|Home | About | Journals | Submit | Contact Us | Français|
Temozolomide (TMZ) administered daily with radiation therapy (RT) for six weeks, followed by adjuvant TMZ for six months, has become standard therapy for patients with glioblastoma multiforme (GBM). After several newly diagnosed patients at our institution developed severe (grade 3–4), prolonged thrombocytopenia, we conducted a retrospective review to define the incidence, depth, and duration of thrombocytopenia associated with this therapy. We reviewed the medical records and laboratory data of all adult patients with newly diagnosed high-grade gliomas who started treatment with this regimen between June 2004, when the regimen was first used at our institution, and August 2005. Of the 52 patients who met the criteria for this review, grade 3–4 thrombocytopenia occurred in 10 (19%; 95% CI, 10%–33%). In eight patients, the thrombocytopenia was attributable to concurrent daily TMZ and RT. The median duration of grade 3–4 thrombocytopenia was 32 days (range, 1–389 days). Five patients (10%) required platelet transfusions, two (4%) have required continued biweekly platelet transfusions for over six months, and nine (17%) discontinued therapy because of thrombocytopenia. Grade 3–4 thrombocytopenia occurred in 25% of women and 14% of men. Grade 3–4 neutropenia and anemia were noted in 10% and 8% of patients, respectively, and were not clinically significant. Between 15% and 20% of our newly diagnosed patients receiving TMZ and RT developed severe (grade 3–4) and potentially irreversible thrombocytopenia. The factors that predispose patients to this toxicity have yet to be determined. This toxicity should be considered when (1) prescribing this regimen to patient populations where a clinical benefit has yet to be shown, (2) contemplating empirical escalations of the dose or duration of TMZ, or (3) combining it with other potentially myelosuppressive therapies.
A recently reported phase III clinical trial demonstrated a significant benefit of combination therapy with temozolomide (TMZ)3 and radiation therapy (RT) compared with RT alone for patients with newly diagnosed glioblastoma multiforme (GBM) (WHO grade IV astrocytoma) (Stupp et al., 2005). The study, conducted by the European Organisation for the Research and Treatment of Cancer (EORTC), used a regimen of concomitant daily TMZ and RT for six weeks, followed by six cycles of adjuvant monthly TMZ. Patients receiving this treatment had an increased median survival (14.6 vs. 12.1 months) and an increased two-year survival rate (26.5% vs. 10.4%). The regimen was well tolerated. Sixteen percent of patients developed grade 3–4 myelosuppression. Grade 3–4 thrombocytopenia, the most common hematologic toxicity, was noted in 3% of patients during concomitant TMZ and RT and in 11% of patients during adjuvant TMZ therapy.
This treatment regimen has quickly become the standard therapy for patients with newly diagnosed GBM. In addition, its documented clinical benefit in selected patients with GBM, its low toxicity profile, and its convenient oral dosing schedule have led clinicians to use TMZ and RT for a variety of other indications. This therapy is now being administered to individuals who were specifically excluded from the EORTC trial, such as patients older than 70 years, patients with diminished renal and hepatic function, and patients with poor performance status. Additionally, many clinicians are currently using the same treatment approach in patients with related tumors for which the efficacy of concurrent TMZ and RT has yet to be documented. Examples include patients with anaplastic astrocytoma (WHO grade III astrocytoma), anaplastic oligodendroglioma (WHO grade III oligodendroglioma), and even low-grade gliomas.
A number of patients treated with TMZ and RT at our institution have developed prolonged myelosuppression, particularly thrombocytopenia. Because standard toxicity assessments used in clinical trials focus on depth of nadir, the duration and clinical consequences of thrombocytopenia in these patients have not been fully described. We therefore conducted a retrospective review of adults with high-grade gliomas who started treatment with TMZ and RT in the Department of Oncology at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins between June 2004 and August 2005.
This study was approved by the institutional review board of the Johns Hopkins Medical Institutions. A computerized database was searched for patients with newly diagnosed, histologically proven, high-grade gliomas who started treatment between June 1, 2004, and August 31, 2005. (This regimen was first used routinely at the Johns Hopkins Medical Institutions in June 2004, following presentation of the regimen at the 2004 annual meeting of the American Society of Clinical Oncology [Stupp et al., 2004].) Patients treated with the regimen of daily TMZ (75 mg/m2) and RT for six weeks, followed by six cycles of adjuvant TMZ (150–200 mg/m2 on days 1–5 of each 28-day cycle), were identified. Radiation was generally given using 46 Gy delivered to the T2 volume with a 1-cm margin, followed by an additional 14 Gy to the tumor bed or T1-contrast enhancing area. Information was recorded for laboratory data, tumor type, demographics, medical comorbidities, RT and chemotherapy doses and treatment dates, concurrent medications, type and number of transfusions, use of hematopoietic growth factors, and complications of myelosuppression. Laboratory data through the completion or discontinuation of therapy, up to April 30, 2006, were evaluated. These included baseline serum chemistries and weekly complete blood counts.
