A Multicenter National Study of AL Amyloidosis in China: Insights Into Real‐World Diagnosis, Treatment, and Prognosis

ABSTRACT

This national, multicenter, retrospective study analyzed 1070 patients with newly diagnosed systemic light‐chain (AL) amyloidosis in China (2008–2025). Findings reveal increasing annual diagnoses and a marked shift in first‐line therapy from proteasome inhibitor (PI)‐based regimens to predominantly daratumumab‐based regimens since 2024. Monthly kinetic assessments demonstrated that daratumumab‐based induction produced deeper and faster hematologic responses than PI‐based therapy [≥ very good hematologic partial response (HemVGPR): 80.3% vs. 70.8%, p = 0.012; median time to ≥ HemVGPR: 1.2 vs. 1.8 months, p = 0.001], with significantly superior cardiac overall response (63.1% vs. 53.3%, p = 0.030). Early mortality rates at 1, 3, and 6 months were 5.5%, 13.2%, and 16.2%, respectively, with daratumumab‐based therapy identified as an independent protective factor for 6‐month mortality. Median event‐free survival (EFS) was 44.4 months, while median overall survival (OS) was not reached. Achievement of hematologic complete response (HemCR) conferred superior EFS and OS over HemVGPR (both p < 0.001), with minimal residual disease negativity further improving EFS in HemCR patients (p = 0.043). Concurrent hepatic and cardiac involvement defined a high‐risk subgroup with poor outcomes (median EFS 12.8 months, OS 51.0 months). Multivariable analysis confirmed Eastern Cooperative Oncology Group (ECOG) performance status (PS) > 2, liver involvement, and gain1q as adverse prognostic factors for EFS, whereas ≥ HemVGPR was protective. For OS, liver involvement, concomitant multiple myeloma, and ECOG PS > 2 were independent risk factors, while ≥ HemVGPR and cardiac response predicted improved survival. This study establishes a contemporary benchmark for AL amyloidosis management in China, confirming the superior real‐world efficacy of daratumumab‐based frontline therapy.

1. Introduction

Light‐chain (AL) amyloidosis is a rare plasma cell disorder with a global incidence of 8–12 cases per million person‐years [1, 2], though data specific to China are lacking. The disease is caused by misfolded immunoglobulin light chains that form toxic amyloid fibrils, leading to multi‐organ dysfunction [3], primarily affecting the heart (70%–80%) and kidneys (60%–70%) [4]. Diagnosis is often delayed by 6–12 months [5, 6] due to nonspecific symptoms, contributing to high symptom burden, early mortality, and poor prognosis [7, 8]. Recent advancements, particularly the introduction of anti‐CD38 monoclonal antibodies like daratumumab (Dara), have improved outcomes. In the ANDROMEDA trial [9], Dara, combined with bortezomib, cyclophosphamide, and dexamethasone (Dara‐CyBorD), achieved hematologic very good partial response or better in 78.5% of patients at 6 months, compared to 49.2% with CyBorD, with superior cardiac and renal response rates. Real‐world studies report a 6‐month early mortality rate below 10% with Dara‐based therapy [10], compared to approximately 25% in the pre‐Dara era [11]. Current data on AL amyloidosis in China are limited, with most evidence derived from single‐center studies or Western cohorts, leaving gaps in understanding treatment patterns, outcomes, and prognostic factors. Moreover, key questions regarding the impact of response depth, minimal residual disease (MRD) status, and the relevance of cytogenetic abnormalities and other survival predictors have not been systematically evaluated in large, real‐world Chinese populations. This multicenter, national study aims to address these gaps by providing real‐world evidence on diagnosis, treatment paradigms, the effectiveness of Dara‐based regimens, and identifying independent prognostic factors to optimize AL amyloidosis management in China.

2. Patients and Methods

2.1. Study Design and Participants

This large, retrospective, observational study enrolled AL amyloidosis patients from 18 Chinese institutions between October 2008 and April 2025. Diagnosis was confirmed via Congo‐red‐positive biopsy and/or immunoelectron microscopy of amyloid deposits in abdominal fat, minor salivary gland, bone marrow, or involved organs, with kappa or lambda light chain typing verified by immunohistochemistry, immunofluorescence, immunoelectron microscopy, or mass spectrometry. The study adhered to the Declaration of Helsinki and was approved by the Medical Ethics Committee of Zhongshan Hospital Fudan University (B2023‐185R). Patients consented to the use of their electronic medical records for research. Retrospective data included demographics, ECOG performance status (PS), clinical and laboratory findings, and treatment details. Bone marrow plasma cell percentage (BMPC) was obtained from bone marrow aspirate smear and/or biopsy; when both were available, the higher value was used. Fluorescence in situ hybridization (FISH) was performed on CD138‐selected bone marrow plasma cells. Organ involvement was defined per National Comprehensive Cancer Network (NCCN) guidelines [12], with cardiac involvement assessed using the 2015 European Mayo 2004 [13] and revised Mayo 2012 models [14]. Follow‐up data were obtained from medical records and telephone contact.

2.2. Outcomes

This study evaluated hematological and organ response rates to first‐line therapy, MRD status, event‐free survival (EFS), and overall survival (OS) in AL amyloidosis patients. First‐line therapy was defined as the initial regimen, regardless of modifications. Responses were assessed per AL consensus criteria [15, 16, 17]. MRD was detected via flow cytometry at sensitivities of 10–5 or 10–6. EFS was defined as the time from diagnosis to hematologic progression, major organ deterioration, or death, whichever comes first. OS was defined as the duration from diagnosis until death from any cause or the last follow‐up.

2.3. Statistical Analysis

Patient demographics and disease characteristics were summarized using descriptive statistics: categorical variables were presented as frequencies and percentages, while continuous variables were expressed as medians and ranges. Group comparisons were performed using the Student's t‐test (for normally distributed data with homogeneous variance) or the Mann–Whitney U test for continuous variables, and the Pearson's chi‐squared test or Fisher's exact test for categorical variables. Survival outcomes (EFS and OS) were analyzed using the Kaplan–Meier method, with subgroup comparisons conducted by the log‐rank test. Cutoff values for continuous variables were determined by receiver operating characteristic (ROC) analysis against the corresponding binary endpoint (early mortality at 1, 3, and 6 months) or survival outcome (EFS and OS), with optimal cutoffs selected by the Youden index. These thresholds are exploratory and cohort‐derived and require external validation. Univariate and multivariate logistic regression models were used to identify prognostic factors for early mortality, while univariate and multivariate Cox proportional hazards models were applied for EFS and OS. Variables with a p value ≤ 0.1 in univariate analysis were entered into the multivariate models. A two‐sided p value < 0.05 was considered statistically significant. All analyses were performed using SPSS (version 27.0) and R (version 4.4.1).

