Postoperative intraperitoneal 5-fluoro-2'-deoxyuridine added to chemoradiation in patients curatively resected (R0) for locally advanced gastric and gastroesophageal junction adenocarcinoma

PURPOSE

Chemoradiation after surgery for locally advanced gastric cancer improves overall and relapse-free survival compared with observation. However, locoregional recurrences remain high. Accordingly, we instituted this pilot/feasibility study, including intraperitoneal 5-fluoro-2'-deoxyuridine (IP FUDR) as part of the treatment.

METHODS

Gastric/gastroesophageal junction adenocarcinoma stage Ib-IV (M0) patients who underwent R(0) resection were eligible and had IP catheters inserted at time of surgery. IP FUDR (3 g/dose/day) was given during study days 1-3 and 15-17 before combined 5-fluorouracil, leucovorin, and external beam radiation (45 Gy). Endpoints included toxicity, completion rate, locoregional recurrence, and survival.

RESULTS

Twenty-eight patients (22 men) were enrolled from 2002-2006 at two institutions; their median age was 59.5 years. After R(0) resection, a median 22 (range, 8-102) lymph nodes were examined, and 22 patients had positive nodes. AJCC stages were IB (n = 8), II (n = 10), IIIA (n = 5), IIIB (n = 1), and IV (n = 4). Full-dose IP FUDR and chemoradiation treatment was completed in 20 and 25 patients, respectively. At nearly 4-year median follow-up, 11 patients were disease-free, 5 were alive with disease, 7 were dead of disease, and 1 was dead from other cause; 4 have been lost to follow-up. Recurrences were local in one, intra-abdominal in six, distant in two, multiple sites in two, and unknown in one. The median relapse-free survival is 65.3 months, and the median overall survival has not yet been reached.

CONCLUSIONS

IP FUDR before chemoradiation after R(0) gastric cancer resection is well tolerated without compromising completion of postoperative adjuvant treatment. Larger randomized trials studying IP FUDR as part of gastric cancer multidisciplinary treatment are needed to prove efficacy in reducing regional recurrence and improving survival.

Intraperitoneal(IP) 5’-fluoro-2’deoxyuridine(FUDR): Safety and outcome when administered prior to adjuvant chemoradiotherapy(chemoRT) following R0 resection for gastric adenocarcinoma

4627

Background: ChemoRT after surgery for locally advanced gastric cancer improves overall and relapse-free survival (OS and RFS) compared to observation (NEJM 2000,345:725–30). However, loco-regional recurrences (>50%) remain high and we hypothesized that adding IP FUDR would further improve outcome. Methods: Patients (pts) ECOG performance status (PS) 0–2, gastric/gastroesphogeal(GEJ) adenocarcinoma stage Ib-IV (M0) undergoing R0 resection were eligible, and had insertion of IP catheters at surgery. IP FUDR(3gm/dose/day) was given on protocol days 1, 2, 3 and 15, 16, 17 prior to 5-FU/LV and external beam RT (45Gy) as in cited study. Simon 2-stage optimum design was used to demonstrate safety. Endpoints also included were loco-regional recurrence and survival. Results: 28 pts with gastric/GEJ adenocarcinoma (25/3) were enrolled from 2002 to 2006 at 2 institutions: median age 59.5 years (range 39–81), M /F (21/7). R0 gastric resection was performed with dissection of median 22 (range 8–102) lymph nodes(LN’s). 22/28 pts were lymph node positive. Full dose IP FUDR was completed in 20/28 pts. 4 pts required dose reduction (1 for grade(gr) 2 hepatic enzyme elevation, 2 gr 2 neutropenia, 1 gr 4 neutropenia), 3 discontinued therapy (1 gr 3 abdominal pain, 1 GI abscess, and 1 bleeding arterial pseudoaneurysm). One pt received no IP treatment due to catheter failure. 24/28 pts completed chemoRT and had toxicity comparable to that previously reported in the Intergroup 0116 trial. At 26 month median follow up (range 2.8–43.4), of the 26 pts evaluable for response, 16 pts are NED, 6 alive with disease, 3 dead of disease, and 1 dead from other cause. 5 recurrences were intra-abdominal, 1 local, 2 distant, and 1 at multiple sites. At present analysis, the median RFS is 32.5 months. Conclusions: IP FUDR prior to chemoRT after R0 gastric cancer resection is well tolerated. A randomized study to test its role in reducing regional recurrence and improving outcome is warranted. (FDA Orphan Products grant# FD-R-2150–04)

No significant financial relationships to disclose.

Neoadjuvant chemotherapy, surgery, and adjuvant intraperitoneal chemotherapy in patients with locally advanced gastric or gastroesophageal junction carcinoma: a phase II study

A phase II trial, using neoadjuvant chemotherapy and intraperitoneal (IP) consolidation, was conducted in patients with locally advanced, potentially resectable gastric cancer or cancer of the gastroesophageal junction, both staged as T3N0, T4N0, or any TN1 or TN2 disease. Preoperative chemotherapy consisted of two cycles of irinotecan 75 mg/m(2) with cisplatin 25 mg/m(2)/week for 4 weeks followed by a 2-week break. Unless disease progression was encountered, surgery was performed and followed by two courses of adjuvant therapy with IP floxuridine 3 g x 3 days plus IP cisplatin 60 mg/m(2) on day 3. Of 32 evaluable patients, 29 (90.6%) underwent surgery, and 25 (86.2%) had R0 on resection. Evidence of primary-tumor downstaging was documented in at least one half of the patients. Toxicity of induction therapy was primarily grade 3/4 neutropenia (38.2%/8.8%), grade 3 diarrhea (20.6%), and grade 3 nausea/vomiting (14.7%). Except for three catheter complications, toxicities with IP therapy were infrequent. After a median follow-up of 28.0 months in 32 patients, 10 patients (31.3%) had no evidence of disease, 4 (12.5%) were alive with disease, 13 (40.6%) had died from disease, and 5 (15.6%) died from unrelated causes. Among 25 patients who underwent R0 resection, there were no local recurrences. Sites of first recurrences were outside the abdominal cavity in seven patients, in the liver in two, and in the abdominal cavity in four patients. Median overall survival for all 32 patients was 36.5 months from the start of treatment after median follow-up of 28 months, whereas median disease-specific survival had not been reached at the time of this analysis. For patients with R0 resection, median overall survival was 48 months after median follow-up of 35 months. The data suggest that an approach consisting of systemic induction therapy, curative surgery with high R0 resection rates, and IP adjuvant therapy has acceptable toxicity and encouraging survival outcomes.