Medical records were analyzed to assess the onset, duration, and severity of myelosuppression (anemia, neutropenia, and thrombocytopenia). Durations of these toxicities were recorded as the interval between the date initially documented and the next date a laboratory value in a higher range was documented. Grading of hematologic toxicity was based on the NCI Common Terminology Criteria for Adverse Events version 3.0. Grade 3 (severe) and grade 4 (life-threatening) hematologic toxicities were noted as follows: hemoglobin (grade 3, <8 g/dl–6.5 g/dl; and grade 4, <6.5 g/dl); neutrophils (grade 3, <1000/mm3–500/mm3; and grade 4, <500/mm3); and platelets (grade 3, <50,000/mm3–25,000/mm3; and grade 4, <25,000/mm3). Patients receiving transfusions were classified as having grade 3–4 cytopenias during the transfusion period. This adjustment was not made for hematopoietic growth factor administration. Cytopenias were attributed to concomitant daily TMZ and RT administration if they occurred at any point before the start of adjuvant TMZ (usually four weeks after the completion of concomitant therapy). In patients who developed hematologic toxicity, TMZ administration was delayed, dose adjusted, or discontinued according to the manufacturer’s guidelines (Schering Corporation, 2005). Additionally, patients developing hematologic toxicity underwent a careful medication review. Other medications potentially contributing to myelosuppression (e.g., certain anticonvulsants, H2-blockers, or trimethoprim-sulfamethoxazole) were discontinued.
Using patient demographic and laboratory data, the following clinical parameters were determined: creatinine clearance (using the Cockcroft-Gault  formula), body surface area (using the Mosteller  formula), and ideal body weight (using the Devine  formula for men and the Robinson et al.  formula for women). Confidence intervals were calculated using exact binomial distribution.
A total of 52 consecutive patients with newly diagnosed high-grade gliomas treated with concomitant daily TMZ and RT, followed by adjuvant monthly TMZ, were identified. Their clinical characteristics are listed in Table 1. All patients had adequate baseline blood counts (neutrophils, >1500/mm3; and platelets, >100,000/mm3). No patient had received prior chemotherapy. Most patients were receiving corticosteroids, trimethoprim-sulfamethoxazole (as described in the EORTC regimen), and anticonvulsants, as needed.
At a median follow-up of 252 days after the start of treatment, 12 patients (23%) had completed the full course of therapy (six weeks of daily TMZ and RT, followed by six monthly cycles of adjuvant TMZ), 35 (67%) had discontinued treatment, and 5 (10%) had been lost to follow-up. Twelve patients (23%; 95% CI, 13%–37%) stopped therapy because of myelosuppression: Of these 12 patients, 9 had thrombocytopenia and 3 had neutropenia. A total of 23 patients stopped treatment for other reasons: progressive disease (15), medical comorbidities (6), and patient preference (2).
Grade 3–4 hematologic toxicities seen among the 52 patients are listed in Table 2. In total, 13 patients (25%; 95% CI, 14%–39%) developed grade 3–4 myelosuppression. Four patients (8%; 95% CI, 2%–19%) developed grade 3–4 anemia, and five (10%; 95% CI, 3%–21%) developed grade 3–4 neutropenia. Thrombocytopenia was the most common and most clinically significant form of myelosuppression, with 10 patients (19%; 95% CI, 10% – 33%) developing grade 3–4 thrombocytopenia. In some cases, multiple cytopenias occurred in a single patient. Five patients received platelet transfusions (range, 1–106 transfusions), and six received packed red blood cell transfusions (range, 2–58 units). There were no complications directly attributable to cytopenias. One patient with normal white blood cell and neutrophil counts developed aspergillus pneumonitis. Four patients with normal platelet counts had bleeding events: Two had gastrointestinal bleeding, one had an intracranial hemorrhage, and one had hematuria. Three of these patients were receiving anticoagulation therapy.
Figure 1 shows the platelet counts of the 10 patients with grade 3–4 thrombocytopenia over the first 150 days of treatment. The median onset date of grade 3–4 thrombocytopenia was day 52 of therapy (approximately one week after the completion of daily RT and TMZ and three weeks before the start of monthly adjuvant TMZ). Severe thrombocytopenia was attributable to concomitant daily TMZ and RT administration in eight patients (80%) and to adjuvant TMZ administration in two (20%). The median duration of thrombocytopenia was 32 days. The duration of grade 3–4 thrombocytopenia for each patient is shown in Fig. 2.