3. Results

3.1. Demographic and Clinical Characteristics

This study enrolled 1070 newly diagnosed AL amyloidosis patients from 18 Chinese hospitals (October 2008–April 2025) (Figure 1). The median time intervals between symptom onset and diagnosis of AL amyloidosis were 6.1 months. Before a definitive diagnosis of AL amyloidosis, 42.4%, 35.1%, and 2.2% initially presented to the department of cardiology, nephrology, and pneumology, respectively. Baseline characteristics are summarized in Table 1. The median age at diagnosis was 62 years (range 30–91), with 66.5% males. ECOG PS ≥ 2 was observed in 19.1% of patients. Immunoglobulin subtypes included IgG (34.1%), IgA (20.4%), IgM (2.1%), IgD (0.8%), and light‐chain (26.9%), while 15.7% (161/1023) had negative results on both serum and urine immunofixation electrophoresis (IFE). FLC data were available in 154 of these 161 patients; 16.2% (25/154) had a normal serum FLC ratio. Crucially, all 25 patients had detectable clonal bone marrow plasma cells and biopsy‐confirmed light‐chain amyloidosis, thus meeting diagnostic criteria for AL amyloidosis independent of conventional paraprotein detection.

FIGURE 1.

Distribution of institutions and patients of the study. The numbers on the map represent the number of hospitals from each province participating in this research, and the colors represent the number of patients. [Color figure can be viewed at wileyonlinelibrary.com]

TABLE 1.

Baseline demographic, clinical characteristics, and treatment of patients with AL amyloidosis.

	N = 1070
Male, no. (%)	712 (66.5)
Age (years), median [IQR]	62 (55–69)
ECOG performance status, no. (%)
0–2	825/1020 (80.9)
> 2	195/1020 (19.1)
IFE, no. (%)
IgG	349/1023 (34.1)
IgA	209/1023 (20.4)
IgM	21/1023 (2.1)
IgD	8/1023 (0.8)
Light chain	275/1023 (26.9)
Negative	161/1023 (15.7)
Lambda restriction, no. (%)	762/999 (76.3)
dFLC (mg/L), median [IQR]	188.8 (79.0–445.7)
BMPC (%), median [IQR]	10.0 (5.0–20.0)
Concomitant multiple myeloma, no. (%)	351 (32.8)
Serum hemoglobin (g/L), median [IQR]	121 (104–134)
Serum albumin (g/L), median [IQR]	33.6 (26.9–38.8)
Serum alkaline phosphatase (U/L), median [IQR]	86 (67–119)
Serum creatinine(μmol/L), median [IQR]	91 (72–127)
eGFR (mL/min/1.73 m2), median [IQR]	70 (47–91)
LDH(U/L), median [IQR]	220 (182–274)
β2‐microglobulin (mg/L), median [IQR]	3.7 (2.7–5.7)
Coagulation factor X (%), median [IQR]	66.5 (53.3–83.0)
NT‐proBNP (pg/mL), median [IQR]	3095 (813.5–7728)
cTNT (ng/mL), median [IQR]	0.065 (0.032–0.128)
LVEF (%), median [IQR]	62 (53–66)
Proteinuria (g/24 h), median [IQR]	1.36 (0.34–4.48)
Organ involvement, no. (%)
Heart	894 (83.6)
Kidney	498/866 (57.5)
Liver	112/1000 (11.2)
Soft tissues	511/702 (72.8)
Bone marrow	367/792 (46.3)
Nerve	78 (7.3)
Gastrointestinal tract	44 (4.1)
Others a	11 (1.0)
Number of involved organs, median [IQR]	2 (2–3)
2015 European modification of 2004 Mayo model, I/II/IIIa/IIIb (%)	13.6/23.7/38.3/24.4
Revised 2012 Mayo model, I/II/III/IV (%)	2.9/40.2/47.5/9.4
Cytogenetic abnormality
gain1q	261/799 (32.6)
t(11;14)	235/768 (30.6)
del13q	208/751 (27.6)
t(4;14)	47/766 (6.1)
del17p	33/793 (4.2)
t(14;16)	6/767 (0.8)
First‐line induction therapy, no. (%)
Dara‐based	332/979 (33.9)
Dara+PI+IMiD	22/979 (2.2)
Dara+PI	254/979 (25.9)
Dara+IMiD	30/979 (3.1)
Dara	26/979 (2.7)
PI‐based	509/979 (52.0)
PI + IMiD‐based	104/979 (10.6)
IMiD‐based	22/979 (2.2)
Others b	12/979 (1.2)
Cycles of first‐line induction therapy, median [IQR]	4 (1–6)
ASCT, no. (%)	47 (4.8)

Abbreviations: ASCT, autologous stem cell transplantation; BMPC, bone marrow plasma cell; cTNT, cardiac troponin T; Dara, daratumumab; dFLC, difference in involved and uninvolved free light chain; ECOG, Eastern Cooperative Oncology Group; eGFR, estimated glomerular filtration rate; IFE, immunofixation electrophoresis; IMiD, immunomodulatory drug; LDH, lactate dehydrogenase; LVEF, left ventricular ejection fraction; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide; PI, proteasome inhibitor.

^a Including space‐occupying lesions in the lung, spinal canal, and brain.

^b Including venetoclax‐based therapy, chemotherapy (including cyclophosphamide and melphalan), and glucocorticoid‐only.

The median difference between involved and uninvolved free light chains (dFLC) was 188.8 mg/L, with lambda restriction in 76.3%. Median bone marrow plasma cell infiltration was 10%. Concomitant multiple myeloma (MM, per CRAB‐SLiM criteria [18]) was present in 351 patients (32.8%). Of these, 55 (15.7%) met SLiM criteria solely on the basis of an involved/uninvolved serum FLC ratio > 100 without CRAB features, 217 (62.0%) had CRAB features only, and 78 (22.3%) fulfilled both SLiM and CRAB criteria. Amyloid deposits in bone marrow were positive in 46.3% of confirmed cases.

According to the Consensus Criteria [12], organ involvement included heart (83.6%), kidney (57.5%), liver (11.2%), soft tissues (72.8%), nerve (7.3%), and gastrointestinal tract (4.1%), with 41.1% having > 2 organs affected. Besides, among patients with cardiac involvement, 25.2% were misdiagnosed as hypertrophic cardiomyopathy, coronary heart disease, restrictive cardiomyopathy, or arrhythmia. Over half were staged as IIIa/IIIb (62.7%) or III/IV (56.9%) under the 2015 European Mayo model and the Revised Mayo 2012 system, respectively.

