A Retrospective Comparison of Matched Elderly Patients Treated with Laromustine (Cloretazine®) or Best Supportive Care or Low Dose Ara-C in the LRF AML14 Trial

Background: A substantial proportion of older patients with AML are considered unlikely to benefit from an intensive treatment approach. They often receive either best supportive care (BSC), low dose treatment such as Low Dose Ara-C (LDAC), or clinical trials of novel agents. In one of the few randomised studies where patients were prospectively considered likely to be unfit for intensive therapy, LDAC was superior to BSC with 18% v 1% patients achieving CR. No patients with high risk cytogenetics (Grimwade 1998), achieved CR (Burnett 2007). Laromustine (Cloretazine^®) is a novel sulfonylhydrazine alkylating agent which preferentially targets the O6 position of guanine resulting in DNA cross-links. Laromustine has previously shown clinical activity in patients with de novo AML and high risk MDS (Giles et al. JCO 2007). A confirmatory phase II study of single agent laromustine was conducted in previously untreated patients ≥ 60 years old with de novo AML, prospectively considered likely to be unfit for intensive chemotherapy. Patients had at least one poor risk factor, defined by age ≥70, performance status 2, unfavorable cytogenetics, or cardiac, pulmonary or hepatic dysfunction. Eighty-five patients received induction therapy with 600 mg/m² laromustine. Second induction cycles were administered in 14 patients after partial response or hematologic improvement. Eighteen patients received at least one consolidation cycle of cytarabine 400 mg/m²/day CIV for 5 days.

Methods: A retrospective non-randomised comparison was performed between the 85 patients treated with laromustine, and 121 patients satisfying the same entry criteria, treated in the AML 14 trial with either BSC or LDAC. Outcomes were compared using Mantel-Haenszel and logrank methods for unadjusted comparisons, and regression methods for adjusted analyses.

Results: Patients in AML14 were slightly older than those treated with laromustine (median age 75 v 73), and tended to have higher white blood cell counts; by contrast, there were significantly fewer cardiac or respiratory comorbidities reported in the AML14 population. Other important risk factors such as performance status and cytogenetics were similar between the groups. Responses overall (CR/CRp) were seen in 33% (28/85) of patients treated with laromustine, compared with 2% (1/60) and 23% (14/61) in patients treated with BSC and LDAC (p<0.0001, p=0.2, respectively). In particular, 1 patient with −5/del(5q), and 3 patients with −7/del(7q) cytogenetics experienced a CR with laromustine; patients in AML 14 with adverse cytogenetics saw no remissions. Survival was significantly improved in the laromustine group compared to BSC (1 year survival 20% v 8%, unadjusted HR 0.58 [0.40–0.84] p=0.004), and roughly comparable to that of LDAC (1 year survival 20% v 25%, HR 1.04 [0.73–1.49] p=0.8). Analyses adjusted for differences in baseline demographics, and using propensity scores gave consistent figures.

Conclusions: Retrospective comparison of unrandomised data has significant limitations even though care has been taken to match for factors known to be predictive for survival. Laromustine was able to achieve a higher CR rate than LDAC or BSC, and produced remissions in groups where no remissions have previously been seen with LDAC or BSC. Laromustine gave significantly better survival than BSC, and demonstrated similar survival to LDAC.