Complications of gastrectomy following CPT-11-based neoadjuvant chemotherapy for gastric cancer

Potential benefits of neoadjuvant therapy for locally advanced gastric cancer include tumor downstaging and an increased R0 resection rate. Potential disadvantages include increased surgical complications. This study assesses postoperative morbidity and mortality by comparing patients undergoing gastrectomy with and without neoadjuvant chemotherapy. From October 1998 to July 2002, a total of 34 patients with locally advanced gastric cancer were placed on a phase II neoadjuvant chemotherapy protocol consisting of two cycles of CPT-11 (75 mg/m(2)) with cisplatin (25 mg/m(2)). Demographic, clinical, morbidity, and mortality data were compared for these patients (CHEMO) versus 85 patients undergoing gastrectomy without neoadjuvant chemotherapy (SURG). The CHEMO patients were more likely to be less than 70 years of age (P< or =0.01), have proximal tumors (P< or =0.01), and undergo proximal gastrectomy (P< or =0.025). Fifty-two percent of SURG patients had T3/T4 tumors compared to 19% of CHEMO patients, consistent with tumor downstaging. The R0 resection rate was similar (80%). Morbidity was 41% in CHEMO patients and 39% in SURG patients. There were five postoperative deaths (4.4%), two in the CHEMO group and three in the SURG group (P=NS). It was concluded that neoadjuvant chemotherapy with CPT-11 and cisplatin is not associated with increased postoperative morbidity compared to surgery alone. CPT-11-based neoadjuvant chemotherapy should be tested further in combined-modality treatment of gastric cancer.

Phase I and pharmacologic study of i.p. 9-aminocamptothecin given as six fractions over 14 days

We sought to define the tolerance of 9-amino-20(S)-camptothecin (9-AC) when given by the i.p. route to patients with cancer in the peritoneal cavity consisting of nodules that did not exceed 1 cm in maximum diameter. 9-AC was given in six fractions over 12 days, at doses ranging from 1.25 to 13.5 mg/m(2) in cycles repeated every 28 days. Dose escalations after the first two dose levels took place in cohorts of three patients, with expansion of the dose level once a dose-limiting toxicity (DLT) was encountered. All patients had blood and i.p. pharmacokinetic (PK) analysis during cycle 1 of each dose level. Topoisomerase (Topo) I signal was serially measured in peripheral blood mononuclear cells (PBMCs) in blood and cells collected in i.p. cytologic washings. Twelve patients received 31 cycles of 9-AC. Tolerance to repeated i.p. drug administration was generally excellent. The DLT was neutropenia encountered at the highest dose level in two patients, whereas the dose of 9 mg/m(2) was well tolerated. The DLTs were associated with peak plasma levels ranging from 47 to 81 ng/ml and also depletion of Topo I in PBMCs. The i.p.:plasma AUC ratio (+/-SD) was 11.5 (+/-3.8). Two patients had objective evidence of clinical benefit and only one of seven patients deemed evaluable for response had progressive disease. We conclude that i.p. 9-AC demonstrates excellent local tolerance at a dose and schedule associated with evidence of systemic effects. A dose of 9 mg/m(2)/cycle administered in a schedule of six divided fractions is suitable for further evaluation against tumors involving primarily the peritoneal cavity.

Neoadjuvant chemotherapy with CPT-11 and cisplatin downstages locally advanced gastric cancer

We examined the role of neoadjuvant therapy in downstaging locally advanced gastric cancer. Preoperative staging was performed with a combination of CT scans, endoscopic ultrasonography and/or laparoscopy, and laparoscopic ultrasonography. Patients with T > or =3 tumors and/or node-positive disease by preoperative clinical staging were eligible for entry. Neoadjuvant therapy consisted of two cycles of CPT-11 (75 mg/m(2)) with cisplatin (25 mg/m(2)) weekly four times every 6 weeks. This was followed by resection with D2 lymph node dissection and two cycles of intraperitoneal chemotherapy with floxuridine and cisplatin. Twenty-two patients were entered into the study (4 with T3N0 disease and 18 with T3N1 disease). Induction chemotherapy was well tolerated with major toxicities being neutropenia and diarrhea. A median of 78%/75% of the planned dosage of CPT-11/cisplatin was delivered. Two patients withdrew consent during the first cycle and were lost to follow-up. One patient progressed to stage IV disease during induction chemotherapy and did not undergo surgery. Nineteen patients underwent surgery. One patient had undetected stage IV disease (liver) and underwent a palliative R2 resection. Of the 18 remaining patients, 17 had curative R0 resections and one had a palliative R1 resection. A median of 21 lymph nodes (range 1 to 121) were examined histologically. There was one postoperative death. Surgical morbidity did not appear to increase after the neoadjuvant regimen. The median postoperative length of hospital stay was 9 days (range 3 to 75 days). Postoperative pathologic staging yielded 16% T3 lesions compared to 85% before treatment based on clinical staging; postoperative American Joint Committee on Cancer staging yielded 37% stage IIIA disease compared to 70% stage IIIA before treatment. With a median follow-up of 15 months, median survival has not yet been reached. We conclude that CPT-11-based neoadjuvant therapy downstages locally advanced gastric cancer. Further follow-up is necessary to determine the ultimate impact of this combination therapy on recurrence and survival.

Intraperitoneal topoisomerase-I inhibitors. Preliminary findings with 9-aminocamptothecin

The i.p. administration of topoisomerase I (Topo I) inhibitors has a pharmacologic advantage over intravenous application, including preservation of the biologically active lactone form. In our ongoing study, patients have received 9-amino-20(S)-camptothecin (9-AC) i.p. on days 1, 3, 5, 8, 10, and 12, repeated every 4 weeks. The daily dose has been escalated to level IV of 1.5 mg/m2 (9.0 mg/m2 per course), median of 3 cycles, range 1-4, with a reversible Grade 3 neutropenia in one patient. Responses included one CR (resolution of a pleural effusion), two patients without progressive disease (PD), two not evaluable, and two patients too early for evaluation. The area under the curve (AUC)i.p./AUCpl ratio (pharmacologic advantage) ranged from 7.6 to 16.5 on average, and, using nonlinear modeling, the pharmacologic decay data were fit to one- or two-compartmental models. Overall, a 9-AC i.p. application is well tolerated and anticipated to be an active regimen against i.p. malignancies, particularly those known to be sensitive to systemic Topo-I inhibitors.

