Venetoclax induces deep hematologic remissions in t(11;14) relapsed/refractory AL amyloidosis

Venetoclax is efficacious in relapsed/refractory t(11;14) multiple myeloma, thus warranting investigation in light-chain amyloidosis (AL). This retrospective cohort includes 43 patients with previously treated AL, from 14 centers in the US and Europe. Thirty-one patients harbored t(11;14), 11 did not, and one t(11;14) status was unknown. Patients received a venetoclax-containing regimen for at least one 21- or 28-day cycle; the median prior treatments was three. The hematologic response rate for all patients was 68%; 63% achieved VGPR/CR. t(11;14) patients had higher hematologic response (81% vs. 40%) and higher VGPR/CR rate (78% vs. 30%, odds ratio: 0.12, 95% CI 0.02–0.62) than non-t(11;14) patients. For the unsegregated cohort, median progression-free survival (PFS) was 31.0 months and median OS was not reached (NR). For t(11;14), median PFS was NR and for non-t(11;14) median PFS was 6.7 months (HR: 0.14, 95% CI 0.04–0.53). Multivariate analysis incorporating age, sex, prior lines of therapy, and disease stage suggested a risk reduction for progression or death in t(11;14) patients. Median OS was NR for either subgroup. The organ response rate was 38%; most responders harbored t(11;14). Grade 3 or higher adverse events occurred in 19% with 7% due to infections. These promising results require confirmation in a randomized clinical trial.


Introduction
Systemic monoclonal immunoglobulin light-chain amyloidosis (AL) is an acquired form of amyloidosis, in which unstable light chains are produced by a neoplastic monoclonal plasma cell population. These light chains form oligomers and bind to cellular and extracellular matrix proteins causing direct toxicity and amyloid deposition in organs such as the heart and kidneys 1,2 . Studies are limited regarding the annual incidence of AL, but estimates range from 3 to 13 cases per million person-years [2][3][4][5][6][7][8][9] .
While there has been researching into targeting different steps of amyloid formation 10 , the backbone of treatment in patients with AL remains suppression of the neoplastic plasma cell clone. Treatment paradigms are borrowed from multiple myeloma (MM) and include combination chemotherapy regimens [11][12][13][14][15][16] though to date no FDA-approved regimen for the treatment of AL exists. Unfortunately <20% of patients are transplant-eligible 17 and patients receiving high-dose melphalan with autologous stem cell transplant (ASCT) face a hematologic relapse rate of over 30% with a time to relapse of 2-4 years 18 . Additionally, recent data suggest high hospitalization rates, progression to dialysis, and early mortality from ASCT in AL patients with renal dysfunction 19 . Daratumumab, a monoclonal antibody targeting CD38 on plasma cells, has shown promising efficacy in the relapsed/refractory (RR) setting as monotherapy and in combination therapy [20][21][22][23] . Daratumumab is also currently being studied in a randomized phase 3 trial in combination with CyBorD (cyclophosphamide, bortezomib, and dexamethasone) versus CyBorD alone in the upfront setting (NCT03201965). Unfortunately, patients with AL develop significant toxicities when treated with immunomodulatory medications (such as thalidomide and lenalidomide), limiting their use 24,25 . As such, options remain bleak for patients who have progressed on chemotherapy, ASCT, and antibody therapy-based approaches.
Venetoclax (ABT-199) is an oral selective B-cell lymphoma 2 (BCL-2) inhibitor FDA-approved for the treatment of chronic lymphocytic leukemia (CLL) and acute myelogenous leukemia (AML) [26][27][28] with biological rationale for evaluation in plasma cell disorders. Preclinical data demonstrated that all human myeloma cell lines sensitive to venetoclax were restricted to the Cyclin D1 (CCND1) subgroup (80% harboring t (11;14)) and expressed elevated BCL-2:myeloid cell leukemia-1 (MCL-1) ratios 29 . High expression of MCL-1 leads to venetoclax resistance, however, this can be overcome with the addition of bortezomib, which is known to inhibit MCL-1 [30][31][32] . Dexamethasone favorably alters sensitivity to venetoclax by increasing expression of both BCL-2 and BCL-2-like protein 11 (Bim) and shifting binding of Bim to BCL-2 33 . Together, this provided a clinical rationale to study venetoclax monotherapy and in combination with bortezomib and steroids in MM. In a phase I study, venetoclax monotherapy was found to have a favorable safety profile and the greatest efficacy was noted in patients with heavily treated t(11;14) MM where 40% of patients had a hematologic response with a 27% VGPR/ CR rate 34 . The results from a phase 1b study examining the use of venetoclax and bortezomib in patients with RR MM showed a hematologic response rate of 67% in all patients and 94% in patients with high BCL-2 expression (with a VGPR/CR rate of 66%) 35 .
While this led to excitement over a potential targeted therapy in MM, preliminary results of the BELLINI trial (NCT02755597), which examined the use of bortezomib and dexamethasone + /− venetoclax in RR MM, showed an increased hazard ratio (HR) for all patients of 2.207 (95% CI: 1.042-3.945) for overall survival (OS) in the venetoclax arm 36 , despite a significantly higher progression-free survival (PFS) in patients receiving venetoclax. The alarming impact on OS led to an FDAhold on the study of venetoclax-bortezomib combination therapy in MM. Interim results presented at the American Society of Hematology 2019 annual meeting demonstrated an improved HR for PFS of 0.095 (95% CI: 0.020-0.458) and a trend toward OS improvement with a HR of 0.649 (95% CI: 0.129-3.253) in t(11;14) patients 37 , renewing interest of venetoclax in t(11;14) RR MM.
Although present in only 17% of patients with MM, up to 60% of AL patients harbor t (11;14) [38][39][40][41][42] . Given the biological rationale, many clinicians have been using venetoclax-containing regimens in patients with RR AL, with anecdotal evidence of efficacy demonstrated in case reports and small case series [43][44][45][46] . Unlike MM, t (11;14) AL patients are traditionally less sensitive to bortezomib 47 ; however, a recent case series showed a very good partial response or complete response (VGPR/CR) rate of 88% (seven out of eight patients) in t(11;14) AL patients treated with ventoclax-containing regimens, with all three patients who were treated with a venetoclax/bortezomib backbone achieving VGPR/CR 48 . While promising, larger datasets are needed to validate response and provide a strong foundation of evidence for a prospective clinical trial. Here, we report a multicenter, international, retrospective cohort study regarding treatment with venetoclax monotherapy and combination therapies for RR AL.

