The key evidence for this protocol comes from two consecutive studies, CALGB 8811 and CALGB 9111.rr CALGB 8811 studied the efficacy of a 5 drug combination regimen for the induction, of adult patients with de novo acute lymphoblastic leukaemia followed by intensive consolidation. CALGB 9111 studied the addition of G-CSF to the 8811 regimen.
In CALGB 8811, 197 patients aged 16 to 80 years, were enrolled. All received induction chemotherapy comprising of cyclophosphamide, daunorubicin, vincristine, prednisolone and L-asparaginase. 85% (167) achieved complete remission (CR), 7% (13) had persistent disease and 9% (17) died during induction. Patients who achieved CR received consolidation with a multi-agent regimen, CNS prophylaxis, late intensification and maintenance therapy for a total of 2 years. The study concluded that this intensive regimen confers a high remission rate with a high proportion of long-term remissions in adult patients with ALL. After the first 76 patient were treated it was noted that patients older than 60 years had very high rates of death during induction (6 of 10). Subsequently, patients greater than 60 received reduced doses of cyclophosphamide, daunorubicin and prednisone during induction with a resultant reduction in early death rate.r
In CALGB 9111, G-CSF was added to the 8811 protocol. 198 adult patients (aged 16 to 83) with de novo ALL were randomised to receive placebo or G-CSF 5 micrograms/kg/day subcutaneously from day +5 until the ANC was > 1 x 109/L for 2 consecutive days. The study concluded that although patients who received G-CSF had higher remission rates and lower mortality, its use did not impact on the disease free survival. All CALGB ALL protocols now include the use of G-CSF in the induction phase only.
In CALGB 8811, CR rates were age dependent - 94% in those patients less than 30 years old, 85% in those aged 30 to 59 and 39% for those patients equal to or greater than 60 years (p-<0.001). Patients who had a mediastinal mass (100%) or blasts of T-cell origin (97% vs 80% for those with B-cell lineage) demonstrated a higher CR rate than other patients. After a median follow-up of 43 months the median survival was 36 months. For those patients who achieved CR, median remission duration was 29 months.
In CALGB 9111, there were 41 patients over 60 years old. The CR rate was 87% for patients less than 60 years and 77% for patients 60 years and older.r There was no statistically significant differences in CR rates with the addition of G-CSF however the older patients who received G-CSF had more rapid platelet recovery (median 17 vs 26 days, p=0.04).
With a median follow up of 4.7 years after CALGB 9111, there was a median overall survival (OS) of 2.3 years and disease free survival (DFS) of 2.4 years in those patients who received G-CSF. In the placebo arm, median OS was 1.7 years and median DFS was 1.8 years, but these differences were not statistically significant.
In the CALGB 8811 studyr myelosuppression and infection were the most frequent major toxicities. 9% (17) patients died during induction mostly from infection (gram-negative, Streptococcus pnemonia, Candida); 9 were over 60 years old. One patient died from tumour lysis syndrome-induced renal failure during induction.
During the consolidation or maintenance phases a further 11 patients died, including 3 who died of haemorrhagic events. The major toxicities for this study are summarised in the table below. It was found that the addition of G-CSF from day 5 of chemotherapy (CALGB 9111), did not significantly reduce the non-haematological toxicities and also, because of the incidence of infections, did not enable patients to complete the first 3 months of chemotherapy any more rapidly than those who received the placebo.r
Toxicity from Larsen et al:r
© Blood 1995