Antiangiogenic effects of camptothecin analogues 9-amino-20(S)-camptothecin, topotecan, and CPT-11 studied in the mouse cornea model

Angiogenesis has been correlated with increased invasion and metastases in a variety of human neoplasms. Inadequate inhibition of the growth of tumor microvessels by anticancer agents may result in treatment failure, rated clinically as progressive or stable disease. We have investigated the antiangiogenic properties of three camptothecin analogues, 9-amino-20(S)-camptothecin, topotecan, and camptosar (CPT-11), currently under investigation in clinical settings. Angiogenesis was induced by basic fibroblast growth factor in the cornea of inbred Swiss-Webster mice, with the aim of exploring the suppression of neovascularization by the analogues injected into the mice daily over a period of 6 days. The dose range chosen is known to inhibit, in the mouse model, the growth of various human tumor xenografts or murine tumors. The statistical analysis evaluated the association between the area of neoangiogenesis and the dose of the drugs tested and correlated the effects with observed drug toxicity. It was established that, as the drug doses increased, the area of neovascularization decreased, appearing to approximate a negative exponential curve. 9-Amino-20(S)-camptothecin at 6.89 and 8.26 micromol/kg (2.5 and 3.0 mg/kg) and topotecan at 8.31 micromol/kg (3.5 mg/kg), both drugs being delivered over a 6-day period, had statistically significant reduction (47.2-72.5%) of neoangiogenesis and acceptable toxicity. At higher doses of the two analogues, toxic body-weight losses and deaths were observed. CPT-11 showed statistically significant reduction of neoangiogenesis at a dose of 359 micromol/kg (210 mg/kg) delivered over a 6-day course. Unlike camptothecin analogues, the nontoxic dose of vincristine did not induce a statistically significant inhibition of angiogenesis, and there was no dose-dependent escalation of antiangiogenic effects. The results indicate that camptothecins are most likely cytotoxic against two tumor compartments: in addition to tumor cells of epithelial origin, the drugs act against endothelial cells and prevent the growth of the tumor microvessels. We have hypothesized that treatment failure in some patients is due to incomplete or inadequate inhibition of the microvessel growth by camptothecins. Presumably, an intensive inhibition of the remaining tumor microvasculature in such patients could be achieved by combining a camptothecin with another antiangiogenic anticancer agent or with a highly selective angiogenic inhibitor exerting minimal dose-limiting toxicity. Such treatment by a camptothecin plus a less toxic inhibitor of angiogenesis can improve antitumor efficacy. To validate this concept, preclinical studies followed by clinical trials are planned.

Pharmacodynamics of topoisomerase I inhibition: Western blot determination of topoisomerase I and cleavable complex in patients with upper gastrointestinal malignancies treated with topotecan

Analogues of camptothecins are specific inhibitors of eukaryotic DNA topoisomerase I (topo I) that lead to DNA damage and, eventually, cellular cytotoxicity. Camptothecin analogues bind to this target enzyme in the course of its normal function and stabilize the DNA-enzyme adduct to form a "cleavable complex." Preclinical experiments using Western blot analyses have shown cleavable complex formation to be the key intermediate step in topo I inhibition. In this series of experiments, it was our goal to convert this laboratory technique into a useful clinical assay, allowing measurement of the target enzyme and detection of the key intermediate in clinical specimens taken from patients being treated with the topo I inhibitor topotecan. Because available antibodies were not sufficiently sensitive at the start of this project, we identified a highly specific human SCL-70 antibody from a patient with scleroderma, which allowed quantitative determination of topo I copy number in HeLa and HT-29 cell lines. Additional refinements of the Western blot technique were accomplished to improve signal:noise ratio. In surgical tumor specimens, we found the median topo I level to be 30.1 x 10(5) copies/cell for gastric adenocarcinomas, compared to 18.4 x 10(5) copies/cell for normal gastric mucosae in the same samples. For lung adenocarcinoma, the median protein level was 21.5 x 10(5) copies/cell, compared with the normal tissue counterpart protein level of 12.7 x 10(5) copies/cell. The median tumor:normal ratios from paired samples of these tumor types were 1.51 and 1.84, respectively. As part of a Phase II study evaluating the efficacy of topotecan (1.5-2.0 mg/m2 daily for 5 days) in upper gastrointestinal malignancies, we obtained tumor and normal mucosa biopsies in 11 patients with gastric or esophageal cancer, 30 min after administration on day 4 or 5. Three patients with gastric adenocarcinoma had stable disease as their best response, with the remainder of patients progressing. Improvement in Western blotting methodology allowed the quantitation of topo I levels in these gastric and esophageal cancer biopsies, which could be augmented by brief heating to release complexed topo I. We were also able to directly visualize high molecular weight topo I-containing bands, which were shown to be cleavable complexes by heat reversal, with restoration of the topo I Mr 100,000 band. Using this heat reversal technique, we determined the presence of cleavable complex in a total of 7 of 11 patient biopsy samples (5 tumors and 2 normal mucosae). In patients treated with topotecan on this dose and schedule, we determined that a median of 73% of the total tumor topo I was involved in cleavable complex (range, 18.3-91%). The intensity of the Mr 100,000 topo I band in biopsy specimens of patients receiving topotecan represented "free" or noncomplexed topo I. The median copy number for the residual, noncomplexed topo I (n = 11) was 7.36 x 10(5) copies/cell, significantly less than the median of 30.1 x 10(5) copies/cell for random tumor specimens from patients with gastric adenocarcinomas (P < 0.001). Pharmacodynamic analysis demonstrated a negative correlation between the noncomplexed topo I copy number and topotecan area under the curve (Spearman rank test: r(s) = -0.81, P = 0.003). Nonlinear regression analyses of these data were best fit with an inhibitory maximum effect model, yielding parameter estimates for Emax and EC50 of 29.3 x 10(5) copies/cell (coefficient of variation = 22%) and 43.1 ng x h/ml (coefficient of variation = 27%), respectively. Through a series of careful modifications and refinements, we have improved the Western blot assay for topo I for use in clinical monitoring. We have demonstrated the ability to directly visualize cleavable complex in patients being treated with topo I inhibitor therapy and have directly quantitated free topo I, as well as the key topo I intermediate (cleavable complex), in biopsy specimens obtained from pat

Drug-induced apoptosis in B-cell chronic lymphocytic leukemia: relationship between p53 gene mutation and bcl-2/bax proteins in drug resistance