Among AL patients with available FISH results, 67.8% had ≥ 1 cytogenetic abnormality. Presence of gain1q, t(11;14), and del13q were frequently observed at diagnosis in the entire group, with an incidence of 32.6%, 30.6%, and 27.6%, while the incidence of t(4;14), del17p, and t(14;16) were rare (6.1%, 4.2%, and 0.8%, respectively). A distinct cytogenetic profile was observed between AL patients with and without concomitant MM. Compared to patients with concurrent MM, those with AL alone exhibited higher t(11;14) frequency (32.2% vs. 28.5%, p = 0.280) but significantly lower rates of gain1q (26.1% vs. 45.2%, p < 0.001), del13q (21.3% vs. 38.1%, p < 0.001), t(4;14) (5.2% vs. 9.7%, p = 0.019), and del17p (2.7% vs. 7.0%, p = 0.004) (Table S1).

3.2. Treatment

Of the 1070 newly diagnosed patients, 979 (91.5%) received first‐line treatment (Table 1), while 91 (8.5%) received supportive care only. The latter group predominantly comprised patients with advanced‐stage disease (Mayo 2012 stage ≥ II in 96.4%; Mayo 2015 stage ≥ II in 89.7%), significant cardiac and multi‐organ comorbidities, or those who were lost to follow‐up or elected to return to local centers for care.

The median number of frontline treatment cycles was 4 (range: 0.25–18). A total of 175 patients (16.4%) underwent second‐line therapy, and 77 (7.2%) received three or more lines of treatment.

First‐line regimens were categorized as follows: Dara‐based regimens (332/979, 33.9%), PI‐based regimens (509, 52.0%), PI + IMiD‐based regimens (104, 10.6%), IMiD‐based regimens (22, 2.2%), and other regimens (12, 1.2%), including Venetoclax‐based regimens, conventional chemotherapy, and glucocorticoids. Among Dara‐based regimens, the most common was Dara‐CyBorD (Dara‐bortezomib‐cyclophosphamide‐dexamethasone, n = 134, 40.4% of Dara group), followed by Dara‐bortezomib‐dexamethasone (n = 117, 35.2%). Quadruplet therapy with IMiD (Dara‐Bor‐Len‐Dex, n = 16; Dara‐Bor‐Tha‐Dex, n = 4; Dara‐Ixa‐Len‐Dex, n = 2; total 6.6%) and Dara‐IMiD‐Dex (Dara‐Len‐Dex, n = 22; Dara‐Pom‐Dex, n = 8; total 9.0%) were less frequently used, while Dara‐dexamethasone alone was administered in 26 patients (7.8%), primarily in frail or rapidly deteriorating patients. During the treatment period, 47 patients (4.8%) underwent autologous stem cell transplantation.

Moreover, we assessed the temporal trends in patient diagnoses and first‐line treatment regimens (Figure 2). A notable increase in the number of diagnosed cases has been observed since 2016. Before 2023, PI‐containing regimens served as the mainstay of first‐line therapy, accounting for the majority of patients. The utilization of daratumumab has steadily increased since its introduction as a first‐line option in 2020. From 2024 onwards, over half of AL patients received Dara‐based regimens as frontline treatment. The proportion of patients receiving only supportive care has declined since 2017, reaching its lowest rate in 2025.

FIGURE 2.

The evolving landscape of patient diagnosis and first‐line treatment patterns. [Color figure can be viewed at wileyonlinelibrary.com]

3.3. Treatment Response and Response Kinetics of First‐Line Therapy

Following the exclusion of patients due to early mortality, loss to follow‐up, or unavailable data, best hematologic response to first‐line therapy was assessable in 622 patients, with rates of 46.5% for hematologic complete response (HemCR), 74.1% for hematologic very good partial response or better (≥ HemVGPR), and 88.6% for ≥ HemPR (Table 2). Non‐evaluable patients (n = 357) had significantly poorer baseline characteristics, reflecting real‐world practice where sicker patients are more prone to loss to follow‐up (Table S6). Organ response was assessable in 537, 281, and 34 patients with cardiac, renal, and hepatic involvement, respectively. The cardiac response rates were 11.5% for cardiac complete response (CarCR), 33.5% for ≥ CarVGPR, and 56.8% for ≥ CarPR. Corresponding renal response rates were 14.9% for RenCR, 42.7% for ≥ RenVGPR, and 66.5% for ≥ RenPR. In patients with hepatic involvement, a hepatic partial response (HepPR) was achieved in 61.8% of patients. The median time intervals from initiation of treatment to the best hematologic, cardiac, and renal responses during first‐line therapy were 2.7, 4.5, and 4.0 months, respectively. Besides, MRD negativity was attained in 49.1% (192/391) of patients during frontline therapy.

TABLE 2.

Best treatment response during first‐line therapy in assessable patients with AL amyloidosis.

	All (N = 979)	Dara‐based (N = 332)	PI‐based (N = 613)	p (Dara‐ vs. PI‐based)
Hematologic response	n = 622	n = 213	n = 404
HemCR	289 (46.5%)	97 (45.5%)	191 (47.3%)	0.734
≥ HemVGPR	461 (74.1%)	171 (80.3%)	286 (70.8%)	0.012
≥ HemPR	551 (88.6%)	204 (95.8%)	342 (84.7%)	< 0.001
Cardiac response	n = 537	n = 198	n = 334
CarCR	62 (11.5%)	24 (12.1%)	37 (11.1%)	0.779
≥ CarVGPR	180 (33.5%)	72 (36.4%)	107 (32.0%)	0.343
≥ CarPR	305 (56.8%)	125 (63.1%)	178 (53.3%)	0.030
Renal response	n = 281	n = 81	n = 199
RenCR	42 (14.9%)	12 (14.8%)	29 (14.6%)	1.000
≥ RenVGPR	120 (42.7%)	40 (49.4%)	79 (39.7%)	0.145
≥ RenPR	187 (66.5%)	60 (74.1%)	126 (63.3%)	0.095
Hepatic response	n = 34	n = 9	n = 25
HepPR	21 (61.8%)	7 (77.8%)	14 (56.0%)	0.427
MRD negativity	192/391 (49.1%)	75/131 (57.3%)	116/259 (44.8%)	0.024

Abbreviations: CarCR, cardiac complete response; CarNR, cardiac no response; CarPR, cardiac partial response; CarVGPR, cardiac very good partial response; Dara, daratumumab; HemCR, hematologic complete response; HemNR, hematologic no response; HemPR, hematologic partial response; HemVGPR, hematologic very good partial response; MRD, minimal residual disease; PI, protease inhibitor; RenCR, renal complete response; RenNR, renal no response; RenPR, renal partial response; RenVGPR, renal very good partial response.

