Systemic ALK-Positive Anaplastic Large-Cell Lymphoma (ALCL): Final Analysis of an International, Individual Patient Data Study of 263 Adults

Introduction: ALK-positive ALCL comprises 0.5% of adult lymphomas. Because of disease rarity, prospective trials specifically dedicated to adult ALK-positive ALCL are lacking. In published adult studies, mainly retrospective, ALK-positive ALCLs were most often pooled with ALK-negative ALCLs and/or other PTCLs, which are distinct entities, preventing a correct assessment of specific prognostic factors and impact of different treatments for ALK-positive ALCL. A retrospective study from the DSHNHL and a population-based study from the Nordic lymphoma registries have suggested a possible benefit of etoposide use, but definitive conclusions are lacking and the role of etoposide is still a matter of debate. Moreover, the respective impact of upfront HDT-ASCT or consolidative radiotherapy (RT) is unknown.

A pooled analysis of individual patient data (IPD) would provide greater statistical power to identify prognostic factors and treatment effects. Therefore, we conducted a pooled analysis of IPD from studies of first-line therapy in adult ALK-positive ALCL.

Methods: Eligibility criteria were as follows: study including patients (pts) with a diagnosis of systemic ALK-positive ALCL after initial centralized pathologic review, age ≥ 18 years, HIV-negative serology, first-line treatment including at least one cycle of systemic chemotherapy, sufficient sample size (at least 10 cases of ALK-positive ALCL) and publication between 2000 and 2015. A minimum dataset was required, comprising characteristics at diagnosis (individual IPI and PIT factors), and first-line chemotherapy regimen.

We assessed PFS, OS, EFS24, cumulative incidence (CI) of relapse, and the associations between patient characteristics or treatment type (etoposide, upfront HDT-ASCT and consolidative RT) and PFS or OS. These associations were analyzed by Cox survival models stratified by cohort and assessed by Log-Likelihood Ratio (LLR) tests with predictive C-index measurement.

Results: IPD of 263 pts diagnosed between 1985 and 2008 were collected and pooled from six published studies: 2 from the LYSA, 1 from the DSHNHL, 1 from Japan, 1 from the IPTCLP and 1 from the Mayo Clinic. Median follow-up was 4.9 years. Median age was 34 years (18-76). Overall, 261 (99%) pts received an anthracycline-based regimen and 92 (35%) received an etoposide-based induction. Thirty-four (13%) pts underwent planned upfront HDT-ASCT, and 54 (21%) received consolidative RT.

The respective CR/CRu, PR and SD/PD rates to first-line therapy were 85%, 5% and 10%. The 5-year PFS and OS rates were 69% (95% CI, 64% to 75%) and 81% (95% CI, 76% to 86%), respectively. The respective 1-, 2- and 5-year CI of relapse were 22%, 25% and 30%. For the 82 pts with a first relapse, the respective 1-, 2- and 5-year OS rates from this event were 38%, 35% and 28%, with a median OS of only 3.2 months. Pts achieving EFS24 had a 5-year OS from this time point of 95% (Fig 1).

The IPI identified different risk groups; 5-year PFS was 80% for IPI 0 to 1, 63% for IPI 2, 54% for IPI 3, and 40% for IPI 4 to 5 (p<0.001); 5-year OS was 92% for IPI 0 to 1, 73% for IPI 2, 59% for IPI 3, and 62% for IPI 4 to 5 (p<0.001, Fig 2). IPI score had the best predictive value for PFS and OS (C-index 0.64 and 0.72, respectively), without significant heterogeneity between cohorts (LLR p=0.12 for PFS and p=0.39 for OS). In a stratified model adjusting for the IPI, CD3 positivity (HR 2.3 for PFS and 3.1 for OS) and hemoglobin ≤ 12 g/dl (HR 2.5 for PFS and 3.2 for OS) remained independent predictors of PFS and OS.

Pts who received an etoposide-based induction chemotherapy had significantly improved 5-year PFS (83% vs 62%, p=0.001) and OS (93% vs 74%, p=0.001, Fig 3), independent of IPI, age and cohort. Etoposide did not increase significantly the response rate at the end of treatment, but significantly decreased the 5-y CI of relapse (17% vs 37%, p=0.001). Finally, CHOEP (n=38), compared to CHOP (n=98), significantly improved 5-year PFS (89% vs 57%, p=0.017) and OS (97% vs 69%, p=0.049, Fig 4).

In stratified Cox models, there was no independent impact of upfront HDT-ASCT or consolidative RT on PFS or OS.

Conclusion: To our knowledge, this is the largest analysis of patients with ALK-positive ALCL. IPI, CD3 positivity and anemia are strong independent predictors of outcome. Etoposide use in induction is associated with important improvement of PFS and OS independent of IPI and age, whereas upfront HDT-ASCT and consolidative RT have no independent impact.

View largeDownload slide

View largeDownload slide

Close modal