MZL

MALT-IPI – a simple and effective tool generated from phase III IELSG-19 trial data

On 18th July 2017, Catherine Thieblemont from Saint Louis Hospital, Diderot University-Sorbonne Paris Cité, Paris, France, and colleagues published their study, which aimed to develop and validate a specific prognostic tool for Mucosa-Associated Lymphoid Tissue (MALT) Lymphoma, in the journal Blood.

The group used final analysis data from the international, randomized, phase III IELSG-19 trial (NCT00210353), which investigated chlorambucil versus rituximab versus chlorambucil plus rituximab in patients with MALT Lymphoma; the Lymphoma Hub reported on this trial earlier this year (read more here).

To develop a MALT Prognostic Index (MALT-IPI), 401 patients from IELSG-19 were analyzed. The median follow-up time was 7.4 years and 22.9% of patients (n=92) were aged over 70 years. Primary gastric MALT Lymphoma was diagnosed in 42.6% of patients (n=171), disseminated disease was observed at diagnosis in 43.6% (n=175), and 123 patients (30.7%) had more than one extranodal site. Lymph node involvement was observed in 35.4% of patients (n=142). Elevated Lactate Dehydrogenase (LDH) and poor Performance Status (PS) were uncommon (10.5% and 1.5%, respectively).

Key Highlights:
MALT-IPI elaboration:
  • A stepwise Cox regression including 400 pts with no missing data identified three independent predictors of shorter Event-Free Survival (EFS):
    • Ann Arbor stage III/IV: HR = 1.79
    • Age ≥ 70 years: HR = 1.72
    • Elevated LDH levels: HR = 1.87
  • MALT-IPI was built with three risk groups defined by the presence of 0 (low risk, n=167), 1 (intermediate risk, n=165), or ≥2 (high risk, n=68) of the three variables, respectively
  • The index distinguished pts with significantly different EFS and PFS, and also identified pts with significantly shorter OS and Cause-Specific Survival (CSS)
  • 5-year survival estimates for low, intermediate, and high risk pts:
    • EFS: 69.8% vs 7% vs 28.7% (P < 0.0001)
    • PFS: 76.0% vs 1% vs 32.5% (P < 0.0001)
    • CSS: 100% vs 1% vs 84.5% (P < 0.0001)
    • OS: 98.7% vs 1% vs 64.3% (P < 0.0001)
MALT-IPI evaluation:
  • In pts with gastric lymphoma, the number of low risk pts was significantly higher (52% vs 34%, P = 0.001) and the number of high risk pts was lower (12% vs 21%)
  • The model (n=400) retained its efficiency in distinguishing pts with poorer prognosis in both the gastric (n=171) and the non-gastric patient (n=229) subgroups
  • The model also maintained statistical significance in each treatment arm (chlorambucil, n=130; rituximab, n=138; R-chlorambucil, n=132) for all four survival endpoints
  • Histologic transformation occurred in 10 pts: 7 were high risk and 3 were intermediate risk according to the new MALT-IPI
External validation:
  • The validation set was created from three independent MALT Lymphoma cohorts (n=633): the IELSG-01 cohort, the Bellinzona-Novara cohort, and the MUV cohort
  • MALT-IPI retained its ability to significantly discriminate different prognostic groups with respect to each survival endpoint:
  • 5-year survival estimates for low, intermediate, and high risk pts:
    • EFS: 76.0% vs 4% vs 15.7% (P < 0.0008)
    • PFS: 56.8% vs 0% vs 22.7% (P < 0.0001)
    • CSS: 98.2% vs 7% vs 74.3% (P < 0.0001)
    • OS: 96.7% vs 7% vs 64.9% (P < 0.0001)
MALT-IPI compared to existing prognostic indices:
  • In the validation set (n=633), a definite risk group allocation was achieved in 87% of pts (n=551) according to the MALT-IPI, in 90% of pts (n=563) according to IPI, and in 90% of pts (n=564) according to R-IPI, respectively
  • The performance of all three models was very similar: MALT-IPI was slightly superior (highest k value) in the survival endpoint analysis; however, the standard errors indicate overlapping CI

The authors stated that MALT-IPI, which encompasses age 70 years or older, elevated LDH levels, and Ann Arbor stage III/IV, was validated in a large, external cohort of patients. They concluded that MALT-IPI is a “simple, accessible and effective tool” to aid the identification of high risk patients with MALT Lymphoma who are likely to experience disease relapse or death.

Abstract:

There are no widely accepted prognostic indices for extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT lymphoma). This study aimed to develop and validate a specific prognostic tool to personalize and optimize lymphoma treatment of patients with MALT lymphoma. A prognostic index was built by Cox regression (stepwise selection) using data from 401 patients enrolled in the international randomized IELSG-19 trial (NCT 00210353). A validation set, including 633 patients, was obtained by merging three independent cohorts of MALT lymphoma patients. The three individual features maintaining the greatest prognostic significance for event-free survival (EFS, the main endpoint of the IELSG-19 trial) were age ≥70 years (HR 1.72, 95% CI 1.26-2.33), Ann Arbor stage III or IV (HR 1.79, 95% CI 1.35-2.38), and elevated LDH level (HR 1.87, 95% CI 1.27-2.77). The prognostic index (MALT-IPI) constructed using these three parameters identified three groups: low, intermediate and high risk (corresponding to the presence of 0, 1 or ≥2 of these factors, respectively). The 5-year EFS rates in the low, intermediate and high risk groups were 70%, 56% and 29%, respectively. The MALT-IPI also significantly discriminated between patients with different progression-free, overall and cause-specific survival. The prognostic utility was retained in gastric and non-gastric lymphomas, in each treatment arm (chlorambucil, rituximab, rituximab plus chlorambucil), and was confirmed in the validation set. The new index, MALT-IPI, is a simple, accessible and effective tool to identify MALT lymphoma patients at risk of poor outcomes. It may help define appropriate treatment approaches for individual patients.

References
  1. Thieblemont C. et al. A MALT lymphoma prognostic index generated from the dataset of the IELSG-19 prospective clinical trial. Blood. 2017 Jul 18. pii: blood-2017-03-771915. DOI: 10.1182/blood-2017-03-771915. [Epub ahead of print]