Moreover, we compared treatment response between the two main first‐line regimens: Dara‐based versus PI‐based (i.e., PI ± IMiD) therapies. Comparison of baseline characteristics of the two groups is listed in Table S2. Patients receiving frontline Dara‐based regimens had more adverse baseline clinical features than PI‐based group, including a higher proportion with ECOG PS > 2 (23.1% vs. 15.7%, p = 0.007), elevated levels of dFLC (median: 231.1 vs. 165.4 mg/L, p < 0.001) and NT‐proBNP (median: 3786 vs. 2573 pg/mL, p = 0.002), a greater frequency of cardiac involvement (86.7% vs. 80.6%, p = 0.017), and more advanced cardiac stage (Mayo 2015 stage IIIb: 27.6% vs. 20.1%, p = 0.013). The observed baseline differences between Dara and PI groups are not intrinsic to the cohort but reflect the reimbursement landscape in China. Prior to the inclusion of the subcutaneous formulation in the National Reimbursement Drug List (NRDL) in November 2024 (effective January 1, 2025), the high cost of Dara led to selective prescribing for higher‐risk patients; following national reimbursement, this selection bias was largely eliminated, and baseline characteristics between groups became balanced.

Despite the inferior baseline, the Dara‐based regimen resulted in a comparable HemCR rate to the PI‐based regimen (45.5% vs. 47.3%, p = 0.734), but significantly higher rates of ≥ HemVGPR (80.3% vs. 70.8%, p = 0.012) and ≥ HemPR (95.8% vs. 84.7%, p < 0.001). Regarding organ response, significantly more patients in the Dara‐based group achieved a cardiac response (≥ CarPR; 63.1% vs. 53.3%, p = 0.030). No statistical difference in best overall renal response during first‐line therapy was observed between the Dara‐ and PI‐based group (≥ RenPR, 74.1% vs. 63.3%, p = 0.095). The rate of MRD negativity was also superior in the Dara‐based group (57.3% vs. 44.8%, p = 0.024).

Furthermore, a longitudinal analysis of treatment response kinetics during first‐line therapy demonstrated that the daratumumab‐based group sustained a superior response rate at each assessment time point compared to the PI‐based group (Figure 3). The median time to ≥ HemPR in Dara‐ versus PI‐based group was 0.9 versus 1.1 months (p < 0.001), and to ≥ HemVGPR was 1.2 versus 1.8 months (p = 0.001). The shorter time to HemCR (2.7 vs. 2.9 months) also favored the daratumumab group but did not reach statistical significance (p = 0.088). The Dara‐based group also achieved cardiac responses more rapidly than the PI‐based group, demonstrating a median time to ≥ CarPR of 1.8 versus 2.3 months (p = 0.091), to ≥ CarVGPR of 3.0 versus 4.7 months (p = 0.023), and to CarCR of 4.4 versus 4.6 months (p = 0.792). Renal response kinetics were statistically comparable between Dara‐ and PI‐groups (median time of achieving ≥ RenPR, 2.1 vs. 2.6 months, p = 0.129; ≥ RenVGPR, 2.4 vs. 3.1 months, p = 0.125; RenCR, 2.9 vs. 3.2 months, p = 0.213). Collectively, the findings suggest that the greater efficacy of Dara‐based therapy in inducing deep hematologic response is associated with more substantial cardiac improvement.

FIGURE 3.

Response kinetics of Dara‐ versus PI‐based first‐line therapy. (a) Kinetics of deep hematologic response (≥ HemVGPR) in Dara‐ versus PI‐based first‐line therapy. (b) Time to hematologic response during Dara‐ versus PI‐based first‐line therapy. (c) and (d) Kinetics of cardiac and renal response in Dara‐ versus PI‐based first‐line therapy. [Color figure can be viewed at wileyonlinelibrary.com]

3.4. Survival

A total of 1052 patients had follow‐up information. With a median follow‐up of 16.1 months (range 0.03–128 months), 229 patients (21.8%) died. The primary cause of death was cardiac‐related (72.9%), followed by infection (13.2%) and renal failure (4.9%). Regarding early mortality, the rates within 1, 3, and 6 months of diagnosis were 5.5%, 13.2%, and 16.2%, respectively. Most early deaths within 1 month were due to cardiac causes, mainly sudden cardiac arrest (30.9%), heart failure (28.6%), and malignant arrhythmia (19.0%). The median EFS was 44.4 months, while the median OS was not reached, with estimated 1‐, 3‐, and 5‐year OS rates of 80.7%, 70.8%, and 65.2%, respectively. When stratified by the 2015 European modification of the Mayo 2004 model, median EFS for stages I, II, IIIa, and IIIb was not reached, 51.9, 33.5, and 8.3 months, respectively. Median OS for stage IIIb patients was significantly shorter at 30.4 months, in contrast to stages I‐IIIa where it was not reached (p < 0.001) (Figure 4). The Mayo 2012 staging system also showed significant prognostic value (EFS: log‐rank p = 0.0049; OS: p = 0.0081), with median EFS of not reached, 50.7, 28.4, and 36.5 months for Stages 1–4, respectively (Figure 4). However, survival curves for Mayo 2012 Stage 3 and Stage 4 overlapped, suggesting limited discriminatory power between these advanced stages compared with the Mayo 2015 system.

FIGURE 4.

Event‐free survival (EFS) and overall survival (OS) according to disease stage, treatment response, and minimal residual disease status (MRD) during first‐line treatment. Kaplan–Meier curves for EFS and OS are stratified by the Mayo 2015 staging system (a, b), Mayo 2012 staging system (c, d), best hematologic response (e, f), best cardiac response (g, h), and MRD negativity (i, j) achieved with first‐line therapy. *p < 0.05, **p < 0.01, ***p < 0.001, ns = no significant difference. [Color figure can be viewed at wileyonlinelibrary.com]

Regarding hematologic response, the median EFS for patients achieving HemCR, HemVGPR, HemPR, and HemNR was not reached, not reached, 49.2 months, and 32.0 months, respectively; the median OS was not reached, not reached, 49.2 months, and 36.5 months, respectively (Figure 4). Notably, among patients who achieved a deep hematologic response, those with HemCR had significantly longer EFS (p < 0.001) and OS (p < 0.001) compared to those with HemVGPR. In terms of cardiac response, patients achieving CarCR, CarVGPR, CarPR, and CarNR had a median EFS of 65.8, 59.6, not reached, and 6.8 months, respectively. The median OS was not reached for patients with ≥ CarPR, and was 52.1 months for CarNR (Figure 4).

3.5. MRD Assessment and Prognostic Value

MRD status was evaluable in 391 patients during first‐line therapy. MRD negativity was attained in 49.1% (192/391) overall, with a significantly higher rate in the Dara‐based group than the PI‐based group (57.3% [75/131] vs. 44.8% [116/259], p = 0.024) (Table 2). Patients achieving MRD negativity had significantly longer EFS (median not reached vs. 33.4 months; p < 0.001) and improved OS (p < 0.001), although median OS was not reached in either group (Figure 4).

