Introduction

Outcomes for patients with multiple myeloma (MM) have improved significantly over the past decade with several new approved modalities [1]. However, MM remains an incurable disease with a continued need for novel and curative therapies [2]. Overexpression of BCL-2 can be found in up to 20% of patients with MM, making these patients potential candidates for treatment with venetoclax, a BCL-2 inhibitor. Myeloma cells, particularly in the presence of t(11;14), demonstrate relatively higher expression of BCL-2 protein which has been associated with a survival advantage [3,4,5,6].

BCL-2 belongs to a family of anti-apoptotic proteins, alongside others like MCL-1 and BCL-XL. For apoptosis to occur, the cellular balance must shift in favor of pro-apoptotic proteins (e.g., BAX and BAK), which are typically inhibited by the anti-apoptotic proteins [7]. Therefore, therapies that inhibit anti-apoptotic proteins such as venetoclax targeting BCL-2 promote cell death. This is particularly true in myeloma cells that rely on BCL-2 for survival, either due to overexpression of BCL-2 or an acquired reduction in MCL-1 or other anti-apoptotic proteins [8, 9]. MM patients with t(11;14) are especially susceptible to venetoclax therapy with response rates ranging from 40-50%, likely because the t(11;14) subtype induces BCL-2 dependency which enhances venetoclax effectiveness [3, 5]. Similarly, studies have shown venetoclax efficacy in patients with high BCL-2 gene expression, but clinical trial data demonstrate much lower efficacy in the non-t(11;14) cohort with high BCL-2 expression [10].

A phase 2 study of venetoclax, carfilzomib and dexamethasone VenKd, in patients with RRMM, with a median of 1 prior line of therapy, demonstrated an impressive response rates of 80%, and a median PFS of 24.8 months in the t(11;14) subgroup and 22.8 months in the non-t(11;14) subgroup [11]. Similarly, a phase I study assessed the combination of venetoclax, daratumumab, and dexamethasone (VenDd) in patients with RRMM with t(11;14) and VenDd with bortezomib (VenDVd) in cytogenetically unselected patients. Both VenDd and VenDVd demonstrated high efficacy, with ≥very good partial response (VGPR) rates of 96% and 79%, respectively, and an 18 month PFS of 90.5% and 66.7% respectively [12]. Despite these impressive response rates, concerns remain about the utility of venetoclax in the management of relapsed/refractory MM (RRMM). The CANOVA study, an open-label phase 3 randomized controlled trial comparing venetoclax-dexamethasone and pomalidomide-dexamethasone in RRMM receiving ≥2 prior lines of therapy, failed to meet its primary endpoint of progression-free survival (PFS) in patients with t(11;14), although there were concerns about informative censoring confounding the results of this study [13, 14]. The BELLINI trial demonstrated an improvement in PFS with venetoclax-based combination, with a pronounced PFS benefit in the t(11;14) subgroup. However, inferior Overall Survival(OS) noted in the venetoclax arm, possibly due to a higher rate of infections, raised caution regarding its use [10]. These studies highlight the impressive efficacy of venetoclax-based combinations, but were largely studied in less heavily pretreated cohorts and did not isolate the impact of secondary cytogenetic abnormalities (SCAs) on outcomes with venetoclax combinations.

Additionally, there is currently no data on the impact of secondary cytogenetic abnormalities on the efficacy of venetoclax-based regimens. Given this, we aimed to evaluate the efficacy and outcomes of venetoclax-based combination therapies in RRMM.

Methods

Ethics approval and consent to participate

The study was conducted in accordance with Declaration of Helsinki and all methods were performed in accordance with relevant guidelines and regulations. The mayo Clinic Institutional Review Board approval was obtained for conducting the study (24-003869). individual informed consent was obtained per institutional requirements and guidelines.

Study cohort and statistical considerations

We included patients treated with Venetoclax (Ven)-based combinations at Mayo Clinic sites between January 2015 and December 2023 in this retrospective study. Patient data was extracted from electronic medical records. Triple-class refractoriness was defined as resistance to at least one proteasome inhibitor (PI), one immunomodulatory drug (IMID), and one anti-CD38 monoclonal antibody. Penta-drug refractory disease was defined as resistance to at least two IMIDs, two PIs, and one anti-CD38 monoclonal antibody. Data for fluorescence in-situ hybridization (FISH)-detected cytogenetic abnormalities at diagnosis as well as within 1 year prior to initiation of venetoclax-based therapy was included. FISH was performed by the Mayo Clinic Genomics laboratory as described by Gagnon et al. [15] In patients without t(11;14), BCL-2 expression was tested in subset of patients by immunohistochemistry (IHC) (supplementary material).

