Effect of consolidation with arsenic trioxide (As2O3) on event-free survival (EFS) and overall survival (OS) among patients with newly diagnosed acute promyelocytic leukemia (APL): North American Intergroup Protocol C9710

Reviewer: Eric Shinohara MD, MSCI
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 5, 2007

Presenter: B. L. Powell
Presenter's Affiliation: Wake Forest University
Type of Session: Plenary


  • Acute promyelocytic leukemia (APL) was originally described in 1957. APL is defined by a fusion gene unique to this disease, which most commonly involves the fusion of the retinoic acid receptor-alpha (RAR a) and the promyelocytic leukemia genes (PML). APL comprises approximately 10% of all acute myelogenous leukemia (AML).  Since APL's original description, several therapeutic advances have been made.
  • In 1973 APL was found to be sensitive to treatment with anthracyclines. This was followed by the discovery that APL was sensitive to treatment with all-trans retinoic acid (ATRA). Most recently, APL has been found to be sensitive to arsenic trioxide (AT) in 1997.
  • Treatment with anthracyclins has been shown to induce complete response (CR) rates of approximately 70% with a 5 year disease free survival of approximately 35-45%. ATRA has been shown to induce a complete response in up to 90% of patients with APL.
  • This study investigated the early addition of AT as consolidation therapy in adults and children with APL.

Materials and Methods

  • The present study is a randomized, phase III trial that evaluated the benefit and toxicity of two courses of AT as the first post remission therapy in patients with newly diagnosed APL.
  • The primary endpoint of this study was event free survival (EFS).
  • Patient eligibility:
    • All patients had to have the clinical diagnosis of APL which was then confirmed by the presence of the PML- RAR a translocation using real time PCR at one of three designated labs.
    • Patients could not have received any systemic therapy other than hydroxyuria previously.
  • Patients underwent randomization and then were both given the same induction chemotherapy. The study was designed such that all but two children (defined as patients less than 15 years of age) were assigned to the non-AT arm.
  • Induction Chemotherapy was ATRA given orally at a dose of 45 mg/m2 per day from day one until complete remission, daunorubicin 50 mg/ m2 IV on days 3-6 and cytarabine at a dose of 200 mg/ m2 continously by IV on days 3-9. Patients under the age of 3 received 1.5 mg/kg of daunorubicin as a continuous infusion.
  • Following induction treatment, patients who had a complete or partial response received consolidation treatment with or without AT. Patients who were assigned to the AT arm received AT at a dose of 0.15 mg/kg per day, five days a week for five weeks. After a two week break the patients in the AT arm received a second cycle. Patients in the AT arm received the AT in the interim between induction and the consolidation therapy. Subsequently all patients received consolidation with two courses of ATRA at a dose of 45 mg/m2 for seven days and daunorubicin at a dose of 50 mg/m2 for three days, except in children less than 15 years of age who received two days of daunorubicin.
  • Patients with a complete remission were then randomized to maintenance therapy with one year of ATRA at 45 mg/m2 orally daily for one year or ATRA at the same dose plus daily 6-mercaptopurine at 60 mg/m2 per day orally and weekly methotrexate at a dose of 20 mg/m2 for one year.