To determine whether MRD adds prognostic value beyond depth of hematologic response, we analyzed patients by response category. Among patients achieving ≥ HemVGPR, MRD negativity was associated with superior EFS (median not reached vs. 59.6 months, p = 0.020), whereas OS did not differ significantly (p = 0.269) (Figure 5). Similarly, among those achieving HemCR, MRD negativity identified a subgroup with significantly longer EFS than MRD‐positive patients (both not reached, p = 0.043), with no significant OS benefit (Figure 5). These findings indicate that MRD discriminates outcomes within patients who have achieved deep hematologic responses, supporting its role as a refinement tool.

FIGURE 5.

Event‐free survival (EFS) and overall survival (OS) according to MRD status. Kaplan–Meier curves are stratified by MRD negativity in the following cohorts: All patients with ≥ HemVGPR (a, b); all patients with HemCR (c, d); all patients with AL amyloidosis alone (without MM) (e, f); AL‐alone patients with ≥ HemVGPR (g, h); and AL‐alone patients with HemCR (i, j). The hematologic responses were evaluated during first‐line therapy. [Color figure can be viewed at wileyonlinelibrary.com]

To exclude MM‐related confounding, we analyzed the AL‐alone subgroup (n = 250 with MRD data). MRD negativity predicted significantly improved EFS (median not reached vs. 49.0 months, p < 0.001) but comparable OS (p = 0.061) (Figure 5). In deep responders within this subgroup, MRD negativity showed a similar EFS benefit for ≥ HemVGPR (p = 0.011) and a trend for HemCR (p = 0.056) (Figure 5). Notably, the prognostic discrimination of MRD appeared stronger in AL‐alone patients than in the overall cohort, suggesting that concomitant MM may dilute its prognostic value.

In multivariable Cox regression, MRD negativity did not retain independent significance for EFS (HR 0.619, p = 0.092) or OS (HR 0.669, p = 0.372) (Table 4), likely due to collinearity with deep hematologic response (≥ HemVGPR) and cardiac response (≥ CarPR). This indicates that MRD captures information overlapping with these established response endpoints, rather than providing fully independent prognostic value.

TABLE 4.

Univariate and multivariate Cox regression analysis of event‐free survival (EFS) and overall survival (OS) among the overall cohort with newly diagnosed AL amyloidosis.

(a) EFS
Variable	Univariate Cox regression analysis			Multivariate Cox regression analysis
	Hazard ratio	95% confidence interval	p	Hazard ratio	95% confidence interval	p
ECOG performance status > 2	3.074	2.440–3.872	< 0.001	2.365	1.293–4.328	0.005
dFLC > 180 mg/L	1.634	1.325–2.014	< 0.001	1.153	0.708–1.879	0.567
LDH > 250 U/L	1.445	1.169–1.787	< 0.001	0.993	0.588–1.678	0.980
β2‐microglobulin > 4.30 mg/L	1.634	1.301–2.050	< 0.001	0.849	0.496–1.451	0.549
gain1q positivity	1.512	1.199–1.907	< 0.001	1.764	1.050–2.964	0.032
Mayo 2015 stage IIIa/b	2.220	1.759–2.800	< 0.001	1.200	0.653–2.205	0.556
Hepatic involvement	1.459	1.077–1.978	0.015	2.368	1.115–5.025	0.025
Concomitant MM	1.745	1.424–2.139	< 0.001	1.083	0.612–1.918	0.784
Dara‐based first‐line therapy	0.688	0.535–0.884	0.004	0.534	0.273–1.046	0.067
MRD negativity during first‐line therapy	0.440	0.305–0.635	< 0.001	0.619	0.354–1.082	0.092
≥ HemVGPR during first‐line therapy	0.364	0.277–0.479	< 0.001	0.473	0.255–0.878	0.018
≥ CarPR during first‐line therapy	0.356	0.276–0.484	< 0.001	0.676	0.395–1.154	0.151

(b) OS
Variable	Univariate Cox regression analysis			Multivariate Cox regression analysis
	Hazard ratio	95% confidence interval	p	Hazard ratio	95% confidence interval	p
ECOG PS > 2	3.477	2.626–4.603	< 0.001	2.514	1.075–5.883	0.034
dFLC > 180 mg/L	1.927	1.460–2.543	< 0.001	1.464	0.745–2.876	0.268
LDH > 250 U/L	2.095	1.600–2.742	< 0.001	1.032	0.499–2.131	0.933
β2‐microglobulin > 4.30 mg/L	2.359	1.746–3.153	< 0.001	0.723	0.357–1.463	0.367
Mayo 2015 stage IIIa/b	3.841	2.685–5.496	< 0.001	1.648	0.652–4.166	0.291
Hepatic involvement	1.823	1.263–2.632	0.001	6.231	2.243–17.307	< 0.001
Concomitant MM	1.897	1.462–2.463	< 0.001	2.689	1.314–5.502	0.007
Dara‐based first‐line therapy	0.658	0.469–0.924	0.016	0.758	0.255–2.253	0.618
MRD negativity during first‐line therapy	0.291	0.152–0.559	< 0.001	0.669	0.278–1.615	0.372
≥ HemVGPR during first‐line therapy	0.256	0.171–0.382	< 0.001	0.223	0.101–0.493	< 0.001
≥ CarPR during first‐line therapy	0.159	0.100–0.253	< 0.001	0.258	0.125–0.535	< 0.001

Abbreviations: CarPR, cardiac partial response; Dara, daratumumab; dFLC, difference in involved and uninvolved free light chain; ECOG, Eastern Cooperative Oncology Group; HemVGPR, hematologic very good partial response; LDH, lactate dehydrogenase; MRD, minimal residual disease.

3.6. Prognostic Factors of Early Mortality

Univariable and multivariable logistic regression analyses were conducted to identify predictors of early mortality. The multivariable analysis revealed distinct sets of predictors for early mortality at 1, 3, and 6 months, indicating time‐dependent heterogeneity. For 1‐month mortality, elevated NT‐proBNP > 7000 pg/mL [odds ratio (OR), 4.581; 95% confidence interval (CI), 1.478–14.197; p = 0.008] and BMPC > 10% (OR, 3.336; 95% CI, 1.180–9.434; p = 0.023) were the main risk factors (Table S3). In the 3‐month mortality model, significant predictors included ECOG PS > 2 (OR, 3.624; 95% CI, 1.815–7.237; p < 0.001), NT‐proBNP > 7000 pg/mL (OR, 2.934; 95% CI, 1.394–6.175; p = 0.005), β2‐microglobulin > 4.30 mg/L (OR, 3.086; 95% CI, 1.529–6.227; p = 0.002), LDH > 250 U/L (OR, 2.443; 95% CI, 1.215–4.912; p = 0.012), and BMPC > 7.5% (OR, 3.577; 95% CI, 1.326–9.646; p = 0.012). Notably, for 6‐month mortality (Table 3), the strongest adverse prognostic factors were ECOG PS > 2 (OR, 5.561; 95% CI, 2.543–12.158; p < 0.001), NT‐proBNP > 3500 pg/mL (OR, 2.916; 95% CI, 1.185–7.177; p = 0.020), β2‐microglobulin > 4.30 mg/L (OR, 2.556; 95% CI, 1.239–5.275; p = 0.011), and hepatic involvement (OR, 4.398; 95% CI, 1.564–12.366; p = 0.005). Importantly, first‐line Dara‐based therapy was identified as an independent protective factor against 6‐month mortality (OR, 0.321; 95% CI, 0.113–0.912; p = 0.033), corroborating the significantly lower mortality rate observed in the Dara‐based group compared to the non‐Dara group (10.7% vs. 16.1%, p = 0.031).