Patients treated with daratumumab (Dara) based combinations with venetoclax were grouped as Dara-Ven, including patients that received Dara-Ven in combination with PI/IMiDs. Patients treated with a PI-based combination without Dara were included in the PI-Ven cohort and venetoclax monotherapy or with dexamethasone were included in the Ven-Dex cohort. A small proportion of patients not fitting these cohorts were classified in the miscellaneous (Ven-misc) cohort.

Response was determined according to International Myeloma Working Group (IMWG) criteria [16, 17]. Progression-free survival was calculated from initiation of Ven-based therapies to progression or death. For event-free survival (EFS), apart from progression or death, stopping treatment due to adverse effects was considered as an event. The OS for venetoclax-based therapies were calculated from initiation of venetoclax till last follow-up or death. Overall response rate (ORR) was defined as the percentage of patients who achieved partial response (PR)/better. Duration of response (DOR) was defined as the time from the first observation of a partial (PR) or better until the earlier of disease progression or death.

Categorical variables were compared using the Chi-square test or Fisher’s exact test, as appropriate. Time-to-event analyses were conducted with Kaplan-Meier estimates and compared using the log-rank test. To evaluate the effects of clinical and laboratory variables on progression-free survival (PFS), a Cox regression model was employed. Associations between clinical and laboratory variables and response to venetoclax were examined using logistic regression analyses. All statistical analyses were performed using BlueSky Statistical Software (version 10.3.4) and the R package (version 8.95), with p-values < 0.05 considered statistically significant.

Results

A total of 232 patients treated with Ven-based combinations were included in this study. The median age at the diagnosis of MM was 62 years (range: 32–87 years), and 65%(n = 152) of patients were male. The median time from MM diagnosis to the initiation of Ven-based therapy was 55 months (95% CI: 28–89 months), and the median follow-up from the initiation of Ven-based treatment was 17.2 months (95% CI: 14–22 months). According to R2-ISS staging at diagnosis, 24.4%(n = 40) of patients were stage I, 26.2% (n = 43) were stage II, 39.6% (n = 65) were stage III, and 9.8%(n = 16) were stage IV. Elevated lactate dehydrogenase (LDH) levels (>upper limit of normal) were observed in 21.4%(n = 36) of patients. Baseline characteristics and FISH findings are depicted in Table 1.

Table 1 Baseline characteristics of the study cohort.
Full size table

The t(11;14) chromosomal abnormality was present in 82.3% (n = 190). Among the 17.7% (n = 41) of patients without t(11;14), BCL-2 expression by IHC was tested in 18 patients and was positive in 94.4%(n = 17) of these patients. Of the 190 patients with t(11;14), 44.7%(n = 85) harbored either a 1q gain/amplification or del(17p) abnormality [deletion 17p in 19.5%(n = 37), 1q gain 31.1%(n = 59) and 1q amplification in 2.6%(n = 5)]. There is no difference in PFS and OS between patient with 1q gain and 1q amplification following venetoclax treatment, Table S1.

Patients had received a median of three prior lines of therapy (IQR: 4–6) before starting the venetoclax regimen. Of the total participants, 65%(n = 152) were triple-class refractory (TCR), and 27.2%(n = 63) were penta-drug refractory. The rates of refractoriness to individual drug classes were as follows: bortezomib – 62%(n = 144), carfilzomib – 50%(n = 116), lenalidomide – 75%(n = 174), pomalidomide – 61%(n = 141), and daratumumab – 82%(n = 190). Additionally, 71%(n = 163) had undergone autologous stem cell transplantation, and 7%(n = 16) had received prior immune effector therapy before initiating Ven-based therapy. The Ven-based combinations included Dara-Ven (19.0%, n = 44), PI-Ven (30.2%, n = 70), Ven-Dex (48.3%, n = 112), and Ven-misc (2.6%, n = 6). A comparison of prior treatments and cytogenetic features in these cohorts of patients is described in Table S2.

