- © 2001 by American Society of Clinical Oncology
Effective Dosing of Topotecan With Carboplatin in Relapsed Ovarian Cancer: A Phase I/II Study
- A. Bowman,
- T. Rye,
- G. Ross,
- A. Wheatley and
- J. F. Smyth
- From the Imperial Cancer Research Fund Medical Oncology Unit, Western General Hospital, Edinburgh; and SmithKline Beecham UK, Welwyn Garden City, Hertfordshire, United Kingdom.
- Address reprint requests to John F. Smyth, MD, Imperial Cancer Research Fund Medical Oncology Unit, Western General Hospital, Crewe Rd, Edinburgh EH4 2XU, Scotland; email: j.smyth{at}icrf.icnet.uk
Abstract
PURPOSE: This phase I/II study was performed to evaluate the feasibility of administering the topoisomerase inhibitor topotecan in combination with carboplatin.
PATIENTS AND METHODS: Topotecan was given as a 30-minute infusion daily for 5 days, with carboplatin given immediately after topotecan on day 5. Treatment was repeated every 21 days. Carboplatin and then topotecan were escalated in sequential cohorts of three to six patients. Four dosage combinations of topotecan days 1 to 5 and carboplatin (day 5) were tested: 0.5 mg/m2/d and carboplatin area under the curve (AUC) of 4, topotecan 0.5 mg/m2/d and carboplatin AUC of 5, topotecan 0.75 mg/m2/d and carboplatin AUC of 5, and topotecan 1.0 mg/m2/d and carboplatin AUC of 5.
RESULTS: Grade 3 and 4 neutropenia was common at doses of 0.75 mg/m2/d and above, but dose-limiting hematologic toxicity occurred in only one patient. The most common reason for dose reduction or delay was failure of myelosuppression to resolve by day 21. Nonhematologic toxicity was generally mild. The maximum-tolerated dose as defined in the protocol was not reached, but topotecan dose escalation was stopped at 1.0 mg/m2/d, because delayed neutrophil recovery precluded re-treatment on a 21-day schedule.
CONCLUSION: Hematologic toxicity was common but rarely serious, and the combination of topotecan with carboplatin on this schedule was safe and well tolerated. Giving carboplatin to patients after topotecan on day 5, rather than on day 1, allowed dose escalation beyond the levels reported in other studies. The recommended doses for previously treated patients are topotecan 0.75 mg/m2/d, days 1 to 5, with carboplatin at an area under the curve (AUC) of 5 following topotecan on day 5. The combination of topotecan 1 mg/m2/d, days 1 to 5, followed on day 5 by carboplatin at an AUC of 5, merits further examination in untreated patients.
EPITHELIAL OVARIAN cancer is the most frequently fatal gynecologic malignancy. Treatment depends on surgical resection and the use of cytotoxic drugs, and although there have been significant advances in recent years, particularly with the use of chemotherapy, to improve the quality and duration of survival, there has been little impact on the death rate from this disease. The majority of women with advanced disease initially treated with surgery and chemotherapy experience relapse. There is therefore great interest in developing effective second-line treatments for ovarian cancer.
