From toxicity to immunity: Evaluating CAR-T and conventional therapies in multiple myeloma through quality-adjusted life year (QALY) outcomes Authors: Guluzar Gulnur Itez, Asuman Sunguroğlu Published: 2026 ABSTRACT AbstractMultiple myeloma (MM) remains characterized by recurrent relapse and cumulative treatment burden despite major therapeutic advances. This study compared conventional stepwise regimens with CAR-T cell therapy in terms of QoL, toxicity, cost, and ethical considerations within a model-based comparative framework. This comparative pharmacoeconomic modeling analysis was based on published clinical trial data. Kaplan-Meier survival curves were digitized to estimate mean overall survival (OS) and progression-free survival (PFS) using area-under-the-curve integration. Treatment costs were calculated based on published pricing and trial-derived treatment durations. A simple and transparent ECOG-based utility model was developed to enable clinicians, researchers, and health policy authorities to estimate quality-adjusted life years (QALY) and incremental cost-effectiveness ratios (ICER) across regimens spanning heterogeneous treatment lines. The present analysis synthesizes outcome data from 19 pivotal trials and real-world cohorts (N = 7,793 patients). Among heavily pretreated patients, CAR-T therapy approximately doubled OS and PFS compared with other late-line regimens and yielded higher QALY estimates with lower cumulative treatment burden. Daratumumab-based combinations improved outcomes but reached very high costs (∼USD 1 million/patient), while carfilzomib-based regimens remained costly but clinically important for high-risk disease. VMP represented a practical lower-cost option for transplant-ineligible or resource-limited patients. In treatment-line–stratified analysis, later-line/RRMM regimens had higher median ICER-equivalent values than early-line/induction regimens (USD 739,050/QALY vs USD 218,687/QALY; 3.4-fold higher), whereas CAR-T regimens showed a median ICER-equivalent estimate of USD 299,723/QALY, approximately 2.5-fold lower than later-line/RRMM conventional regimens. Within this model-based comparative framework, CAR-T provided substantial survival and quality-adjusted outcome gains after three or more prior therapy lines, with promising potential for earlier use. Despite its upfront single-payment structure, CAR-T did not show a disproportionate ICER-equivalent burden compared with later-line conventional regimens. However, limited global availability raises ethical concerns regarding access, infrastructure, reimbursement, and equity. Further studies incorporating longer follow-up and patient-level QoL data are needed to refine its role in multiple myeloma care.Keywords: multiple myeloma, CAR-T cell therapy, drug toxicity, quality-adjusted life years (QALY), cost-effectiveness, Eastern Cooperative oncology group (ECOG) performance status, value-based medicine 1. Introduction Multiple myeloma (MM) is a multisystemic malignancy characterized by the uncontrolled proliferation of plasma cells in the bone marrow, leading to disruption of bone, renal, and immune integrity. The monoclonal immunoglobulins produced by neoplastic plasma cells correlate with disease severity and alter the hematopoietic balance by disturbing the bone marrow microenvironment. Clinically, the disease defined by anemia, hypercalcemia, bone lesions, and renal impairment (CRAB criteria), presents with variable progression and therapeutic response due to its heterogeneous nature. 1 Over the past two decades, therapeutic strategies have evolved from conventional chemotherapy to proteasome inhibitors, immunomodulatory agents, monoclonal antibodies, bispecific antibodies, antibody-drug conjugates, and most recently, CAR-T cell therapies.2–4 These advances have markedly extended survival, yet introduced new dimensions requiring assessment toxicity, quality of life, and cost. Today, therapeutic success is measured not only by survival outcomes but also by quality-of-life indicators, cost-effectiveness, and equitable access. 5 Despite numerous pharmacological, immunological and technical interventions developed over the past two centuries, myeloma continues to relapse and develop resistance to therapy.6,7 In recent years, CAR-T cell therapy has introduced a paradigm shift by re-engineering the patient’s immune system to target malignant plasma cells. Although applied in advanced disease stages, CAR-T therapy has demonstrated significant improvements in overall survival (OS) and progression-free survival (PFS), stimulating trials that explore its potential in earlier treatment lines. Current research focuses on minimizing toxicity, shortening manufacturing time, and enhancing antigen-targeting capacity.8–14 This study provides a comprehensive overview of the clinical, economic, and ethical dimensions of conventional and CAR-T therapies. It also introduces a simple pharmacoeconomic model based on ECOG performance scores, designed to allow clinicians, non-economist researchers, policymakers, and health authorities to easily evaluate cumulative toxicity across treatment lines. This approach serves not only as a numerical tool for cost-effectiveness but also as a patient-centered, ethically grounded, and translational framework for evaluating therapeutic value. 2. Myeloma therapies The treatment of multiple myeloma has undergone a long evolution, progressing from the early historical practices of the 1800s, including rhubarb, opium, and supportive measures such as corset use, to today’s sophisticated therapeutic approaches. Over time, the introduction of corticosteroids, alkylating agents, proteasome inhibitors (PIs), immunomodulatory drugs (IMiDs), and monoclonal antibodies (mAbs) has substantially reshaped the therapeutic paradigm. More recent advances include bispecific antibodies, antibody-drug conjugates (ADCs), and CAR-T cell therapies.3,15 Contemporary myeloma therapy is individualized according to disease stage, cytogenetic risk profile, performance status, and comorbidities. Patients are typically stratified by transplant eligibility. In eligible candidates, high-dose melphalan followed by autologous stem cell transplantation (ASCT) remains the gold standard. Eligibility is determined not only by chronological age but by biological fitness, performance status, and organ function. Patients with adequate cardiac, renal, hepatic, and pulmonary reserve (ECOG ≤2) are considered suitable for transplantation, whereas frail or organ-compromised individuals are treated with more tolerable regimens aimed at long-term disease control. 16 As summarized in Figure X, MM treatment typically follows a sequential strategy including induction therapy, ASCT in eligible patients, maintenance therapy, and increasingly diversified options in the relapsed/refractory setting, such as next-generation PIs and IMiDs, monoclonal and bispecific antibodies, ADCs, and CAR-T cell therapies 17 (see Figure 1). Simplified treatment algorithm for multiple myeloma according to transplant eligibility. Eligibility for ASCT is based on biological rather than chronological age. Transplant-ineligible patients receive non-transplant regimens aiming to sustain long-term disease control and quality of life. The figure highlights that ASCT eligibility is determined by a multi-dimensional functional assessment encompassing organ reserve, performance status, comorbidities, and psychosocial capacity rather than age alone, and that this single eligibility decision shapes the entire downstream sequence of therapy and its cumulative cost-toxicity trajectory (Created with BioRender.com). 2.1. Induction therapies Induction therapy represents the first step in the management of multiple myeloma. Its primary goals are to reduce tumor burden, achieve deep remission, and prepare eligible patients for autologous stem cell transplantation (ASCT).2,3 Currently, the VRd regimen (bortezomib, lenalidomide, dexamethasone) is regarded as the gold standard for transplant-eligible patients in both Europe and the United States, achieving high response rates (≥VGPR above 70%) and prolonged progression-free survival (PFS). 