Fax: (011) 39 6963487
Second-line treatment with carboplatin for recurrent or progressive oligodendroglial tumors after PCV (procarbazine, lomustine, and vincristine) chemotherapy
A Phase II study
Article first published online: 15 JAN 2004
Copyright © 2004 American Cancer Society
Volume 100, Issue 4, pages 807–813, 15 February 2004
How to Cite
Soffietti, R., Nobile, M., Rudà, R., Borgognone, M., Costanza, A., Laguzzi, E. and Mutani, R. (2004), Second-line treatment with carboplatin for recurrent or progressive oligodendroglial tumors after PCV (procarbazine, lomustine, and vincristine) chemotherapy. Cancer, 100: 807–813. doi: 10.1002/cncr.20042
- Issue published online: 3 FEB 2004
- Article first published online: 15 JAN 2004
- Manuscript Accepted: 20 NOV 2003
- Manuscript Revised: 12 NOV 2003
- Manuscript Received: 25 AUG 2003
The efficacy of second-line chemotherapy for patients with recurrent or progressive oligodendroglial tumors is limited. In the current study, the authors investigated the use of carboplatin as a second-line chemotherapeutic agent against these types of tumors.
Twenty-three patients with recurrent or progressive oligodendrogliomas or oligoastrocytomas after first-line PCV (procarbazine, lomustine, and vincristine) chemotherapy were enrolled in a single-institution Phase II study of second-line carboplatin chemotherapy. All patients had undergone surgery, and most also had undergone conventional radiotherapy. Carboplatin was administered at a dose of 560 mg/m2 intravenously every 4 weeks. Responses were evaluated according to conventional criteria, based on magnetic resonance imaging (MRI) findings.
Three of 23 patients (13%) had partial responses, with neurologic improvement. Twelve patients (52%) had stable disease; in 2 of these 12 patients, a minor response was seen on MRI. Eight patients (35%) had progressive disease. The median time to tumor progression was 3 months for all patients and 9 months for patients who experienced responses to treatment. Progression-free survival rates at 6 and 12 months were 34.8% and 8.7%, respectively. Among the salvage treatment plans followed after carboplatin chemotherapy were supportive care alone, radiotherapy, third-line chemotherapy, and reoperation. The median survival duration from the start of carboplatin administration was 16 months. Myelotoxicity was severe, with Grade 3 or 4 thrombocytopenia in 60% of patients and Grade 3 or 4 neutropenia in 48% of patients.
When administered according to a monthly schedule, carboplatin exhibited modest activity in adult patients with recurrent or progressive oligodendroglioma or oligoastrocytoma who experienced treatment failure after PCV chemotherapy; the current treatment regimen also was associated with severe toxicity. Further improvement of second-line chemotherapy for the patient group examined in the current study is necessary. Cancer 2004;100:807–13. © 2004 American Cancer Society.
Anaplastic and recurrent or progressive low-grade oligodendroglial tumors are sensitive to chemotherapy, with 60–70% rates of response to PCV (procarbazine, lomustine, and vincristine)1–4 and 50–60% rates of response to temozolomide.5, 6 Nonetheless, almost all patients who experience responses (including long-lasting responses) eventually have recurrent disease. Scant data are available on the efficacy of second-line chemotherapy for patients with oligodendroglial tumors. PCV has been reported to be an effective salvage treatment regimen for tumor recurrences following non-PCV chemotherapy.7 Other agents, either alone (e.g., paclitaxel, irinotecan, temozolomide) or in combination (e.g., VP16 + cisplatin), have yielded modest results in patients who experienced treatment failure after first-line chemotherapy with PCV.7–12
Carboplatin is an analog of cisplatin that has less associated nephrotoxicity, ototoxicity, and neurotoxicity compared with the parent compound.13 Carboplatin has exhibited activity against malignant glioma in vitro,14 and it has been widely used to treat both pediatric15–17 and adult18–21 brain tumors. Friedman et al.22 preliminarily reported that carboplatin was active in the treatment of adults with progressive oligodendroglioma. The objectives of the current Phase II study were to determine the activity of carboplatin as second-line treatment for patients with recurrent or progressive oligodendroglial tumors after first-line PCV and to evaluate the toxicity of carboplatin in this pretreated patient population.