We investigated the relationship among chemosensitivity to drug-induced apoptosis in vitro, the presence of p53 gene mutations, and the expression of bcl-2 and bax proteins in B-cells from B-cell chronic lymphocytic leukemia (B-CLL) patients. Apoptosis was induced with a camptothecin analogue, 9-amino-20(s)-camptothecin, or a purine analogue, fludarabine. Cell death was monitored by propidium iodide staining and FACS analysis. Drug-induced apoptosis in B-CLL cells was p53-independent. Immunoblot analysis of bcl-2 and bax expression revealed a correlation between drug-induced apoptosis and the ratio of endogenous levels of bcl-2 to bax proteins. B-CLL cells with none to low bcl-2/bax ratios were drug-sensitive as compared to cells with intermediate to high ratios that were drug-resistant (P = 0.015). Prior to drug treatment, bax protein migrated as a single species of 21 kDa. Following drug-induced apoptosis, anti-bax specific protein complexes of 36-42 kDa were up-regulated. Using two-dimensional gel electrophoresis, bax complexes were disrupted under reducing conditions to reveal homo- and heterodimers of 18 and 21 kDa suggesting that disulfide interactions were required for complex formation. The de novo appearance of the 18 kDa anti-bax specific protein together with its increased expression in drug-sensitive B-CLL B-cells undergoing cell death suggests a role for this protein in the regulation of apoptosis.

Growth inhibition of human cancer metastases by camptothecins in newly developed xenograft models

Several metastatic models have been developed using clonal selection of human malignant cells metastasizing into a specific organ in NIH-I Swiss immunodeficient mice. The organs of choice were the central nervous system (CNS), targeted by metastases of malignant melanoma, and the liver, with metastases of colon adenocarcinoma. Additional models of adrenal metastases by malignant melanoma, and CNS involvement by implanted human lung squamous carcinoma or lymphoblastoid cells, are also available. Organ metastases, as well as the effects of treatment, were confirmed by autopsies and histological examination of the tissues or by a surgical inspection of the liver. The treatment end points were established as the increases in survival times of treated mice relative to placebo-treated controls. Camptothecins injected i.m. or delivered via gastrointestinal tract inhibit the growth of CNS metastases and increase the survival of treated animals. 9-Amino-20(S)-camptothecin was effective in the CNS model and in the model of liver metastases. The drug increased 3.3- and 5.7-fold the survival rates relative to untreated controls with metastases of colon adenocarcinoma to the liver, and all camptothecins were significantly more effective than 5-fluorouracil, currently a drug of choice in treatment of this disease. The xenograft models of metastases are available for studies of drug passage through the blood-brain barrier optimization of drug delivery to the liver, and for the development of new camptothecin-based treatment strategies.

Frequent homozygous deletions of D13S218 on 13q14 in B-cell chronic lymphocytic leukemia independent of disease stage and retinoblastoma gene inactivation

A correlative study has been performed to delineate further the role of the Rb gene in the disease B-cell chronic lymphocytic leukemia (B-CLL). First, we examined DNAs from B cells from 140 B-CLL patients representing all Rai stages of disease for the loss of 13q14 using two microsatellite markers mapping distal to the Rb locus. Loss of heterozygosity (LOH) of D13S133 was infrequent, occurring in 5 of 140 (4%) patients. The frequency for LOH of D13S218 was 33 of 140 (24%) samples and was independent of Rai stage of disease. Rb protein was detected in 19 of 23 (83%) samples. Of 4 patients lacking detectable Rb gene expression, only one showed LOH of D13S218. Rb protein levels varied from undetectable to high in samples with or without LOH for D13S218, and the levels were also independent of Rai stage of disease. Our findings support the role of DBM on 13q14, rather than Rb, as the candidate tumor suppressor gene that is frequently targeted for deletion in B-CLL. In addition, the data suggest that other mechanism(s) contribute to altered Rb expression detected in one-fourth of B-CLL B-cells.

Chemosensitivity of lymphocytes from patients with B-cell chronic lymphocytic leukemia to chlorambucil, fludarabine, and camptothecin analogs

Chemosensitivity of B lymphocytes, obtained from 65 patients with B-cell chronic lymphocytic leukemia (B-CLL), Rai stages 0 through IV, was determined using the MTT assay. The results were expressed by the drug concentration required for 50% inhibition of cell viability (IC50). The cytotoxicity of chlorambucil (CLB) was compared with that of fludarabine and the DNA topoisomerase I inhibitors, camptothecin, 9-aminocamptothecin, 10,11-methylenedioxy-20(S)-camptothecin (10,11-MDC) and 9-amino-10,11-methylenedioxy-20(S)-campthothecin (9-A-10,11-MDC), and topotecan. Considerable heterogeneity in sensitivity to CLB was observed, with a median IC50 of 40.5 mumol/L in untreated patients. B-CLL cells from patients treated with CLB had a significantly higher median IC50 of 86.0 mumol/L (P < .01). Untreated as well as CLB-treated patients were divided into two subsets. For the purpose of this study, B-CLL lymphocytes with an IC50 CLB of less than 61.0 mumol/L were designated as "sensitive" and those with an IC50 CLB of > or = 61.0 mumol/L were designated as "resistant." After baseline assays, 15 untreated patients received CLB; after treatment, the IC50 increased in B-CLL lymphocytes from 13 of 15 patients. The response to CLB treatment, determined by its effect on the absolute lymphocyte count and by the Eastern Cooperative Oncology Group clinical criteria, was significantly better in patients whose lymphocytes had an IC50 CLB of less than 61.0 mumol/L before therapy (P < .01). B-CLL lymphocytes also had a variable degree of sensitivity in vitro to each of the other drugs. There was significant cross-resistance between CLB and fludarabine (P < 0.01). Whereas only 29% of CLB-resistant B-lymphocyte specimens obtained from individual patients were sensitive to fludarabine in vitro, 52% and 67% of CLB-resistant lymphocyte samples were sensitive to 10,11-MDC and 9-A-10,11-MDC, respectively. We have previously reported that p53 gene mutations were associated with aggressive B-CLL and a poor prognosis. B lymphocytes from seven patients with these mutations were resistant to CLB, and five of six were resistant to fludarabine. Lymphocytes from four of seven were resistant to 10,11-MDC, and three of four were resistant to 9-A-10,11-MDC. This study implies that the MTT assay may be useful in identifying subsets of CLL patients resistant to conventional chemotherapy. However, definitive conclusions can not be drawn in view of the small number of patients studied prospectively. In addition, these results suggest the potential of camptothecin-based therapy for patients unresponsive to standard treatment.