18 The KRd regimen (carfilzomib, lenalidomide, dexamethasone) offers deeper responses in patients with high-risk cytogenetic profiles. 19 The addition of CD38-targeted monoclonal antibodies has further enhanced induction efficacy. In the GRIFFIN trial, Dara-VRd significantly improved minimal residual disease (MRD) negativity and response depth compared with conventional VRd. 20 Similarly, Dara-KRd has shown promising results in high-risk patients. 21 For transplant-ineligible, elderly, or highly comorbid patients, more tolerable regimens such as D-Rd (daratumumab, lenalidomide, dexamethasone) and VRd-lite are preferred. 22 The D-VMP regimen (daratumumab, bortezomib, melphalan, prednisone) became the standard of care in this population following the ALCYONE trial, in which daratumumab addition significantly improved PFS and OS18,23 The Rd regimen (lenalidomide, dexamethasone) remains the mainstay for frail or very elderly patients.24,25 Common adverse events include peripheral neuropathy related to bortezomib, cardiotoxicity associated with carfilzomib, and hematologic toxicity from lenalidomide. In contrast, daratumumab is generally well tolerated, with only mild infusion-related reactions such as fever, chills, or nasal congestion. 26 The incorporation of anti-CD38 antibodies enhances treatment depth without increasing toxicity, markedly improving MRD negativity rates and extending PFS (see Table 1 and Appendix Table A1). Table 1.Current induction regimens and their main adverse effects.RegimenPatient groupMain advantagesMain toxicitiesVRdTransplant-eligible or ineligible (standard-risk)High response rate, prolonged PFSPeripheral neuropathy, hematologic toxicityD-VRdTransplant-eligibleIncreased MRD negativity, well toleratedInfusion reactions, hematologic toxicityKRdYounger, fit, high-risk cytogeneticsDeeper response, increased MRD negativityCardiotoxicity, hypertensionD-KRdTransplant-eligible, high-riskDeep and durable MRD negativityCardiac and hematologic toxicityD-VMPTransplant-ineligible, elderlyOS benefit, effective in elderlyHematologic toxicity, infectionsD-RdTransplant-ineligible, frail or elderlyLong OS, oral administration convenienceImmunosuppression, hematologic toxicityRdFrail or relapsed patientsOral administration, convenienceMyelosuppression, fatigueVRd-LiteElderly, frail, or comorbidBetter tolerabilityPotential lower efficacyV: Velcade(Bortezomib), R: Revlimid(Lenalidomide), D: Darzalex(Daratumumab), K: Kyprolis(Carfilzomib), M: Melphalan, P: Prednisone, d: Dexamethasone. MRD: Minimal Residual Disease. OS=Overall Survival, PFS: Progression Free Survival. Current induction regimens and their main adverse effects. V: Velcade(Bortezomib), R: Revlimid(Lenalidomide), D: Darzalex(Daratumumab), K: Kyprolis(Carfilzomib), M: Melphalan, P: Prednisone, d: Dexamethasone. MRD: Minimal Residual Disease. OS=Overall Survival, PFS: Progression Free Survival. 2.2. Autologous stem cell transplantation Autologous stem cell transplantation (ASCT) is a therapeutic procedure in which the patient’s own hematopoietic stem cells are reinfused following high-dose melphalan conditioning. For many years, ASCT has been regarded as the standard of care for transplant-eligible patients with multiple myeloma. The process includes several key steps: clinical preparation; stem-cell mobilization using cyclophosphamide and granulocyte colony-stimulating factor (G-CSF, e.g., filgrastim); CD34+ cell collection by apheresis; cryopreservation of the harvested cells in 10% dimethyl sulfoxide (DMSO); high-dose melphalan conditioning chemotherapy; and subsequent reinfusion of the stem cells. Once reinfused, the stem cells home to the bone-marrow microenvironment, initiate proliferation, and restore hematopoietic and immune functions within approximately 2-4 weeks 27 (see Figure 2). Schematic representation of the autologous stem cell transplantation (ASCT) process. The procedure comprises sequential stages: patient preparation, hematopoietic stem cell mobilization using cyclophosphamide (Day 1) and granulocyte colony-stimulating factor (G-CSF, Days 4-9), peripheral stem cell collection through apheresis (Days 9-10), and cryopreservation of harvested stem cells in 10% DMSO. Following a resting interval, high-dose melphalan conditioning (Day 15) eradicates residual myeloma cells and prepares the bone marrow niche for reinfusion. Autologous CD34+ cells are then reinfused (Day 17). After transplantation, the patient is monitored in a protective environment to prevent infection during the neutropenic phase. Engraftment occurs as reinfused stem cells home to the marrow, proliferate, and reconstitute hematopoiesis and immune function within approximately 2-4 weeks. The figure provides a stepwise visual reference for the standard ASCT workflow as it is applied in transplant-eligible multiple myeloma patients (Created with BioRender.com). When combined with modern induction regimens such as VRd or Dara-VRd, ASCT further reduces tumor burden, deepens response, and significantly prolongs progression-free survival (PFS).28,29 However, the procedure is associated with substantial biological stress and transient toxicities. During mobilization and conditioning, increases in cytokines, oxidative stress, and mucosal injury are commonly observed.30,31 Administration of G-CSF can trigger transient inflammatory responses; some studies have reported p53-mediated DNA damage and reduced engraftment efficiency.32,33 Melphalan induces DNA damage and apoptosis in rapidly dividing cells, contributing to mucositis, pancytopenia, and infection risk, and has also been linked to long-term bone fragility and secondary malignancies.34–36 DMSO, used during cryopreservation, may cause transient cardiovascular or gastrointestinal side effects during reinfusion. 37 Patients are closely monitored for infections during neutropenic and immunosuppressive phases, with appropriate supportive therapy provided. Despite these challenges, long-term survival outcomes following ASCT remain remarkable. Transplantation enhances the depth of hematologic remission and nearly doubles median overall survival.18,23,24,38 Although transplantation-related toxicities may create clinical uncertainty 39 and novel agents are showing increasing efficacy in transplant-ineligible populations.18,24 ASCT continues to offer unmatched benefits in survival, quality of life, and durable disease control when applied to appropriately selected patients. 2.3. Maintenance therapies The primary goal of maintenance therapy is to sustain remission achieved after first-line treatment or transplantation and to suppress residual malignant plasma cells. Currently, lenalidomide remains the standard maintenance agent, while bortezomib or monoclonal antibodies such as daratumumab represent effective options for high-risk patients. 40 Despite prolonged maintenance, the emergence of drug-resistant clones and an immunosuppressive bone marrow microenvironment often set the stage for inevitable relapse.6,7 Upon relapse, treatment selection depends on the patient’s overall condition, comorbidities, cytogenetic risk profile, and prior treatment response. In the relapsed/refractory (R/R) setting, second-line regimens are typically employed, which may involve introducing a new drug class or modifying the existing protocol 2 . However, with successive relapses, therapeutic efficacy diminishes due to clonal evolution, cumulative toxicity, and emerging resistance mechanisms. At this stage, less effective agents are discontinued or replaced with alternatives, while each subsequent line of therapy becomes progressively more complex and less tolerable.41–43 2.4. Relapsed/refractory phase In the relapsed/refractory (R/R) phase, proteasome inhibitors (bortezomib, carfilzomib, ixazomib), immunomodulatory agents (lenalidomide, pomalidomide), CD38 monoclonal antibodies (daratumumab, isatuximab), SLAMF7-targeted antibodies (elotuzumab), and newer modalities such as bispecific antibodies and antibody-drug conjugates (ADCs) are used sequentially. 