  • 518 adults and 64 children (patients under the age of 15; 11% of the study cohort) with previously untreated APL were accrued by five cooperative groups (CALGB, SWOG, ECOG, COG, NCIC-CTG) from June of 1999 to March of 2005. After real time PCR analysis 45 patients were found to be ineligible and were excluded from analyses.
  • Patient characteristics were well matched between the two arms of the study. Approximately 80% of the patients were Caucasian and 80% had a performance status of 0 or 1. The majority of patients had intermediate risk disease.
  • Median follow up was 29 months. In adults, 90% had a remission (89% had a CR) and this did not differ between the two arms. Children had a response rate of 88%. CR rates were noted to be lower in high risk patients.
  • Of 243 patients in the AT arm, 218 had a CR or partial response (PR) (90% of patients) and were eligible for consolidation. Of these patients, 202 went on to receive at least one dose of AT.
  • EFS at 3 years was 81% in the AT arm and 66% in the non-AT arm, which was found to be a significant difference (p=0.0007). Pediatric patients who did not receive AT had a similar EFS compared with adults who did not receive AT (62%; p=0.5667). AT's benefit was seen across all groups of patients.
  • Overall survival (OS) at 3 years was 86% in the AT arm and 79% in the non-AT arm, however, this difference did not reach statistical significance (p=0.063). Pediatric patients who did not receive AT had a similar OS compared with adults who did not receive AT (86%; p=0.193).
  • Toxicity:
    • There were 41 deaths (8%) within 60 days of the start of treatment. The majority of these deaths were in high risk patients (defined as a white blood cell count of greater than 10,000) with 19% of high risk patients dying during induction. Other factors, including PFS, age, and gender were not associated with death during induction.
    • There were no deaths during consolidation.
    • During induction grade 3 hematologic toxicity was seen in 7% of patients and grade 4 toxicity was seen in 79% of patients. Grade 3 hematologic toxicity was seen in 65% of patients and grade 4 was seen in 22% of patients.
    • There was no increase in hematologic toxicity in the AT arm compared with the non-AT arm. Non-hematologic toxicity was higher in the AT arm compared with the non-AT arm.

Author's Conclusions

  • The authors conclude that the addition of AT as the first agent used in consolidation improved EFS in all groups of patients and trends towards a significant improvement in OS. They concluded that the toxicity associated with AT was minimal and acceptable and that AT should be used in consolidation treatment in APL.
  • Toxicity during induction is a problem in high risk patients and ways to decrease this toxicity are essential. Increased rates of relapse are also more common in high risk patients and newer protocols or agents may be needed to improve outcomes in these patients.

Clinical/Scientific Implications

  • Results from the present study suggest that AT is safe and effective in consolidation treatment of APL. Despite the advances that have been made in the treatment of APL, there are still several treatment hurdles that need to be overcome. First, early deaths related to complications from APL, such as hemorrhage occurs in 10-15% patients, predominantly those that are at high risk. Additionally, 5-30% of patients ultimately relapse, again, primarily in high risk patients. The present study did not have any effect on early death related to APL, but AT's use in consolidation did appear to improve EFS, even in high risk patients.
  • Prior studies have investigated AT as a mono-agent, used in induction, consolidation and maintenance (Mathews, Blood 2006). This prior study demonstrated that AT was effective as a single agent and may be one of the most effective agents in the treatment of APL. This prior study also suggested that AT did not exhibit cross resistance with conventional agents. Results from the present study are in agreement with these findings and suggest that AT can be combined with current therapies to produce a therapeutic advantage.
  • However, in the present study, is the improvement seen with AT due to the effectiveness of AT or due to inferior performance of the control arm? To answer this question the control arm from the present study (C9710) was compared with control arms from previous studies, including currently unpublished work by the PETHEMA APL group. From this comparison it appeared that the control arm in the present study had worse EFS survival compared with previous studies. It appeared that the EFS in the AT arm of the C9710 study was comparable to the control arm of the PETHEMA study. One possible reason for this is that the doses of chemotherapy used in C9710 were relatively low. This suggests that the addition of AT can compensate for the use of lower does chemotherapy in the treatment of APL. This suggests that the use of AT may allow the reduction of anthracyclin doses without compromising outcomes. Further studies of AT's use with higher doses of chemotherapy may reveal a further therapeutic advantage to AT use.
  • In the future, studies with expanded use of AT, perhaps in induction as well as studies investigating the use of higher doses of AT would be of great interest. Further studies to address ways to improve the early complications in high risk patients as well as better treatment regiments in high risk patients to reduce relapse could potential help this group of patients with a worse prognosis.