Meaningful change in patient-reported outcomes after CAR-T for multiple myeloma: differences by race and ethnicity

TO THE EDITOR:

Chimeric antigen receptor T-cell therapy (CAR-T) has revolutionized treatment for relapsed/refractory multiple myeloma (RRMM). In clinical trials and real-world practice, CAR-T remarkably improves progression-free and overall survival relative to other therapies.1, 2, 3, 4 CAR-T trials also demonstrated meaningful improvements in patient-reported outcomes (PROs).^5,6 PROs are health information reported directly by patients without interpretation by another person⁷ and include health-related quality of life (HRQOL) and symptom burden. PROs are a high priority in the context of prolonged remission after CAR-T and are key outcomes in clinical trials.⁷ However, few real-world studies of patients with RRMM who received CAR-T have examined PROs, and no studies have evaluated differences by race and ethnicity, despite known disparities in MM incidence and care.⁸ This study compared meaningful changes in PROs among patients with RRMM receiving standard-of-care CAR-T by race and ethnicity using minimally important differences (MIDs), which represent the smallest numerical change that is meaningful for how patients feel.

Advarra Institutional Review Board deemed the protocol exempt/low risk (protocol no. Pro00065891). Eligibility criteria were: (1) ≥18 years old, (2) scheduled for standard-of-care CAR-T for RRMM, (3) able to speak and read English or Spanish, (4) without psychiatric or neurologic diagnoses that could preclude participation, and (5) able to provide informed consent. Participants completed PROs electronically via Research Electronic Data Capture in English or Spanish (individual preference) prelymphodepleting chemotherapy (baseline) and 90 days after CAR-T (D90). Baseline was completed a median 6 of days prelymphodepleting chemotherapy (range, 1-91). The median window around D90 was 0 days (range, 0-4). Participants were compensated US $10 per assessment. Participants self-reported their demographics. Clinical characteristics, CAR-T information, and safety/clinical outcomes were abstracted from medical records. This study was conducted in accordance with the Declaration of Helsinki.

The 27-item Functional Assessment of Cancer Therapy-General (FACT-G) assessed overall HRQOL and 4 subscales: physical, social, emotional, and functional well-being.⁹ Higher scores indicated better HRQOL. Evidence-based thresholds were applied to identify clinically low individual scores (overall HRQOL ≤ 62, physical well-being ≤ 15, social well-being ≤ 16, emotional well-being ≤ 13, and functional well-being ≤ 11) and group-level averages (overall HRQOL ≤ 70, physical well-being ≤ 18, social well-being ≤ 19, emotional well-being ≤ 15, and functional well-being ≤ 14). MIDs were 4-point changes for overall HRQOL and 2-point changes for each subscale.¹⁰

The 31-item PRO Measurement Information System (PROMIS)-29+2 Profile v2.1 assessed fatigue, depression, anxiety, sleep disturbance, pain interference (with daily life), physical function, social function, cognitive function, and pain intensity.¹¹ Participants rated their pain intensity from 0 to 10. For all other domains, raw scores were converted to T-scores (normative mean = 50, standard deviation = 10). Higher scores indicated worse fatigue, depression, anxiety, sleep disturbance, pain interference (mild, 56-60; moderate, 61-70; severe > 70), and pain intensity (mild, 1-2; moderate, 3-5; severe ≥ 6). Lower scores indicated worse physical, social, and cognitive function (severe < 30; moderate, 30-39; mild, 40-44). MIDs were 2-point changes for pain intensity and 5-point changes for all other domains.¹²

Participants were categorized as non-Hispanic White or as racial and ethnic minority (minority) based on self-report. χ², Fisher exact tests, independent sample t tests, and Wilcoxon rank-sum tests were used to compare participant characteristics and PROs by race and ethnicity. For each participant, PRO changes from baseline to D90 were categorized as meaningfully improved, no meaningful change, or meaningfully worsened using MIDs. χ² or Fisher exact tests were used to examine whether the proportion of participants with meaningful change differed by race and ethnicity. Multinomial logistic regression was conducted to confirm differences while controlling for a priori high-risk/priority characteristics, including age, marital status, Eastern Cooperative Oncology Group performance status, high-risk cytogenetics, extramedullary disease, high bone marrow burden, and prior anti–B-cell maturation antigen therapy. Analyses were conducted in R version 4.1.2 (lcmm and lme4 packages). P < .05 indicated significance.

