Scoring System To Predict the Risk of Early Death during Induction Chemotherapy for Primary Acute Myeloid Leukemia (AML).

Complications during induction chemotherapy (CT) for AML cause mortality in 5–20% of cases. It is relevant to define the risk of early death in this setting. In patients with a high probability of dying due to toxicity, the intensification of the supportive measures and/or alternative antineoplastic approaches could be appropriate. The aim of this study was to detect the features associated with lethal complications during induction CT for AML. We defined early deaths (ED) as those occurring before 42 days after the start of induction in the absence of evident leukemia. We analyzed all consecutive patients diagnosed with AML between June 1998 and February 2006 in 20 Spanish hospitals. These cases were treated according two multicenter trials CETLAM 99 (n=326) and CETLAM 2003 (n=248). Both schemes included idarubicin 12 mg/m2 days 1,3 and 5, etoposide 100 mg/m2 days 1–3 and cytarabine 500 mg/m2/12 hours days 1, 3, 5 and 7 as front-line treatment. In the CETLAM 2003, G-CSF (150 mg/m² days 0–7) was added as priming therapy. The series included 574 patients, 248 (43%) female, with a median age of 48 years. Creatinine level was elevated (> 1,2 mg/dL) in 11% of patients. The overall mortality rate during induction (ED) was 12% (n=69). 335 (58%) patients achieved a complete remission with a single course of CT, 108 (19%) a partial remission and 62 (11%) were refractory. The most common causes of death were infection (n= 28, 46%), bleeding (n=7, 11%), pulmonary failure not due to infection (n=6, 10%), and multiorgan failure (n=4, 7%). Univariate analysis showed that age older than 50 years old, male gender, M4 or M5 FAB subtype, leukocyte count higher than 100x10⁹/L, blasts in the marrow >70% and creatinine level above 1,2 mg/dL were associated with more frequent ED. In multivariate analysis, elevated creatinine level [hazard ratio (HR) 2.6 (1.3–5.2); P=0.009], leukocytosis >100x10⁹/L [HR 2.3 (1.1–4.6) P=0.021] and age > 50 years old [HR 2.1 (1.4–3.9) P=0.018] were independent risk factors. The remaining parameters, including among others the use of G-CSF, gender, blasts in the marrow, treatment protocol, cytogenetics at diagnosis and FLT-3 mutational status were not associated with higher incidence of ED. Taking into account the three significant variables identified as risk factors, we developed a scoring system to predict the probability of ED during induction. The probability of ED in the low risk (none risk factor, n=212), intermediate risk (1 risk factor, n=239) and high risk (2 or 3 risk factors, n=57) categories were 3%, 11% and 28% respectively (P<0.001). The HR of ED for patients in the intermediate and high risk groups were 3.6 (1.5–8.1, P=0.003) and 7.9 (3.1–19.7; P<0.001) as compared with the low risk group. In conclusion, impaired kidney function, advanced age and hyperleukocytosis are the relevant variables increasing mortality during induction CT for AML. With these three factors it is possible to define a scoring system which predicts the probability of ED. A different approach for patients in the high-risk group should be investigated.