TABLE 3.

Univariate and multivariate logistic regression analysis of early mortality in 6 months among the overall cohort with newly diagnosed AL amyloidosis.

Variable	Univariate logistic regression analysis			Multivariate logistic regression analysis
	Odds ratio	95% confidence interval	p	Odds ratio	95% confidence interval	p
ECOG performance status > 2	4.768	3.228–7.042	< 0.001	5.561	2.543–12.158	< 0.001
NT‐proBNP > 3500 pg/mL	6.836	4.344–10.757	< 0.001	2.916	1.185–7.177	0.020
cTNT > 0.060 ng/mL	7.240	3.969–13.209	< 0.001	2.400	0.925–6.232	0.072
β2‐microglobulin > 4.30 mg/L	3.091	2.054–4.652	< 0.001	2.556	1.239–5.275	0.011
LDH > 250 U/L	2.768	1.909–4.014	< 0.001	1.946	0.926–4.088	0.079
dFLC > 180 mg/L	2.379	1.609–3.518	< 0.001	1.106	0.532–2.298	0.787
BMPC > 7.5%	2.934	1.857–4.637	< 0.001	2.229	0.840–5.916	0.107
Hepatic involvement	2.631	1.604–4.314	< 0.001	4.398	1.564–12.366	0.005
Number of organs involved > 2	1.920	1.344–2.742	< 0.001	1.102	0.491–2.475	0.814
Dara‐based first‐line therapy	0.627	0.409–0.961	0.032	0.321	0.113–0.912	0.033

Abbreviations: BMPC, bone marrow plasmacytosis; cTNT, cardiac troponin I; Dara, daratumumab; dFLC, difference in involved and uninvolved free light chain; ECOG, Eastern Cooperative Oncology Group; LDH, lactate dehydrogenase; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide.

3.7. Prognostic Factors of EFS and OS

We identified several prognostic factors of interest beyond established predictors (Figure S1). Regarding organ involvement, while cardiac involvement was independently associated with adverse outcomes, concurrent cardiac and hepatic involvement defined a particularly high‐risk subgroup. These patients exhibited the shortest median EFS (12.8 months) compared to those with only cardiac involvement (33.1 months, p = 0.029), only hepatic involvement (not reached, p = 0.016), or involvement of other organs (97.9 months, p < 0.001). Similarly, median OS was significantly shorter in the dual‐involvement group (51.0 months vs. not reached vs. not reached vs. 97.9 months, p < 0.001). Cytogenetic profiling by FISH demonstrated that gain1q positivity was associated with significantly inferior EFS (median, 25.1 vs. 49.2 months; p < 0.001), while it did not significantly impact OS (median, both not reached; p = 0.107). No other cytogenetic abnormalities, including t(11;14), showed prognostic significance in our cohort. Additionally, the presence of concomitant MM in AL amyloidosis was associated with significantly inferior EFS (median, 18.6 vs. 53.5 months; p < 0.001) and OS (both not reached; p < 0.001). Among MM subgroups (CRAB‐only vs. SLiM‐only vs. CRAB+SLiM), EFS did not differ significantly (overall p = 0.234), whereas OS differed across subgroups (overall p = 0.017), with CRAB‐only patients showing inferior survival compared with SLiM‐only patients (p = 0.011). However, given the short median follow‐up (8.80 months for SLiM and 12.27 months for CRAB+SLiM), these survival data are immature and require longer observation. Furthermore, the first‐line treatment regimen emerged as another critical determinant of prognosis. Patients receiving daratumumab‐based therapy had markedly longer EFS (median not reached vs. 39.5 months; p = 0.007) and OS (median, both not reached; p = 0.029) than those receiving PI‐based regimens. However, these survival comparisons should be interpreted with caution given the substantially shorter median follow‐up in the Dara group (12.0 vs. 27.4 months) and the immaturity of OS data. Landmark analyses at 3 and 6 months consistently demonstrated a clinical benefit favoring daratumumab‐based regimens, though the differences did not reach statistical significance.

In the multivariable analysis of EFS adjusting for all covariates (Table 4a), ECOG PS > 2 (HR, 2.365; 95% CI, 1.293–4.328; p = 0.005), liver involvement (HR, 2.368; 95% CI, 1.115–5.025; p = 0.025), and gain1q positivity (HR, 1.764; 95% CI, 1.050–2.964; p = 0.032) remained independent adverse prognostic factors. Achievement of ≥ HemVGPR during frontline therapy retained significance as a protective factor (HR, 0.473; 95% CI, 0.255–0.878; p = 0.018). For OS, the multivariable Cox regression analysis (Table 4b) identified liver involvement (HR, 6.231; 95% CI, 2.243–17.307; p < 0.001), concomitant MM (HR, 2.689; 95% CI, 1.314–5.502; p = 0.007), and ECOG PS > 2 (HR, 2.514; 95% CI, 1.075–5.883; p = 0.034) as independent adverse prognostic factors. Achievement of ≥ HemVGPR (HR, 0.223; 95% CI, 0.101–0.493; p < 0.001) and ≥ CarPR (HR, 0.258; 95% CI, 0.125–0.535; p < 0.001) during frontline therapy were significantly associated with improved OS.

3.8. Prognostic Factors in Subgroups Stratified by Concomitant MM

Given the distinct clinical features of AL amyloidosis with and without concomitant MM, we performed separate prognostic analyses in these subgroups (Tables S4 and S5). In patients with concomitant MM (n = 351), multivariable analysis identified ECOG PS > 2, hepatic involvement, and β2‐microglobulin > 3.5 mg/L as independent predictors of early mortality, while hepatic involvement, ECOG PS > 2, and achievement of ≥ HemVGPR or ≥CarPR independently predicted EFS and OS. Notably, neither Mayo 2015 stage nor daratumumab‐based therapy retained independent significance in this subgroup, likely due to the dominant adverse impact of MM‐related tumor burden and organ dysfunction.