2 Each treatment line provides temporary disease control, yet relapse remains almost inevitable. This resistance process is linked to (1) the persistence of refractory clones, (2) the immunosuppressive myeloma microenvironment, and (3) the survival of stem-like cell populations.6,43,44 Over time, clonal evolution, cumulative toxicity, and declining organ reserve progressively reduce therapeutic efficacy. Ineffective agents are therefore discontinued or replaced with drugs of different classes. Minimal residual disease (MRD) positivity and high-risk cytogenetic profiles are key indicators for early treatment modification.45,46 Prolonged therapy imposes biological, psychological, and economic burdens on patients. 47 Exhaustion of hematopoietic reserve leads to progenitor suppression and pancytopenia. 48 Immunomodulatory and antibody-based therapies suppress T/NK-cell function, increasing the risk of hypogammaglobulinemia and opportunistic infections.49,50 Moreover, microenvironmental disruption reduces stromal support, enhances osteoclast activity, and weakens the hematopoietic niche. Consequently, each new line of therapy not only reduces tumor burden but also erodes biological resilience, immune defense, and overall quality of life.27,51 Advanced treatment lines impose substantial toxicity not only on the hematopoietic system but also on the liver, kidneys, and central nervous system. Hepatic enzyme induction, impaired renal elimination, and neuro-inflammatory changes further compromise treatment tolerance and drug metabolism. This multi-organ load is one of the major factors limiting the clinical benefit of combination regimens in late-line therapy.52–54 The resulting cycle of cumulative exhaustion progressively diminishes efficacy and renders the patient increasingly fragile (see Figure 3). Schematic representation of disease progression from premalignant stages (MGUS, SMM) to active multiple myeloma, and the cumulative impact of sequential therapy lines. Each treatment line contributes to cumulative toxicity, bone marrow damage, and declining patient tolerance, ultimately leading to poor outcomes. The figure conveys that, beyond the natural disease trajectory, sequential lines of therapy entrap the patient within a hospital–drug–toxicity cycle whose multi-organ burden ultimately defines myeloma as a fatal, incurable disease (Created with BioRender.com). Although multiple myeloma does not progress as aggressively as acute leukemias, its chronic and relentless nature gradually exhausts patients biologically, psychologically, and economically. Since the 19th century, therapeutic advances have dramatically extended survival, allowing many patients today to live five years or longer after diagnosis. Yet this progress comes at a cost: with each subsequent line of therapy, progression-free survival (PFS) shortens, cumulative toxicity increases, and survival curves narrow. By occupying the bone marrow niche, myeloma not only disrupts hematopoiesis but also destabilizes systemic homeostasis. Anemia, immune deficiency, bone destruction, and organ dysfunction often reduce quality of life more than they contribute directly to mortality.47,55 In this study, we comprehensively evaluated the clinical and economic impact of multiple myeloma treatment regimens by comparing estimated costs and mean overall survival (OS) values derived from published Kaplan-Meier curves. To quantitatively assess quality of life, quality-adjusted life years (QALYs) were calculated. These QALY values were not extracted directly from prior literature but were instead modeled using ECOG performance scores and survival data, in accordance with the study’s central hypothesis (see Table 2, Supplemental Table S1, Supplemental Table S2 and see section “4.Quality-Adjusted Life Year (QALY) and ECOG-Based Modeling”). Table 2.Estimated costs, mean overall survival (OS), and model-derived QALY values for major myeloma therapy regimens used in clinical practice.TherapySetting/LinesMean OS (Months)*Mean PFS* (Months)Cost** USDQALY***Source (Trial/Cohort; N)Rd 24 First Line/Frail5740995,924 USD3.6195MAIA: N=737D-Rd 24 First Line/TIE7054.11,924,803 USD4.434MAIA: N=737D-VRd38,56First Line/TE159.1283.752,184,110 USD9.9874Cohort Study: N=326VRd38,56First Line/TE128.937.98508,940 USD7.145Cohort Study: N=1000KRd 57 2-5th Line56.923795,518 USD2.608ATLAS: N=180D-VMP18,23First Line/TIE8766.7766,416 USD4.9778ALCYONE: N=706OCTANS: N=220VMP18,24First Line/TIE6842.472,731 USD3.0866ALCYONE: N=706OCTANS: N= 220PVd 58 2-4th+ Lines4324633,028 USD0.7478 a OPTIMISMM: N=599Vd 58 2-4th+ Lines381870,333 USD0.6727 a OPTIMISMM: N=599D-Kd59,602-4th Lines61372,153,071 USD2.103CANDOR: N=466D-Kd=399D-Vd59–622-4+Lines6656611,243 USD1.9338CASTOR: N=498D-Pd59,60,632-4th Lines73321,176,157 USD2.4164D-Pd: N=399Kd59,602-4th Lines5520935,796 USD1.7618CANDOR: N=466Elo-Rd 64 2-4th Lines47.131995,764 USD0.9655ELOQUENT-2: N=646Elo-Pd 42 3-5th+ Lines3422835,752 USD0.8139ELOQUENT-3: N=117Elo-PVd 65 2-10th Lines3518700,628 USD0.875PHASE 2: N=48Isa-Kd59,662-4th Lines54341,700,907 USD2.511IKEMA: N=302Isa-Pd 67 3+ Lines3823.3907,989 USD0.95ICARIA-MM:N=307CAR-T Legend-268,694+ Lines54 +30640,000 USD1.7433Legend 2: N=74CARTITUDE-168,704+ Lines85 +43.85640,000 USD2.7384CARTITUDE-1:N=97X-Vd 71 2-4th Lines34.9818.5839,679 USD1.575BOSTON: N=402*Mean OS values were calculated by digitalizing Kaplan–Meier survival curves and estimating the area under the survival percentage curve. Values represent fully observed cohorts only; extrapolation for censored data was performed using a simple geometric projection model.**Cost estimation reflects total drug acquisition and administration expenses up to treatment discontinuation, based on list prices obtained from https://drugs.com. Supportive care and concomitant therapy costs were included within the same time horizon.***QALY values were derived from mean OS and ECOG-based utility adjustments within the hypothesis framework of this study (see Methods and Supplemental Material for details).aQALY could not be modeled due to insufficient follow-up data.V: Velcade(Bortezomib), R: Revlimid(Lenalidomide), D: Darzalex(Daratumumab), K: Kyprolis(Carfilzomib), M: Melphalan, P: Prednisone(VMP), d: Dexamethasone., P:Pomalyst (Pomalidomide), X:Xpovio (Selinexor), Elo:Empliciti(Elotuzumab), Isa:Sarclisa(Isatuximab). OS=Overall Survival. PFS: Progression Free Survival. USD: United States Dollar. QALY: The quality-adjusted life-year. Estimated costs, mean overall survival (OS), and model-derived QALY values for major myeloma therapy regimens used in clinical practice. *Mean OS values were calculated by digitalizing Kaplan–Meier survival curves and estimating the area under the survival percentage curve. Values represent fully observed cohorts only; extrapolation for censored data was performed using a simple geometric projection model. **Cost estimation reflects total drug acquisition and administration expenses up to treatment discontinuation, based on list prices obtained from https://drugs.com. Supportive care and concomitant therapy costs were included within the same time horizon. ***QALY values were derived from mean OS and ECOG-based utility adjustments within the hypothesis framework of this study (see Methods and Supplemental Material for details). aQALY could not be modeled due to insufficient follow-up data. V: Velcade(Bortezomib), R: Revlimid(Lenalidomide), D: Darzalex(Daratumumab), K: Kyprolis(Carfilzomib), M: Melphalan, P: Prednisone(VMP), d: Dexamethasone., P:Pomalyst (Pomalidomide), X:Xpovio (Selinexor), Elo:Empliciti(Elotuzumab), Isa:Sarclisa(Isatuximab). OS=Overall Survival. PFS: Progression Free Survival. USD: United States Dollar. QALY: The quality-adjusted life-year. This approach allows comparison of treatment regimens not only in terms of clinical efficacy, but also through the lens of cost-effectiveness and ethical sustainability (see Supplementary Material for details). However, model-based QALY estimates presented here should not be directly compared with classical QALY values derived from prospective patient surveys. Although absolute QALY values are lower, the model provides internally consistent and reliable comparative insights across regimens (see Methods and Supplemental Methods). 