RASGRP1/APTX Ratio Strongly Correlates with Clinical Response and Survival in AML Patients Treated with Tipifarnib-Bortezomib Combination.

Outcome of elderly acute myeloid leukemia (AML) patients is dismal. Targeted-therapies might improve current results by overcoming drug-resistance and reducing toxicity. In particular, the farnesyl-transferase inhibitor Tipifarnib (Zarnestra^®), and the proteasome inhibitor Bortezomib (Velcade^®), appeared synergistic in AML cells ex vivo, and their association was shown to be safe in vivo in a phase I trial by our group.

We conduced a phase II study aiming to assess efficacy and toxicity of Tipifarnib-Bortezomib association in AML patients >18 years, unfit for conventional therapy, or >60 years, in relapse. Furthermore, we aimed to identify biological features potentially predictive of clinical response. In particular, we focused on the RASGRP1/APTX ratio, which was previously found to be effective in predicting treatment response in patients treated with Tipifarnib alone.

Bortezomib (1.0 mg/m²) was administered as weekly infusion for three consecutive weeks (days 1, 8, 15). Tipifarnib was administered at dose of 300-600 mg BID for 21 consecutive days. Response was assessed at the end of each cycle (28 days). Patients' withdrawn was planned in case of progression or stable disease after six cycles. Real-time quantitative-PCR (q-PCR) was used for RASGRP1/APTX quantification.

Eighty patients were enrolled (47 male). Median age was 71 years (43-89) and WBC at diagnosis was 4.2 × 10⁹/L (0.5- 42.1). Thirty-two out of 80 patients had a secondary-AML, 14 had a high risk cytogenetic and 42 were previously untreated. Seventy-five patients actually initiated the treatment, 62 completed at least the first cycle while 13 early dropped out for non-leukemia related adverse event. Nine patients achieved complete remission (CR), 1 patients obtained a partial response (PR) and in 2 cases an hematological improvement (HI) was documented for an overall response rate (ORR) of 19%. Eighteen had progressive disease (PD) and the remaining showed stable disease (SD). Median time to response was 112 days, corresponding to 4 cycles (range 2-14). Marrow response (CR+PR) was significantly associated with overall survival (OS) (p<0.0001). RASGRP1/APTX was evaluated before treatment initiation on bone marrow (BM) and/or peripheral blood (PB). The median RASGRP/APTX value on BM was 15.3 (15-19.8) in responder patients and 2.2 (0.5-25.9) in non responders, respectively (p=0.00006). Its median value on PB was 31.6 (19.3-35.5) in responders and 6.4 (0.5-27.1) in non responders, respectively (p=0.00001). Interestingly, no marrow responses were recorded in patients with marrow RASGRP1/APTX ratio <8, while the response rate was 43% in patients with RASGRP1/APTX >8 (p<0.0001). Finally, RASGRP1/APTX levels significantly correlated with OS (p=0.001) with a median OS of 490 days and 162 days in patients with RASGRP1/APTX >8 and <8 respectively. Conversely, there was no correlation between cytogenetics, secondary AML, previous treatment and response or overall survival. Toxicity was overall mild, the most common adverse event being febrile neutropenia. Permanent treatment interruption due to Tipifarnib-Bortezomib related adverse events occurred in 13/75 (17%) of patients. With a median follow-up of 122 days (range 9-737), 57/75 (76%) patients are dead and 18/75 (24%) are alive, six of which in CR.

We conclude that the clinical efficacy of the combination Tipifarnib-Bortezomib was similar to what reported for Tipifarnib alone. However, noteworthy, we could confirm that the RASGPR1/APTX BM or PB level is an effective predictor of response. Though higher RASGRP1/APTX is relatively rare (∼10% of cases), Tipifarnib (±Bortezomib) may represent an important option in a subset of high risk/frail AML patients.

Supported by BolognAIL, AIRC, European LeukemiaNET, COFIN, FIRB 2006, Fondazione del Monte di Bologna e Ravenna.

No relevant conflicts of interest to declare.