Of 161 eligible patients approached, 99 enrolled (61% consent rate) (supplemental Figure 1). The top reasons for nonparticipation were “not interested in research” and “overwhelmed prior to CAR-T.” Three participants died, 2 withdrew, and retention was high (n = 89, 93%). Participants were an average of 67 years old and 61% male, with a median of 5 prior lines of therapy (Table 1). Half received idecabtagene-vicleucel and ciltacabtagene-autoleucel, each. One-third (32%) were a minority (n = 16 non-Hispanic Black, n = 15 Hispanic, n = 1 “other”). A larger proportion of participants belonging to the minority subgroup had an annual household income <$40 000 (46% vs 15%) and were employed (31% vs 9%). Best overall response rate was 88%, and 65% had a complete response or better (supplemental Table 1).

Table 1.

Participant demographic and clinical characteristics overall and by race and ethnicity

	Overall (N = 99)	Non-Hispanic White (n = 67)	Racial and ethnic minority (n = 32)	P value∗
Age, mean (SD)	67.17 (9.62)	68.38 (9.74)	64.63 (9.00)	.11
Range	46-84	46-84	46-82
Sex, n (%)				.41
Female	39 (39)	24 (36)	15 (47)
Male	60 (61)	43 (64)	17 (53)
Race/ethnicity, n (%)				—
Non-Hispanic/Latino White	67 (68)	67 (100)	—
Non-Hispanic/Latino Black	16 (16)	—	16 (50)
Hispanic/Latino	15 (15)	—	15 (47)
Other	1 (1)	—	1 (3)
Marital status, n (%)
Married	79 (80)	53 (79)	26 (81)	>.99
Not married	20 (20)	14 (21)	6 (19)
Educational status, n (%)				.12
College degree or more	53 (54)	40 (60)	13 (41)
Less than college degree	46 (47)	27 (40)	19 (59)
Household income, n (%), $				.004
<40 000	21 (26)	8 (15)	13 (46)
≥40 000	61 (74)	46 (85)	15 (54)
Missing	17	13	4
Employment status				.01
Employed	16 (16)	6 (9)	10 (31)
Not employed	83 (84)	61 (91)	22 (69)
CAR-T treatment				>.99
Idecabtagene-vicleucel	49 (49)	33 (49)	16 (50)
Ciltacabtagene-autoleucel	50 (51)	34 (51)	16 (50)
CCI, mean (SD)	1.33 (1.55)	1.28 (1.56)	1.46 (1.55)	.49
Range	0-7	0-7	0-6
Missing	6	2	4
High marrow burden, n (%)	22 (24)	15 (23)	7 (25)	>.99
Missing	6	2	4
Extramedullary disease, n (%)	15 (15)	12 (18)	3 (9)	.37
Missing	1	1	0
ECOG performance status ≥2, n (%)	9 (9)	6 (9)	3 (9)	>.99
High-risk cytogenetic abnormalities,†n (%)
Any	35 (37)	24 (37)	11 (38)	>.99
del(17p)	22 (23)	14 (22)	8 (28)	.71
t(4;14)	8 (9)	8 (12)	0	.06
t(14;16)	8 (9)	4 (6)	4 (14)	.25
Missing	5	2	3
Prior therapies, median (range)	5 (3-10)	5 (3-9)	5 (3-10)	.62
Prior anti-BCMA treatment, n (%)	7 (7)	5 (8)	2 (6)	>.99
Pentarefractory, n (%)	14 (14)	8 (12)	6 (19)	.55

High marrow burden was defined as ≥50% CD138⁺ plasma cells in pretreatment bone marrow core biopsy. High-risk cytogenetics were defined as the presence of del(17p), t(4;14), and/or t(14;16) at any time before CAR-T infusion. Pentarefractory was defined as refractory to immunomodulatory agents, proteasome inhibitors or anti-CD38 antibodies. Percentages may not sum to 100 due to rounding.