3. CAR-T therapies A major milestone in multiple myeloma therapy was the identification of B-cell maturation antigen (BCMA) as a therapeutic target. To date, two autologous CAR-T cell products targeting BCMA have received regulatory approval. Idecabtagene vicleucel (ide-cel, Abecma) was approved by the FDA in April 2021 following the KarMMa trials and demonstrated remarkable efficacy in heavily pretreated patients (≥3 prior lines including PI, IMiD, and anti-CD38). In this population, the overall response rate (ORR) reached 76% versus 32% with standard of care; ≥VGPR was 58% versus 14%; one-year progression-free survival (PFS) was 55% versus 30%; and minimal residual disease (MRD) negativity was approximately 20%. The most frequently reported toxicities included neutropenia, leukopenia, and cytokine release syndrome (CRS), with grade ≥3 CRS observed in roughly 14% of patients.14,72 Ciltacabtagene autoleucel (cilta-cel, Carvykti), a second-generation BCMA-directed CAR-T construct with dual-epitope binding capacity, received FDA approval in February 2022. According to 2025 analyses, the median overall survival (OS) is approximately 56 months and median PFS about 18 months, with an MRD-negativity rate of 68%. Reported toxicities include low-grade CRS, immune effector cell-associated neurotoxicity syndrome (ICANS), cytopenias, and infections.70,73 3.1. Treatment process and clinical application The CAR-T therapy process involves the collection of a patient’s T lymphocytes by apheresis, their genetic modification ex vivo, expansion in culture, and reinfusion as a single intravenous dose. Before infusion, lymphodepleting chemotherapy (fludarabine + cyclophosphamide) is administered to facilitate in vivo CAR-T expansion, reduce immune suppression within the tumor microenvironment, and enhance persistence. Manufacturing typically requires 2-4 weeks and includes the following main steps 74 (see Figure 4): • Activation: stimulation using anti-CD3/CD28 microbeads and cytokines such as IL-2, IL-7, and IL-15. • Activation: stimulation using anti-CD3/CD28 microbeads and cytokines such as IL-2, IL-7, and IL-15. • Genetic modification: transduction of the CAR construct, most commonly via lentiviral vectors. • Genetic modification: transduction of the CAR construct, most commonly via lentiviral vectors. • Expansion: large-scale proliferation in controlled culture systems to generate millions of CAR-T cells. • Expansion: large-scale proliferation in controlled culture systems to generate millions of CAR-T cells. • Quality control and cryopreservation: verification of cell viability, microbial safety, and CAR expression, followed by freezing and storage until infusion. • Quality control and cryopreservation: verification of cell viability, microbial safety, and CAR expression, followed by freezing and storage until infusion. Schematic representation of the CAR-T cell manufacturing process. Autologous T cells are harvested from the patient via leukapheresis, then enriched and activated through CD3/CD28 co-stimulation with cytokine support (IL-2, IL-7, IL-15). The CAR construct is introduced using lentiviral vectors, followed by large-scale expansion, quality control, and cryopreservation. After bridging therapy and lymphodepleting chemotherapy, the engineered CAR-T cells are thawed and administered intravenously. The figure provides a stepwise visual reference for the standard CAR-T cell manufacturing workflow as it is applied in autologous cellular immunotherapy (Created with BioRender.com). The prepared cells are thawed and administered as a single intravenous infusion. The most common adverse events are cytokine release syndrome (CRS) and neurotoxicity, both of which require management in experienced centers with intensive-care support. 3.2. Clinical efficacy, innovations, and limitations of CAR-T cell therapy CAR-T cell therapy achieves high response rates (≥70%) and prolonged remissions in patients with relapsed/refractory multiple myeloma (RRMM). Clinical trials such as CARTITUDE-4 have demonstrated that CAR-T therapy can be used as early as the second line, aiming to improve both survival and quality of life with earlier intervention. 13 Its single-dose, potentially curative nature positions CAR-T as one of the most promising modalities in modern oncology. However, high cost, lengthy manufacturing, and the need for intensive toxicity management currently limit its widespread application. The comparative analysis in this study highlights that ciltacabtagene autoleucel (cilta-cel) offers unexpectedly favorable outcomes in terms of both economic sustainability and biological toxicity profile. 74 Cilta-cel’s single-infusion design reduces cumulative toxicity associated with multiple treatment lines and improves not only survival but also quality-of-life metrics (QALY). 60 Although upfront costs are high, the long-term reduction in complications, lower hospitalization rates, and extended treatment-free intervals make CAR-T therapy a relatively cost-effective investment 74 (see Supplemental Material for details). Recent advances have focused on moving CAR-T therapy into earlier lines of treatment, shortening manufacturing time, and enhancing antigen-binding efficiency (“docking efficiency”). Traditional autologous CAR-T production takes several weeks, during which many patients may experience disease progression before infusion. To address this, rapid manufacturing platforms such as FasT CAR-T have been developed. For instance, BCMA/CD19 dual-target GC012F CAR-T cells can be generated within 48-72 hours. Similarly, the NEX-T process integrates non-viral gene-transfer systems (e.g., Sleeping Beauty transposons) and automated quality-control algorithms, potentially reducing production time to just a few days.9,10 In addition, “off-the-shelf” allogeneic CAR-T products aim to eliminate wait times and improve accessibility. 11 At the molecular level, optimization of antigen-binding domains including linker length and conformational stability has improved docking efficiency. Dual-target constructs (e.g., BCMA + GPRC5D or BCMA + CD38) show promise in preventing antigen escape. 12 Early-line clinical testing (before third-line therapy) suggests these approaches may preserve immune-cell function and extend durable remission.13,14 However, as manufacturing time decreases, careful evaluation of cell quality, CRS incidence, and long-term efficacy remains critical. Despite impressive response rates, CAR-T therapy faces major limitations due to severe toxicities and restricted T-cell persistence. The most frequent and potentially life-threatening adverse event is cytokine release syndrome (CRS), triggered by excessive release of pro-inflammatory cytokines such as IL-6, IFN-γ, and TNF-α, leading to systemic inflammation, fever, hypotension, hypoxia, and multi-organ dysfunction. 75 CRS typically occurs within the first few days post-infusion and is managed with tocilizumab or corticosteroids. A second key toxicity is immune effector cell-associated neurotoxicity syndrome (ICANS), commonly observed with CD19-and BCMA-directed CAR-T therapies. Symptoms include confusion, aphasia, tremor, or seizures, linked to endothelial activation and disruption of the blood-brain barrier76. Prolonged neutropenia and hypogammaglobulinemia further increase the risk of secondary infections and delay immune reconstitution. 