Daratumumab plus bortezomib and dexamethasone (Dara‐VD) in newly diagnosed Mayo 2004 stage IIIA and IIIB light‐chain amyloidosis: Long‐term follow‐up results from a prospective phase 2 study

Summary

Anti‐CD38 monoclonal antibodies dramatically improve the prognosis in immunoglobulin light‐chain (AL) amyloidosis, yet patients with end‐stage (Mayo 2004 IIIB) disease are typically excluded from prospective trials. To evaluate the daratumumab plus bortezomib and dexamethasone (Dara‐VD) regimen in Mayo 2004 stage III patients, we conducted a prospective phase 2 trial including 20 stage IIIA and 20 stage IIIB patients. The long‐term follow‐up results are reported here. The 24‐month haematological complete response (CR) and very good partial response (VGPR) rate were 52.5% and 12.5% respectively. Organ responses improved after 12 months; the 24‐month cardiac CR and VGPR rates were 15.0% and 35.0%. After a median follow‐up of 50.9 months, 16 patients died and 10 patients had progressive diseases. The median overall survival (OS) had not been reached in either group, with a 4‐year OS rate of 60.0% (95% confidence interval [CI]: 45.2%–75.2%). The median event‐free survival (EFS) was 33.5 months (95% CI 21.0–46.0 months), with a 4‐year EFS rate of 37.5% (95% CI: 24.3–54.7). No significant difference in OS or EFS was observed between stage IIIA and IIIB patients. In conclusion, the strong anti‐plasma cell regimen Dara‐VD provides comparable long‐term responses and prognosis for Mayo stage IIIB patients.

INTRODUCTION

Immunoglobulin light‐chain (AL) amyloidosis is a systemic disease caused by the deposition of misfolded amyloidgenic light chains produced by clonal plasma cells. Cardiac involvement is both common and a primary determinant of prognosis. ¹ The Mayo 2004 staging system, which utilizes cardiac troponin and N‐terminal pro‐B‐type natriuretic peptide (NT‐proBNP) levels, provides an accurate assessment of cardiac severity. ² This system defines high‐risk stage III disease. In 2013, a European modification further stratified these patients into subgroups IIIA and IIIB to better distinguish outcomes. ³ Patients with stage IIIB disease have the poorest prognosis, experiencing early mortality rates approaching 40% within 3 months of treatment initiation and a median overall survival of only 3–6 months. ³

The proteasome inhibitor and anti‐CD38 monoclonal antibody have dramatically improved the response rates and outcomes for patients with AL amyloidosis. The phase 3 ANDROMEDA trial, which compared the efficacy of cyclophosphamide, bortezomib and dexamethasone with or without daratumumab (Dara‐CyBorD vs. CyBorD), demonstrated that the daratumumab combination achieved higher rates of haematological complete response (CR) and improved cardiac and renal organ responses in newly diagnosed patients. ⁴ The Dara‐CyBorD regimen is now an approved first‐line treatment in newly diagnosed AL amyloidosis patients. ¹ However, patients with Mayo stage IIIB disease were excluded from this study and most other prospective clinical trials. The treatment of stage IIIB patients, therefore, remains a significant challenge in clinical practice. Studies have demonstrated that rapid haematological and early cardiac responses can improve outcomes in this high‐risk group. ⁵ , ⁶ Several retrospective studies have indicated that the addition of daratumumab to front‐line treatment can enhance both haematological and cardiac responses in newly diagnosed Mayo stage IIIB patients. ⁷ , ⁸ , ⁹ , ¹⁰

Thus, we designed a prospective phase 2 study to evaluate daratumumab in combination with bortezomib and dexamethasone (Dara‐VD) for newly diagnosed Mayo 2004 stages IIIA and IIIB patients (ClinicalTrials.gov identifier: NCT04474938). Alkylating drugs were abandoned to minimize myelosuppression, based on previous evidence that the addition of cyclophosphamide to bortezomib did not improve efficacy or survival. ¹¹

In a prior publication, we reported favourable efficacy and tolerability after a median follow‐up of 24 months. ¹² The 3‐month rates of haematological CR, very good partial response (VGPR) and overall response (ORR) were 47.5%, 20.0% and 82.5% respectively. The 2‐year overall survival (OS) rates were 74.3% (95% confidence interval [CI], 48.7%–88.4%) and 65.0% (95% CI, 40.3%–81.5%) in the IIIA and IIIB subgroups respectively. Those outcomes appear more favourable than those reported in the prospective phase 2 study of daratumumab monotherapy for stage IIIB patients (EMN22 study) ¹³ and are comparable to the results of a retrospective study of Dara‐VCD in stage IIIB patients. ⁹

However, the duration of responses and long‐term outcomes for patients treated with front‐line daratumumab remain unclear. Herein, we report the long‐term follow‐up data from this prospective phase 2 study and highlight the favourable outcomes observed in these advanced stage AL amyloidosis patients.