The ORR for the entire cohort was 57%. In patients with t(11;14), the ORR was 64% (n = 122), compared to 27% (n = 11) in those without t(11;14) (p < 0.001). Within the t(11;14) cohort, patients with 1q gain/amplification or del(17p) had an ORR of 57% (n = 49), whereas those without these abnormalities had an ORR of 71% (n = 73) (p = 0.04). In the non-t(11;14) cohort, patients with 1q gain/amplification or del(17p) had an ORR of 15% (n = 4), whereas those without these abnormalities had an ORR of 43% (n = 6; p = 0.06). Among non-t(11;14) patients with BCL-2 positivity, the ORR was 24% (n = 4). Within the t(11;14) cohort, ORR was 75% (n = 33) for Dara-Ven, 66% (n = 35) for PI-Ven, and 57% (n = 51) for Ven-Dex (p = 0.2). Details of comparison of ORR among treatment regimen and cytogenetic cohort is shown in Tables S3 and S4.

The presence of TCR and secondary cytogenetic abnormalities, including 1q gain/amplification or del(17p), was associated with significantly lower odds of achieving a partial response (PR) or better (Table 2). Within the t(11;14) cohort, ORR was 57.4% (n = 70) in patients with TCR, compared to 76.5% (n = 52) in those without TCR.

Table 2 Predictors of achieving a higher Overall Response Rate with venetoclax-based regimens.
Full size table

Impact of Cytogenetic Abnormalities on Outcomes

The median PFS for the entire cohort was 9.4 months (95% CI: 8.4–12). In patients with t(11;14), median PFS was 11.8 months (95% CI: 9.5–15.3) compared to 2.9 months (95% CI: 1.9–8.4) in those without t(11;14) (p < 0.001, Fig. 1). This difference remained significant in multivariate analyses adjusting for TCR status, combination regimen, and the presence of 1q gain/amplification or del(17p) (Table 3).

Fig. 1
figure 1

Progression Free Survival in patients with t(11;14) vs non-t(11;14): Patients with t(11;14) demonstrated a significantly better median PFS of 11.8 months for t(11;14) patients and 2.9 months for those without t(11;14), p < 0.0001.

Full size image
Table 3 Cox Regression Analysis for Progression-Free Survival (95% Confidence Interval) with venetoclax-based therapies.
Full size table

The median DOR for t(11;14) patients was 12.6 months (95% CI: 10.7–21.2). Within the t(11;14) cohort, median PFS was longest with Dara-Ven at 18.9 months (95% CI: 15.1–43.1), followed by PI-Ven at 12.9 months (95% CI: 8.6–21.7), and Ven-Dex at 8.5 months (95% CI: 6.8–11.9) (p = 0.042). PFS in non-t(11;14) patients is shown in Fig. S2, with detailed comparisons across cytogenetic abnormalities and treatment regimens in Tables S3 and S4.

The proportion of patients with TCR disease and secondary cytogenetic abnormalities (1q gain/amplification, del(17p)) was comparable across treatment groups. However, the Dara-Ven cohort had a lower proportion of penta-refractory patients, a higher proportion of t(11;14), and a lower R2-ISS stage (Table S2).

Among the non-t(11;14) cohort with BCL-2 positivity on IHC, the median PFS was 1.8 months (95% CI: 1.2-3.4), Fig. S3. There was no difference in PFS and OS between patients reported as ‘positive’ or strongly positive for BCL-2 (Table S5). In the cohort of patients with t(11;14), the presence of secondary cytogenetic abnormalities [either del(17p) or 1q gain/amplification (n = 85/190)] was associated with a significantly inferior PFS [median PFS 7.7 months (95% CI: 5-13)] vs. 15.1 months [(95% CI: 10.8–22.2), p = 0.013] in patients without del(17p) and/or 1q gain/amplification (Fig. 2, Figs. S4 and S5). We identified 1q gain/amplification/ del(17p) (HR 1.8; 95%CI 1.3-2.4), advanced R2-ISS stage (HR-1.7 95%CI 1.1-2.5) and TCR status (HR 1.5, 95%CI 1.03–2.4) to be independent predictors of inferior PFS, Table 3 The median PFS was 6 months (95%CI 4.8–9.2) in patients who were TCR vs in patient without TCR disease 17.3 months (95%CI 11.9-28.4), p < 0.0001. To further delineate the impact of secondary cytogenetic abnormality (SCA) [del(17p) and/or 1q gain/amplification], we evaluated the PFS among patients with TCR disease within the t(11;14) cohort. Patients with TCR + /SCA- had a median PFS of 10.8 months (95% CI: 9.1–20.6) and TCR + /SCA+ had median PFS of 5 months (95% CI: 4.1–10). (p = 0.038), Fig. S6. In patients who had extended PFS (≥24 months), 50% (n = 22) had TCR before venetoclax compared with 72%(n = 108) in patients with PFS < 24 months, p = 0.006. Similarly, 33%(n = 14) of patients with extended PFS had del(17p) or 1q gain/amplification compared with 53% in patient with PFS < 24 months, p = 0.02, other characteristics of patients with extended PFS is shown in Table S6. The median event free survival for the entire cohort was 9.3 months (95% CI 8.2-11.8).

