DISCUSSION
This study shows a significant improvement in 3-year EFS estimates of children diagnosed with ALL at MTRH (from 18% to 41%). This could be attributed to a reduction in the abandonment of therapy and relapse. However, death was the most common cause of treatment failure, with a significant proportion of deaths occurring during induction in the two cohorts.
Treatment abandonment as a cause of treatment failure reduced significantly (from 27% to 14%) in this study. It is modestly lower than Honduras, 22% and Indonesia, 35%.4-5 However, this was higher than observed in Pakistan, 7.5%, and Rwanda, 10%.3,7 Abandonment has been directly associated with low socioeconomic status and long travel time in some LMICs.3,6 Lack of medical insurance has been associated with higher rates of treatment abandonment and lower overall survival of cancer patients in previous studies in Kenya.16 With most patients abandoning treatment during maintenance, it is probable that because patients look relatively well during maintenance, parents may think they are cured. The long duration of ALL treatment is also likely to cause families to become lethargic over time. The reduction in abandonment in this study can be attributed to continuous counseling of parents, assessing for the economic challenges, and mitigating them whenever possible. That includes assisting parents in enrolling in health insurance (NHIF) and parent support groups by the social-work services department.
The relapsed leukemia decreased significantly as a first treatment outcome. Out of those that achieved remission, the proportion of children that had a relapse (19%) was similar to what has been found in other studies where relapses are reported to range between 12% and 20%.17 Due to decreased abandonment, there were probably more patients at risk of relapse in this study. Like other centers in LMIC, our patients are treated with a standard treatment protocol without risk stratification. That means sub-optimal intensity post-induction therapy for high-risk ALL children and corresponding poor EFS. In Rwanda, in a study by Rubagumya et al. children with ALL who were treated with low-intensity treatment had EFS of 22%, with relapse as the leading cause of treatment failure (53%). However, treatment-related toxicity was acceptably low.7Therefore, there is a need to adopt a risk stratification criterion at MTRH, with more intensification of therapy for the high-risk group to reduce relapses.
Historically, WBC and age at diagnosis, sex, initial treatment response determined by end induction bone marrow morphological remission, and immunophenotype have been used to stratify patients’ risk status and predict the outcome. In this study, we showed that ALL patients with an initial WBC <50 x 109/L and patients in the age group 1 to 9 years had better event-free and overall survival with the current treatment protocol. Patients with CNS involvement were at high risk of death despite intensified intra-thecal chemotherapy alone, as there were no radiotherapy services available during the period, and few patients were referred out for craniospinal radiation. Studies from HIC have shown that determination of end-induction measurable residual disease (MRD) by flow-cytometry or PCR has proven to be a reliable predictor for relapse and overall outcome. Post-induction treatment intensification for end-induction MRD-positive patients improved overall survival.18,19 However, treatment intensification for high-risk patients demands more intensive supportive care that is not readily achievable in most centers in LMICs. Also, we do not have the resources to monitor MRD in our setting.
The proportion of patients who died during induction decreased (from 24% to 17%) between the two periods, although the difference was not statistically significant. Most of the deaths were treatment-related. These induction deaths approximate what was observed in other LMICs. In a study by Jabeen, K et al. (2016) in Pakistan, Induction deaths accounted for 19.2%, most of which were infection-related.6 In Egypt, in a study by Abdelmabood S et al. (2020), induction deaths were 23%, the majority related to infections.3 One factor for the decrease in deaths in our cohort is likely due to dropping doxorubicin from the induction regimen when L-asparaginase was more available. Using a less intensive three-drug induction regimen (prednisone, asparaginase, vincristine) has been associated with decreased early mortality (toxic deaths) and improved short-term survival compared to a more intensive anthracycline-containing induction regimen.20,21 The other factor includes prompt recognition of neutropenic fever and initiation of empiric antibiotics. The use of antimicrobial prophylaxis has been shown to reduce significantly induction deaths related to infections.22 In our setting, we have used cotrimoxazole, fluconazole, and ciprofloxacin prophylaxis during severe neutropenia (<0.5 x 109/L). Invasive infections are a common complication and cause of death during induction, requiring more effective infection prevention strategies.23 Improved supply of blood products, especially platelets, has also contributed to the supportive care of the patients during induction. At MTRH, there is an urgent need to create more space to allow for better isolation, develop antibiotic therapy guidelines based on local antimicrobial sensitivity patterns, and improve supplies of blood products to reduce induction deaths further.
Delay in diagnosis and initiation of treatment is also among the reasons for poor outcomes in LMICs.24 Although the duration of illness did not significantly influence the outcome in the two periods, more than half of the patients had had symptoms for more than 1-month before presentation at MTRH. In India, prolonged symptom-to-diagnosis interval in childhood acute ALL adversely impacted survival, with those less than eight weeks having a better median survival of 59 +/- 5 months.25 Dai Q et al. (2022) found that diagnosis delay was an independent predictor for treatment outcomes in childhood ALL.26 Longer time interval to the diagnosis of ALL has been associated with more extensive disease, increased toxicity from chemotherapy, and death. Almost all the patients diagnosed at MTRH are referred from lower-level County health facilities, with no adequate expertise and diagnostic capacity, resulting in a delay in referral, diagnosis, and initiation of treatment.
Like in the earlier cohort, further analysis of event-free survival estimates of those with NHIF at diagnosis and those without did not reveal any statistically significant difference. This could be attributed to the fact that 40 out of the 43 who did not have health insurance were assisted to be enrolled through the social work services department during the first weeks of therapy. Moreover, there is a possibility of waiver for the poorest.
Being a retrospective study, the limitations experienced included missing medical records and lack of documentation of information in some of the retrieved records.
In conclusion, there was an improvement in event-free survival and a reduction in treatment abandonment and relapse. However, induction deaths did not reduce significantly. Although the survival outcomes are much lower compared to HIC, this study highlights the importance of collaboration with institutions in HIC and LMIC.27 and continuous quality improvement strategies that are adapted to available resources. Future strategies to reduce induction mortality should include working in collaboration with county lower-level facilities to improve early recognition, referral, diagnosis, and initiation of ALL treatment. Improved availability of blood products, especially platelets, adequate ward space for isolation, and other infection prevention strategies, including handwashing and prophylactic antimicrobial use, need to be implemented. There is a need for advocacy for universal health care through NHIF coverage and improved parental education and counseling to reduce treatment abandonment further. Proper risk stratification, use of risk-adapted protocols, and treatment of relapsed disease in resource-limited settings such as MTRH need to be studied. Collaboration with institutions in HIC through the existing twinning programs needs to be enhanced for knowledge and skills transfer. Implementing all or most of this will help us to achieve a 60% cure of patients with ALL by 2030, as stated by the WHO Global initiative on Childhood Cancer.1