75 Despite these remarkable outcomes, limited accessibility remains a major challenge. Ide-cel is currently available only in the United States, France, Switzerland, Japan, and Germany, while cilta-cel is approved solely in the United States and Germany. Manufacturing delays, logistical barriers, and strict eligibility criteria continue to restrict broader use. The autologous nature of most CAR-T products limits functional capacity in older or heavily pretreated patients, reducing T-cell persistence.77,78 Additional barriers include antigen loss, immunosuppressive microenvironmental effects (e.g., TGF-β, IL-10, accumulation of MDSCs), and high production costs. In allogeneic (“off-the-shelf”) CAR-T approaches, safety concerns such as graft-versus-host disease (GVHD) and immune rejection remain unresolved. 79 To overcome these challenges, next-generation strategies such as armored CAR-T, self-destruct switches, and regulated CAR systems are being developed to enhance persistence and minimize toxicity. 3.3. CAR-T cell trials The first-generation BCMA-targeted CAR-T therapy, idecabtagene vicleucel (ide-cel; bb2121), represents the pioneering autologous CAR-T construct evaluated in the KarMMa trials. Its single-epitope binding design limited both antigen affinity and cellular persistence. In vivo proliferation peaked early, leading to rapid but short-lived responses.80,81 Consequently, KarMMa-1 reported a median progression-free survival (PFS) of 8.8 months and a median overall survival (OS) of 19.4 months values substantially lower than those achieved by later dual-epitope designs. 81 By contrast, ciltacabtagene autoleucel (cilta-cel; LCAR-B38M), a second-generation BCMA-directed CAR-T product developed through the LEGEND-2 and CARTITUDE programs, demonstrated markedly improved outcomes.12,69 Its dual-epitope binding configuration enhances antigen affinity, strengthens T-cell proliferation, and promotes durable immunologic memory. 70 Among clinical trials evaluating CAR-T efficacy, LEGEND-2 and CARTITUDE-1 provide the most comprehensive long-term data. CARTITUDE-1 confirmed that deep and durable responses can be achieved even in heavily pretreated patients. After 61 months of follow-up, median OS reached 60.7 months, while the mean OS (area-under-curve analysis) was estimated at ∼85 months. 70 LEGEND-2 (NCT03090659), the first-in-human, open-label phase 1 study of the dual-epitope construct LCAR-B38M, enrolled 74 patients with RRMM previously treated with at least three therapy lines including proteasome inhibitors and immunomodulatory agents. Following lymphodepletion with fludarabine and cyclophosphamide, patients received 0.5-1.0 × 106 CAR+ T cells/kg as single or split doses. Response rates were high: overall response rate (ORR) 88-90%, complete response (CR) 74%, and MRD negativity ∼68%. 69 Median PFS was 20 months and median OS 36 months. CRS occurred in most patients but was predominantly grade 1-2 and manageable; neurotoxicity was rare. This study provided the first clinical evidence that BCMA-directed CAR-T therapy could induce deep and durable remissions. Following these successes, CARTITUDE-2, a multicenter phase 2 trial, explored CAR-T efficacy across diverse patient subgroups: Cohort A: triple-class-refractory patients (≥3 prior lines), Cohort B: lenalidomide-refractory patients with 1-3 prior lines, 82 Cohort C: patients at first relapse, 83 Cohort D: transplant-ineligible, lenalidomide-refractory patients, 84 Cohort E: high-risk or MRD-positive early-stage disease. 73 In CARTITUDE-4, cilta-cel demonstrated statistical superiority over standard regimens in early-relapse settings in terms of progression-free survival 85. The ongoing CARTITUDE-5 and CARTITUDE-6 studies85,86 are now evaluating whether CAR-T therapy can replace the conventional ASCT + maintenance paradigm as a first-line treatment option, independent of transplant eligibility. Together, these consecutive trials establish CAR-T therapy not merely as a “last resort,” but as an emerging early-line, immune-rejuvenating strategy in multiple myeloma (see Table 3). Table 3.Clinical summary of cilta-cel and ide-cel based CAR-T trials.Study namePhase/DesignPatient profileTreatment lineRegimen/ComparatorKey messageLEGEND-2 (NCT03090659) 69 Phase 1, open-label, single-arm≥3 prior lines (PI + IMiD-treated RRMM)4+ lineLCAR-B38M single infusionFirst-in-human BCMA-CAR-T; deep and durable responsesCARTITUDE-1 (NCT03548207) 70 Phase 1b/2, multicenter≥3 prior lines (PI + IMiD + anti-CD38)4+ lineCilta-cel single infusionGlobal validation; high depth and long PFSCARTITUDE-2 (NCT04133636) 82 Phase 2, multi-cohortCohorts A-E: RRMM, len-refractory, 1-3 prior lines2nd–3rd lineCilta-cel single infusionHigh efficacy maintained in early-line settingsCARTITUDE-4 (NCT04181827) 87 Phase 3, global1–3 prior lines, len-refractory2nd–3rd lineCilta-cel vs PVd/DPdPhase 3 confirmation of superiority vs SOCCARTITUDE-5 (NCT04923893) 85 Phase 3 (ongoing)Newly diagnosed, transplant ineligable1st lineVRd+Cilta-cel vs VRd+Rd maintenanceCAR-T as potential maintenance alternative in frontlineCARTITUDE-6 (NCT05257083) 86 Phase 3, comparativeNewly diagnosed, transplant-eligible1st lineD-VRd+Cilta-cel + Lenalidomide vs D-VRd+ASCT+DVRd+LenalidomideImmune-regenerative CAR-T as ASCT replacementKarMMa-1 (NCT03361748) 88 Phase 2, single-arm≥3 prior lines RRMM4+ lineIde-cel (bb2121)First approved BCMA-CAR-T (US)KarMMa-3 (NCT03651128) 14 Phase 3, randomized2nd–4th line-Ide-cel vs SOCIde-cel effective but with shallower depth of responseBCMA: B-cell maturation antigen; SOC: Standard of Care; NDMM: Newly Diagnosed Multiple Myeloma; RRMM: Relapsed Refractory Multiple Myeloma; CAR: Chimeric antigen receptor; PFS: Progression Free Survival. Clinical summary of cilta-cel and ide-cel based CAR-T trials. BCMA: B-cell maturation antigen; SOC: Standard of Care; NDMM: Newly Diagnosed Multiple Myeloma; RRMM: Relapsed Refractory Multiple Myeloma; CAR: Chimeric antigen receptor; PFS: Progression Free Survival. 4. Bispecific antibodies Bispecific antibodies have recently emerged as an important therapeutic class in relapsed/refractory multiple myeloma (RRMM). These agents redirect cytotoxic T cells toward malignant plasma cells by simultaneously binding CD3 on T cells and tumor-associated antigens on myeloma cells, such as BCMA or GPRC5D, thereby inducing T-cell activation independent of MHC antigen presentation. This interaction promotes targeted cytotoxicity through perforin- and granzyme-mediated apoptosis of myeloma cells.4,89,90 Currently approved bispecific antibodies include the BCMA-directed agents teclistamab, elranatamab, and linvoseltamab, as well as the GPRC5D-directed agent talquetamab. Their clinical activity has been demonstrated in pivotal studies, including MajesTEC-1 for teclistamab, MagnetisMM-3 for elranatamab, MonumenTAL-1 for talquetamab, and LINKER-MM1 for linvoseltamab.89–92 Teclistamab was first approved for RRMM in 2022, followed by talquetamab and elranatamab in 2023, while linvoseltamab received accelerated approval in 2025. In addition, the regulatory landscape continues to evolve rapidly, with expanded indications and combination strategies, including teclistamab plus daratumumab, receiving approval as recently as 2026. 