DISCUSSION
The introduction of effective treatment for childhood leukemia in a lower-middle income country is challenging. Reports from Southeast Asia illustrate limiting factors such as infectious and bleeding complications in a program established in neighboring Laos10. In Indonesia high rates of treatment refusal were documented already before and during induction chemotherapy due to socio-economical limitations 11, but also difficulty to cope with treatment toxicity and pessimism with respect to chances of cure are furthermore reported 12 . Similar observations were made in Latin America with almost 50 % of events due to abandoned of treatment or treatment related toxicity4. However, encouraging examples have been reported in lower-middle income countries in Central America using a BFM-adapted treatment protocol in five countries achieving an overall 3 year survival rate of 68 % and an event free survival rate of 59 % including treatment abandonment 13. In Vietnam, 5 year relapse-free survival was 47.8 % using a French protocol (FRALLE 2000), excluding patients who had abandoned treatment. Furthermore, only children from families that could afford treatment costs could be included 5. The approach of the Kantha Bopha Foundation was to provide free care building on trained Cambodian personnel and infrastructure. Taking the reported incidence for Europe of 46.7 cases of ALL per million children under 15 years old, and given that 31.9% of the 16.7 million Cambodian population is aged under 15 years, we estimate that we could likely reach about one sixth of the children with ALL in the country. In this setting, a slightly modified version of the protocol ALL-MB91, combining treatment elements from the BFM and Dana-Farber consortia was used, with the intention to provide a moderately toxic regimen in a context in which the control of infections was expected to be a major limitation. Analyzing a most recent three-year period (2015-2017), we show that the estimated 3-year overall survival of 110 patients was 34.9%. These patients remained disease-free. We documented a total of 13 relapses in this cohort, but this may underestimate the number of relapses hidden in other lethal complications. In this unselected population, adherence to first line treatment was remarkable. No family abandoned treatment. The experience in Russia, where the protocol was initially established, showed that an event-free survival of 67 % could be achieved with this protocol under challenging socio-economic conditions 7. In Cambodia, more than half of the patients presented with a large leukemia burden, which is likely explained by a very delayed access to care. Importantly, in patients with a lower leukemia burden, overall and disease-free survival was 50.5 %, implicating that education of care providers and early access to diagnostic testing is of great importance. Furthermore, malnutrition was also expected to impact treatment related morbidity as discussed in ALL studies in India 14.
The majority of events in our cohort included infections and bleeding complications. Despite a reduction of treatment intensity compared to ALL protocols in high-income countries, SR patients were still exposed to total of 8 weeks of 6 mg/m2 dexamethasone and 8 additional 2-day pulses during maintenance, and to cumulative 120 mg/m2 of daunorubicin. In a similar geographical region, but with a better financial background, a modified BFM-ALL treatment was developed for Malaysia and Singapore, with a risk-stratification based on molecular minimal residual disease. In this study 8, non-HR patients (85.8% of all patients) were treated without anthracyclines in induction and with cumulative 60 mg/m2 doxorubicin in reinduction. They were exposed to a total of 6 weeks of 6 mg/m2 dexamethasone after one week of a prednisone pre-phase and obtained up to 6 pulses with 6mg/m2 dexamethasone for 7 days during maintenance. Treatment related complications occurred at similar rates compared to major protocols in the US and in Europe. Only 17 (3 %) non-HR patients died of a lethal infection. Six-year EFS was 80.6% on this protocol. SR patients despite significant deintensification of the chemotherapy had an EFS of 93%, which provides first strongest evidence that a massive reduction of anthracyclines is possible without compromising outcome for most of the patients. This concept is further supported by 2 studies in Brazil and Egypt, developed in cooperation with the St. Judes Children’s Research Hospital, where patients with very-low-risk ALL based on their response to induction were treated with a low-intensity regimen including a maximum of 50 mg/m2 of doxorubicin. In both studies, 5 year-overall survival was estimated over 95%15,16.
Clearly, further reduction of both anthracyclines and glucocorticoids will have to be implemented in our treatment strategy in Cambodia.
In order to reduce treatment intensity, an affordable diagnostic set up will have to be established to reliably identify patients with favorable leukemia biology. Introduction of flow cytometry (FCM) with a robust panel to diagnose and follow ALL is warranted. FCM was shown to identify most HR patients based on response to treatment at day 15 on a BFM regimen 17. FCM was successfully used to risk-stratify patients for low-intensity chemotherapy in the studies that we mentioned above 15,16. This readout can be achieved at reasonable costs. Improvement of genetic- and molecular classification will hardly be affordable in the current setting. However, polymerase chain reaction (PCR) based multiplex ligation-dependent probe amplification may provide a robust way to determine gene copy number variations. With this approach, a classifier was established and validated recently across a number of different ALL study cohorts that enables to detect patients with favorable risk cytogenetics readily18. Thus it is feasible to improve risk stratification with the introduction of affordable technology in order to reduce treatment intensity for most patients.
To reduce treatment related mortality supportive care is also of central importance. In this cohort bacterial infections during aplasia under chemotherapy pose the predominant problem. Empirical treatment of febrile neutropenia should be revisited according to recent guidelines19. The option of antibiotic prophylaxis may possibly be considered during induction chemotherapy. For the reduction of bleeding complications, fractionation of platelets will have to be introduced, the key infrastructure being available on site.
In the near future, promising CD-19 and CD-22 directed immunotherapy could be explored in a first line setting where patients may clearly benefit most from reduction of treatment intensity. The chemoimmunotherapeutic inotuzumab ozogamicin shows promising results in elderly patients with ALL and the CD19-directed bispecific monoclonal antibody blinatumomab markedly improved outcome in children with relapsed ALL, with a clear reduction of treatment related toxicity in the bridge to stem cell transplantation 20. There is a strong interest to develop clinical studies also in this region of the world with the aim to reduce toxicity of chemotherapy by replacing it at least in part by immunotherapy. The current focus in Cambodia remains to tailor an even less intensive chemotherapy backbone and to integrate the team into an international working group in the region, possibly together with experts of St. Jude VIVA forum to foster the development of pediatric oncology in Cambodia.
In conclusion, our retrospective analysis shows that the introduction of a complex chemotherapy protocol is feasible in the context of a charity-funded public institution in Cambodia but also underscores current socio-economical, geographical and institutional limitations. The outcome for patients with SR ALL is very encouraging. Our data provide the basis to justify modifications of the treatment strategy and major investments in supportive care to improve outcome for children with leukemia in Cambodia. We will aim at a further reduction of treatment intensity for non-HR patients and optimization of supportive care.