Topotecan is a water-soluble semi-synthetic analog of the alkaloid camptothecin. As a specific inhibitor of topoisomerase I, topotecan causes lethal DNA damage during replication. Topotecan is active in recurrent ovarian cancer, both in platinum-sensitive and platinum-refractory patients.1-8 Used as a single agent, doses of 1.5 mg/m2 given intravenously on 5 consecutive days gave results at least equivalent to those of paclitaxel at 175 mg/m2 over 3 hours, both in terms of a higher response rate (20.5% v 14%) and longer time to progression (P = .72).5 The major dose-limiting toxicity of topotecan is myelosuppression, which may be severe in some patients, but neutropenic infections are uncommon. Given the single-agent activity of topotecan in cancers such as ovarian, head and neck, and small-cell lung cancer, it has been logical to examine the combination of topotecan with platinum drugs that have a similar spectrum of disease sensitivity and differing principle toxicities. It has, however, proved difficult to combine effective doses of topotecan with cisplatin after the finding of profound and unexpected myelotoxicity. In a phase I study combining topotecan with cisplatin, Rowinsky et al9 administered cisplatin either on day 1 of a 5-day schedule of topotecan or on the fifth day. Myelotoxicity was influenced by the sequence of treatment, where cisplatin given before topotecan induced significantly worse neutropenia and thrombocytopenia than the alternate sequence. Even with the use of granulocyte colony-stimulating factor (GCSF), dose-limiting neutropenia and thrombocytopenia resulted from doses of 0.75 mg/m2 of topotecan and 50 mg/m2 of cisplatin. In a phase I study recently reported by de Jonge et al,10 topotecan was administered orally for 5 days in escalating doses, and a fixed dose of 75 mg/m2 of cisplatin was given either on day 1 or day 5 of a 3-week cycle. As with the study by Rowinsky et al, the administration of cisplatin on day 1 produced more severe myelosuppression than when given on day 5. The maximum-tolerated doses were 1.25 mg/m2/d of topotecan if cisplatin was given on day 1, and 2 mg/m2 followed by 75 mg/m2 of cisplatin on day 5 (the single-agent oral dose of topotecan is 2.3 mg/m2 daily × 5). Carboplatin is recognized to be more myelosuppressive than cisplatin, but in view of its increasing use in the treatment of ovarian cancer both as first-line and second-line therapy, there is an obvious interest in the potential to combine topotecan with carboplatin.
In a study reported by Simpson et al,11 a dose of topotecan at 0.5 mg/m2/d given intravenously for 5 consecutive days with carboplatin at an area under the curve (AUC) of 4 given on day 1 produced dose-limiting neutropenia. We therefore began a study of intravenous (IV) topotecan given daily for five days with carboplatin given on the fifth day, repeated on a three-weekly cycle, to evaluate whether, by analogy with cisplatin, it was possible to administer higher doses of both drugs when the platinum was given at the end of the topotecan sequence.
PATIENTS AND METHODS
Patients
Patients with advanced ovarian cancer who had experienced relapse after platinum-based therapy were eligible for the study. Patients were required to have a creatinine clearance (measured by isotope scan) of at least 50 mL/min, adequate bone marrow reserve, and good liver function. Written informed consent was obtained from each patient. Standard exclusion criteria applied, including extensive prior radiotherapy and concurrent treatment with anticancer therapy or any recent treatment with investigational drugs.
Methods
The primary objective of the study was to establish the maximum-tolerated dose of the combination of topotecan with carboplatin. Secondary objectives were to evaluate the response rate, time to progression, and toxicities of the combination. Simpson’s maximum-tolerated doses were taken as the starting doses for topotecan and carboplatin.11 Topotecan was administered as a 30-minute IV infusion on days 1 to 5, with carboplatin given as a 30-minute infusion immediately after topotecan on day 5. Treatment was repeated three-weekly for up to six cycles. Patients were re-treated on day 21 if nonhematologic toxicities had resolved and neutrophil count was greater than 1,000/mm3, with platelets greater than 100,000/mm3 and hemoglobin greater than 9.0 g/dL. The first cohort of three to six patients was to receive topotecan at a dose of 0.5 mg/m2 for 5 days, with carboplatin given at an AUC of 4 (estimated by isotope clearance and the Calvert formula) on day 5.12 If tolerated, the second cohort would receive the same dose of topotecan with carboplatin at an AUC of 5, and subsequent cohorts were planned with a fixed dose of carboplatin at an AUC of 5 and dose increments of topotecan of 0.25 mg/m2/d. Patients received standard antiemetic prophylaxis of domperidone on days 1 to 4 and dexamethasone with granisetron before treatment on day 5, with dexamethasone and domperidone for a further 3 days. Full blood counts were performed weekly throughout the study, and patients were assessed using clinical examination and CA-125 at each cycle. Tumor assessment by computed tomography scan was performed after cycles 3 and 6 in those patients with measurable disease.
