibited greater than a one log cell kill at 10.0g/ml and 15.0g/ml BCNU, respectively. The modal chromosome number and BCNU sensitivity followed a similar pattern in the 30 clones that were isolated; 21 clones with near-diploid and pseudodiploid chromosome numbers were all resistant to BCNU doses at or greater than 10/ml. In contrast, 9 clones isolated from the 3 malignant gliomas with 3n ± and 4n ± modal chromosome numbers were sensitive to this concentration of BCNU. The karyotypes of the hyperdiploid clones were more complex; they contained 5 or more ploidy changes and/or had marker chromosomes. These studies confirm the association of diploidy and BCNU-resistance in freshly resected malignant gliomas.     

Combined Modality Therapy for Poorly Differentiated Gliomas of the Posterior Fossa in Children: A Children's Cancer Group Report

Purpose: To study the effectiveness of combined chemotherapy and radiotherapy for children with high-grade astrocytomas of the posterior fossa. Patients and Methods: In the CCG-945 study, 250 patients were treated by members of the Children's Cancer Group (CCG). Sixteen children were randomly assigned to one of two chemotherapy regimens, vincristine, lomustine, and prednisone or 8-in-1, using the same involved-field irradiation in both. Six infants received 8-in-1 chemotherapy before involved-field irradiation. All pathologic specimens had central review. Results: Twenty-two patients with an institutional diagnosis high-grade posterior fossa tumors received chemotherapy and/or irradiation. Fifteen were confirmed by central review to have high-grade gliomas. Overall survival for confirmed high-grade astrocytoma of the posterior fossa was approximately 36 ± 13% at 5 years for the children (n = 11) and 25 ± 15% at 5 years for the infants (n = 4). Conclusions: Involved-field irradiation with chemotherapy appeared to prevent extraneural and subarachnoid metastases. We also confirmed the rarity of the tumor (6% of patients registered). Further Phase III trials are necessary to improve survival in this aggressive tumor.