METHODS

Patients

The study protocol was published previously. ¹² Briefly, the inclusion criteria were age ≥18 years old, biopsy‐proved newly diagnosed Mayo 2004 stage III patients, and with a measurable haematological disease (difference between involved and uninvolved free light chain [dFLC] >50 mg/L). Patients with co‐diagnosed multiple myeloma or Waldenström macroglobulinaemia, uncontrolled infection, severe cardiac arrhythmia, severely compromised renal and liver function would be excluded. All patients provided informed consent to participate in this study. This study was approved by the Institution Ethics Committee of Peking Union Medical College Hospital (HS‐2414) on 23 June 2020, in accordance with the Declaration of Helsinki.

Treatment and follow‐up

Daratumumab was administered at 16 mg/kg weekly during cycles 1–2, every 2 weeks during cycles 3–6 and every 4 weeks for cycles 7–12. Bortezomib (1.3 mg/m²) and dexamethasone (20 mg) were given weekly for six cycles, of a 28‐day cycle. Drug doses could be tapered based on individual patient risks and adverse events. Haematological and organ responses were assessed every 3 months after treatment completion. For patients who achieved a haematological VGPR or CR at treatment completion, bone marrow minimal residual disease (MRD) was measured by EuroFlow‐based multiparametric flow cytometry, ¹⁴ with a sensitivity of 10⁻⁵.

Definition and evaluation of response

The primary end‐point of our trial was haematological response ≥VGPR at 3 months after treatment initiation. Secondary end‐points included: haematological response (assessed at day 8 of cycle 1 and at 3, 6, 12 and 24 months), MRD negativity, organ response (assessed at 3, 6, 12 and 24 months), early mortality (within 1 and 3 months from treatment initiation), OS and treatment‐related adverse events (TRAEs).

Haematological responses were evaluated according to current validated criteria. ¹⁵ , ¹⁶ Stringent serum free light chain (FLC) response was defined as a reduction in dFLC to <10 mg/L ¹⁷ or the involved FLC ≤20 mg/L. ¹⁸ Cardiac response was assessed with a recently published four‐level grading system. ¹⁹ Renal response was defined as either a ≥30% reduction in proteinuria or decrease in proteinuria to <0.5 g/24 h in the absence of renal progression (≥25% decrease in estimated glomerular filtration rate (eGFR)). ²⁰ Hepatic response was defined as a ≥50% decrease in abnormal alkaline phosphatase value level. ²¹ Patients who died prior to response assessment and lack any post‐baseline evaluation were categorized as the no‐response group.

Event‐free survival (EFS) was defined as the time from treatment initiation to any of the following events occur first: death, major organ progression (i.e. cardiac, hepatic and renal) and haematological progression as per consensus guidelines or initiation of second‐line treatment. Specifically, haematological progression was defined as: for CR patients, any detectable monoclonal protein or abnormal FLC ratio (involved FLC must double and be above the upper limit of normal value); for VGPR or PR patients, a 50% increase in serum monoclonal protein to >5 g/L or increase in urine monoclonal protein to >200 mg/day; or a 50% increase in involved FLC to >100 mg/L. ²¹ Cardiac progression criteria included any of the following ¹⁹ : a >30% and >300 pg/mL increase in NT‐proBNP (or BNP progression >30% and >70 pg/mL) from the nadir, not related to infection, cardiac arrthymia or creatinine elevation; a ≥10% decrease in EF from the best‐recorded value; or a >33% increase in troponin I from the nadir. Hepatic progression was defined as a ≥50% increase in alkaline phosphatase from the nadir. ²² Renal progression was defined as a ≥25% worsening of eGFR. ²³ OS was defined as the time from treatment initiation to death from any cause.

Statistical analysis

The intention‐to‐treat (ITT) population included all enrolled patients. Haematological and organ responses were assessed on an ITT analysis. Statistical analyses were performed using SPSS 29.0 (Inc., Chicago, IL, USA) and R software (Version 4.4.2). Survival curves were plotted by the Kaplan–Meier methods, the 95% confidence intervals (CIs) and difference comparison were done by the log‐rank test. A competing risk model was applied to evaluate the duration of maintenance of stringent FLC response, with death from other causes prior to loss of stringent FLC response defined as a competing event. The difference comparison was performed by Gray's test. A p‐value less than 0.05 was considered statistically significant.

