Abstract 4317


APL is widely accepted as a curable leukemia with most multi-institutional studies showing very low treatment related mortality. This is in contrast to treatment in clinical practice outside the study population where the treatment related mortality is higher. A few recent population based studies show that mortality maybe as high as 30% in APL patients during induction. A recent analysis of SEER data from 13 population-based cancer registries with 1400 APL patients in the US showed that 17% of all patients and 24% of patients greater than 55 years of age die within one month of diagnosis. Swedish registry data and Brazilian data also show this high mortality during induction. The most common causes of death are bleeding, infection, differentiation syndrome and multi-organ failure. Patients who survive induction have an excellent cure rate with few late relapses. Hence, decreasing early deaths is a high priority both at experienced as well as smaller centers with limited leukemia treatment experience in this highly curable disease.


At Georgia Health Sciences University, between 7/2005 and 6/2009, 19 patients were diagnosed with APL. Seven patients (5 high-risk and 2 low-risk) died during induction resulting in an unusually high mortality rate of 37%. All patients who survived induction are still in remission at present. The high early death rate prompted us to develop a simple, 2 page treatment algorithm that focuses on quick diagnosis, prompt initiation of therapy, and proactive and aggressive management of all the major causes of death during induction. We also developed a network of physicians in smaller community based treatment centers and gave them access to our protocol and helped them manage these patients in the induction period with the hypothesis that this standardized treatment approach will result in decreasing induction mortality.


From 11/2010 to 7/2012, we treated 5 patients at GHSU and helped manage 4 patients at 2 outreach sites. The age range was 30 to 60; two patients were high-risk, 6 intermediate- and one low-risk. In the pre-algorithm cohort the cumulative survival was 63.1% at 1 year with all deaths happening within 31 days. In contrast, after the implementation of a standardized algorithm the cumulative survival was 100% with no deaths during the induction or subsequent follow-up period, log rank p-value=0.05, with a median follow-up of more than 4-years in surviving patients.


While we recognize that this is a small cohort, our own experience and a similar approach pioneered by investigators in Brazil clearly shows that this centralized, algorithm-based management under the direct supervision of a leukemia expert can be an effective intervention to decrease early deaths in APL. Based on the Brazilian experience an international consortium was formed to reduce the mortality and interim data show a reduction in early mortality to 7.5% with this networking of treatment centers. We believe our experience warrants large scale implementation with development of a network of physicians and standardization of treatment in the United States to improve early outcomes in this highly curable leukemia.


Awan:Allos Therapeutics: Speakers Bureau.

Author notes


Asterisk with author names denotes non-ASH members.