p53 gene mutation in B-cell chronic lymphocytic leukemia is associated with drug resistance and is independent of MDR1/MDR3 gene expression

We studied 53 patients with B-cell chronic lymphocytic leukemia (B-CLL) and found mutations of the p53 gene in 15%. Patients with p53 gene mutations were found to have an aggressive form of B-CLL disease characterized by advanced Rai stage, rapid lymphocyte doubling time (LDT), and resistance to chemotherapy. While 27 of 29 treated patients (93%) without p53 mutations achieved a partial remission, only one of seven treated patients (14%) with p53 mutations achieved a partial remission (P = .00009). Adjusting for prognostic factors (age, sex, race, and Rai stage), patients with p53 gene mutations had a 13-fold greater risk of death than patients without p53 mutations (P = .013). In addition to examining the clinical relevance of p53 gene mutations in B-CLL, we investigated the possible role of p53 gene regulation in the expression of the multidrug resistance genes MDR1 and MDR3. We quantitated MDR1 and MDR3 mRNA expression by reverse transcription-polymerase chain reaction (RT-PCR). Expression of both the MDR1 and MDR3 genes was independent of p53 gene mutation or prior drug treatment, and did not predict for clinical response. Our findings indicate that p53 gene mutations in B-CLL are associated with a poor clinical outcome and may be a prognostic indicator for drug resistance.

Camptothecin overcomes MDR1-mediated resistance in human KB carcinoma cells

In order to understand the high efficacy of camptothecin derivatives against human colon tumor xenografts in nude mice, we have studied the transport properties of camptothecin derivatives across cellular membranes of MDR1-overexpressing cells. MDR1 overexpression was shown to have little effect on camptothecin cytotoxicity; camptothecin was equally cytotoxic to both the drug-sensitive parental cell line, KB 3-1, and its multidrug-resistant derivative, KB V1. The ability of camptothecin to overcome MDR1-mediated resistance is most likely due to unimpaired accumulation of camptothecin in MDR1 cells as suggested from the following experiments: (a) cytotoxicity of camptothecin against KB V1 cells was not altered by the known MDR1-reversing agent, verapamil; (b) camptothecin was ineffective as compared with vinblastine in competing with [3H]azidopine for photoaffinity labeling of MDR1; (c) camptothecin was equally efficient in trapping cellular topoisomerase I molecules on chromosomal DNA in the form of cleavable complexes in both KB 3-1 and KB V1 cells. The mechanism by which camptothecin overcomes MDR1-mediated resistance has been further studied using a number of uncharged and charged camptothecin derivatives. In contrast to the uncharged camptothecin derivatives, such as 9-amino-camptothecin and 10,11-methylenedioxy-camptothecin, the charged camptothecin derivative, topotecan, showed reduced cytotoxicity against MDR1-overexpressing KB V1 cells. The reduced cytotoxicity of topotecan in KB V1 cells was due to the overexpression of MDR1 in KB V1 cells since verapamil restored both topotecan accumulation and cytotoxicity. These results suggest that the charge on camptothecin can affect the drug's sensitivity to MDR1. The possible effect of membrane permeability in determining drug selectivity of MDR1 is discussed.

Complete growth inhibition of human cancer xenografts in nude mice by treatment with 20-(S)-camptothecin

20-(S)-Camptothecin (CAM), a plant alkaloid, was tested against 13 human cancer xenograft lines carried by immunodeficient (nude) mice. The drug, formulated in 20% intralipid and given i.m., was more effective than any other clinically available drug tested. It was found that: (a) CAM, at nontoxic doses, suppressed growth and induced regression of cancer of the colon (3 lines), lung (4 lines), breast (2 lines), stomach (1 line), ovary (1 line), and malignant melanoma (2 lines); (b) the drug was equally effective administered i.m. or p.o. Both routes are significantly better than i.v. administration; (c) CAM is substantially more effective and less toxic than its sodium salt, which was unsuccessfully tested in cancer patients. CAM should be further tested against responsive cancers as a drug which is easy to isolate and formulate for large-scale studies.

DNA topoisomerase I--targeted chemotherapy of human colon cancer in xenografts

Drug development is needed to improve chemotherapy of patients with locally advanced or metastatic colon carcinoma, who otherwise have an unfavorable prognosis. DNA topoisomerase I, a nuclear enzyme important for solving topological problems arising during DNA replication and for other cellular functions, has been identified as a principal target of a plant alkaloid 20(S)-camptothecin. Significantly increased concentrations of this enzyme, compared to that in normal colonic mucosa, were found in advanced stages of human colon adenocarcinoma and in xenografts of colon cancer carried by immunodeficient mice. Several synthetic analogs of camptothecin, selected by tests with the purified enzyme and tissue-culture screens, were evaluated in the xenograft model. Unlike other anticancer drugs tested, 20(RS)-9-amino-camptothecin (9-AC) induced disease-free remissions. The overall drug toxicity was low and allowed for repeated courses of treatment.

DNA topoisomerase II-mediated interaction of doxorubicin and daunorubicin congeners with DNA

Three groups of doxorubicin and daunorubicin analogues, differing by their substituents on the chromophore and sugar moieties, were used in this study. The 3'-N-unsubstituted (Group 1), 3'-N-acyl (Group 2), and 3'-N-alkyl (Group 3) analogues were tested for: (a) in vivo antitumor activity and in vitro cytotoxicity; (b) cellular or tissue uptake and metabolic conversion; (c) strength of DNA intercalation; and (d) interaction with DNA topoisomerase II (topo-II). Compounds of Group 1 were cytotoxic, were strongly intercalative, and, except for those with C-14 side chain substitution, induced the formation of topo-II-DNA cleavable complexes. As shown previously, esterolysis of C-14-acyl substituents was required to yield a metabolite which can interact with topo-II in the purified system. The C-14-substituted compounds of Group 2 and their C-14-unsubstituted metabolites were cytotoxic. These drugs were weak intercalators, and the C-14-unsubstituted cogeners induced cleavable complex formation in the purified system, but with reduced potency relative to doxorubicin. The type of the 3'-N-position substituent determined whether Group 3 analogues were cytotoxic and strong intercalators, or less active and nonintercalating. Although C-14-unsubstituted intercalators of Group 3 did not form cleavable complexes in the purified system, they were cytotoxic. The study shows that DNA intercalation is required but not sufficient for the activity by topo-II-targeted anthracyclines. In addition to the planar chromophore which is involved in intercalation, two other domains of the anthracycline molecule are important for the interaction with topo-II: (a) substitution of the C-14 position totally inhibits drug activity in the purified system, but enhances cytotoxicity by aiding drug uptake and presumably acting on other cellular targets; and (b) substitutions on the 3'-N position of the sugar ring can, depending on the nature of the substituent, inhibit intercalation and/or topo-II-targeting activity. These findings may provide guidance for the synthesis and development of new active analogues.