BCMA, B-cell maturation antigen; CCI, Charlson Comorbidity Index; ECOG, Eastern Cooperative Oncology Group; SD, standard deviation.

^∗Wilcoxon rank-sum tests for age, prior therapies, and CCI, Fisher exact tests for categorical variables with cell counts <5, and χ² tests for all other categorical variables.

^†Categories are not mutually exclusive.

Average PRO scores were within normal ranges at baseline and D90 for both racial and ethnic groups (supplemental Table 2). As exceptions, PROMIS pain intensity was moderate at baseline and mild at D90 for both groups, and PROMIS physical function was mildly impaired for both groups at baseline and D90. There were no racial and ethnic differences in average PRO scores at either time point. At baseline, PROs with the largest proportion of participants reporting meaningful deficits were PROMIS pain intensity (83%), PROMIS pain interference (58%), and PROMIS physical function (58%), and at D90 were PROMIS pain intensity (65%), PROMIS physical function (57%), and PROMIS social function (45%) (supplemental Figures 2-4). There were no racial and ethnic differences in the proportion of participants reporting meaningful PRO deficits at either time point.

PROs with the largest proportion of participants reporting meaningful improvement from baseline to D90 were FACT-G overall HRQOL (49%), FACT-G physical well-being (48%), FACT-G emotional well-being (46%), and PROMIS pain interference (44%). A larger proportion of minority participants had meaningfully improved PROMIS fatigue (47% vs 26%; P = .01) and meaningfully worsened PROMIS fatigue (27% vs 16%; P = .03) (Figure 1A). A larger proportion of minority participants had meaningfully improved PROMIS pain intensity (50% vs 19%; P < .001) (Figure 1B) and meaningfully worsened PROMIS social function (40% vs 20%; P = .05) (Figure 1C). There were no other racial and ethnic differences in meaningful PRO changes (supplemental Figures 5 and 6). Findings were consistent in multinomial logistic regression analyses.

Figure 1.

Proportion of participants with meaningfully improved, no meaningful change, and meaningfully worsened PROs from baseline to D90 by race and ethnicity, NHW vs minority. (A) PROMIS fatigue, (B) PROMIS pain intensity, (C) PROMIS social function. MIDs were 5-point changes for fatigue and social function and 2-point changes for pain intensity. ∗Significantly different between racial and ethnic groups. NHW, non-Hispanic White.

In summary, most patients with RRMM receiving standard-of-care CAR-T had meaningfully improved or unchanged PROs from before CAR-T infusion to D90 after infusion. Almost half had meaningfully improved overall HRQOL, physical well-being, emotional well-being (FACT-G), and pain interference (PROMIS), indicating a substantial proportion of CAR-T recipients perceive meaningful benefits for important patient-centered outcomes. A larger proportion of minority participants had meaningful change in fatigue (PROMIS), both improvement and worsening, whereas most non-Hispanic White participants had no meaningful change. Fatigue is one of the most ubiquitous symptoms after CAR-T,¹³ and our findings suggest CAR-T may especially affect fatigue among minority patients. Future studies should examine predictors of meaningful changes in fatigue to identify potentially modifiable protective and risk factors that could inform approaches to supportive care. In addition, a larger proportion of minority participants had meaningfully improved pain intensity (PROMIS), suggesting minority patients may derive unique benefit from CAR-T. However, minority participants were also more likely to report meaningfully worsened social function (PROMIS), which warrants further investigation.

This study expands prior work showing racial and ethnic differences in real-world safety and efficacy of CAR-T for RRMM but no subsequent differences in survival.^14,15 Limitations include a 61% enrollment rate and a single time point post–CAR-T infusion. The sample included a relatively small number of minority participants, which necessitated the grouping of all minorities and impeded more nuanced subgroup comparisons. Larger studies with longer follow-up and more racially and ethnically diverse samples are needed to confirm these findings and allow for comparisons between minority subgroups.