MATERIALS AND METHODS
Eligibility requirements for the current study were as follows: age ≥ 18 years and ≤ 70 years; histologic diagnosis of Grade II or III oligodendroglioma or mixed oligoastrocytoma (according to the World Health Organization classification system) at first or second surgery; evidence of tumor progression at least 6 weeks after first-line chemotherapy with PCV; measurable enhancing tumor on magnetic resonance imaging (MRI); Karnofsky Performance Status (KPS) score ≥ 60; and written informed consent. Adequate hematologic, hepatic, and renal functioning was required and was defined by the following criteria: absolute neutrophil count (ANC) ≥ 1500/mL; platelet count ≥ 100,000/mL; hemoglobin concentration ≥ 10 g/dL; aspartate aminotransferase (AST), alanine aminotransferase (ALT), and bilirubin levels < 1.5 times the upper limit of normal; serum creatinine levels < 1.5 mg/dL; and creatinine clearance > 60 mL per minute. A biopsy of the recurrent tumor was not required, but patients who underwent surgical resection before study entry were eligible only if they still had measurable disease on MRI performed within 3 days of surgery. Patients were ineligible if they had previously been treated with chemotherapy other than PCV; if they had any other neurologic, psychiatric, or medical illness that precluded compliance to or tolerance of treatment; if they had a history of malignancy other than carcinoma in situ of the cervix or nonmelanoma skin cancer; or if they were pregnant or lactating or did not wish to use effective contraception.
Treatment and Monitoring
Carboplatin was administered at a dose of 560 mg/m2 intravenously, in 5% dextrose (500 mL), over 1 hour once every 4 weeks. Carboplatin infusion was preceded by antiemetic prophylaxis with ondansetron (8 mg intravenously, diluted in 100 mL saline solution). Retreatment with chemotherapy was not initiated until the ANC was > 1500/mL, the platelet count was > 100,000/mL, and the creatinine concentration was < 1.5 mg/dL. Patients received a 25% dose reduction if the previous course resulted in an ANC nadir of < 500/mL or a platelet count nadir of < 50,000/mL. Cytokine use was permitted. Toxicities were graded according to the National Cancer Institute Common Toxicity Criteria.
Treatment was discontinued if the tumor progressed or if unmanageable toxicity developed. Pretreatment evaluation included medical history; physical and neurologic examination; complete blood cell counts; measurement of AST, ALT, and bilirubin levels; measurement of serum creatinine and urea levels and creatinine clearance; measurement of serum magnesium levels; urinalysis; audiogram; and MRI with and without contrast medium. All tests were performed within 2 weeks of treatment. In addition, complete blood cell counts were obtained weekly during treatment; creatinine clearance and serum creatinine, AST, ALT, bilirubin, and magnesium levels were measured before each course of therapy; and neurologic examination, audiogram, and MRI were performed before each course of therapy. All patients received the lowest corticosteroid dose determined to be necessary for neurologic stability. If required, anticonvulsants were prescribed to control seizures.
The current protocol was approved by the Institutional Review Board of the University of Turin–San Giovanni Battista Hospital (Turin, Italy).
Patients were considered to be assessable for response if they received at least one cycle of carboplatin and underwent posttreatment neurologic and radiologic evaluation. Response evaluation was based on measurable changes in tumor size as observed on MRI, with the corticosteroid requirement and the neurologic examination taken into consideration.23 Tumor size was defined as the maximum area of the enhancing mass and was calculated by multiplying the largest cross-sectional diameter (in cm) by the largest diameter perpendicular to that diameter. Responses were defined as follows: complete response (CR), disappearance of all enhancing tumor on 2 consecutive scans that were at least 4 weeks apart, discontinuation of corticosteroids, and neurologic improvement or stabilization; partial response (PR), a decrease of ≥ 50% in the size of the enhancing tumor on 2 consecutive scans that were at least 4 weeks apart, stabilization or reduction of the corticosteroid dose, and neurologic improvement or stabilization; progressive disease (PD), an increase of ≥ 25% in the size of the enhancing tumor or the appearance of any new tumor, stabilization or augmentation of the corticosteroid dose, and neurologic stabilization or deterioration; and stable disease (SD), any other situation. In evaluating patient responses, surgical defects, areas of calcification, and nonenhancing abnormalities were not considered.
