DISCUSSION
Our study showed a predominance of T-LLy phenotype similar to what is reported in other areas of the world, however, the proportion of B-LLy was higher. 3-5 In our Latin American cohort, 47.6 % of the patients had T-cell phenotype and 38.9% with B-LLy phenotype, while reported series from Europe and North America indicate a different distribution of phenotypes (around 75% and 20%, respectively).3,5 This tendency to a balanced proportion of T-cell over B-cell phenotypes in Latin American LLy has been previously suggested in a smaller Brazilian report.4 In our cohort most of the cases of B precursor phenotype LLy are diagnosed in Central America and the Andean countries. (Table 3). Similar distributions of T and B immunophenotypes have been observed in acute leukemias in some countries of Latin America.6 These findings must be confirmed in further epidemiological studies and possible biological explanations and implications of these differences further investigated.
Toxic deaths (including induction death and death in complete remission) occurred in 6.3% of our patients. This is higher than the observed rates in high income countries which describe toxic mortality between 1 to 3.7% in lymphoblastic lymphomas. However these rates are lower than the recently reported 6.4% induction mortality rate and 3.8% deaths in complete remission for patients with acute lymphoblastic leukemia (ALL) in Latin America. 6,7 Several factors may explain this finding, seventy-six percent of our cohort presented with advanced high-risk disease, many had severe clinical complications like mediastinal compression or tumor lysis syndromes, and 3.2% of the patients required dialysis. In contrast, tumor lysis was reported in only 4% of patients included in a recent prospective European study, and impaired renal function occurred in 2.8% of the cases.7 Notably our institutions had limited resources for diagnosis and supportive care, and this may partly explain the higher risk of fatal toxic events and the delay in diagnosis confirmation. Local physicians were keen to use empirical tratments prior to the diagnosis in 18.4% of our cases, a higher proportion compared to other reports and more frequent than in ALL in a Latin American cohort that reported a 5%.6 This prephase approach may explain the relative low percentage of induction deaths observed especially in T-LLy.
Diagnosis of LLy still constitutes a challenge in our setting.8 A delay between biopsy and definitive diagnosis greater than 15 days was reported in 23.8% of our patients. This factor influenced the therapeutic decisions that were made in cases with high risk disease. Approximately one fourth of our patients had the diagnosis confirmed by immunocytology in fresh tissue or fluids, presenting the opportunity to establish a more timely diagnosis. However, in approximately one quarter of the cases, the diagnosis of LLy was achieved after a second pathological consultation.
The fact that in our cohort, patients presented with higher stages of the disease and with severe clinical complications may be explained by multiple contextual factors. One of the possible causes is the absence of optimal local diagnostic techniques like flow cytometry or molecular biology in some institutions delaying diagnosis and risk stratification assignment. Additionally, there is insufficient supportive care treatment in some centers, with drugs like rasburicase not being available, lack of access to urgent hemodialysis implementation , restricted ICU admissions, and/or insufficiently trained pediatricians and surgeons in the management of oncologic diseases and their emergency complications.
Treatment regimens used were diverse and were categorized as BFM and non BFM based ones. There was no significant difference in outcomes, but there was a tendency favoring BFM-based protocols (Figure 2). The CNS relapse rate was 5.5%, higher than reports from Europe and North America reporting less than 3%.7,9 There was wide variation among institutions participating in our study regarding CNS-directed therapy and prophylactic cranial radiotherapy was not used routinely.9 In fact, CNS radiotherapy was applied to only 11 children (8.7%), including the 4 cases with initial CNS involvement at diagnosis, and the remaining as CNS cranial prophylaxis with 1200Gy.
Survival results of patients treated for LLy in our cohort can be considered acceptable but lower than that observed in high income countries.7,9-11 Moreover, considering the abandonment-adjusted outcome it is imperative that we develop strategies to decrease abandonment rates. Recent work by the PAHO regional working group as part of the Global Initiative for Childhood Cancer recommends strategies that could be implemented to address this problem.12 Among the events observed in this cohort, there was an increased proportion of non-disease related mortality and abandonment and may be explained by lack of optimal resources and socio-economic and cultural factors. Abandonment is a relevant and prevalent problem in LA countries, occurring in 8.7 % of our cases, representing a quarter of all clinical events, and has a heterogeneous distribution being less prevalent in Brazil and Uruguay, as reported previously for ALL or mature B-cell malignancies.6,13,14
Lastly, despite being a retrospective study that did not include patient sensitive information and was considered exempt by USA regulations, 9 of the 16 centers in Latin America initially invited to participate in the study had difficulties in obtaining local IRB approval. Inconsistency and non-harmonized IRB approaches to the review and approval of retrospective data to answer important research questions constitutes a barrier to research initiatives in Latin America, even for the collection of non-sensitive data like in the present study.15
Limitations of our study include the retrospective design that may have missed cases and incomplete clinical data, heterogeneous resources among participating centers and no uniform therapeutic approach. With these limitations prognostic factors could not be assessed. Despite these limitations, our transnational study provides important findings to establish the landscape of lymphoblastic lymphoma in children across a wide-range of representative countries in Latin America.
Acknowledgment: This work was supported by the American Lebanese Syrian Associated Charities (ALSAC).
Conflict of interest: none to declare
Table 1: Clinical characteristics of patients with LLy.