Is protracted low-dose temozolomide feasible in glioma patients?

Temozolomide, an alkylating drug, is widely used in chemotherapy for glioma patients. The expression of O⁶-alkylguanine DNA alkyltransferase (AGAT) causes resistance to the antitumoral effects of alkylating agents. Prolonged exposure to temozolomide leads to AGAT consumption, thus exhausting the cellular repair mechanism and possibly overcoming inherent tumoral chemoresistance.¹ This fascinating preclinical model has prompted investigators to test the 21-days every-28-days schedule, which seems the most promising.¹ Protocols are under way to use this schedule in low-grade glioma (European Organization for Research and Treatment of Cancer [EORTC]/National Cancer Institute of Canada [NCIC]) and glioblastoma (Radiation Therapy Oncology Group [RTOG]/EORTC) patients. The maximum tolerated dose (MTD) of 100 mg/m² established in a Phase I study² was based on findings after two courses of treatment. However, because toxicity was detected only after two courses of treatment, no definitive conclusions were drawn regarding the long-term safety of this dosage. Moreover, it is reasonable to assume that if more than two courses are delivered, toxicity will be greater. The aim of the current study was therefore to investigate toxicity after this schedule given at a lower dosage than the above MTD, as is sometimes done in clinical practice.

Methods.

Inclusion criteria were initial histologic diagnosis of high-grade or low-grade glioma with radiologic evidence of recurrence or disease progression; no previous chemotherapy; age of 18 years or older; Karnofsky performance score of 60 or greater; and adequate hematologic, renal, hepatic, and cardiac function. All patients signed a form giving their fully informed consent to participate in the study, which was approved by the Institutional Review Board of Bellaria Hospital, Bologna, Italy.

Treatment consisted of temozolomide 75 mg/m²/day for 21 days every 28 days associated with prophylaxis against Pneumocystis carinii pneumonia using oral trimethoprim/sulfamethoxazole (80 mg/400 mg) three times a week.³

All adverse events were recorded and graded according to the Common Toxicity Criteria, Version 3 (http://ctep.cancer.gov/forms/CTCAEv3.pdf). Hematology was performed weekly, and complete biochemistry was undertaken before recycling. Chemotherapy was given only if absolute neutrophil count was 1,500/mL or greater, absolute lymphocyte count was 500/mL or greater, and platelet count was 100,000/mL or greater; otherwise, treatment was delayed for a maximum of 3 weeks. Chemotherapy was withheld for irreversible Grade 3 or 4 nonhematologic toxicity; the temozolomide dose was reduced by 25% in cases of reversible hematologic Grade 4 toxicity.

Results.

The study ran from November 2003 through May 2005 in the GICNO (Gruppo Italiano Cooperativo Neuro-Oncologia) network. The characteristics of patients and toxicities are reported in tables 1 and 2. The most common adverse event was lymphopenia, which occurred in 52.9% of patients. On analyzing toxicity within time frames of 0 to 3 cycles (100% of the patients), 4 to 6 cycles (57%), 7 to 9 cycles (37%), and more than 10 cycles (22%), a clear trend toward an increase in cumulative lymphopenia was observed as the cycles were delivered, being 25%, 72%, 84%, and 91% in the four groups. In agreement with the high percentage of lymphopenia, 12% of the patients had Grade 1 or 2 infections, 1 patient had herpes zoster concomitant with lymphopenia Grade 2, 3 patients had upper respiratory tract infections concomitant with Grade 2 or 3 lymphopenia, 1 patient had fungal pneumonia concomitant with Grade 4 lymphopenia, and 1 patient had a urinary infection concomitant with Grade 1 lymphopenia. No patients had Pneumocystis carinii pneumonia, and none with infection had neutropenia. The most frequent nonhematologic toxicity was transaminase increase, found in 18 patients (35.3%), being more frequent during the first 3 months (21.4%) or 6 months (27.6%) of therapy. Other toxicities had no cumulative trend.

One patient with anaplastic astrocytoma died of infection concomitant with Grade 4 pancytopenia after the eighth cycle of chemotherapy and had not recovered after 2 months. At autopsy, a fungal pneumonia, refractory to appropriate antibiotic therapy, was detected. Chemotherapy was discontinued in 27 patients due to disease progression and in 6 patients for adverse events that did not resolve within 3 weeks (Grade 5 hematologic toxicity, Grade 5 pulmonary thromboembolism, Grade 4 neutropenia, Grade 3 lymphopenia, Grade 2 skin allergy, Grade 2 neutropenia). Twenty-two cycles (7.5%) were delayed for a median time of 1 week (range 1 to 4 weeks) due to lymphopenia (54.6%), infection (22.7%), neutropenia (9.1%), nausea (4.5%), and deep venous thrombosis (9.1%).

Response to therapy was not an end point of the current study: a classic Phase II study calls for more rigorous and homogeneous histologic inclusion and evaluation criteria. Nor would our statistical design have allowed us to reliably achieve this end point. However, of five patients with anaplastic gliomas, three had disease stabilization, one had a partial response, and one had a complete response. Of the nine patients with glioblastoma who started chemotherapy 3 months after the end radiation therapy, five had disease stabilization. Nine patients with low-grade gliomas and no previous radiation therapy started temozolomide: seven of them obtained stabilization, and one obtained a partial response.

Discussion.

Gliomas often have an inherent or acquired resistance to chemotherapy. Low- and continuous-dose temozolomide regimens have been suggested as a novel strategy to overcome drug resistance, which leads to AGAT depletion. So far, no large-scale study on the long-term feasibility and toxicity of temozolomide administered for 3 weeks on/1 week off has appeared in literature, although trials using this schedule are being planned. The standard temozolomide schedule^4,5 incurs higher Grade 3/4 thrombocytopenia rates (7% to 10% vs 3.9%) and lower Grade 3/4 neutropenia rates (4% to 4.5% vs 9.8%) than the schedule given by us, using a 3 weeks on/1 week off regimen. Lymphopenia occurred in 52.9% of our patients. The standard temozolomide schedule rarely leads to lymphopenia, no cases being reported in 250 glioblastoma patients treated in Phase II/III trials.^4,5 The six patients (12%) who developed infections were lymphopenic, not neutropenic; this may have been due to the concomitant administration of steroids. Fungal and respiratory infections have also been observed in patients treated with the standard schedule, although the incidence is lower (6.5%).⁴ In our study, an extended low-dose temozolomide schedule was followed by a transaminase increase in 35.3% of patients, and this percentage is higher than the 6% reported after the standard schedule. The increased creatinine level observed by us has not been described by other authors after either standard or protracted temozolomide schedules.^4–8 In some patients, continuous administration could lead to an increase in creatinine levels due to a protracted increase in peak plasma concentrations. All our patients received trimethoprim-sulfamethoxazole prophylaxis, which may have determined or enhanced renal dysfunction.

The findings reported have important clinical implications. First, interruption of treatment due to toxicity occurred in 12% of our patients compared with 2.3% of patients given the standard schedule.⁴ Second, one of these six patients died of fungal pneumonia after severe hematologic toxicity, which presented after 7 months of chemotherapy. Third, five patients (9.8%) refused to continue with the schedule, and four of these patients opted for the standard schedule, which involves fewer hematologic controls and days of therapy. These aspects must be considered when evaluating quality of life in patients with malignant glioma.

These protracted schedules should be reserved for carefully controlled clinical trials and should be used in clinical practice only if proven to be more effective than standard schedules. Otherwise, they might be considered for patients with unmethylated MGMT promoter, in whom intensive and prolonged temozolomide schedules may help to overcome drug resistance.