Conflict-of-interest disclosure: L.B.O. obtained research funding from the National Institutes of Health, National Cancer Institute and the Department of Defense. B.D.G. provided consultancy for Janssen; reports advisory board/honoraria from Elly Health; has patents, royalties, or other intellectual properties; and reports receiving licensing fees from Moffitt Cancer Center. T.L.C. obtained research funding from the National Institutes of Health, National Cancer Institute, and National Institute of Diabetes and Digestive and Kidney Diseases. R.C.B. reports serving on the advisory board of Janssen, Bristol Myers Squibb, Pfizer, and GSK; and receives research funding from Bristol Myers Squibb, Janssen, AbbVie, Karyopharm, and Regeneron. K.H.S. receives honoraria from AbbVie, Adaptive, Amgen, Bristol Myers Squibb, Janssen, Karyopharm, Kite/Arcellx, Regeneron, Sanofi, Sebia, and Takeda; and receives research funding to the institution outside the submitted work from AbbVie, Pfizer, and Karyopharm. B.B. provided consultancy for Pfizer Pharmaceutics, Janssen Pharmaceutics, Oncopeptides, Kite Pharmaceuticals, and Sanofi Pharmaceutics; and receives honoraria from AbbVie. A.F.G.-C. reports serving on the advisory board of Bristol Myers Squibb, Cellectar, Janssen, Pfizer, and Sanofi; and serves on the speakers’ bureau of Amgen, Janssen, Pfizer, and Sanofi. F.P. obtained research funding from the National Institutes of Health, National Cancer Institute. M.A. reports serving on the advisory board of Janssen, Bristol Myers Squibb, and Sanofi; and obtained research funding from Janssen, Bristol Myers Squibb, Pfizer, and Sanofi. C.L.F. receives honoraria/consulting fees from Bristol Myers Squibb, Seattle Genetics, Celgene, AbbVie, Sanofi, Incyte, Amgen, ONK Therapeutics, and Janssen; and obtained research funding from Bristol Myers Squibb, Janssen, and Roche/Genentech. O.C.P. receives honoraria/consulting fees from Legend Biotech Inc, Bristol Myers Squibb, and Janssen Biotech Inc. T.N. receives clinical trial support from Novartis and Karyopharm. F.L.L. reports receiving advisory/consulting fees A2, Allogene, Amgen, bluebird bio, Bristol Myers Squibb/Celgene, Calibr, Caribou Biosciences, Cellular Biomedicine Group, Cowen, Daiichi Sankyo, EcoR1, Emerging Therapy Solutions, GammaDelta Therapeutics, Gerson Lehrman Group, Iovance, Kite Pharma, Janssen, Legend Biotech, Novartis, Sana, Takeda, Wugen, Umoja, and Pfizer; reports contracts for service from Kite Pharma (institutional), Allogene (institutional), CERo Therapeutics (institutional), Novartis (institutional), bluebird bio (institutional), Bristol Myers Squibb (institutional), National Institutes of Health, National Cancer Institute, and the Leukemia and Lymphoma Society; reports patents, royalties, and other intellectual property; several patents are held by the institution in F.L.L.’s name (unlicensed) in the field of cellular immunotherapy; and reports education or editorial activity from Aptitude Health, American Society of Hematology, BioPharma Communications CARE Education, Clinical Care Options Oncology, Imedex, and the Society for Immunotherapy of Cancer. H.S.L.J. reports consulting from SBR Biosciences; and obtained research funding from Kite Pharma. D.K.H. reports consulting or advisory role from Bristol Myers Squibb, Janssen, Legend Biotech, Pfizer, Kite Pharma/Gilead Sciences, AstraZeneca, and Karyopharm; and obtained research funding from Bristol Myers Squibb, Janssen, Karyopharm, Kite Pharma, Adaptive Biotech, and the National Institutes of Health, National Cancer Institute. L.C.P. obtained research funding from Bristol Myers Squibb, Karyopharm, Janssen, the National Institutes of Health, National Cancer Institute, and the Department of Defense; and reports support for attending Tandem American Society for Transplantation and Cellular Therapy/Center for International Blood & Marrow Transplant Research Annual Meeting (2023). The remaining authors declare no competing financial interests.