Fig. 2
figure 2

Impact of Secondary cytogenetic abnormalities in t(11;14): In patients with t(11;14), presence of 1q gain/amplification/ del(17p) is associated with inferior progression free survival (PFS), median PFS 7.7 months vs. 15.1 months, p = 0.013.

Full size image

Discontinuation, adverse effects and venetoclax dosing

At the last follow-up, forty-two patients were receiving ongoing treatment with venetoclax-based combinations. The most common reasons for discontinuation were disease progression (77.0%, n = 140), adverse effects (12.0%, n = 22), and patient/provider preference (9.0%, n = 17). Additionally, 2.0%(n = 4) of patients died without disease progression (3 due to infection and 1 due to an unclear cause).

The most frequently utilized maximum dose of venetoclax was 400 mg (42.0%). Among the 232 patients included in this study, dose reductions were required in 15.0%(n = 35), and 23.0%(n = 54) underwent dose ramp-up. Among those who required dose reduction, 5.0%(n = 11) was due to cytopenia, 4.0%(n = 9) due to infection, and 3.0%(n = 7) due to diarrhea. No cases of tumor lysis syndrome were observed.

Infectious complications were reported in 23.0% (n = 52) of patients and was the reason for discontinuation of venetoclax in 2% (n = 4) of patients. The incidence varied across treatment groups, occurring in 16.3% (n = 7) of patients receiving Dara-Ven, 27% (n = 19) of those on PI-Ven, and 21% (n = 23) of those on Ven-Dex (p = 0.3). Within the PI-Ven subgroup, infection rates were 20.0% (n = 2) for Ven-Ixa, 31.8% (n = 7) for Ven-K, and 24.3% (n = 9) for Ven-V (p = 0.4 for PI-Ven overall; p = 0.5 for Ven-K vs. Ven-V). Of the fifty-two patients who developed infections while on venetoclax, 35 experienced respiratory tract infections, including 28 cases of pneumonia. Among these, 7 were attributed to COVID-19, 4 to other viral pneumonias, 13 to bacterial pneumonias (including 3 cases complicated by sepsis) and 4 were of unclear etiology. Seven patients developed urinary tract infections. Other infections observed included cytomegalovirus (CMV) viremia (n = 4), pulmonary aspergillosis (n = 2), cellulitis (n = 1), listeria meningitis (n = 1), dental infection with neck abscess (n = 1) and Clostridioides difficile diarrhea (n = 1).

Discussion

In this study, we report efficacy and outcomes with venetoclax-based regimens in a cohort of heavily pretreated patients, majority of whom were triple-class refractory. Despite improvements in therapeutic strategies for MM, heavily pretreated disease continues to present a challenge particularly in frail patients not eligible for immune effector therapies. Even in candidates eligible for CAR-T therapy, effective bridging strategies can be limited given the high likelihood of drug refractoriness to various classes of antimyeloma therapy. Here, we demonstrate excellent ORR of 57% with venetoclax based therapies in triple-class refractory patients harboring a t(11;14). This is better than historical ORRs of around 30–45% without various combination therapies in triple-class refractory patients, highlighting utility of venetoclax- based combinations in heavily pretreated patients [18, 19]. The ORR of 64% and median PFS of 11.8 months in patients with t(11;14) are consistent with previously reported studies [20, 21]. This response rate and durability is comparable to that seen with bispecific antibodies in a similar population, highlighting the utility of this approach in clinical practice [22].