Toxicities were graded according to the National Cancer Institute common toxicity criteria (1994). Dose-limiting toxicity was defined as grade 4 neutropenia or thrombocytopenia lasting for 7 days or complicated by infection or bleeding. Escalation to the next dose level was undertaken if the incidence of dose-limiting toxicity was less than 33% in the previous cohort of three to six patients. Topotecan doses were reduced by 0.25 mg/m2/d in any patient who had experienced grade 4 thrombocytopenia, prolonged or complicated grade 4 neutropenia, or grade 3 neutropenia lasting until day 21 of the previous treatment course. Topotecan was supplied by SmithKline Beecham Pharmaceuticals, Mundells, Welwyn Garden City, United Kingdom.
RESULTS
Patient Characteristics and Dose Escalation
Twenty patients entered the study (Table 1). All had received prior platinum treatment, and 10 had received paclitaxel. The median number of previous regimens was two, and the time from the last treatment to starting topotecan was a median of 6.5 months (range, 1 to 43 months). Five patients were thought to have platinum-sensitive disease, but 15 had experienced relapse during or within 6 months of platinum.
Patient Characteristics
Four patients were treated at the initial dose level of topotecan 0.5 mg/m2/d on days 1 to 5 with carboplatin at an AUC of 4 on day 5. The carboplatin dose was escalated to an AUC of 5 and the topotecan dose remained the same in the second cohort of four patients. Six patients received treatment at the third dose level of topotecan 0.75 mg/m2/d on days 1 to 5, with carboplatin at an AUC of 5, and the final cohort of six patients was treated with topotecan 1.0 mg/m2/d on days 1 to 5, and carboplatin at an AUC of 5 on day 5.
Twenty patients received a total of 83 cycles (median, 4.5; range, one to six cycles per patient), and all cycles were assessable for toxicity. Eight patients completed six cycles of therapy, three of them at the highest dose level. Twelve stopped prematurely because of disease progression (four patients), hematologic toxicity (three at the third dose level), intercurrent illnesses (three), surgery (one), or patient preference (one). We intended to treat on a 21-day cycle, but delays were frequent; 27 cycles (33%) were delayed for a total of 40 weeks in all. Nineteen cycles (23%) were delayed to allow recovery of myelosuppression, and the other delays were for reasons unrelated to drug or toxicity. The dose of carboplatin was not reduced in any patient, but nine patients required topotecan dose reductions for hematologic toxicity.
These delays and dose modifications resulted in treatment being delivered at a dose-intensity that was considerably less than planned at all dose levels (Table 2). Although the maximum-tolerated dose as defined in the protocol had not been reached, dose escalation was stopped after cohort 4, when it was apparent that the majority of cycles could not be given at the planned doses on a 21-day schedule.
Planned and Actual Dose-Intensity Per Dose Level
Hematologic Toxicity
Hematologic toxicity was common and increased with topotecan dose (Table 3). Grade 4 neutropenia did not occur at topotecan doses less than 0.75 mg/m2/d but was experienced by four of six patients treated at the highest dose level. The proportion of cycles with grade 3 or 4 neutropenia was seven (33%) of 21, 19 (63%) of 30, and seven (78%) of nine, when topotecan doses of 0.5, 0.75, and 1.0 mg/m2/d, respectively, were administered with carboplatin at an AUC of 5 on day 5. Grade 3 or 4 thrombocytopenia was less common, occurring in two (10%) of 21 cycles at dose level 2 and six (20%) of 30 at dose level 3. There was no serious thrombocytopenia at the highest dose administered. Anemia was common, but the incidence of grade 3 anemia was only 7% (six of 83 cycles). The median neutrophil and platelet count nadirs (× 109/L) for cycles administered at topotecan doses of 0.5, 0.75, and 1.0 mg/m2/d with carboplatin at an AUC of 5 were 1.1 and 104, 0.9 and 96, and 0.4 and 80, respectively.
