The outcome of adult patients with AML who are primary non-responders to two courses of induction chemotherapy is poor. However, the utility of a 3rd induction for a select subgroup of these patients is uncertain. Here, we evaluated the rates of response and survival after a 3rd course of induction chemotherapy for primary non-responders with AML.

We identified 98 patients from the Princess Margaret Cancer Centre between May 1999 and March 2015 who were non-responders to induction and reinduction chemotherapy. No-response to re-induction chemotherapy was defined according to the Revised Recommendations of the International Working Group for AML (JCO, 2003) as patients who survived > 7 days post re-induction and had persistent AML in blood or bone marrow (>5%). Median age was 58.3 years [range: 20-76.6]. 50 (51%) were male. 2% had favorable, 18% normal, 18% intermediate, and 48% adverse cytogenetics. 50% had de novo AML, 23% had AML secondary to MDS or MPN, and 17% had therapy-related AML. Induction chemotherapy consisted of "7+3" (n =88), Nove-HiDAC (n=1), Flag-Ida (n= 2), or similar variants (n=7). Reinduction chemotherapy consisted of Nove-HiDAC (n=70), Flag-Ida (n=7), "7+3" (n=1) or other similar variants (n =20). No patients received the same regimen for both induction and reinduction.

Of the 98 primary non-responders, 15 received a 3rd induction regimen, while the others received supportive/palliative care ± low-dose chemotherapy (57 pts), or a non-induction clinical trial (26 pts). Average age was 56.4 (sd: 12.9) for patients who received supportive/palliative care and 47.0 (sd: 17.5) for patients who received a 3rd induction (p=0.008). Other baseline characteristics including gender, cytogenetic risk, marrow blast count post 2nd induction, and time between 1st and 2nd induction, did not differ between patients who did and did not receive a 3rd induction. Time to 3rd induction was a median of 54 days [range:36-126] from the start of the 2nd induction. Of the 15 third inductions, 7 were clinical trials evaluating novel agents in combination with induction chemotherapy, while the other 8 were combinations of standard chemotherapeutics (Flag-Ida n=1), AMSA+HiDAC (n=2), Daunorubicin+ HiDAC (n=1), Nove-HiDAC (n=4).

Of the 15 patients who received a 3rd induction, 3 (20%) achieved a CR following Nove-HiDAC and Flag-Ida or AMSA+HiDAC chemotherapy, where the Ara-C was given as continuous infusion. 1 patient underwent allogeneic stem cell transplant (SCT) approximately 3.7 months after 3rd induction and remains alive 4.6 years post CR. 2 patients relapsed 2.3 and 4.7 months post CR without having received alloSCT. None of the 12 other patients responded to the 3rd induction and none had prolonged aplasia. 2 of 15 (13%) died during 3rd induction. Among the 83 patients who did not receive a 3rdinduction, 1 achieved a CR after a phase 1 clinical trial (MDM2 inhibitor) and remains in CR 3.6 years following an alloSCT.

For patients who survived the immediate post induction period and were discharged from hospital median overall survival from the start of the 2nd induction did not differ between patients who did and did not receive a 3rd induction (276 days [range: 78-1304] vs 181.5 days [range: 47-1855] respectively p= 0.14). Median duration of hospital stay (including subsequent admissions) was longer for patients receiving a 3rd induction compared to those who did not (94 days following start of the 2nd induction [range: 47-169] vs 57 days [range: 51-181], respectively;(p= 0.003)).

In summary, remissions after 3rd inductions for primary non-responders are uncommon, and short-lived, suggesting that 3rd inductions should be considered with caution and only when an SCT strategy is in place.


Gupta:Incyte Corporation: Consultancy, Research Funding; Novartis: Consultancy, Honoraria, Research Funding. Schuh:Amgen: Membership on an entity's Board of Directors or advisory committees. Yee:Novartis Canada: Membership on an entity's Board of Directors or advisory committees, Research Funding. Schimmer:Novartis: Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.