Abstract
INTRODUCTION: In the “Stupp” EORTC 26981-22981/NCIC CE3 trial, neutropenia (<1.0x10^9/L) or thrombocytopenia (<50x10^9/L) occurred in 4% and 3% respectively of patients during the concurrent phase, and in 4% and 11% respectively of the 78.5% who proceeded to adjuvant temozolomide. This audit was undertaken to measure our haematological toxicity and identify predictors. METHOD: All patients with Glioblastoma or Gliosarcoma first seen by neuro-oncology between 01/01/2014 and 31/12/2015 with a new patient Webform on our electronic medical record system, and who commenced the Stupp protocol, were included. Age at diagnosis, gender, ACE comorbidity score and ECOG performance status were extracted from Webforms. Chi-squared tests were used to investigate whether neutropenia or thrombocytopenia were independent of gender, comorbidity score and performance status, and unpaired sample t-tests to investigate whether mean age varied between those who did and did not become neutropenic or thrombocytopenic. RESULTS: 111 patients were included. Ninety-six patients (86.5%) completed six weeks of concurrent temozolomide. Fifteen interrupted or ceased temozolomide, due to thrombocytopenia(9), elevated ALT(3), pancytopenia(1), surgery(1) or patient choice(1). During the concurrent phase, the percentage (of 111) experiencing reduced counts were: haemoglobin ≤100g/L and ≤80g/L, 7.2% and 0.9%; platelets ≤100x10^9/L and ≤50x10^9/L, 13.5% and 8.1%; neutrophils ≤1.0x10^9/L and ≤0.5x10^9/L, 9.0% and 6.3%; lymphocytes ≤0.8x10^9/L and ≤0.5x10^9/L, 82.0% and 55.9%. Fifteen patients were recent without adjuvant data, and 20 did not receive adjuvant temozolomide due to poor performance status(9), progression(3), myelosuppression(3), death(2), patient choice(2) or liver function tests(1). Thus 76 (76/96=79.2%) proceeded to adjuvant temozolomide. Nine received >6 adjuvant cycles and eleven had not yet completed 6 cycles. Thirty-one stopped early, due to progression(12), poor performance status(11), myelosuppression(5), death(1) and other side effects(2). During adjuvant therapy, the percentage (of 76) experiencing reduced counts were: haemoglobin ≤100g/L and ≤80g/L, 1.3% and 0%; platelets ≤100x10^9/L and ≤50x10^9/L, 31.6% and 3.9%; neutrophils ≤1.0x10^9/L and ≤0.5x10^9/L, 2.6% and 0%; lymphocytes ≤0.8x10^9/L and ≤0.5x10^9/L, 65.8% and 40.8%. Overall, 7/8 patients with platelets ≤100x10^9/L and 3/4 patients with platelets ≤50x10^9/L concurrently, who received adjuvant temozolomide, also experienced platelets ≤100x10^9/L adjuvantly. Of those who did not drop platelets ≤100x10^9/L concurrently, 25% experienced platelets ≤100x10^9/L adjuvantly. However, all three patients with platelets ≤50x10^9/L adjuvantly had only reduced lymphocytes previously. Five of the adjuvant phase patients had neutrophils ≤1.0x10^9/L and two ≤0.5x10^9/L concurrently; none dropped neutrophils ≤1.0x10^9/L in the adjuvant phase, but two dropped platelets ≤100x10^9/L. Twelve patients received platelet transfusions. There were no clinically significant haemorrhages and one non-fatal neutropenic sepsis due to myelosuppression. Incidence of myelosuppression was independent of gender, ACE comorbidity score and ECOG performance status. Mean age did not differ between the groups who did and did not become myelosuppressed. CONCLUSIONS: Compared to the Stupp data, this audit showed rates of thrombocytopenia and neutropenia which were higher in the concurrent phase but equivalent (neutrophils) or lower (platelets) in the adjuvant phase. Concurrent thrombocytopenia predicted for platelets ≤100x10^9/L in the adjuvant phase, but not for platelets ≤50x10^9/L, probably due to appropriate dose reductions. Clinical factors were not associated with myelosuppression in univariate analysis.