Implementation of the Hunger 2 regimen for acute lymphoblastic leukaemia in the Butaro Cancer Centre of Excellence: toxicity and interim outcomes

Background: Outcomes in acute lymphoblastic leukaemia (ALL) in low income countries are poor, with only
20-50% of patients successfully treated. Many of the failures are due to relapsed disease but
mortality from complications of therapy is also increased. Hunger et al. have proposed regimens
of graduated intensity for ALL in resource-limited centres so that capacity for managing toxicity,
particularly neutropenic fever can be built over time without excessive treatment-related deaths
(Hunger, Sung, & Howard, 2009).
The Butaro Cancer Center of Excellence (BCCOE) is located in a Rwandan district hospital with
limited antibiotics, no microbiology and no intensive care. The centre began treating ALL in
2012 with the Hunger 1 regimen; few deaths from toxicity were experienced but the cure rate
was unacceptably low (Rubagumya, Xu, et al., 2017). In October 2016 the centre transitioned to
using the Hunger 2 regimen, which is more intense and incorporates a delayed intensification
phase.

Objective: The objective was to assess the toxicity and interim outcomes at 24 weeks of the first 25 patients
treated with the Hunger 2 regimen for ALL at BCCOE in Rwanda starting from October 2016.

Methods: The sample included all patients, both adult and paediatric, with a pathology-confirmed
diagnosis of ALL, from October 2016 to September 2017, who were treated with the Hunger 2
regimen. Patients who received previous treatment with chemotherapy were excluded.
A retrospective cohort study design was used, with patient status at week 24 as the key outcome
variable, as well as survival of episodes of neutropenic fever. Routine data was collected by
manual chart review.

Results: Twenty-four patients had a bone marrow diagnosis of ALL during the study period, but only 17
of those survived to start treatment with the Hunger 2 regimen. The cohort included a high
Page 7 of 41
proportion of patients with high-risk features such as older age, T-cell subtype and white blood
count (WBC) greater than 50,000 at presentation.
Most treatment delays were caused by neutropenia and thrombocytopenia, with negligible delays
due to medication stockouts or social factors. There was a high use of red blood cell and platelet
transfusions during the induction phase.
Patients had a median of one episode of neutropenic fever. Median time to antibiotics was 10.5
hours. Identified infections included malaria, pyoderma, pneumonia and enterocolitis.. The most
common antibiotics used were ceftriaxone and ceftazidime. All episodes resolved with no patient
deaths.
The majority of patients (63%) who commenced treatment achieved remission and were still
alive and in treatment at week 24. Two patients failed induction, one patient died during
induction, another patient abandoned treatment and another transferred out.

Conclusion: This study suggests that neutropenic fever may be less of a limiting factor in ALL treatment in
resource-poor settings than has been supposed, as all patients survived this complication despite
limited resources. The high number of deaths prior to treatment was surprising; this may be due
to delayed presentation but deserves further study. The ability to report on ultimate outcomes of
the treatment and associations with risk factors was limited by the short duration of follow-up
and small cohort size.