Hematologic Toxicity: Worst CTC Grade Per Patient
Although myelosuppression was common, dose-limiting hematologic toxicity was not. Four episodes of grade 4 thrombocytopenia were documented at topotecan doses of 0.5 mg/m2/d (one of 21 cycles) and 0.75 mg/m2/d (three of 30 cycles) both with carboplatin at an AUC of 5. One patient required platelet support; this patient also experienced febrile neutropenia during the same cycle requiring GCSF, and withdrew from treatment. There were no toxic deaths and no other episodes of prolonged or complicated neutropenia. None of the other patients required GCSF or platelets.
The duration of myelosuppression seemed to be related to topotecan dose. Recovery by day 21 was the rule at the first dose level, but delays beyond day 21 were necessary in two of four patients and four of six patients entering the second and third dose levels, respectively, often after three or four cycles. At the highest dose, four of six patients had grade 3 neutropenia lasting beyond day 21 of the first cycle, necessitating a dose reduction. Treatment was delayed beyond day 21 because of myelosuppression in 19 cycles (23%) in all.
The dose of topotecan was reduced in nine patients. Seven of these reductions were for patients with grade 3 neutropenia who did not recover to grade 2 by day 21, and two others were because of grade 4 thrombocytopenia. A total of 27 cycles were given at a reduced dose of topotecan: zero of 19 cycles at the first dose level, four of 19 at the second, five of 18 at the third, and 18 of 27 cycles administered to patients who entered at the highest dose level.
Nonhematologic Toxicity
Toxicity was generally mild, and there were no dose-limiting nonhematologic toxicities. Alopecia occurred in 10 of 20 patients, and was as high as grade 2 in only one of these. Mild to moderate fatigue was reported by most patients and was graded as severe in three. Grade 3 emesis occurred in five patients despite prophylactic antiemetics. No serious adverse events occurred that were thought to be related to the combination, but two patients experienced cerebrovascular accidents while on treatment, and another had a myocardial infarction.
Efficacy
Partial responses were documented in five of 20 patients (25%; 95% confidence interval, 6% to 44%). Four patients had stable disease for at least 8 weeks, and the disease was progressive in four others. Seven patients were not assessable for objective response; three had no measurable disease on scan, two withdrew after one cycle (one for toxicity, one for surgery), and two patients were taken off the study without repeat abdominal scans, having suffered a cerebrovascular accident. Three of the responses occurred at the highest topotecan dose level, at which point four patients were assessable for response. Marker reductions of more than 50% were seen in 10 of 18 (56%; 95% confidence interval, 33% to 79%) patients whose CA-125 levels were significantly elevated at study entry, including four who had a complete marker response. Three of the patients who were not assessable for objective response had a marker response.
No patient had prior topotecan, but all had received at least one platinum-based regimen before. Two of the objective responses were in tumors expected to be sensitive to platinum, and three occurred in patients with platinum-refractory (two patients) or platinum-resistant (one) disease. Two of 10 patients who had received prior paclitaxel responded, including one who had experienced disease progression with paclitaxel treatment. There were no responses among patients who had had three prior regimens, but four of 10 patients who had received two prior regimens responded. There was no clear relationship between the probability of response and time from previous therapy.
Nine patients were alive at the time of the analysis, with a median follow-up of 63 weeks (range, 24 to 96 weeks). The disease progressed in 17 patients, and the median time to progression was 24 weeks (range, 7 to 46 weeks); one patient died of a cerebro-vascular accident while on treatment, and two others had not progressed at 24 and 25 weeks of follow-up. Nine patients died of ovarian cancer, and two died of other illnesses. Median survival was 37 weeks (range, 8 to 82+ weeks).
