Prognostic factors of early mortality in children and adolescents with relapsed/refractory solid tumors participating in dose-finding trials in the targeted and immune therapies era: An ITCC study

Introduction: Phase I/II trials are essential to introduce novel agents for children with cancer. Defining risk factors of early mortality could maximize the efficiency of such trials. 

Methods: Patients < 18 years with relapsed/refractory solid tumors in their first phase I/II trial were eligible in retrospect. Mortality at 30 and 90 days on treatment (30-DM, 90-DM) were calculated. Clinical/laboratory parameters and adult prognostic scores (Royal Marsden Hospital -RMH-, MD Anderson Cancer Center -MDACC-) were assessed at baseline and correlated with 90-DM (univariate analysis, logistic regression) to devise a pediatric-specific prognostic score (ITCC). 

Results: N = 507. Median age 11.6 years (range 0.5–17.9); 45 % females. 30-DM and 90-DM (95 %CI) were 4.7 % (3.1–7.0 %) and 22.9 % (19.3–26.8 %), respectively. RMH (n = 348) and MDACC (n = 345) scores correlated with 90-DM (p < 0.001). Performance status ≤ 80 %, no school attendance and lactate dehydrogenase (LDH) above normal levels strongly correlated with higher 90-DM, constituting the ITCC score (1 point each). The 90-DM with ITCC score (n = 306) of 0, 1, 2 and 3 was 2.7 %, 10.7 %, 36.4 % and 80.0 %, respectively. Odds ratios (95 %CI) for 90-DM with 1, 2 and 3 points were 4.23 (1.28–19.1); 20.0 (6.55–87.4); and 140 (37.4–720), respectively. Among patients with predicted risk of 90-DM ≥ 75 %, those who ultimately died within 90 days represented 1.4 % (RMH, MDACC) versus 7.8 % (ITCC) of the sample; p < 0.001. 

Conclusions: The early mortality rates reported here will serve as a reference for future phase I/II trials. Risk scoring based on performance status, school attendance and LDH levels can estimate 90-DM in oncology phase I/II trials.