Study population and design
We conducted a retrospective cohort study involving 43 patients with RR AL that was approved by the Columbia University Institutional Review Board (primary Institutional Review Board). Data Use Agreements were carried out with participating institutions as applicable. Data were collected from 14 centers in the United States and Europe (see Supplementary Information).
Inclusion criteria consisted of patients with an established diagnosis of AL who had progressed on at least one prior line of therapy and were treated with a venetoclax-containing regimen for at least one 21-or 28-day cycle (Fig. 1).
Primary outcomes included OS, PFS, VGPR/CR rate, and toxicity (as assessed by Common Terminology Criteria for Adverse Events version 4.0) with pre-planned stratification by t(11;14) status. Other outcomes included hematologic response rate, organ response rate, and reason for discontinuation of therapy.

Procedures
De-identified data extracted by local investigators and collected included patient demographics and clinical characteristics such as age, gender, light-chain subtype, t(11;14) status, the presence of high-risk cytogenetic features (del(17p) or monosomy 17, t(4;14), t(14;16), and gain (1q)), prior regimens, stage according to the Revised Prognostic Staging System for Light Chain Amyloidosis 49 , venetoclax-containing regimen along with information regarding the depth of hematologic response, organ response, toxicity, PFS, and OS following initiation of a venetoclax-based regimen. Patients were considered to harbor t (11;14) if it was detected on karyotyping or fluorescent in situ hybridization irrespective of the proportion of plasma cells harboring t (11;14).
Patient demographic and clinical characteristics information were then stratified based on t(11;14) status.
Treatment response was reported as determined by local investigators using the "New Criteria for Response to Treatment in Immunoglobulin Light Chain Amyloidosis Based on Free Light Chain Measurement" 50 . A prespecified comparison of response and outcome between patients with and without t (11;14) was planned prior to data analysis. Organ responses were evaluated by criteria set forth by the Roundtable of Clinical Research in Immunoglobulin Light Chain Amyloidosis 51 .

Statistical analysis
We created a 2 by 2 contingency table for VGPR/CR hematologic response outcomes by t(11;14)-positive and -negative subgroups. The odds ratio was reported, and then a Fisher's Exact test was used to analyze the subgroups. For survival outcomes, PFS and OS curves were estimated using the Kaplan-Meier product limit estimator with medians and 95% confidence interval bounds. A data cutoff date was set for January 15, 2020, and patients were censored at their last follow-up date. PFS is defined as the time from the first day of treatment until hematologic progression or death. OS is defined as the time from the commencement of the venetoclaxcontaining regimen to death or last follow-up. A twotailed log-rank test was used to perform survival comparisons between the patient groups. Hazard ratios (HRs) and 95% confidence intervals (CIs) between the patient groups were reported for survival outcomes using Cox regression methods. A multivariate analysis using Cox regression modeling was used to assess the effect of hematologic response and t(11;14) status on OS and PFS after adjusting for the effects of age, sex, stage, and prior lines of therapy. Chi-square tests and logistic regression were used to assess if t(11;14) status is significantly different between patient demographics and characteristics (Table 1). A P value of less than 0.05 was considered to indicate statistical significance. Data analyses were performed using R version 3.6.3 (R Foundation) and SAS 9.4 (SAS Institute).