Our findings demonstrate that patients with t(11;14) have higher response rates and a longer PFS compared to those without t(11;14). Notably, BCL-2 expression by immunohistochemistry in patients without t(11;14) was a poor marker of efficacy. In the BELLINI trial, quantitative PCR of bone marrow aspirate samples was used to assess BCL-2 expression, categorizing it into high and low levels. The study reported better responses in patients with high BCL-2 expression who did not have t(11;14) than what is noted in our cohort [10]. Results from biomarker analyses of the Phase 3 BELLINI study confirmed a strong correlation between BCL-2 gene expression and IHC staining [23]. Notably, the median PFS and OS were comparable in patients with BCL2high versus BCL2low expression by qPCR in a post-hoc analysis of the CANOVA trial [13]. Given that venetoclax targets the BCL-2 protein, one would expect better outcomes in patients with positive BCL-2 staining on IHC. One possible explanation for the poor outcomes observed in our study is the co-expression of other anti-apoptotic proteins, such as MCL-1, which may lower the BCL-2/MCL-1 ratio. This contrasts with t(11;14) myeloma cells, which typically have a higher BCL-2/MCL-1 ratio and, consequently, better venetoclax sensitivity. Another possibility is the relatively low discriminatory ability of the IHC, which is often subjective and semiquantitative compared with PCR, which can be more precise, and likely explains the difference seen in studies using PCR to measure and classify patients based on BCL-2 expression and the current study. Further validation of venetoclax efficacy in BCL-2 expression by IHC in the absence of a t(11;14) is needed, and based on our data, BCL-2 expression by IHC alone does not warrant use of venetoclax.

In patients with t(11;14) and high-risk cytogenetic abnormalities (e.g., del(17p) and 1q gain/amplification), PFS with venetoclax-based regimens was inferior compared to t(11;14) patients without these abnormalities. Previous studies have reported variable impact of the presence of high-risk cytogenetics on response to venetoclax [10, 20]. In our subset analysis of t(11;14) patients who typically respond well to venetoclax, the response rates and PFS were significantly inferior in the presence of del(17p) or 1q gain/amplification. Presence of 1q gain/amplification has been shown to increase the expression of MCL-1, an anti-apoptotic protein whose overexpression can reduce the sensitivity of myeloma cells to venetoclax [9]. We noted that only a minority of patients discontinued treatment due to adverse effects. This is likely a reflection of the fact that a large proportion of patients had the venetoclax dose capped at 400 mg. Unlike in clinical trials, where doses of 800-1200 mg were utilized, our practice patterns demonstrate that 400 mg may be a sufficient dose for venetoclax.

The treatment combinations were understandably heterogeneous given the varied prior treatment exposures. A comparison of different combination therapies is likely confounded by prior treatment exposures, although notably, the proportion of patients that were TCR in these subgroups were comparable. We observed that Dara-containing regimen had better outcomes compared to proteasome inhibitor plus venetoclax. This is similar to the findings reported by Bahlis et al., who observed deep responses with venetoclax combined with daratumumab, with or without bortezomib [12]. Daratumumab exerts a unique mechanism of action involving antibody-dependent cytotoxicity and T-cell immunomodulation through T-cell activation and expansion [24]. Our findings also align with previous reports on the efficacy of venetoclax when combined with bortezomib compared to venetoclax monotherapy. The synergistic effect of venetoclax and bortezomib is likely due to bortezomib induced increase in NOXA, which decreases activity of anti-apoptotic protein MCL-1, thereby increasing myeloma cells dependency on BCL-2 for survival [9, 25, 26]. Additionally, We observed better outcomes in non-triple-class refractory (non-TCR) patients compared to those with TCR, suggesting a potential benefit of venetoclax in earlier lines of therapy. While the combination of venetoclax with carfilzomib rather than bortezomib demonstrated a trend toward better outcomes, the numbers at risk are small numbers and there is heterogeneity with regard to prior treatment. In the study cohort, 2%(n = 4) of patients discontinued treatment due to infection and there were 3 infection-related deaths while on treatment with a venetoclax-based regimen. Infection prophylaxis strategies followed standard guidelines and were individualized for patients based on the treating physician’s judgement. No unique or uniform infection prophylaxis approach was adopted for patients being treated with a venetoclax-based regimen.

Our study is limited by the challenges of a retrospective analysis, including investigator reported responses and inherent variability in follow-up schedules. Treatment combinations utilized were at the discretion of the treating physician and contribute to the heterogeneity of the cohort. Nonetheless, venetoclax based treatments likely continue to hold value in patients that are heavily pretreated and harbor a t(11;14).

In conclusion, venetoclax-based regimens continue to represent an important treatment option for patients harboring t(11;14) in relapsed/refractory MM, offering substantial benefit in PFS and OS, although secondary cytogenetic abnormalities such as del(17p) and 1q gain/amplification may reduce the efficacy. Our findings suggest that BCL-2 expression by IHC alone, without t(11;14), may not be a sufficient biomarker for using venetoclax and alternative treatment options may be explored in this setting. Further studies focusing on role of venetoclax in earlier lines of therapy, particularly in combination with CD38-targeted agents, could optimize treatment strategies and improve long-term outcomes in this patient population.