In contrast, among patients with AL amyloidosis alone (n = 719), ECOG PS > 2 was the sole independent predictor of early mortality, and ≥ CarPR during first‐line therapy emerged as the strongest independent protective factor for both EFS (HR 0.42, p = 0.005) and OS (HR 0.15, p < 0.001), with hepatic involvement additionally predicting OS (HR 3.21, p = 0.042). These findings suggest that treatment response depth, particularly cardiac response, is the primary driver of long‐term outcomes in pure AL amyloidosis, whereas concomitant MM introduces additional complexity that obscures the prognostic value of traditional AL staging systems. Notably, β2‐microglobulin loses independent prognostic significance when patients with concomitant MM are excluded, supporting its role as a marker of MM‐related tumor burden rather than AL‐specific disease activity. The divergent prognostic landscapes between these subgroups underscore the need for separate risk stratification approaches in clinical practice.

4. Discussion

This study represents the largest multicenter investigation of Chinese patients with AL amyloidosis, spanning both pre‐ and post‐daratumumab eras, and offers a comprehensive analysis of diagnosis, clinical features, treatment patterns, response kinetics, and prognosis.

The integration of daratumumab into first‐line therapy has significantly transformed the treatment landscape. Following its FDA approval in January 2021 and inclusion in China's national medical insurance in 2022, daratumumab accessibility improved markedly, with over half of AL amyloidosis patients receiving Dara‐based regimens since 2024 according to the real‐world data from this study. The ANDROMEDA trial [9] demonstrated superior efficacy of Dara‐CyBorD over CyBorD, with higher rates of deep hematologic response (HemCR 53.3% vs. 18.1%; ≥HemVGPR 78.5% vs. 49.2%) and improved cardiac (41.5% vs. 22.2%) and renal (53.0% vs. 23.9%) responses. For the high‐risk Mayo stage IIIb subgroup, Dara‐based schemes also improved hematologic responses and survival outcomes [19, 20]. Consistent with these trial findings, our real‐world study demonstrates that Dara‐based therapy induces rapid and profound hematologic responses, with significantly faster time to ≥ HemVGPR (median 1.2 vs. 1.8 months, p = 0.001) and earlier organ responses (≥ CarPR: 1.8 vs. 2.3 months; ≥ RenPR: 2.1 vs. 2.6 months), despite less favorable baseline characteristics in the Dara group. These response kinetics advantages are clinically meaningful, as early and deep hematologic response is associated with superior survival and improved organ outcomes in AL amyloidosis [21]. However, we acknowledge important limitations in interpreting the survival data. The median follow‐up for Dara‐treated patients was only 12.0 months compared with 27.4 months for PI‐treated patients, reflecting the recent shift toward Dara‐based regimens. While we observed a significant EFS benefit (p = 0.007) and a trend toward improved OS (p = 0.029), these survival comparisons are immature and require longer observation. The most robust evidence supporting Dara‐based therapy in our cohort is the significant reduction in early mortality (6‐month mortality: 10.7% vs. 16.1%, p = 0.031; multivariable OR 0.321), which is not dependent on long‐term follow‐up and aligns with the rapid response kinetics observed. Long‐term survival benefits will need to be confirmed as follow‐up matures.

Despite advancements in therapy, early mortality remains a significant challenge, with 10%–20% of patients dying within 6 months of diagnosis [22], primarily driven by cardiac involvement [22, 23, 24]. In our cohort, early mortality rates at 1, 3, and 6 months were 5.5%, 13.2%, and 16.2%, respectively. Multivariable analysis identified ECOG PS > 2, elevated NT‐proBNP, β2‐microglobulin > 4.30 mg/L, LDH > 250 U/L, and hepatic involvement as key risk factors for early mortality. Notably, first‐line Dara‐based therapy emerged as an independent protective factor, with significantly lower 6‐month mortality (10.7% vs. 16.1%, p = 0.031). These findings suggest that extreme early mortality is primarily driven by the baseline burden of malignancy and organ dysfunction, whereas the choice of effective frontline therapy becomes a critical determinant of survival by the 6‐month timepoint.

Accurate risk stratification at diagnosis is critical for guiding treatment intensity and predicting outcomes. In our cohort, both the Mayo 2012 and Mayo 2015 European modification staging systems demonstrated significant prognostic discrimination. However, the Mayo 2012 system showed limited separation between Stage 3 and Stage 4, with overlapping survival curves for EFS (median 28.4 vs. 36.5 months) and similarly non‐reached median OS, suggesting modest incremental value of Stage 4 over Stage 3 in the context of modern therapy. In contrast, the Mayo 2015 system's Stage IIIb category identified a distinctly high‐risk subgroup with markedly shortened EFS (8.3 months) and OS (30.4 months) compared with Stages I–IIIa (both not reached; p < 0.001). This observation aligns with prior reports that the NT‐proBNP threshold of 8500 ng/L defining Stage IIIb captures patients with the most advanced cardiac involvement, and supports the continued preference for the Mayo 2015 system in contemporary clinical practice and trial design.

The depth of hematologic response is a crucial determinant of patient outcomes. While current guidelines recommend ≥ HemVGPR as a treatment goal [23, 25], our findings support the superiority of HemCR over HemVGPR for EFS, OS (both p < 0.001; Figure 4) and organ response. Patients achieving HemCR demonstrated higher rates of CarCR (20.8% vs. 8.9%, p = 0.004), ≥ CarVGPR (51.1% vs. 30.6%, p < 0.001), and ≥ CarPR (77.4% vs. 51.6%, p < 0.001). In light of these significant survival and organ response benefits, we propose that the goal of first‐line therapy for AL amyloidosis should be to achieve HemCR in the contemporary era of novel agents.

The prognostic value of MRD negativity has been well established in MM [26], and it has evolved from a prognostic indicator to a validated treatment endpoint. However, the role of MRD in AL amyloidosis remains unclear, and its assessment is not yet routinely incorporated into clinical trials or practice. Previous studies have demonstrated that persistent MRD negativity is associated with improved organ response and MRD positivity with shorter progression‐free survival [27, 28, 29, 30], while these findings are based on limited patient numbers.

Our study provides the largest real‐world analysis of MRD in AL amyloidosis to date (n = 391). We demonstrate that MRD negativity is strongly associated with superior EFS across all analyzed subgroups, including the overall population, deep responders, and AL‐alone patients. These collective results underscore the prognostic relevance of MRD status in AL amyloidosis. Crucially, MRD assessment refines prognostication by distinguishing MRD‐positive patients with inferior outcomes from those with true deep remissions, supporting its routine clinical use to guide response‐adapted strategies in patients who attain a deep hematologic response. The higher MRD negativity rate with Dara‐based therapy (57.3% vs. 44.8%) supports the deeper and more durable responses achieved with anti‐CD38 monoclonal antibodies, consistent with findings in the ANDROMEDA trial. However, whether MRD‐guided maintenance strategies can further improve outcomes in AL amyloidosis requires prospective validation.