93 Although these agents represent a major advance in RRMM, their clinical use is associated with immune-mediated toxicities such as cytokine release syndrome, neurotoxicity, infections, cytopenias, and antigen-specific off-tumor effects, including plasma cell depletion. Importantly, because much of the available evidence derives from relatively recent, non-randomized or single-arm studies with still-maturing long-term OS/PFS and economic data, their inclusion in the present OS/PFS-based cost-effectiveness framework would have limited methodological comparability with more established regimens.89–92 5. Quality-adjusted life year (QALY) and ECOG-Based modeling The Quality-Adjusted Life Year (QALY) is an integrative measure that captures both the quantity and quality of life. One QALY represents one year lived in perfect health; for an individual experiencing 50% health quality, one year corresponds to 0.5 QALY. This concept allows treatment outcomes to be compared not only by survival duration but also by their impact on quality of life. In multiple myeloma, cumulative toxicity and treatment-related complications affect patients as profoundly as survival itself. Physical limitations, fatigue, immune suppression, and infection risk directly influence therapeutic success. Modern oncology therefore emphasizes combined metrics of longevity and quality most notably, the QALY as standard of value-based care. Quality of life is commonly assessed using validated questionnaires such as EORTC QLQ-C30, EQ-5D, or FACT-MM. The resulting utility scores, ranging from 0 to 1, are multiplied by the number of life-years gained to calculate QALY values 94 : In parallel, the ECOG Performance Status (0-4) provides a clinician-rated, objective measure of functional capacity and independence: 0 = fully active; 1 = mild restriction; 2 = capable of limited activity; 3 = restricted to self-care; 4 = completely disabled; 5 = death. Multiple studies have demonstrated a strong correlation between ECOG scores and HRQoL or EQ-5D utility values.94–98 However, this relationship has rarely been formalized into a simple, quantitative model. In this study, based on 99 the following ECOG-to-utility conversion was applied: ECOG 0 = 0.85, 1 = 0.75, 2 = 0.65, 3 = 0.55, 4 = 0.45. Each one-point increase in ECOG corresponds to an average utility decrement of 0.05-0.10, representing a clinically meaningful change (MID). This method directly links performance status to quality of life and enables standardized, transparent, and practical QALY estimation even in the absence of patient-reported data. Supplementary materials detail how this ECOG-based model was adapted to specific treatment arms and integrated with cost data. Furthermore, the model incorporates Incremental Cost-Effectiveness Ratio (ICER) analysis, which quantifies the additional cost required to gain one extra QALY compared with an alternative treatment: • Low ICER → higher cost-effectiveness • Low ICER → higher cost-effectiveness • High ICER → lower value or higher cost per QALY 100 • High ICER → lower value or higher cost per QALY 100 This framework assesses not only the economic burden of therapy but also the value gained per unit of quality-adjusted survival. In our analysis, CAR-T therapies particularly those from CARTITUDE-1 showed lower ICER values than most conventional regimens. This finding suggests that CAR-T represents not a “high-cost innovation,” but a high-value therapy whose long-term gains in quality of life justify its initial expense. Ultimately, this study proposes a simple, mathematically consistent, and ethically grounded model that enables clinicians, policymakers, and researchers to evaluate myeloma therapies not only in clinical and economic terms but also through a humanistic and justice-oriented lens. 6. Methods (summary) This analysis was designed as a model-based, hypothesis-generating comparative evaluation of multiple myeloma treatment regimens from a healthcare system perspective. Published outcome data were synthesized from 19 pivotal phase II/III clinical trials and large real-world cohorts in newly diagnosed and relapsed/refractory multiple myeloma, comprising 7,793 patients across 21 treatment regimens (Table 2; Supplemental Material). Trial-level survival inputs were reconstructed through manual digitization of published Kaplan–Meier curves and area-under-the-curve integration. Regimen-specific costs, ECOG-derived utility weights, quality-adjusted life-years (QALYs), and ICER-equivalent estimates were then harmonized within a unified comparative framework across induction-phase, relapsed/refractory, and CAR-T cellular therapy strategies. The resulting estimates are intended to provide a transparent and reproducible basis for relative cross-regimen comparison, while acknowledging the inherent limitations of trial-level reconstruction and model-based inference. QALY estimates were calculated by weighting survival duration with utility values derived from ECOG performance scores. The utility coefficients were obtained according to the model described in the hypothesis section (ECOG 0 = 0.85; 1 = 0.75; 2 = 0.65; 3 = 0.55; 4 = 0.45). In studies where ECOG data were not reported, estimated utilities were derived by optimizing toxicity (TEAE) and mortality data. Since all regimens were analyzed within the same methodological framework, results are proportionally comparable across treatment lines (Supplemental Material Table S1). Mean overall survival (OS) was estimated by calculating the area under the Kaplan–Meier curve (AUC). Published survival curves were digitized using WebPlotDigitizer, and the area beneath each curve was estimated by geometric integration. Progression-free survival (PFS) was extracted using the same approach. Missing curve segments were interpolated geometrically according to slope trajectory and directional trends. Mean OS, rather than median OS, was used as the primary survival metric because health-economic evaluations require expected survival, represented by the AUC, for cost-effectiveness and QALY-based modeling; by contrast, median OS reflects only the time to 50% mortality and does not capture total expected life-years. To support treatment-line–stratified interpretation, regimens were grouped descriptively as early-line/induction regimens (Rd, D-Rd, VRd, D-VRd, KRd, D-VMP, and VMP), later-line RRMM regimens (PVd, Vd, D-Kd, Kd, D-Pd, D-Vd, Elo-Rd, Elo-Pd, Elo-PVd, Isa-Kd, Isa-Pd, and XVd), and CAR-T cellular therapy regimens (LEGEND-2 and CARTITUDE-1). Median ICER-equivalent values were calculated for each group and compared as fold-differences to contextualize relative cost-per-QALY patterns across treatment settings. Beyond this methodological rationale, the systematic computation and harmonization of mean OS across all evaluated regimens represents a distinct contribution of this work, since standardized AUC-derived survival estimates are not routinely reported in trial publications and are rarely compiled in a comparable regimen-by-regimen format. The resulting dataset (Supplemental Material, Table S2) may therefore serve as a methodological reference for future pharmacoeconomic analyses in multiple myeloma. Reported median OS values from the original trials remain available in the corresponding publications and can be used for conventional clinical interpretation. Cost estimation was aligned with the primary endpoint reported in each trial. Drug costs were calculated according to dose, administration frequency, and treatment duration, using unit prices adapted from recent literature and national reimbursement databases (Supplemental Material, Table S2). Details of the model structure, parameter sources, dose calculations, and sensitivity analyses are provided in the Supplementary Methods. Cost calculations focused primarily on drug acquisition and administration; supportive care, long-term follow-up, and differential hospitalization costs, including CAR-T-specific procedures such as apheresis, conditioning chemotherapy, and inpatient post-infusion monitoring, were not separately itemized. This scope was applied consistently across all regimens to preserve internal comparability, with related limitations addressed in the Limitations section. Finally, included regimens spanned a broad range of treatment lines, from second to tenth line and including heavily pre-treated populations with four or more prior therapies. Accordingly, the comparisons presented herein should be interpreted as conceptual and contextual rather than direct head-to-head evaluations. 7. Discussion The primary aim of this study was to evaluate the advantages of CAR-T cell therapies over conventional regimens in terms of toxicity, progression-free survival (PFS), and overall survival (OS). Within this scope, QALY calculations and other pharmacoeconomic analyses were developed to interpret this important paradigm shift in hematology through an ethical and value-based framework. The study contrasts traditional approaches characterized by repeated chemotherapy cycles, high costs, and continuous hospital dependency with the potential of CAR-T therapy: a single infusion that offers prolonged survival, reduced toxicity, and sustained remission. The goal, therefore, was not only to demonstrate clinical superiority but also to examine the ethical and human dimensions of modern therapeutic strategies. QALY-based cost analyses were used as a numerical representation of this ethical framework, seeking a clinical and philosophical answer to the question: “Is progress measured by the length of life, or by the quality of living?” Most existing pharmacoeconomic studies in the literature rely on predefined QALY values and Markov-based cost-effectiveness models. 