Results
Forty-three patients with RR AL diagnosed between 2009 and 2019 at 14 institutions were included in the study. Patient demographics and disease characteristics with t(11;14) stratification are shown in Table 1. Patients were predominantly male (65%) with a median age of 66 years (range 49-83). Of all patients, 72% harbored t (11;14). Three patients harbored gain(1q), two of which concomitantly harbored t (11;14), and two patients harbored del(17p) or monosomy 17, one of which also harbored t (11;14).
Patients were heavily pretreated, having received a median of three (range 1-10) prior lines of therapy. Nearly, all patients received prior treatment with a proteasome inhibitor (98%) and the majority had prior treatment with cyclophosphamide (79%) and daratumumab (58%). In addition, 21% of patients had prior ASCT.
Thirty-five patients had full demographic and disease characteristics available and were evaluable for PFS, OS, and hematologic response. When stratified by t(11;14) status and adjusted for age, sex, prior lines of therapy, and disease stage, the data are highly suggestive that patients harboring t(11;14) have a reduction of risk for progression or death (HR 0.292, 95% CI: 0.046-1.855, P = 0.192) compared to non-t(11;14) patients.
Multivariate analysis revealed that hematologic response was significant for predicting for risk of progression or death (P = 0.013) for the follow-up period, and patients with a PR or VGPR/CR showed a reduction Data for cardiac, renal, neurologic, or hepatic organ involvement was available in 36 patients, 32 of which were evaluable for organ response. Of these 32 patients, 12 (38%) patients had responses in at least one listed organ system. Of the 12 patients with organ response, 10 harbored t(11;14) (83%) and two (17%) were non-t (11;14). Of 20 evaluable patients with cardiac involvement, six patients (30%) had organ response. Of 20 evaluable patients with renal involvement, eight patients (40%) had organ response. Of four evaluable patients with neurologic involvement, none had a response. Two patients had hepatic involvement however neither met the criteria for hepatic response. Other organs involved pulmonary (two), skin (two), lip (one), salivary glands (one), and soft tissue (one). One patient had acquired AL-related Factor X deficiency, which reversed following treatment with venetoclax and achievement of VGPR.
Toxicity evaluation revealed that eight of 43 patients (19%) experienced a grade 3 or higher non-hematologic adverse event attributed to therapy ( Table 2) and eight of 43 patients (19%) discontinued treatment due to toxicity. The most common non-hematologic adverse events were infection (35% of patients with 7% grade 3 or higher), diarrhea (26% of patients with 5% grade 3 or higher), and fatigue (19% of patients none of which were grade 3 or higher). The most common hematologic adverse event was thrombocytopenia (9% of patients with 5% grade 3 or higher) ( Table 2). Though not neutropenic, one patient died from sepsis as a consequence of venetoclax therapy. In addition, one death due to heart failure in this cohort occurred within four weeks of starting venetoclax; however, this was not attributed to therapy.