Cytogenetic profiling revealed distinct patterns in AL amyloidosis, with t(11;14), gain1q, and del13q being the most common abnormalities [31, 32, 33]. In our cohort, the most frequent cytogenetic abnormalities in patients with AL‐alone were t(11;14) (32.2%), gain1q (26.1%), and del13q (21.3%), consistent with prior reports; the most common in those with concurrent MM were gain1q (45.2%), del13q (38.1%), and t(11;14) (28.5%) (Table S1). While t(11;14) was associated with inferior outcomes in the bortezomib era [32, 33, 34, 35], daratumumab‐based regimens improved hematologic responses in this subgroup [36]. Conversely, gain1q emerged as a marker of poorer outcomes [37, 38, 39]. In our cohort, patients with gain1q had significantly inferior EFS (median, 25.1 vs. 49.2 months; p < 0.001), while OS did not differ significantly (p = 0.107). The rate of deep hematologic response was also lower in gain1q‐positive patients (HemCR, 41.8% vs. 52.3%, p = 0.028). These trends persisted in the AL‐alone subgroup, where gain1q positivity was associated with reduced EFS (median, 26.5 vs. 53.5 months; p = 0.004) and a lower HemCR rate (41.2% vs. 57.5%, p = 0.011). Further, stratification of AL‐alone patients by frontline therapy uncovered a critical distinction in our patients: gain1q21 positivity was associated with a significantly worse EFS with PI‐based regimens (median, 27.6 vs. 54.4 months; p = 0.019), but only showed a non‐significant trend with Dara‐based regimens (median, 25.1 vs. 36.5 months; p = 0.094). The adverse prognostic impact of gain1q positivity in patients receiving PI‐based regimens contrasts with previous reports that did not highlight this association, necessitating a re‐evaluation of its prognostic significance and demanding further investigation. Moreover, among patients with AL and concomitant MM, no significant differences were observed between gain1q‐positive and ‐negative groups in EFS (median, 16.9 vs. 23.0 months; p = 0.280), OS (median, 65.8 vs. 56.5; p = 0.945), or deep hematologic response (HemCR rate, 43.5% vs. 38.1%, p = 0.472). The influence of cytogenetic abnormalities was likely overshadowed by the dominant, unfavorable clinical features of concurrent MM. Collectively, these findings suggest that gain1q + patients may not fully benefit from current therapies, warranting further investigation into novel frontline immunotherapies.

Liver involvement, present in 10%–20% of AL amyloidosis patients, is associated with poor outcomes and elevated treatment‐related toxicity [40, 41, 42]. Coexistence with elevated NT‐proBNP or hyperbilirubinemia indicates an extremely poor prognosis [41]. In our cohort, 112 patients had liver involvement, with 88.4% also exhibiting cardiac involvement. Among 34 evaluable patients, those achieving HepPR had significantly longer median EFS (51.0 vs. 4.8 months, p = 0.006) but comparable OS (51.0 vs. 44.4 months, p = 0.208). Concurrent liver and cardiac involvement was linked to substantially worse survival. Multivariable analysis identified liver involvement as an independent predictor of early mortality, EFS, and OS, highlighting the need for tailored management in this high‐risk subgroup.

The prognostic landscape differs markedly between AL amyloidosis with and without concomitant MM. In our cohort, patients with concomitant MM exhibited a higher prevalence of adverse risk factors and worse baseline characteristics, consistent with prior reports that MM confers additional tumor‐related morbidity. The consistent independent prognostic value of hepatic involvement across both subgroups underscores the need for heightened vigilance in this high‐risk population, regardless of MM status.

This study has limitations, including its retrospective design, potential selection bias, and incomplete follow‐up data. Response assessment was incomplete (63.5% evaluable for hematologic response), with non‐evaluable patients having poorer baseline features—reflecting real‐world practice where sicker patients are more prone to loss to follow‐up—likely overestimating response rates; early mortality and multivariable analyses offer more robust endpoints. Additionally, the relatively short follow‐up duration, particularly for Dara‐based therapy cohorts, necessitates longer‐term studies to define the natural history of AL amyloidosis in Chinese patients.

In conclusion, this study provides a comprehensive real‐world benchmark for AL amyloidosis in China, highlighting the growing disease burden and the paradigm shift toward frontline Dara‐based therapy. Daratumumab's rapid and deep responses, coupled with its protective effect against early mortality, solidify its role as a cornerstone of first‐line treatment. We advocate for HemCR and MRD negativity as central treatment goals and identify specific high‐risk subgroups through multivariable survival analysis. These findings offer valuable insights to guide clinical practice, optimize resource allocation, and measure future advancements in AL amyloidosis management.

Author Contributions

Peng Liu L, Chunrui Li, and Peilong Lai conceived and performed research. Jing Li, Chenqi Yu, Jinghua Wang and Ning An analyzed data. Jing Li and Chenqi Yu wrote the manuscript and generated figures. Yujun Dong, Lijuan Chen, Jun Luo, Sili Wang, Quanyi Lu, Songfu Jiang, Jin Zhang, Yong Zou, Lihong Shou, Lin Shen, Junmin Chen, Li Zhang, Xuzhang Lu, Jia Feng, Hongyu Zhang, Huifang Jiang and Jihao Zhou, contributed to study design and supervised the project; Tianhong Xu, Wenjing Wang, Yang Yang, Fanjie Zhou, Weiwei Xie, Xuxing Shen and Yang Yang collected clinical data. All authors have reviewed and approved the submitted version of the manuscript.

Ethics Statement

The study adhered to the Declaration of Helsinki and was approved by the Medical Ethics Committee of Zhongshan Hospital Fudan University (B2023‐185R).

Consent

Patients consented to the use of their electronic medical records for research.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Cytogenetic abnormalities in AL patients with and without concomitant multiple myeloma.

Table S2: Comparison of baseline clinical characteristics of patients with AL amyloidosis receiving Dara‐ and PI‐based first‐line therapy.

Table S3: Univariate and multivariate logistic regression analysis of early mortality in one and three months among the overall cohort with newly diagnosed AL amyloidosis.

Table S4: Univariate and multivariate logistic regression analysis of early mortality six months in AL patients with and without concomitant multiple myeloma.

Table S5: Univariate and multivariate Cox regression analysis of event‐free survival (EFS) and overall survival (OS) in AL patients with and without concomitant multiple myeloma.

Table S6: Comparison of baseline clinical characteristics of patients with AL amyloidosis with or without evaluable hematologic response data.

Figure S1: Event‐free survival (EFS) and overall survival (OS) according to prognostic factors. Kaplan–Meier curves for EFS and OS are stratified by (a, b) organ involvement type, (c, d) baseline gain1q status, (e, f) presence of concomitant multiple myeloma (MM), and (g, h) first‐line therapy type (daratumumab [Dara]‐based vs. proteasome inhibitor [PI]‐based). Landmark analyses for EFS and OS compare Dara‐based and PI‐based first‐line therapy at (i, j) 3 months and (k, l) 6 months.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.