#amyloidosisJC 11/4/15 @ 9 pm EST: examining the prognostic and predictive relevance of cytogenetics in AL amyloidosis

CLICK HERE FOR A LINK TO THE FULL TEXT OF THE ARTICLE

Background

Light chain (AL) amyloidosis is a rare plasma cell dyscrasia which causes morbidity and mortality through deposition of toxic amyloid fibrils. Although the prognostic significance of specific cytogenetic aberrations in the plasma cell malignancy multiple myeloma is established, the relevance of cytogenetics in AL amyloidosis is still being defined. Several recent retrospective studies including this study by Bochtler et al., have attempted to characterize the significance of chromosome aberrations in AL amyloidosis. This paper seeks to establish translocation t(11;14), as a predictive marker of poor response to bortezomib. This is important because bortezomib-based regimens (such as Vel/Dex) are often used as front-line treatment for AL amyloidosis. In this disease where patients often have progressive multi-organ damage, selection of an efficacious regimen to rapidly establish disease control is crucial. This paper raises the question of whether cytogenetic studies could guide treatment of AL amyloidosis.

Methods

·      Retrospective, single-center study of 101 consecutive transplant ineligible AL amyloidosis patients who received Vel/Dex as first-line therapy. These patients had high-risk clinical features such as severe cardiac or kidney involvement. Patients with concurrent stage II/III MM or IgM paraprotein were excluded.

·      Outcomes assessed included hematologic response after 3 cycles (by consensus criteria for PR, VGPR), hemEFS (hematologic relapse, hematologic progression, start of a second-line therapy, death) and OS. Early deaths counted as remission failures on intent-to-treat basis.

·       High-risk aberrations were defined as t(4;14), t(14;16) and deletion 17p13.

·      After backward variable selection, validation and calibration of the final multivariable model was done using a CyBorD-treated cohort (32 patients).

Results

Patient Characteristics 

Most patients in Vel/Dex and CyBorD cohorts had cardiac involvement (91% vs 88%) and around half had renal involvement in both cohorts

HemEFS

·      Median hemEFS 4.7 mo with median F/U of 24.0 months and hematologic events in 84/101 patients. The t(11;14)-positive group had inferior hemEFS (median hemEFS 3.4 months vs 8.8 months in t(11;14)-negative group; p=.002). Median hemEFS in high-risk group was 10.3 months vs 3.9 months in patients without high-risk aberrations (P=.15).

Overall Survival (OS)

·      Median OS was 15.7 mo with median F/U time of 24.1 months and 53 observed deaths. t(11;14) predicted for shorter OS with median OS 8.7 months vs 40.7 months in t(11;14)-negative group (P=.05). High risk aberrations conferred favorable prognosis (median OS = NR for high-risk v 10.6 months for absence of high-risk aberrations; P=.04).

Remission Rates

·      95/101 (94%) pt had initial dFLC >50mg/L and thus evaluable for ≥VGPR

·      30 (32%) achieved ≥VGPR after 3 cycles, 24 (25%) attained PR and 21 (22%) did not achieve remission, 20 (21%) had early death before remission assessment. ≥VGPR rates worse in t(11;14)-positive patients compared with t(11;14)-negative patients [14/61 (23%) versus 16/34 (47%) patients, P=.02].  High-risk aberrations had favorable remission rates compared to those without [8/12 (67%) versus 21/80 (26%), P=.008]

Multivariable testing in Vel/Dex Cohort with Clinical and Cytogenetic Factors

·      By Cox regression, t(11;14) and dFLC were the only two statistically significant prognostic markers for both OS and hemEFS. NT-pro-BNP reached statistical signification for OS. Thus t(11;14) is an independent negative risk factor in the Vel/Dex treated cohort. 

    A final model developed: hemEFS, t(11;14), sex, dFLC, NT-proBNP and dose reduction. For OS, t(11;14), dFLC and NT-proBNP. For outcome remission ≥VGPR, t(11;14) and age.

External validation using CyBorD cohort

·      ≥VGPR rate after 3 cycles, 24% (4/17) in t(11;14)-positive compared with 80% (8/10) in t(11;14)-negative patients. Median hemEFS 5.7 months in all 32 patients, 4.0 months in t(11;14)-positive and NR in t(11;14)-negative patients (P=.01). Median OS for both groups NR. Median F/U was 7.5 months. Calibration analysis of final models performed which showed that all trained predictors had a trend for better prediction of the outcome in the CyBorD validation cohort.

Authors’ Conclusions

• Cytogenetic aberrations are important independent prognostic factors in AL.
• Bortezomib is less beneficial to patients with t(11;14) 
• Bortezomib overcomes the poor prognosis of patients with high-risk aberrations.

Our Comments

• Study was of small sample size and would need confirmation with other studies, prospectively and multi-center ideally

• A potential criticism is the validation using a CyBord-treated cohort rather than additional Vel/Dex-treated patients. CyBorD cohort followup duration is a lot shorter than the Vel/Dex cohort. Despite this, results seen in the CyBord patients seem consistent with those described for the Vel-Dex cohort

• Paper suggests that cytogenetics may have predictive as well as prognostic significance.

Special thanks to Sandy Wong (@SandyWong02111) from Tufts University for her efforts preparing this summary. Looking forward to co-moderating this journal club discussion with her.