Randomized Phase III Study Comparing an Early PET Driven Treatment De-Escalation to a Not PET-Monitored Strategy in Patients with Advanced Stages Hodgkin Lymphoma: Interim Analysis of the AHL2011 Lysa Study

BACKGROUND

Escalated BEACOPP (BEAesc) demonstrated a better disease control than ABVD but no overall survival (OS) improvement in patients with advanced Hodgkin Lymphoma (HL) (Federico et al , J Clin Oncol 2009, Viviani et al, N Engl J Med 2011, Mounier et al, Ann Oncol 2014). The superior efficiency of BEAesc is associated to a substantially higher immediate hematological toxicity, and an increased risk of secondary myelodysplasia/leukemia and infertility. So, to better manage HL treatment there is a need to identify early responding patients able to benefit from a strategy of dose intensity de-escalation after upfront BEAesc, without impairing the disease control. PET performed after 2 cycles of chemotherapy (PET2) might identify such a population suitable for receiving ABVD after 2 cycles of upfront BEAesc, and was implemented in the present study.

METHODS

The AHL 2011 trial (NCT01358747) was designed to evaluate in 16 to 60 years old HL patients with Ann Arbor stage III, IV or high risk IIB, a treatment strategy driven by PET after 2 BEAesc cycles (PET2), delivering 4 cycles of ABVD for PET2 negative patients and 4 cycles of BEAesc for PET2 positive patients. This PET-driven strategy was randomly compared to a standard treatment not monitored by PET and delivering 6 cycles of BEAesc. A baseline PET was mandatory before treatment and PET2 were centrally reviewed and interpreted according to Deauville criteria within 48 hours. The allocation of treatment in the experimental arm was based on the PET2 central review results. PFS was the primary endpoint of the study with a hypothesis of non-inferiority of the PET driven arm compared to the standard arm with a margin of 10% (85% 5y-PFS in the standard arm vs >75% in the PET driven arm). An interim analysis of the primary endpoint was planned after 50% (n = 49) of the 97 scheduled events needed for the final analysis.

RESULTS

From May 2011 to May 2014, 823 patients were registered and 782 were eligible for the interim analysis including 401 patients in the arm A, and 381 in arm B. Patients characteristics were well balanced in both arms: median age was 30 years (16 - 60), 64% were male, 81% had nodular sclerosis and 12% mixed cellularity HL, 12% had stage IIB, 28% stage III, 60% stage IV, and 58% had an international prognosis score ≥3.

After 2 cycles of BEAesc PET was positive in 97 (12%) patients and PET2 positivity was similar in the standard and experimental arms (n = 48, 12% and n = 49, 13% respectively). Based on PET2 results, 319 (84%) patients received 4 cycles of ABVD and 49 (13%) 4 additional cycles of BEAesc in the experimental arm.

Grade ≥3 toxicity was significantly higher in patients receiving 6 cycles of BEAesc compared to those who received 2 cycles of BEAesc + 4 cycles of ABVD with more frequent anemia (11% vs 2%), leukopenia (85% vs 72%), thrombocytopenia (44% vs 13%), febrile neutropenia (6% vs 3%), and sepsis (7% vs 4%). 182 serious adverse events (SAE) related to treatment occurred in 108 (24%) patients treated with 6 cycles of BEAesc (leading to death in 4 cases), compared to 72 SAE (leading to death in 1 case) in 50 (15%) patients treated with 2 x BEAesc + 4 x ABVD (p<0.002). In these latter patients most of SAE (67%) occurred during the 2 first cycles of chemotherapy.

With a median follow up of 16.3 months, in an intent to treat basis, the estimated 2y-PFS was similar in the standard (91.6%) and the PET driven arms (88.3%; p = 0.79). PET2 positivity was related to a significantly lower 2y-PFS compared to PET2 negative patients in the whole population (72.9% vs 92.8%; p<0.0001) and in both randomization arms (75.1% vs 94% and 70.8% vs 91.6% in the standard and PET driven arms, respectively; p<0.0001 for both). Overall survival was similar in both randomization arms and in PET2 positive or negative subsets of patients.

CONCLUSIONS

This interim analysis suggests that PET performed after 2 cycles of BEAesc can be safely used to guide subsequent treatment and supports the response-adapted strategy delivering 4 cycles of ABVD for patients with negative PET2 (84%) without impairing the disease control. This approach allows to significantly reduce the treatment-related immediate toxicity in most patients and provides similar patients' outcome compared to standard BEAesc treatment. Still, PET positivity after 2 cycles of BEAesc is related to a higher risk of disease progression, encouraging to develop new treatment options in patients with PET2 positive advanced stage HL.