The primary endpoint of the current study was patient response. Secondary endpoints were time to tumor progression (TTP), progression-free survival (PFS) at 6 and 12 months, and toxicity. Response rate was defined as the sum of the CR and PR rates. TTP was defined as the interval from the initiation of chemotherapy to the first observation of evidence of tumor progression. TTP curves were generated using the Kaplan–Meier method.24 The sample size for the current Phase II study was determined according to a single-stage procedure.25 In the absence of reliable data on the efficacy of second-line treatment when the current study was designed, we assumed 20% to be the largest response rate that implied a negative result and 45% to be the smallest response rate that indicated a positive result. With α set at 0.05 (one-sided), a total of 23 patients were required to achieve statistical power of 0.85.
Between 1994 and 2000, 23 consecutive adult patients were enrolled; patient characteristics at study entry are reported in Table 1. There were 14 males and 9 females, with an age range of 23–66 years (median, 51 years). All patients had undergone surgery at the time of initial diagnosis (macroscopic total resection [n = 7], subtotal resection [n = 14], or stereotactic biopsy [n = 2]), and 6 patients had undergone reoperation at recurrence before carboplatin chemotherapy. The distribution of histologic diagnoses (either at first or second surgery) was as follows: Grade II oligodendroglioma, n = 12; Grade III (anaplastic) oligodendroglioma, n = 3; Grade II mixed oligoastrocytoma, n = 5; and Grade III (anaplastic) mixed oligoastrocytoma, n = 3.
|Patient no.||Age (yrs)||Gender||Karnofsky score||Surgery||Histology||Radiotherapy (dose)||PCV chemotherapya||TTP after PCV chemotherapy (mos)|
|No. of cycles||Response|
|3||66||F||60||STR||OII||Yes (65 Gy)||6||CR||45|
|5||53||F||70||STR||OII||Yes (45 Gy)||6||PR||28|
|7||56||M||70||MTR||OII||Yes (56 Gy)||6||CR||15|
|8||37||F||90||STR||OII||Yes (54 Gy)||6||PR||12|
|9||64||M||70||STR||OII||Yes (56 Gy)||3||SD||8|
|10||40||M||80||STR||OII||Yes (50 Gy)||6||SD||30|
|12||58||M||60||STR||OII||Yes (50 Gy)||6||PR||48|
|13||58||M||60||B||OAII||Yes (56 Gy)||1||PD||2|
|14||41||M||90||MTRSTR||OAIIOAII||Yes (60 Gy)||6||PR||12|
|15||30||F||60||STRSTR||OAIIOAII||Yes (54 Gy)||6||PR||24|
|17||45||M||60||B||OAII||Yes (60 Gy)||6||PR||12|
|19||43||M||60||STR||OIII||Yes (64 Gy)||4||SD||9|
|20||58||M||80||STR||OIII||Yes (55 Gy)||6||CR||50|
|21||60||F||60||STR||OIII||Yes (55 Gy)||5||PR||23|
|22||47||F||80||MTRSTR||OAIIOAIII||Yes (55 Gy)||3||SD||6|
|23||26||M||60||STR||OAIII||Yes (55 Gy)||6||PR||15|
Sixteen patients previously had received radiotherapy, whereas seven had not. Radiotherapy was delivered to limited fields with conventional fractionation (1.8–2 gray [Gy] daily), with a median total tumor dose of 55 Gy (range, 45–65 Gy). Ten patients received adjuvant radiotherapy following first (n = 8) or second (n = 2) surgery; 1 patient received radiotherapy following biopsy and experienced a PR; 1 patient received radiotherapy at first tumor progression (before receiving PCV) and experienced a major PR; and 4 patients received radiotherapy for tumor progression after receiving PCV, with radiotherapy resulting in 1 minor response, 1 case of SD, and 2 cases of PD.