RESULTS

Summary of patient baseline characteristics

Twenty (50.0%) Mayo stage IIIA and 20 (50.0%) stage IIIB patients were enrolled from May 2021 to March 2022. The detailed demographic and baseline characteristics of the stage IIIA and stage IIIB patients were published previously ¹² and in Table S1. In summary, the study comprised 30 (75.0%) males and the median age at diagnosis was 59 years (range, 40–77 years). The median NT‐proBNP level was 7801 ng/L (range, 803–35 000 ng/L), and the median cardiac troponin‐I level was 0.17 μg/L (range, 0.07–3.07 μg/L). Twenty‐four (60%) patients presented as New York Heart Association (NYHA) cardiac class III or IV. Concurrent kidney and liver involvement were present in 12 (30.0%) and 10 (25.0%) patients respectively (Table S1).

The responses after further follow‐up

The 24‐month haematological CR and VGPR rates were 52.5% (95% CI 36.3%–68.7%) and 12.5% (95% CI 1.8%–23.2%) respectively (Table 1). Among the 25 patients who achieved VGPR or better at the end of treatment, 5 (25%) stage IIIA and 10 (50%) stage IIIB patients achieved MRD negativity. Haematological responses plateaued between 12 months and 24 months, with only 1 (2.5%) additional patient improving from VGPR to CR and 2 (5.0%) additional patients improving from PR to VGPR during this interval (Figure 1). The decline in the CR rate at 24 months compared to 12 months in the stage IIIB group was attributed to one patient with a previous haematological CR who died from cardiac progression.

TABLE 1.

Summary of prolonged haematological responses.

Response	MRD negative, N (%) at the end of treatment	Response at 24 months, N (%)	Loss of stringent FLC response at last follow‐up, N (%)
CR	15 (37.5%)	21 (52.5%)	/
VGPR		5 (12.5%)	/
CR + VGPR		26 (67.5%)	/
Stringent FLC response		19 (47.5%)	10 (35.7%)
PR		0 (0.0%)	/
ORR		26 (67.5%)	/
Mayo stage IIIA
CR	5 (25.0%)	10 (50.0%)
VGPR		4 (20.0%)	/
CR + VGPR		14 (70.0%)	/
Stringent FLC response		8 (40.0%)	7 (58.3%)
PR		0 (0.0%)
ORR		14 (70.0%)	/
Mayo stage IIIB
CR	10 (50%)	11 (55.0%)
VGPR		1 (5.0%)	/
CR + VGPR		12 (60.0%)	/
Stringent FLC response		11 (55.0%)	3 (18.8%)
PR		0 (0.0%)	/
ORR		12 (60.0%)	/

Abbreviations: CR, complete response; FLC, free light chain; MRD, minimal residual disease; ORR, overall response; PR, parital response; VGPR, very good partial response.

FIGURE 1.

The haematological response at each fixed time point in Mayo stage IIIA and stage IIIB patients. NR*: Including nonresponse, haematological progression, patients receiving salvage treatment and death. CR, complete response; NR, No response; VGPR, very good partial response.

The 24‐month cardiac CR, VGPR and PR rate were 15.0% (95% CI 3.4%–26.6%), 35.0% (95% CI 19.6%–50.4%) and 10.0% (95% CI 2.0%–15.8%) respectively (Figure 2). Compared with the 12‐month assessment, an additional 3 (7.5%) patients improved from VGPR to CR and 5 (12.5%) patients improved from PR to VGPR by 24 months. One patient with cardiac VGPR and one with cardiac PR died from cardiac progression.

FIGURE 2.

The cardiac response at each fixed time point. NR*: Including nonresponse, cardiac progression, patients receiving salvage treatment and death. CR, complete response; VGPR, very good partial response.

The hepatic response rate at 24 months was 80.0% (8 of 10 patients), compared with the response rate of 20.0% (2 of 10 patients) at 12 months. The renal response rate at 24 months was 33.3% (4 of 12 patients).

Survival after further follow‐up

As previously reported, 12 patients died from sudden cardiac death or heart failure exacerbation after infection. ¹² After a median follow‐up of 50.9 months (95% CI 48.0–53.8 months) until 31 December 2025, an additional four deaths occurred: two from cardiac progression, one from rectal cancer and one from an unknown cause. Ten of the 27 patients who completed 12 months of treatment had disease progressions. In detail, five patients (4 stage IIIA patients and 1 stage IIIB patient) had haematological progression from a haematological CR status. Two stage IIIA patients experienced renal progression (1 while in haematological CR and the other in VGPR status). One stage IIIB patient had cardiac progression while in haematological CR status. One stage IIIA patient had concurrent cardiac and renal progressions during haematological VGPR status. Another stage IIIA patient had simultaneous haematological, cardiac and hepatic progressions.