60 However, such models are often complex, making them difficult for clinicians and biomedical researchers to apply or interpret. In this context, our analysis provides a simpler but systematic framework that allows direct comparison of toxicity and survival outcomes without requiring complex modeling. Moreover, the cumulative toxicity observed with conventional therapies and the biological-economic transformation brought by CAR-T cells suggest a need to reassess therapeutic priorities in hematology. This study responds to that need by evaluating treatment lines in terms of toxicity, survival, and quality of life, proposing an ethical, clinically meaningful, and practical framework for decision-making. During the analysis, significant data gaps and transparency issues were encountered across clinical trials. Research reports should be written not only for clinicians but also for investigators from other disciplines. Yet, many pivotal phase studies compromise scientific clarity through excessive technical terminology, incomplete or unclear data presentation (particularly omission of death counts), and TEAE graphics that fail to reflect the patient’s actual clinical condition. Beyond the methodological implications, these limitations also carry a normative dimension: they blur the boundaries between methodological reliability and ethical accountability. Broadly defined treatment categories such as “2nd-10th line” often encompass heterogeneous patient populations that differ greatly in biological and functional capacity. A patient in the second line of therapy and one in the tenth line are clinically incomparable. Such grouping may distort both positive and negative interpretations of efficacy and safety. From an ethical standpoint, publishers, reviewers, and researchers share a shared obligation to report baseline and final patient status using functional indices (e.g., ECOG, quality-of-life scales) and to provide arm-specific mortality data transparently. Clear definition of heterogeneity, and, when necessary, separate analyses by treatment line, is essential not only for methodological accuracy but also for a patient-centered ethical evaluation. In the transplant-ineligible group, within the assumptions of this comparative model, the VMP regimen remains the most cost-effective option, offering a favorable balance between total cost and utility/QALY gains. Despite the abundance of new agents, the continued clinical efficacy of agents such as melphalan and prednisone drugs whose origins trace back to post-World War II era, suggests that progress in medicine is not always defined by innovation, but may also involve continuity and resilience. 101 Analysis of induction-phase regimens (VRd, D-VRd, D-VMP, VMP, D-Rd, and Rd) revealed the highest QALY and utility values at this early treatment stage (Table 2; Supplemental Table S2). Beyond these quantitative observations, a broader normative question emerges: should we favor well-known, lower-cost regimens with predictable toxicity and management profiles, or prioritize new, high-cost combinations whose long-term outcomes remain uncertain? This may also support further consideration of individualized treatments that strengthen the patient’s own immune system, integrating nutritional and lifestyle-based support with modern immunotherapies. Such a holistic approach (offered here as a contextual reflection rather than a data-derived conclusion) may represent a potentially scientifically sound and ethically sustainable direction for the future.102,103 In transplant-eligible patients, both survival and quality-of-life metrics were higher. Adding daratumumab to VRd, which yielded an average QALY of 7.145, extended survival by more than 2.5 years although at approximately four times the cost. The most striking benefit of daratumumab was observed in progression-free survival (PFS): median PFS increased from 38 months with VRd to 84 months with D-VRd. Similar improvements were observed when daratumumab was added to VMP or Rd regimens, leading to sustained gains in both survival and quality of life. 104 Analysis of carfilzomib-based regimens (e.g., KRd, D-Kd) showed overall-survival advantages often exceeding 50 months, although progression-free survival (PFS) durations were shorter. This finding suggest that carfilzomib provides a mortality-reducing and quality-of-life-enhancing effect, particularly in high-risk patient groups. 105 Despite its short-term toxicity risks, carfilzomib continues to hold a valuable place within risk-adaptive therapeutic strategies. Among late-line therapies, evaluated regimens include Elo-Rd, Elo-Pd, Elo-PVd, Isa-Pd, XVd, and CAR-T cell treatments such as those from the CARTITUDE-1 and LEGEND-2 trials (Appendix (B), Figure B1). When daratumumab intolerance or resistance occurs, elotuzumab-based regimens become clinically relevant due to their low toxicity profile and unique SLAMF7-targeting mechanism. 106 In patients refractory to both lenalidomide and daratumumab, Elo-Pd remains a feasible alternative. Because most of these patients have previously received proteasome inhibitors, combinations such as Elo-PVd which include bortezomib are typically used in later treatment lines (Appendix Figure B1). Although late-line regimens such as PVd, Pd, and XVd achieve median overall survival (OS) approaching 40 months, the reliability of these data remains limited. The referenced studies involve highly heterogeneous patient populations, and the clinical characteristics of survivors are often underreported. In heavily pre-treated patients exposed to cumulative toxicity, the true clinical meaning of these survival durations is uncertain. The derived utility values were nearly ten-fold lower than those observed in the induction phase, with a mean progression-free survival (PFS) of only about 20 months. Treatment costs averaged roughly USD 800 000 per patient. Regimens such as Elo-Rd and Isa-Pd yielded QALY scores approaching 1.0, yet these modest gains came at an additional cost of about USD 100 000. Among them, Elo-Rd offered roughly one extra year of survival compared with other late-line options; however, this difference may partly reflect patient heterogeneity or residual lenalidomide sensitivity. 107 Within this landscape, the most notable finding was the CAR-T cell therapy results. Despite excluding patients from second- or third-line cohorts, cilta-cel demonstrated an estimated mean OS exceeding 85 months, PFS around 44 months, and a substantially lower annual cost (Table 2; Supplemental Table S2). In the ongoing CARTITUDE-1, 70 the mean OS exceeded 85 months, and the calculated QALY reached 2.744 approximately three times higher than conventional late-line regimens. These quantitative findings suggest that, within the broader interpretive framework adopted here, CAR-T therapy may reshape the end-stage paradigm not only biologically, but also economically and ethically. To evaluate how much it costs to provide one additional year of quality-adjusted life, we performed an Incremental Cost-Effectiveness Ratio (ICER) analysis (Figure 5). This approach quantifies the comparative cost of achieving one year in which the patient perceives themselves as “healthy.” It also integrates clinical, economic, and ethical dimensions into a unified assessment of therapeutic value. As stated in the Methods, the included regimens were derived from studies conducted in different treatment lines and patient populations. Accordingly, the results should not be interpreted as direct head-to-head