Discussion
This multicenter, international, retrospective cohort study is the largest study, to our knowledge, that reports on outcomes of patients with RR AL treated with venetoclax-containing regimens. Given the retrospective nature and the heterogeneity of regimens and dosage used, the data reported have certain limitations; however, important insights can be gleaned. First, it is important to note that this is the largest cohort to report on targeted therapy for patients with AL based on cytogenetics. Stratification by t (11;14) demonstrates a large and statistically significant difference in high quality responses (VGPR/CR 78% vs. 30%) favoring t (11;14) patients with comparable hematologic responses in daratumumabrefractory patients. While OS was not reached for either subgroup due to a short follow-up period, the Kaplan-Meier survival curves trended toward a better response in t (11;14) patients. PFS analysis revealed a significant 86% reduction in risk of progression or death in t (11;14) patients (71% after multivariate adjustment).
While patient demographics and characteristics were generally comparable between t(11;14) patients and non-t (11;14) patients, it is important to highlight some potentially important differences. First, while t(11;14) patients had a younger median age than non-t(11;14) patients, a greater proportion of t(11;14) patients were actually ≤75 years old. Additionally, although both subgroups were heavily pretreated, t(11;14) patients had fewer median prior lines of treatment than non-t(11;14) patients (3 vs. 5), though this is not statistically significant. The disease stage was the same for both subgroups and thus did not confound subgroup analysis.
Our analysis showed there was no statistical difference in PFS or OS between older and younger patients, sex, or prior lines of therapy in our cohort. However, disease stage did show a statistically significant increase in the risk of progression or death with each increase in stage (HR 2.45, 95% CI: 1.06, 5.67, P = 0.036).
While there is conflicting data in the literature about the prognostic relevance of t (11;14) in AL 23,52 , our data suggest that the presence of this genetic aberration is a good biomarker for response to venetoclax. In plasma cell dyscrasias, t(11;14) is associated with a lower bone marrow plasma cell count, with the involved clone generally being the main clone in AL compared to a subclone in non-AL plasma cell dyscrasias 53 . In MM, t(11;14) patients are characterized by lower levels of monoclonal protein as well as more mature plasma cell morphology 54 . In addition, MGUS is associated with a higher incidence of t (11;14), suggesting that this translocation alone is not sufficient for plasma cells to acquire full malignant potential and secondary cytogenetic aberrations are required for myeloma progression 53,55,56 . It seems that overexpression of BCL-1 and BCL-2 associated with t(11;14) inhibits apoptosis and BCL-2 sequestration by venetoclax allows plasma cells to undergo apoptosis and more effectively induces deep and long-lasting remissions in "low malignancy" plasma cell dyscrasias like AL. Given the retrospective nature of this study, the results require confirmation by a prospective clinical trial, though given the high proportion of AL patients harboring t(11;14), venetoclax may have larger therapeutic implications than in MM.
Other important findings in this study are OS and PFS for patients with RR AL, regardless of t(11;14) status. Given the rarity of AL, and even more infrequent encounters with RR disease, data are lacking regarding PFS and OS in this patient population. Published data for patients with RR AL treated with daratumumab-based therapy show a 10-month PFS of 89% 21 and a 10-month OS between 80% and 94% 21,57 . More than half of our patients were daratumumab-refractory and had a median of three prior lines of treatment. With a 12-month PFS and OS of 78% (90% in t(11;14) disease) and 93% (97% in t (11;14) disease) in the overall cohort respectively and a 12-month PFS and OS of 80% (84% in t(11;14) disease) and 92% (95% in t (11;14) disease) in daratumumabrefractory patients, we believe our results compare favorably. While overall grade 3 or higher toxicity was noted to be 19% with a large portion being infection-related (7%), other grade 3 or higher toxicities were rare. The impact of administering prophylactic intravenous immunoglobulin and antibiotic regimens for patients in this cohort is not known, and this approach may be a potential way to abrogate infection risk, which could be examined in a randomized clinical trial. Toxicity did not appear to differ based on t (11;14) status. One patient died as a Abdominal pain 1 (2%) 1 (2%) 0 (0%) 0 (0%) 1 (9%) 1 (9%) Unspecified 2 (5%) 0 (0%) 2 (6%) 0 (0%) 0 (0%) 0 (0%)
For patients evaluable for organ response, a high proportion (38%) achieved a response including 30% of patients with cardiac involvement and 40% of patients with renal involvement. It is important to note that some of the evaluable patients had improvement in organ function biomarkers but did not quite meet the criteria for partial organ response. In addition, some of these patients were still receiving treatment with venetoclax and continuing to have hematologic response and reduction in involved-organ biomarkers, so it is unclear at this time if they will eventually meet the criteria for organ response, which is typically delayed in AL. These data suggest that venetoclax can prevent and even improve organ dysfunction in patients with AL.
This study has several limitations aside from its retrospective nature. Given the heterogeneity of venetoclaxcontaining regimens, a wide range of dosages used, and irregular follow-up, identifying the optimal venetoclax combination with drug dosing was not possible. Additionally, the lack BCL-2 expression data make it unclear whether all t(11;14) AL patients overexpressed BCL-2 and if any of the non-t(11;14) patients who had good responses concurrently overexpressed BCL-2, perhaps by another mechanism. The lack of minimal residual disease data also limits the depth of response analysis and prognostic implications. Finally, the toxicities related to venetoclax-based regimens may have been underreported given the lack of a formal process to capture the adverse effects in a cohort that was approached retrospectively.
Taken together, however, the results of our multiinstitutional study indicate a strong rationale toward a biomarker-driven approach for AL patients. Though our findings demonstrate the efficacy of venetoclax in the RR setting, they additionally serve as a platform for targeted treatment in newly diagnosed t(11;14) AL. Larger prospective studies are required to validate our findings that promote a new and convenient therapeutic approach, with a manageable toxicity profile and the potential to favorably alter the clinical course of AL.

Data sharing statement
For original data please contact Vikram Premkumar (vikram.premkumar@gmail.com). Data including all individual participant data following de-identification will be available to researchers who provide a methodologically sound proposal, to achieve the aims of the approved proposal. The study protocol is available upon request. Data will be available beginning 9 months and ending 36 months following publication of the article.
Author details