Three patients received venetoclax plus dexamethasone as a salvage treatment. Two patients with cardiac progression died before salvage therapy. Currently, three patients with isolated haematological progression and two patients with isolated renal progression are under close follow‐up.

The median OS was not reached. The median EFS was 33.5 months (95% CI 21.0–46.0 months). The estimated 4‐year OS and EFS rate were 60.0% (95% CI, 45.2%–75.2%) and 37.5% (95% CI, 24.3%–54.7%) respectively. Subgroup analysis revealed no significant difference in OS between Mayo stage IIIA and IIIB patients (median OS not reached in both groups, p = 0.401). There was no significant difference in EFS (Mayo stage IIIA vs. IIIB, 33.5 months vs. 32.8 months, p = 0.398) (Figure 3A,B).

FIGURE 3.

Kaplan–Meier graph of overall survival (OS) and event‐free survival (EFS) (A) Kaplan–Meier graph of OS by Mayo 2004 stage; the area between the dashed lines indicates the confidence interval of OS; (B) Kaplan–Meier graph of EFS by Mayo 2004 stage; (C) Kaplan–Meier graph of 6‐month landmark analysis of OS by Mayo 2004 stage.

Baseline positive bone marrow t(11;14) translocation status did not significantly correlate with inferior OS (median OS not reached in both t(11;14)‐positive and ‐negative groups, p = 0.331).

Notably, in a 6‐month landmark analysis of OS, the stage IIIB patients showed a trend towards inferior survival compared to stage IIIA patients (p = 0.062) within the first 6 months. The 6‐month early mortality rate was 10% (95% CI 1.6%–27.8%) and 35.0% (95% CI 15.1%–55.8%) in stage IIIA and IIIB group respectively. However, OS was similar between both subgroups thereafter (p = 0.398, Figure 3C).

In the stage IIIB subgroup, patients who achieved haematological CR within 3 months had a significantly lower 6‐month mortality than those who did not (6‐month mortality rate 23.5% vs. 100%, p < 0.001).

Duration of haematological response

Twenty‐eight (70%) patients achieved a stringent FLC response as their best haematological response. Patients who achieved a stringent FLC response had significantly longer OS than those who did not (median OS not reached vs. 10.4 months, p < 0.001).

At the 24‐month follow‐up, a total of 19 (47.5%, 95% CI 31.3%–63.7%) patients maintained a stringent FLC response.

At the last follow‐up, 10 patients (35.7%, including 7 stage IIIA and 3 stage IIIB patients) lost their stringent FLC response. Only three of these patients had achieved MRD negativity at the end of treatment. Conversely, 12 patients (42.9%, including 3 stage IIIA and 9 stage IIIB patients) maintained a stringent FLC response, 10 of whom had previously attained MRD negativity (Table 1). Six patients (21.4%, including 2 stage IIIA and 4 stage IIIB patients) died before losing their stringent FLC response. In a competing risk analysis, the stage IIIB patients had a tendency for a longer median duration of stringent FLC response compared to stage IIIA patients (not reached vs. 40.0 months, p = 0.053, Figure 4).

FIGURE 4.

Competing risk graph of percentage of loss of stringent free light chain (FLC) response.

DISCUSSION

Mayo 2004 stage III AL amyloidosis is characterized by severe cardiac involvement, high early mortality and poor outcomes. While novel anti‐plasma cell therapies such as daratumumab and bortezomib have revolutionized the treatment, patients with stage IIIB disease are frequently excluded from prospective trials, making their management a persistent challenge. We therefore conducted this first prospective study evaluating Dara‐VD regimen in stage IIIA and stage IIIB patients and reported the promising efficacies and high tolerances in those patients, particularly in stage IIIB patients. ¹² Here, we present the long‐term follow‐up results from this trial.