All patients previously had been treated for tumor recurrence with PCV (the standard regimen) for a maximum of six cycles when a response or SD was observed and toxicity was acceptable. Response rates after PCV chemotherapy were as follows: CR, 17% (4 of 23 patients); PR, 57% (13 of 23 patients); SD, 17% (4 of 23 patients); and PD, 9% (2 of 23 patients). The overall response rate (CR + PR) was 74%, and the median TTP was 15 months (range, 2–50 months).
All patients treated with PCV experienced recurrence and were eligible for assessment of response to carboplatin. Both tumors that originally were low-grade lesions and those that originally were high-grade lesions were intensely enhancing and enlarging on MRI. Twenty-one patients had unifocal tumors, and two had multifocal tumors. All patients had supratentorial hemispheric tumors, with a coexisting cerebellar lesion in one patient. No patient had spinal involvement or extranervous metastases.
Twenty-two patients were symptomatic, whereas only one patient was asymptomatic (with radiologic progression). Twelve patients experienced hemiparesis, which was either isolated (n = 5) or associated with homonymous hemianopsia (n = 3), sensory deficit (n = 1), or seizures (n = 3). Four patients had frontal lobe syndrome (characterized by altered personality and impaired executive cognitive functioning), three had headaches, and three had seizures only (without hemiparesis). All patients who had seizures received anticonvulsant medication (phenytoin [n = 2] or valproic acid [n = 4]). All patients who were symptomatic, aside from those with isolated seizures, received steroids (4–16 mg dexamethasone administered daily). KPS score in the current study population ranged from 60 to 90 (median, 70).
Response and Outcome after Carboplatin Administration
A total of 72 cycles of carboplatin were administered (range, 1–7 cycles per patient; median, 3 cycles per patient). Response and outcome data are reported in Tables 2 and 3. Three of 23 patients (13%) had PRs that were evident on MRI within the first 2 cycles. All three of these patients exhibited clear neurologic improvement: one experienced the disappearance of seizures (without any change in the administration of antiepileptic drugs concurrent with chemotherapy), one experienced a reduction in the frequency and intensity of headaches, and one experienced an improvement with respect to hemiparesis. Twelve of 23 patients (52%) had SD, and in 2 of these 12 patients, a minor response (i.e., a reduction of 20–30% in tumor size on MRI) was noted. No neurologic improvement was observed among patients with SD. Eight patients (35%) had PD. Five of these eight patients exhibited neurologic stability but had increased tumor size (rate of increase > 25%) after the first treatment cycle. All 3 patients who had responses to carboplatin previously had responses to PCV (CR [n = 1] or PR [n = 2]).
|Patient no.||No. of carboplatin cycles||Best response||TTP (mos)||Treatment following carboplatin chemotherapy (response)||Survival (mos)|
|2||7||PR||10||Radiotherapy (SD); STR as second surgery||25|
|4||7||PR||9||Radiotherapy (SD); temozolomide (PR)||53|
|18||1||PD||1||Temozolomide (SD); STR as third surgery; radiotherapy (SD)||45|
|20||1||PD||1||Temozolomide (PR); tamoxifen (PD)||51|
|22||6||SD (mR)||7||Temozolomide (PD)||11|
|Endpoint||No. of patients||Rate (%)||95% CI|
|Response (CR + PR)||3/23||13.0||4.5–32.1|
|At 6 mos||8/23||34.8||18.8–55.1|
|At 12 mos||2/23||8.7||2.4–26.8|
|At 12 mos||15/23||65.2||44.9–81.2|
Data on TTP were available for all patients. The median TTP was 3 months (range, 1–15 months). The 3 patients who achieved PR had TTPs of 8, 9, and 10 months, respectively (median, 9 months). The 12 patients with SD had TTPs of 2, 3, 3, 3, 4, 4, 4, 7, 7, 7, 13, and 15 months, respectively (median, 4 months). The PFS rates at 6 and 12 months were 34.8% and 8.7%, respectively.