Two patients (4%; 95% CI, 1%–13%) developed long-term thrombocytopenia lasting more than six months. One of these patients developed grade 3–4 thrombocytopenia on day 27 of treatment and died thrombocytopenic on day 238. A bone marrow biopsy and aspirate performed on day 129 demonstrated selective absence of megakaryocytes but relatively preserved erythroid and myeloid precursors. Earlier bone marrow procedures had been performed but had not yielded specimens adequate for analysis. The other patient developed grade 3–4 thrombocytopenia on day 90 of treatment. Because of a clear downward trend in platelet count apparent shortly after the completion of concomitant daily TMZ and RT administration, the patient had never received adjuvant TMZ therapy. Bone marrow biopsy and aspirate performed on day 127 demonstrated marked hypo-cellularity. The patient was referred to the hematology service and started treatment (antithymocyte globulin and cyclosporine) for aplastic anemia, but continued to have severe cytopenias requiring frequent transfusions. He died thrombocytopenic on day 479.
Table 3 lists the clinical characteristics of the 10 patients who developed grade 3–4 thrombocytopenia. Their median age was 56 years (range, 41–78 years), compared with 52 years for the overall patient cohort. Six patients (60%) were women, a somewhat higher proportion than the 46% of the overall patient group. Nine of 10 patients had normal liver function (bilirubin, <1.3 mg/dl); liver chemistry data were not available for the other patient. The median creatinine clearance was 104 ml/min (range, 78–148 ml/min). The median ratio of actual to ideal body weight was 1.09 (range, 0.83–1.54). Eight patients (80%) had GBM, and two (20%) had anaplastic astrocytoma, a distribution similar to that seen in the overall patient cohort.
This study demonstrates the clinical impact of myelosuppression from RT and TMZ administration. While the rate and depth of myelosuppression reported here are similar to those previously described (Stupp et al., 2005), it is the duration of this toxicity—a variable not typically reported in clinical trials—that had the greatest effect on patients. In general, patients in this study who developed prolonged grade 3–4 thrombocytopenia could no longer receive chemotherapy and required frequent platelet transfusions (usually twice weekly). Among patients in this study, 17% discontinued treatment because of thrombocytopenia. This includes some patients who never met criteria for grade 3–4 thrombocytopenia, but whose platelet counts remained below 100,000/mm3—the threshold for restarting chemotherapy. Two patients (representing 4% of total patients and 20% of patients developing grade 3–4 thrombocytopenia) developed apparently irreversible thrombocytopenia (one died after 211 consecutive days of grade 3–4 thrombocytopenia, and the other died after 389 consecutive days of grade 3–4 thrombocytopenia). Five patients (10%) required platelet transfusions (range, 1–59 transfusions). Patients with severe thrombocytopenia also faced an increased risk of bleeding, most worrisome for intracranial hemorrhage in this population. That such a complication has not occurred in this cohort may reflect the importance of close monitoring and frequent platelet transfusions. Patients also developed other forms of myelosuppression, but these did not affect their clinical course to the same degree. For instance, grade 3–4 anemia persisted for extended periods (median, 86 days; and range, 8–389 days), but it affected fewer patients, necessitated fewer transfusions, and did not lead to the discontinuation of treatment.
Given the clinical experience with related chemotherapeutic agents, prolonged, severe hematologic toxicity from TMZ and RT administration is not entirely unexpected. TMZ belongs to the imidazotetrazinone class of alkylating agents (Fig. 3), which also includes dacarbazine and mitozolomide. Dacarbazine, used in the treatment of malignant melanoma and Hodgkin’s lymphoma, has been associated with profound myelosuppression, including fatal thrombocytopenia and leukopenia (Ahmann et al., 1976; Costanza et al., 1977; Gutterman et al., 1974; Kleeberg and Schreml, 1976). Mitozolomide, an earlier analog of TMZ that entered clinical trials in the 1980s, demonstrated activity against malignant melanoma and lung cancer. However, this drug never advanced beyond phase II clinical trials because of severe, unpredictable myelosuppression, particularly thrombocytopenia (Neijt et al., 1989; Newlands et al., 1985; Schornagel et al., 1990). TMZ itself has been associated with the development of myelodysplastic syndrome, acute myeloid leukemia, and severe lymphopenia (Y.B. Su et al., 2004; Y.W. Su et al., 2005). Doyle et al. (2005) reported that 3 of 16 patients (19%) receiving daily TMZ and RT developed prolonged (duration, 51–155 days) and ultimately fatal myelosuppression.