The 24‐month haematological CR rate (52.5%) in our study was comparable to rates reported in the Dara‐CyBorD group of the ANDROMEDA study (59.5%) ²⁴ and in a retrospective study by Chakraborty et al. (48%). ²⁵ For stage IIIB patients, our haematological CR and VGPR rates were also comparable to those in a retrospective cohort from Columbia University. ⁹ Notably, 70% of our patients achieved stringent FLC responses, and among those with VGPR and CR, 60% attained MRD negativity at the end of treatment. Rapid and deep haematological responses likely underpinned the progressive improvement in cardiac responses, with the cardiac VGPR and CR rate reaching 57.5% at 24 months, which was consistent with findings by Muchtar et al. ¹⁹ and higher than the 33% rate reported by Chakraborty et al. for stage IIIB patients treated with the Dara‐CyBorD regimen. ⁹

Daratumumab‐based treatment provides more favourable outcomes than bortezomib‐based therapy for patients with advanced‐stage disease. The median OS was not reached in both groups in our study, whereas the median OS was 36 and 4 months in stage IIIA and IIIB patients in bortezomib‐based treatment as reported by Manwani et al. respectively. ¹⁷ The median OS in our stage IIIB subgroup appears longer compared to daratumumab monotherapy in the EMN 22 study (median OS 10.3 months). ¹³ Further results of long‐term follow‐up from EMN22 study are anticipated. Notably, no significant difference in long‐term OS and EFS was observed between stage IIIA and IIIB patients in our cohort. This lack of statistical significance may be partly due to limited sample size. However, it may also reflect that stage IIIB patients who survived beyond the initial high‐risk period achieved similar favourable outcomes, potentially driven by deeper haematological response. Whether extending daratumumab maintenance could further deepen responses and improve outcomes in stage IIIA patients requires further investigations.

Despite achieving rapid and high haematological response rates and prolonging survival in stage IIIB patients, Although daratumumab‐based therapy achieves high hematologic response rates and extends survival in patients with stage IIIB disease, this has not translated into the early cardiac improvements necessary to reduce early mortality in this ultra‐high‐risk group. In our study, the 6‐month early mortality rate for stage IIIB patients remained at 35%. This aligns with the rate reported in EMN22 study utilizing daratumumab monotherapy ¹³ and falls within the 6‐month early mortality rate range from 22.8% to 43.7% in previous Dara‐CyBorD studies. ⁷ , ²⁵ , ²⁶ While a marked decline from earlier eras (e.g. EMN23 study ²⁷ and Pavia study ⁶ ), this rate is similar to the 32.3% reported in bortezomib‐based treatment era. ²⁸

Reducing early mortality remains the pivotal challenge for improving outcomes in stage IIIB patients, even in the daratumumab era. We identified that stage IIIB patients achieving haematological CR within 3 months had a significantly lower early mortality rate than those who did not. Furthermore, survival beyond 6 months was comparable between stage IIIA and stage IIIB patients, and the duration of stringent FLC response was actually longer in the stage IIIB subgroup. This sustained haematological response may be attributed to a higher rate of MRD negativity in the stage IIIB patients, which reflects deeper haematological response that could translate into longer OS. Therefore, pursuing faster and deeper haematological responses is crucial in stage IIIB patients. More intensive anti‐plasma cell drugs, such as teclistamab and chimeric antigen receptor T cell, merit further consideration in stage IIIB patients.

In conclusion, our research provides the long‐term follow‐up data from the first clinical trial including Mayo stage IIIB AL amyloidosis patients. The potent anti‐plasma cell regimen Dara‐VD yielded a comparable prognosis for Mayo stage IIIB patients and IIIA patients. This favourable outcomes in stage IIIB patients may be attributed to their higher MRD negativity and longer duration of staying in stringent FLC responses.

AUTHOR CONTRIBUTIONS

X.‐M. Gao was responsible for original draft writing, reviewing and editing, funding acquisition, investigation and methodology; C. Xu was responsible for original draft, visualization and investigation; Y.‐J. Gao was responsible for investigation and methodology; A. Guan, Y. Xu and L. Chang were responsible for investigation; K.‐N. Shen was responsible for original draft writing, reviewing and editing, conceptualization, investigation, methodology, funding acquisition and formal analysis; J. Li was responsible for draft reviewing and editing, conceptualization, investigation, supervision and funding acquisition.

FUNDING INFORMATION

This work was supported by the National Natural Science Foundation of China (82470202 for LJ); the CAMS Innovation Fund for Medical Sciences (CIFMS) (Grant No. 2025‐I2M‐XHXX‐010 for SKN); the National High Level Hospital Clinical Research Funding (Grant No. 2025‐PUMCH‐A‐182 for SKN); the Peking Union Medical College Hospital Talent Cultivation Program (Category D, Grant No. UHB12604 for GXM); and funding from Xi'an Janssen Pharmaceutical Ltd., Beijing, China.

CONFLICT OF INTEREST STATEMENT

The authors declared no conflicts of interest.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov identifier: NCT04474938.

Supporting information

Table S1.

DATA AVAILABILITY STATEMENT

The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.