Lack of drug-induced DNA cross-links in chlorambucil-resistant Chinese hamster ovary cells

Chlorambucil (CLB) is an alkylating agent commonly used in the treatment of several neoplastic disorders. The mechanisms underlying resistance to this drug are not fully defined. We used the DNA alkaline elution technique to study cross-link formation in the wild type (K1) and a CLB-resistant (ChlR) Chinese hamster ovary cell line. [14C]CLB was used to measure drug uptake. The CLB-resistant cells were found to have negligible DNA cross-link formation compared to K1 cells at all time points tested. There was a correlation between the resistance to CLB and the decreased ability of resistant cells to form DNA cross-links. Results of drug uptake experiments excluded altered CLB accumulation as the basis for these findings. Assays of O6-alkylguanine transferase and topoisomerase. II provide evidence against a role of these enzymes in CLB resistance. These studies suggest that the mechanism of CLB cytotoxicity involves the formation of DNA cross-links. Reduced cross-link formation may confer resistance to CLB.

DNA topoisomerase I-mediated DNA cleavage and cytotoxicity of camptothecin analogues

20(S)-Camptothecin, the 20(S)-camptothecin sodium salt, and 12 analogues with substituents on the A ring differ widely in their effectiveness in the treatment of murine L1210 lymphoblastic leukemia in vivo. The drugs were screened in the following systems: System 1, the cleavage of DNA in the presence of purified topoisomerase I; System 2, drug-induced trapping of topoisomerase I in a covalent complex with DNA; and System 3, the induction of protein-associated DNA breaks in drug-treated L1210 leukemia cells. 9-Amino-20(S), 10-amino-20(RS), and 10,11-methylenedioxy-20(RS), drugs effective against murine L1210 leukemia in vivo, stabilize topoisomerase I-DNA cleavable complexes in a purified system and in cultured L1210 cells. Other analogues, inactive against L1210 leukemia in vivo, were totally ineffective in topoisomerase I-directed screens. The rest of the analogues were intermediate in terms of their antitumor and topoisomerase I-directed activities. The study shows that the drug-induced accumulation of enzyme-DNA cleavable complexes is directly proportional to drug cytotoxicity and antitumor activity.

Studies on drug resistance in chronic lymphocytic leukemia

Chronic lymphocytic leukemia is a neoplastic disease in which drug resistance invariably occurs. We have studied the uptake and interaction with molecular targets of two drugs, chlorambucil and adriamycin, in CLL lymphocytes and CHO cell lines. Resistance does not appear related to uptake for either drug. Exposure to CLB causes DNA cross-links in the sensitive but not in the resistant cell line. The GSH content of B-CLL lymphocytes is depleted after a 20-hr incubation. An inability to maintain its GSH content may contribute to this cell's vulnerability to CLB. The resistance of CLL lymphocytes to ADR may be related to the undetectable levels of its target enzyme DNA topoisomerase II. Future approaches may involve study of novel anthracyclines, DNA topoisomerase I inhibitors and the development of in vitro predictive tests.

Resistance of human leukemic and normal lymphocytes to drug-induced DNA cleavage and low levels of DNA topoisomerase II

Adriamycin, amsacrine, and etoposide produce protein-associated DNA breaks in numerous cell types. However, in vitro exposure to Adriamycin (0.1-50.0 micrograms/ml) resulted in no detectable DNA cleavage in lymphocytes from patients with B-cell chronic lymphocytic leukemia (CLL) or in either B- or T-lymphocytes from normal donors. In contrast, DNA cleavage was observed in T-cells from CLL patients. Exposure to amsacrine or etoposide caused at least 50-fold less DNA cleavage in CLL and normal lymphocytes as compared to L1210 cells. These findings cannot be accounted for by differences in drug uptake. An attempt was made to explain the relative resistance of human lymphocytes to drug-induced DNA cleavage. DNA topoisomerase II, an intracellular target of tested drugs, was assayed in CLL and normal human blood lymphocytes by immunoblotting. The enzyme was detected neither in unfractionated lymphocytes nor in the enriched B- and T-cells from 28 untreated patients with CLL (Stage 0-IV) and from seven normal donors. Exponentially growing L1210 cells had approximately 7 x 10(5) enzyme copies per cell, suggesting a 100-fold higher content than that of CLL or normal lymphocytes. There were, however, detectable levels of DNA topoisomerase II in cells obtained from patients with diffuse histiocytic, nodular poorly differentiated and nodular mixed lymphomas, in Burkitt's lymphoma, acute lymphoblastic leukemia and CLL with prolymphocytic transformation. DNA topoisomerase I, a potential target for anticancer chemotherapy, was detectable in CLL and normal lymphocytes, as well as in cells of other malignancies tested. The above results may offer an explanation for the ineffectiveness of Adriamycin in the treatment of CLL. It could be suggested that low levels of DNA topoisomerase II contribute to drug resistance operating in human malignancies with a large compartment of nonproliferating cells.

Effects of N-trifluoroacetyladriamycin-14-O-hemiadipate and radiation on L1210 cells

N-Trifluoroacetyladriamycin-14-O-hemiadipate (AD 143) is the most active among the 14-O-hemiester adriamycin-trifluoroacetamide derivatives and has been selected for preclinical studies. We now report its ability to enhance the kill by ionizing radiation of murine leukemic cells in culture. A 1-h exposure to either 1.28-12.8 micrograms/ml of AD 143 or to 0.16-1.62 micrograms/ml of adriamycin (ADR) was followed at 0 h by graded doses (0-1 Krad) of radiation, and cell viability was assessed by soft agar cloning technique. Regression analyses of the dose-response curve have shown that both compounds, at the concentrations employed, decrease the reciprocal of the slope D0 from 97 rad for radiation alone, to 66-56 rad for AD 143 (1.28-5.12 micrograms/ml) plus radiation, or to 85-61 rad for ADR (0.16-0.65 micrograms/ml) when used with radiation. ADR, however, had a significant "shoulder"-modifying effect. The Dq remained essentially unchanged after AD 143 pretreatment. Quantitation of synergism (superadditivity), additivity, and antagonism was performed by isobologram analysis and by a computerized method based on the "median effect principle." Both approaches have shown that synergism of AD 143 or ADR with radiation becomes apparent with dose escalation. This effect is discernible at significantly lower levels of AD 143 than of ADR, corresponding to less than LD50 measured by the clonogenic assay.