Different treatment options were offered for tumor progression after carboplatin administration: 13 patients received supportive care alone; 4 patients received radiotherapy alone (resulting in PR [n = 2], SD [n = 1], or PD [n = 1]); 2 patients received radiotherapy (resulting in SD in both cases) followed by temozolomide (resulting in PR [n = 1] or SD [n = 1]); 3 patients received temozolomide alone (resulting in SD [n = 2] or PD [n = 1]); and 1 patient received temozolomide (resulting in PR) followed by high-dose tamoxifen (resulting in SD). Two of the six patients who received temozolomide after experiencing treatment failure with carboplatin had a PR; both patients previously had responses to PCV (one PR and one CR), and one also had a PR to carboplatin. Patients 2 and 18 underwent a second and third surgery, respectively. Overall survival in the study population ranged from 2 to 53 months (median, 16 months).
All 23 patients were evaluated for toxicity. Toxicities were primarily hematologic and included Grade 3 or 4 thrombocytopenia in 14 patients (60%), Grade 3 or 4 neutropenia in 11 patients (48%), and Grade 3 or 4 anemia in 9 patients (39%). Dose reductions and treatment delays were required in 8 (11%) and 10 (14%) cycles, respectively. Treatment was discontinued for two patients because of prolonged thrombocytopenia and leukopenia, respectively; these two patients recovered without any additional events. There were no treatment-related deaths. Outside of cases of myelosuppression, Grade 2 emesis was encountered in 2 of 23 patients. No cases of nephrotoxicity, hepatotoxicity, allergic reaction, or hearing loss were observed in the study population.
Available data on second-line chemotherapy for patients with recurrent or progressive oligodendroglial tumors after treatment failure with PCV are limited. Comparison of results among existing studies is made difficult by the heterogeneity of these studies with respect to patient selection (e.g., originally low-grade vs. high-grade tumors, pure oligodendrogliomas vs. mixed oligoastrocytomas, and enhancing vs. nonenhancing tumors) and the timing of previous PCV chemotherapy (adjuvant administration vs. administration at recurrence). All of these factors may have contributed to the high variability in the observed rates of response (ranging from 17–33% to 83–92%) to first-line PCV.7–12
Peterson et al.7 reported a response rate of 40% (4 of 10 patients), which included 1 CR, after treatment with etoposide + cisplatin. Chamberlain and Kormanik8 treated 20 patients with paclitaxel; 3 (15%) experienced a PR without neurologic improvement, and 7 (35%) had SD (median duration of response + SD, 9 months; range, 4–10 months). To our knowledge, there have been three studies on the use of temozolomide to treat anaplastic or aggressive oligodendroglial tumors.9, 10, 12 Response rates varied considerably among these studies, ranging from 25–26%9, 12 to 44%,10 with CR rates of 4–17%. The reported 6-month PFS rates were 29%, 44%, and 51%, while the 12-month PFS rates were 11%, 25%, and 27%.9, 10, 12 Irinotecan yielded a PR rate of 13%, with 6- and 12-month PFS rates of 33% and 0%, respectively.11 Responses and disease stabilization have been reported in individual patients after treatment with 5-fluorouracil + folinic acid26 and after treatment with carboplatin.18, 22
Regarding patient selection, we included patients with histologic diagnoses of low-grade or high-grade oligodendroglioma or oligoastrocytoma. It is possible that some patients who had tumors that originally were low-grade lesions and who did not undergo second surgery before carboplatin treatment may have had anaplastic tumors.
Some patients did not receive radiotherapy before carboplatin chemotherapy; this reflects our institution's strategy of postponing radiotherapy as long as possible for patients with oligodendroglial tumors to avoid the delayed neurologic sequelae associated with irradiation. Most of the remaining patients in the study were referred for chemotherapy after experiencing treatment failure with radiotherapy.