DISCUSSION
In this study we examined the combination of 5 days of IV topotecan with carboplatin administered on day 5 given on a 21-day cycle to try to achieve higher doses of both drugs than in previously reported schedules.11 Our patient population included those with platinum-sensitive and platinum-resistant tumors. The highest dose level that we achieved was 1 mg/m2/d of topotecan, days 1 to 5, followed by carboplatin at an AUC of 5 given on day 5. These patients were not treated with prophylactic GCSF. Do the actual doses matter? By investigating cell lines under laboratory conditions, it is possible to demonstrate positive synergy, additive effects, and less-than-additive effects with different sequences of exposure to these two drugs using different cell lines.9 Although some would argue that doses approximate to the single-agent maximum-tolerated doses would be preferable, the converse argument could be made that if there is synergy with regard to toxicity (myelosuppression), there may similarly be a positive antitumor synergy in schedules where lower doses have been achieved. Rowinsky et al9 suggested that the more marked myelosuppression occurring when cisplatin precedes topotecan could be due to lower renal clearance of the topoisomerase inhibitor, consequent on renal toxicity from cisplatin. Interestingly, in the de Jonge study10 with oral topotecan, no pharmacokinetic sequence-dependent effects were demonstrated.
Although no direct comparison can be made with the trial reported by Simpson et al,11 our results may support the existence of a sequence effect when carboplatin is given with topotecan. We were able to escalate both the carboplatin and the topotecan dose beyond Simpson’s maximum-tolerated doses of 0.5 mg/m2/d and AUC 4, when we gave the carboplatin on day 5 rather than day 1. It seems unlikely that this effect is accounted for by changes in the systemic exposure to topotecan. In a study in which carboplatin was given on day 1 or day 5 with topotecan and etoposide, Faucette et al13 reported that the AUC of topotecan was lower in the more myelotoxic schedule of carboplatin given on day 1. Whatever the mechanism of any possible interaction, the schedule we have used in this study is clearly active, with sustained responses in platinum- and paclitaxel-resistant as well as platinum-sensitive tumors. The CA-125 data suggest that half of the patients had some benefit from treatment.
Severe hematologic toxicity was less than has been reported for many other combinations of topotecan with platinum drugs. The incidence of dose-limiting myelosuppression was acceptable, occurring in 5% of cycles, and only one patient in 20 experienced neutropenic infection. Nadir counts were modest, with no recorded neutrophil count below 200/mm3 and only one platelet count below 15,000/mm3. We had not reached the maximum-tolerated dose as defined in the protocol when dose escalation was discontinued, because it was apparent that the majority of patients could not tolerate the highest dose level when it was given on a 21-day cycle. At a topotecan dose of 1 mg/m2/d, neutropenia failed to resolve to grade 2 by day 21 in four of six patients. Perhaps this slow recovery was not surprising; because carboplatin was given on day 5 of the schedule, we were attempting to re-treat with topotecan on day 16 after the previous carboplatin dose in a population of patients who had received two or three prior regimes. The protocol did not allow us to explore the feasibility of further topotecan escalation given on a 28-day cycle, and we did not wish to use prophylactic GCSF to maintain the dosing interval, but both of these approaches may be worth further examination. Although we cannot extrapolate our results to an untreated population, it may be that patients without prior platinum exposure would tolerate our highest dose level or even a further dose escalation.
Despite the inconvenience of a 5-day dosing schedule, compliance was good, and the regimen was acceptable to patients. Alopecia was mild in the majority, and there was a low incidence of significant nonhematologic toxicity. The recommended doses for previously treated patients are topotecan 0.75 mg/m2/d, days 1 to 5, with carboplatin at an AUC of 5 (estimated by isotope clearance and the Calvert formula) after topotecan on day 5. We believe that, in view of the excellent patient tolerability of the combination of topotecan and carboplatin given in the schedule reported in this study, doses of 1 mg/m2/d of topotecan and carboplatin at an AUC of 5 on day 5 should be explored as first-line therapy for patients who have not received any previous myelotoxic treatment.
- Received August 22, 2000.
- Accepted March 29, 2001.