Other potential causes of thrombocytopenia in these patients were also considered. Almost all patients in this chart review received trimethoprim-sulfamethoxazole for Pneumocystis jirovecii pneumonia prophylaxis, as did patients in the EORTC study. Notably, patients in this series preferentially received oral trimethoprim-sulfamethoxazole, whereas patients in the EORTC study received either trimethoprim-sulfamethoxazole or inhaled pentamidine, which has less reported hematologic toxicity. However, because the proportion of patients receiving each agent is not reported in the EORTC study, it is difficult to assess the potential contribution of Pneumocystis jirovecii pneumonia prophylaxis to myelosuppression in these patients. Additionally, a number of patients in this study were taking H2-blockers (for peptic ulcer prophylaxis) or anticonvulsants, classes of drugs that have been associated with thrombocytopenia (George et al., 1998). In general, once patients developed thrombocytopenia, these medications were discontinued, and an alternative drug was started (e.g., proton-pump inhibitors or a different anticonvulsant). Because the precise timing of drug initiation and discontinuation was not available for this retrospective study, the degree to which these and other medications may have contributed to thrombocytopenia in this patient sample cannot be determined.
The use of concurrent cranial RT may also play a role in myelosuppression. In a study comparing sequential versus concurrent chemotherapy (cisplatin and carmustine) and RT in patients with high-grade astrocytomas, patients receiving concurrent therapy had a significantly higher rate of leukopenia, a significantly lower white blood cell nadir, and an increase in platelet-transfusion requirements. The hematologic toxicity of cranial irradiation was attributed the effect of radiation on circulating hematopoietic stem cells (Kleinberg et al., 1999). In addition, vertex radiation beams, which were used in most patients in this series as a means of limiting RT delivery to the uninvolved, contralateral brain, could theoretically affect the bone marrow of cervical vertebral bodies and lead to increased myelosuppression. However, the hypocellular bone marrow specimens described in this study were obtained from the iliac crest, a region unlikely to be affected by cranial RT.
The clinical characteristics of patients who developed grade 3–4 thrombocytopenia do not suggest any obvious predisposing factors, although the small sample size in this study may limit the ability to detect an association. While women seem more prone to this toxicity (25% of women developed grade 3–4 thrombocytopenia, compared with 14% of men), the two most severe cases occurred in men. Advanced age was not linked with myelosuppression. Although four patients in the treatment population were older than 70 years, only one developed transient grade 3–4 thrombocytopenia. Patients who developed grade 3–4 thrombocytopenia had a median age of 56 years, slightly older than the overall patient cohort (median age, 52 years) but identical to the patients in the EORTC study (median age, 56 years) (Stupp et al., 2005).
The possibility of prolonged thrombocytopenia from TMZ and RT should not change current clinical practice. Treatment with TMZ and RT clearly prolongs overall survival in patients with GBM. For patients with this devastating malignancy, a small chance of severe, possibly permanent, bone marrow toxicity is an acceptable risk. However, this regimen is now being prescribed to patients with anaplastic astrocytomas, patients with anaplastic oligodendrogliomas, and even patients with low-grade gliomas, who may live for years without chemotherapy and in whom the benefits of this regimen are not known. Careful attention to what thus far appears to be an unpredictable toxicity is important, because irreversible thrombocytopenia in a patient with a low-grade glioma would be undesirable. Prospective data from future trials may further define the toxicities of TMZ and RT. Ideally, further research will also determine which patients are predisposed to the myelosuppression of this treatment regimen, allowing clinicians to tailor therapy individually and limit this complication.
Modifications of this regimen are now being studied using higher doses of TMZ for longer periods, and with a tendency to favor daily TMZ regimens rather than the conventional five consecutive days per month. The data presented in this article suggest that these dose escalations will likely be associated with more clinically significant thrombocytopenia. Furthermore, many clinical trials of novel agents for newly diagnosed GBM are now structured such that patients receive new drugs in addition to “standard” TMZ and RT (Butowski et al., 2005; Chang et al., 2004). If the hematologic toxicity of the TMZ and RT regimen is not fully defined, myelosuppression in these patients could be attributed incorrectly to a study drug rather than to standard therapy.
In conclusion, our experience suggests that the 15%–20% of newly diagnosed patients receiving TMZ and RT who develop grade 3–4 thrombocytopenia face a significant risk of prolonged, possibly irreversible, thrombocytopenia. This complication often limits the administration of further chemotherapy, necessitates frequent transfusions, and places patients at long-term risk of bleeding. The factors that predispose patients to this toxicity have yet to be determined. This toxicity should be considered when (1) prescribing this regimen to patient populations where a clinical benefit has yet to be shown, (2) contemplating empiric escalations of the dose or duration of TMZ, or (3) combining it with other potentially myelosuppressive therapies.
We thank Xiaobu Ye, M.D., for assistance with statistical calculations.
1This study was presented in abstract form at the Annual Meeting of the American Association for Cancer Research, Washington, DC, April 1–5, 2006.
3Abbreviations used are as follows: EORTC, European Organisation for the Research and Treatment of Cancer; GBM, glioblastoma multiforme; RT, radiation therapy; TMZ, temozolomide.