Effects of new N-alkyl analogues of adriamycin on in vitro survival and cell cycle progression of L1210 cells

The effects of N-benzyladriamycin-14-valerate (AD198) and N,N-dimethyladriamycin-14-valerate (AD199), two novel lipophilic N-alkyl derivatives of Adriamycin (ADR), on cell growth and cell cycle distribution were investigated in L1210 cells grown in suspension. Following a 1-h exposure to the drug levels selected, growth inhibition was noticeable in all cultures for most or all of the observation period of 96 h. With flow cytometry, an asynchronous cell population was measured with respect to cellular DNA, RNA, and light scatter (size) properties following a 1-h incubation with the various ADR analogues. In addition, flow cytometric techniques were utilized to determine whether drug treatment altered the sensitivity of DNA in situ to acid-induced denaturation or to binding by small DNA-intercalating dyes. Unlike the parent compound ADR or its DNA-nonbinding derivative N-trifluoroacetyladriamycin-14-O-hemiadipate (AD143), the N-alkyl derivatives AD198 and AD199 only slightly affected L1210 cell cycle traverse over the first 5 h posttreatment. However, by 24 h, AD199 (0.62 micrograms/ml) caused an S- and G2 + M-phase accumulation which became more dramatic at 48 and 72 h. AD198 (3.27 micrograms/ml) also caused an accumulation of cells predominantly in G2 + M phase at longer culture times (48 to 96 h). The two half-substituted congeners N-benzyladriamycin (AD288) and N,N-dimethyladriamycin (AD280) affected L1210 cell cycle traverse over a similar time scale at concentrations of 12.3 and 4.17 micrograms/ml, respectively. AD280 blocked cells in G1 and G2 + M whereas AD288 caused predominantly a G2 + M accumulation. While neither ADR nor AD143 interfered appreciably with binding and fluorescence of the intercalating dye acridine orange, all of the N-alkyl analogues tested reduced the fluorescence signal of acridine orange-stained L1210 cells by 26 to 60%. This effect lasted, with decreasing intensity, for at least 48 h following a 1-h exposure to the drugs. In addition, while ADR appeared to stabilize DNA in situ against acid-induced denaturation, all N-alkyl derivatives, to varying degrees, tended to increase DNA denaturability. Thus alkylation at the glycoside amine combined with the lipophilic 14-valerate side chain function accounts for several new biochemical and biological properties of AD198 and AD199, relative to ADR.(ABSTRACT TRUNCATED AT 400 WORDS)

Two mechanisms of adriamycin-DNA interaction in L1210 cells

Among the effects exerted by adriamycin (ADR), interaction with DNA is closely related to cytotoxicity. The interaction results in the formation of protein-associated DNA single-strand breaks (PA-SSB) and, at drug levels greater than or equal to 2.8 X 10(-6) M, also in "direct" (nonenzymatic) DNA single-strand breaks (D-SSB). To characterize the two types of DNA lesions, euoxic mouse leukemia L1210 cells were treated with various antioxidant agents in the presence of 2.8 X 10(-6), X 10(-5), or X 10(-4) M concentrations of ADR. The enzymes superoxide dismutase (200 micrograms/ml) or catalase (250 micrograms/ml), the OH' scavengers dimethyl sulfoxide (70 mM) or ethanol (70 mM), and an inhibitor of superoxide production, 2-deoxy-glucose (1 and 10 mM), reduced the frequency of D-SSB to 18.3 to 68.2% of its level in ADR-treated controls, while the frequency of PA-SSB remained unchanged. These observations seem to indicate that ADR-mediated free radicals cause discernible DNA damage in euoxic cells only at very high drug concentrations, greater than the peak plasma level achievable clinically following i.v. bolus. At lower ADR levels, relevant to clinical use, another type of interaction between the drug and DNA prevails, which apparently does not involve a free-radical mechanism.

Relationship of adriamycin concentrations to the DNA lesions induced in hypoxic and euoxic L1210 cells

Exponentially growing L1210 mouse leukemia cells were incubated with Adriamycin (ADR) under hypoxic (95% N2:5% CO2) or euoxic conditions (95% air:5% CO2) for 1 hr at 37 degrees at a drug concentration ranging from 2.8 X 10(-8) to 2.8 X 10(-4) M, i.e., from levels attained clinically by bolus delivery to the high levels used as an i.p. drug dwell or experimentally, in in vitro conditions. High-pressure liquid chromatography analyses showed diminishing efficiency in drug uptake by the cells as the dose was increased. There were no significant differences between hypoxic and euoxic cells in drug uptake and metabolism. The frequency of DNA protein-associated single-strand breaks and DNA-protein cross-links per 10(6) nucleotides, detected by the alkaline elution technique, increased with the dose in the range of 2.8 X 10(-8) to 2.8 X 10(-6) M in both euoxic and hypoxic cells and declined thereafter. However, the number of DNA lesions relative to a normalized drug level declined steadily, starting with the 2.8 X 10(-7) M concentration. Concentrations greater than 2.8 X 10(-6) M of ADR induced still another type of lesion, direct DNA strand breaks, only in euoxic cells. The results indicate that a common mechanism of interaction between drug and DNA is present in hypoxic and in euoxic cells at low ADR, while an O2-dependent mechanism becomes operational in euoxic cells at high ADR levels.

Adriamycin damage: in vivo repair by lymphoma cells

The in vivo repair of ADR-induced cell damage was investigated in the DBA3 transplantable mouse lymphoma. After a single injection of 5 or 15 mg ADR/kg body weight into DBA/2J mice, the survival fraction of clonogens showed a 2.2- to 4.4-fold decrease at 12 or 18 hours post injection and returned to pretreatment levels within 6 hours. These changes were accompanied by the appearance and disappearance of DNA crosslinks and breaks. Because cell division and/or cell loss could not explain the return of clonogens to pretreatment level, the results strongly suggest repair of ADR damage in tumor cells in situ. Such an efficient repair mechanism, responding to a high toxic dose of ADR, constitutes a therapeutically unfavorable event that may contribute to drug resistance.