The rationale for conducting a Phase II study of carboplatin as second-line therapy for oligodendroglial tumors was twofold: first, laboratory experiments have demonstrated the absence of cross-resistance to nitrosoureas and platinum compounds27; second, clinical studies have demonstrated that carboplatin exhibits activity in patients with malignant astrocytic gliomas previously treated with nitrosoureas.18, 19, 21 We chose to test the monthly schedule of carboplatin because it holds several advantages over the weekly schedule; specifically, the monthly schedule is associated with less prominent myelotoxicity, a lower rate of allergic reactions, and increased convenience for patients and their families.
The observed response to carboplatin in the current study is similar to the responses generated by other agents, although temozolomide may possess somewhat more activity than carboplatin does. The rate of response to carboplatin (13%) was less than the corresponding rate for temozolomide (25–44%). Nonetheless, in terms of TTP and PFS rates, the results yielded by carboplatin in the current study (TTP: all patients, 3 months; patients with responses, 9 months; PFS rate: 6-month, 34.8%; 12-month, 8.7%) are similar to those reported by van den Bent et al.12 in a Phase II study of temozolomide (TTP: all patients, 3.7 months; patients with responses, 8 months; PFS rate: 6-month, 29%; 12-month, 11%).
The existence of an association between response to first-line PCV and response to second-line chemotherapy remains to be established. Such an association is suggested by the current study and by the study conducted by Chinot et al.10 but does not emerge from the data reported by van den Bent et al.9 Anaplastic oligodendrogliomas with 1p loss, either alone or in association with 19q loss, have significantly better responses to chemotherapy.28, 29 Thus, in future trials, the analysis of such molecular markers will be essential for predicting chemosensitivity and for better defining responses to various agents.
Regarding treatment options for progressive disease after carboplatin chemotherapy, two observations are noteworthy. First, some patients had responses to radiotherapy (Patients 1 and 16) or temozolomide (Patient 20) after previously having no response to carboplatin. Second, some patients received salvage treatment (including surgery) on multiple occasions, with responses varying from occasion to occasion. These findings confirm that, as has been suggested by others,7, 8 oligodendroglial tumors may respond to several sequential treatments, even if these treatments are palliative.
Survival times after carboplatin chemotherapy varied considerably. Patients who received effective salvage treatment after carboplatin chemotherapy (Patients 4 and 20) tended to have longer survival durations, but certain patients who experienced failure after salvage therapy (Patient 11) or who received supportive care only (Patients 5, 7 and 21) had relatively long survival times. Thus, the unpredictable possibility of a naturally protracted course of disease in patients with recurrent oligodendroglial tumors hinders evaluation of the impact of treatment on survival.
The hematologic toxicity observed in the current study was severe and probably was related to previous myelosuppressive therapy (namely, PCV); in fact, less myelotoxicity was observed among chemotherapy-naive patients who were treated with carboplatin according to the dosage schedule used in the current study.30 The use of a fixed carboplatin dose, rather than a dose based on the Calvert formula,31 may account for the observation of moderate-to-severe myelosuppression. Hypersensitivity to carboplatin was not observed; because the risk of hypersensitivity appears to increase with repeated exposure to the drug (generally after several courses of therapy),32 our finding may be related to the fact that most patients in the current study did not receive more than three treatment cycles.
In conclusion, monthly carboplatin exhibited modest activity as a second-line chemotherapeutic agent for adult patients who had recurrent or progressive oligodendroglioma or oligoastrocytoma after treatment failure with PCV. In addition, monthly carboplatin was found to be associated with severe myelotoxicity. Based on its activity and toxicity profiles, temozolomide appears to represent the best second-line chemotherapy option at present; however, because only a limited number of patients experience responses to currently available drugs, and because the duration of PFS for these patients is short, further improvement of second-line treatment (through novel agents and combinations of these agents) is necessary.
- 27Resistance to cisplatin and analogues: mechanisms and potential clinical implications. Cancer Chemother Pharmacol. 1988; 21: 325–332., , .