Abstract
Background: Induction therapy is a critical phase in the treatment of acute leukemia (AL), yet mortality during this period remains high in low- and middle-income countries. While early mortality during induction is <5% in high-income settings, significantly higher rates have been reported in Latin America. Understanding trends and associated risk factors in our context is essential for improving outcomes.
Objectives: To estimate induction-related mortality and identify clinical and treatment-related factors associated with death during the induction-to-remission phase in adult patients with acute leukemia in a referral center in Mexico City.
Methods: We conducted a retrospective cohort study including adults (≥18 years) with newly diagnosed AL (ALL, AML, APL, MPAL) who received intensive induction chemotherapy between 2005 and 2022. Demographic, clinical, and treatment-related data were collected from medical records. Patients who died during induction were compared with those who survived. Trends in supportive care practices and infectious complications were analyzed across three time periods. Univariate and multivariate logistic regression were used to identify factors associated with induction mortality.
Results: A total of 345 patients were included; 51.6% were male, with a median age of 37 years (IQR 23–50), and only 12.8% were aged ≥60. The cohort was enriched in acute lymphoblastic leukemia (ALL, 57.4%), followed by AML (34.5%), APL (6.4%), and MPAL (1.7%). Overall induction mortality was 10.4%, with no significant change over time (14.8% in 2005–2010 vs. 10.7% in 2017–2022, p=0.634). Mortality varied by subtype: 8.1% in ALL, 14.3% in AML, and 9.1% in APL. Infection was the leading cause of death (74.3%), followed by hemorrhage (14.3%).
Serious complications during induction were frequent and showed dynamic trends over time. Septic shock decreased from 23.0% in 2005–2010 to 16.3% in 2011–2016, but rose again to 31.5% in 2017–2022 (p=0.011). ICU admissions fluctuated across periods (22.1%, 8.9%, and 19.7%; p=0.239), while the use of mechanical ventilation increased significantly (12.1%, 5.9%, and 18.0%; p=0.029).
Over time, the use of antiviral and antifungal prophylaxis increased significantly (p<0.001), while antibacterial prophylaxis decreased (from 57.4% to 25.7%, p<0.001). The proportion of patients treated in private isolation rooms rose from 54.7% to 87.3% (p<0.001). Despite these changes, the incidence of invasive fungal infections increased (from 3.3% to 18.8%, p=0.010), and Clostridium difficile infection emerged as a relevant complication, rising from 0% to 5.4% (p=0.138).
In multivariate analysis, the following variables were independently associated with induction mortality:
Age ≥45 years (OR 6.46, 95% CI 2.16–19.36, p=0.001)
AML diagnosis vs. other AL (OR 3.90, 95% CI 1.39–10.91, p=0.010)
Mechanical ventilation (OR 16.35, 95% CI 4.76–56.14, p<0.001)
Clostridium difficile infection (OR 7.46, 95% CI 1.55–35.80, p=0.012)
Other clinically relevant variables such as septic shock, ICU admission, and renal replacement therapy were included in the multivariate model due to their strong univariate associations but did not retain statistical significance after adjustment.
Conclusions: Despite improvements in supportive care, including antimicrobial prophylaxis and isolation practices, induction-related mortality in adults with acute leukemia remains unacceptably high. The recent increase in severe infectious complications—particularly septic shock, invasive fungal infections, and Clostridium difficile—highlights critical gaps in early recognition and prevention strategies. These findings support the urgent need for evidence-based national guidelines tailored to resource-limited settings to improve outcomes during induction therapy.