Protein-associated DNA breaks and DNA-protein cross-links caused by DNA nonbinding derivatives of adriamycin in L1210 cells

The effects of Adriamycin derivatives on L1210 mouse leukemia cells were studied with the DNA alkaline elution assay. The exposure of exponentially growing cells to approximately equitoxic concentrations of N-trifluoroacetyladriamycin-14-valerate (13.8 microM) and its metabolites, N-trifluoroacetyladriamycin (9.0 microM) and N-trifluoroacetyladriamycinol (43.7 microM), for 1 hr in vitro resulted in a high frequency of protein-associated DNA breaks and DNA-protein cross-links. These effects were comparable to those observed with Adriamycin (2.8 microM) and with adriamycinol (26.9 microM). In contrast to Adriamycin and its metabolite adriamycinol, N-trifluoroacetyladriamycin-14-valerate and its two major metabolites do not bind to DNA. Despite the absence of this direct interaction, N-trifluoroacetyladriamycin-14-valerate and its metabolites produce alterations in DNA comparable with the effects of intercalating agents. No evidence for conversion of N-trifluoroacetyladriamycin-14-valerate to Adriamycin or adriamycinol was found in L1210 cells. The similar effects on DNA macromolecules, observed between intercalating and non-DNA-binding anthracyclines, are consistent with the concept that mechanisms other than direct interaction with DNA play a role in the toxic effects of these compounds.

Cell kinetics of irradiated experimental tumors: cell transition from the non-proliferating to the proliferating pool

Parenchymal tumor cells of murine mammary carcinomas can be divided into two pools, using nucleoli as morphological 'markers'. Cells with dense nucleoli traverse the cell cycle and divide, thus constituting the proliferating pool. Cells with trabeculate or ring-shaped nucleoli either proceed slowly through G1 phase or are arrested in it. The role of these non-proliferating, G1 phase-confined cells in tumor regeneration was studied in vivo after a subcurative dose of X-irradiation in two transplantable tumor lines. Tumor-bearing mice were continuously injected with methyl[3H]thymidine before and after irradiation. Finally, the labeling was discontinued, mice injected with vincristine sulfate and cells arrested in metaphase were accumulated over a 10-hr period. Two clearly delineated groups of vincristine-arrested mitoses emerged in autoradiograms prepared from tumor tissue at the time of starting tumor regrowth: one group with the silver-grain counts corresponding to the background level, the other with heavily labeled mitoses. As the only source of unlabeled mitoses was unlabeled G1 phase-confined cells persisting in the tumor, this observation indicated cell transition from the non-proliferating to the proliferating pool, which took place in the initial phase of the tumor regrowth. Unlabeled progenitors have apoparently remained in G1 phase for at least 5-12 days after irradiation.

Identification and kinetics of G1 phase-confined cells in experimental mammary carcinomas

Three lines of mouse mammary carcinoma growing in isogenic hosts and differing in degree of histological differentiation and rates of proliferation were used to study parenchymal cells with various types of nucleoli. The relative number of cells possessing trabeculate or ring-shaped nucleoli or nucleolar fragments was closely related to the growth rate and degree of differentiation of tumor lines tested. All three subpopulations increased with increasing age and with decelerated tumor growth. In some cells in late telophase, either trabeculate or ring-shaped nucleoli could be distinguished in mitotic poles. This demonstrated that cells with these nucleoli are detected at the beginning of G1 phase. Even low levels of DNA synthesis, which would indicate that some cells with trabeculate or ring-shaped nucleoli or possessing nucleolar fragments were in S phase, could not be demonstrated. Microfluorometric measurements have indicated that cells with trabeculate and ring-shaped nucleoli have a DNA content close to 2c, whereas cells with dense nucleoli have a DNA content corresponding to either 2c, 2 to 4c, 4c, or greater than 4c. On the basis of these observations, it is concluded that cells with trabeculate and ring-shaped nucleoli and cells with nucleolar fragments either proceed slowly through G1 or are arrested in this phase. Cells with trabeculate nucleoli were replaced steadily, having a transit time of no longer than 84 hr. These cells constituted a "fast" component of cell renewal of G1-confined cells. A "slow" component, cells bearing ring-shaped nucleoli or nucleolar fragments, were replaced after a lag of 24 to 48 hr. with residency time for some of these cells being in excess of 84 hr.

Cell kinetics of irradiated experimental tumors: relationship between the proliferating and the nonproliferating pool

The role of nonproliferating cells in tumor regeneration has been studied after subcurative doses of low L.E.T. irradiation. Radiation was applied in a single dose at three different levels; 0-47, 0-94 and 1-88 krad. Studies included estimation of the absolute number of cells per tumor, differential cell counts, and autoradiographic determination of kinetic variables, employing transplantable mouse mammary adenocarcinoma DBAH. Quantitative changes of morphologically defined proliferating and nonproliferating cell pools were followed at different time intervals after irradiation. Irradiation resulted in reduction of the number of cells in both pools, with apparent sparing of nonproliferating cells. The regenerative period started with a gradual increase in the number of cells in the proliferating pool, whereas the number of cells in the nonproliferating pool continued to fall in tumors irradiated with 0-94 and 1-88 krad. In the late phase of tumor regrowth, the increasing number of cells in the non proliferating pool corresponded to its replenishment by cell transition from the proliferating pool. In an effort to clarify whether cell transition from the nonproliferating to the proliferating pool may take place during the regrowth of radiation perturbed tumors, cell loss rates from both pools were estimated using experimental data. In addition to cell loses from the tumor as a whole, the 'net loss rate' of the nonproliferating pool reflects the rate of cell transition from the nonproliferating to the proliferating pool, minus the rate of transition in the opposite direction. A similar definition applies to cell loss rates from the proliferating pool. The results showed: (1) high losses in both pools, with excess losses in the proliferating during the early phase after irradiation; (2) in the early stage of regrowth after irradiation, the cell net loss rate f-or the nonproliferating pool increased, in contrast to the behavior of cell loss rate for the proliferating pool and the average cell loss rate for the tumor as a whole; (3) in the late stage of regrowth a decrease in net loss rate for the nonproliferating pool reflects the excess production of nonproliferating cells over control tumors. These results suggest that cell transition from the nonproliferating to the proliferating pool takes place at the beginning of tumor regrowth after subcurative single-dose irradiation.