comparisons. Instead, they should be viewed as exploratory, context-generating comparisons within a common analytical framework. ICER distribution across myeloma therapies based on ECOG-adjusted utility approach. Each bar represents the incremental cost-effectiveness ratio (ICER) expressed as cost per QALY gained. Green tones illustrate relative cost-effectiveness proximity among regimens, highlighting CAR-T (CARTITUDE-1) within the lowest ICER range, comparable to D-VRd and other advanced monoclonal antibody combinations. The central message of the figure is that CAR-T therapy, despite being applied in the late-line setting, achieves an ICER comparable to that of earlier-line induction regimens including transplant-based and non-transplant combinations such as VMP, D-VMP, VRd, and D-VRd, while several conventional late-line regimens with which it competes clinically (D-Kd, Elo-Rd, Elo-Pd, Isa-Pd) approach or exceed 1 000 000 USD per QALY gained. In the treatment-line–stratified descriptive comparison, later-line RRMM regimens showed a substantially higher median ICER-equivalent value than early-line/induction regimens (USD 739,050/QALY vs USD 218,687/QALY), corresponding to an approximately 3.4-fold increase. By contrast, CAR-T cellular therapy regimens showed a median ICER-equivalent value of USD 299,723/QALY, which was approximately 2.5-fold lower than the median value of later-line RRMM conventional regimens despite their upfront single-payment structure. In the late-line analysis, within this modeling framework, the calculated ICER values indicate that, despite its seemingly high upfront cost, CAR-T therapy may stand among the most value-based interventions in multiple myeloma. Its single-infusion design, minimal hospitalization time, and lack of cumulative toxicity management may substantially reduce total healthcare expenditures. In contrast, conventional treatments administered “piece by piece” across multiple cycles create substantial long-term economic and biological burdens. Therefore, within the present comparative framework, CAR-T should not be viewed only as an “expensive innovation,” but rather as a potentially high-value and sustainable solution. When cost optimization, insurance coverage, and public reimbursement mechanisms are implemented, its financial impact may become lighter than that of conventional regimens over time. It should nevertheless be acknowledged that the favorable economic profile projected within this analytical framework reflects modeled long-term value rather than real-world affordability. Actual access remains constrained by upfront costs, infrastructural requirements, and substantial reimbursement variability across healthcare systems, particularly in low- and middle-income settings. Current evidence shows that CAR-T therapy provides meaningful improvements in survival and quality of life, even in heavily pretreated patients. Although results from early-line trials (such as the CARTITUDE and KarMMa series) are still maturing, preliminary findings suggest that CAR-T could be administered earlier, with lower toxicity and longer-lasting remission. This potential points toward a new therapeutic paradigm in myeloma, one focused not on multiple sequential treatments, but on single-intervention, durable control. In this sense, CAR-T may represent not only an early intervention candidate, but also a personalized therapeutic approach with the potential to achieve lasting remission with one infusion. 7.1. Ethical and normative considerations Beyond the empirical findings summarized above, the present analysis also invites a normative reflection. From an ethical standpoint, a therapy initially limited by high cost but ultimately capable of providing more life years, less toxicity, and lower overall expenses demands a re-examination of what constitutes justice and value in modern medicine. Today, CAR-T cell therapy remains accessible only in a few countries and primarily to patients who can afford its one-time infusion price. This disparity raises critical questions about how biotechnological innovation can be balanced with equity and accessibility. As a contextual perspective rather than a data-derived conclusion, in a patient-centered and socially sustainable healthcare model, CAR-T should be recognized not only as a scientific breakthrough, but also as a transformative model redefining ethical, economic, and human values in medicine. Expanding its accessibility will not only improve clinical outcomes, but also strengthen equality, humanism, and the principle of “kill the tumor, not the human. See Supplemental Material for details. 7.2. Limitations Several limitations of the present analysis warrant explicit acknowledgment. First, the model relied on ECOG performance status as a surrogate for health-related utility. Although baseline ECOG data were available in most included trials, ECOG status was less consistently reported at later time points or at study completion, and several studies reported aggregated ECOG categories such as ECOG 0–1 or ECOG 0–3 rather than distinct ECOG 0, 1, 2, 3, and 4 subgroups. Accordingly, mean or approximated ECOG-based values were required for utility estimation. While the assigned utility coefficients were consistent with published health-economic literature, they were treated as fixed-point estimates, and patient-level heterogeneity in self-reported quality of life was not modeled. Modest variation in assumed utility weights may therefore shift absolute QALY estimates. However, because all regimens were evaluated using the same ECOG-to-utility mapping framework, relative ranking and proportional differences between treatment strategies are expected to be more stable than absolute QALY values. Future studies incorporating prospectively collected patient-reported outcomes and probabilistic sensitivity analyses using utility distributions would improve the precision of this approach. Second, OS and PFS inputs were derived from published Kaplan–Meier curves using manual digital extraction with WebPlotDigitizer, followed by estimation of mean survival values. Although this approach is commonly used when individual patient-level data are unavailable, it remains vulnerable to graphical and methodological uncertainty, including image resolution, curve overlap, censoring density, axis calibration, and assumptions required when extrapolating survival curves toward the x-axis. Therefore, derived OS and PFS values should be interpreted as approximate reconstructed estimates rather than direct patient-level survival data. Access to individual patient-level data or standardized survival tables from the original trials would enable more robust survival modeling. Third, cost estimation was based on publicly available drug price information obtained from a single source, drugs.com. Actual real-world costs may vary according to country-specific pricing policies, negotiated discounts, reimbursement structures, hospital procurement systems, insurance coverage, and patient assistance programs. Therefore, the absolute cost values reported here should not be interpreted as universal real-world expenditures. Because the same pricing source and calculation framework were applied consistently across all regimens, the cost findings are more appropriately interpreted as proportional differences between treatment strategies rather than precise monetary estimates. Given the heterogeneity of treatment lines, an exploratory Kruskal–Wallis analysis with Dunn’s post hoc comparisons was performed across early-line/induction, later-line RRMM, and CAR-T categories. Later-line RRMM regimens showed significantly higher ICER-equivalent values than early-line/induction regimens (adjusted p = 0.0048), whereas CAR-T regimens did not differ significantly from either early-line/induction regimens (adjusted p > 0.9999) or later-line RRMM regimens (adjusted p = 0.4630). However, the lack of statistical significance for CAR-T should not be interpreted as evidence of equivalence, because the CAR-T group included only two regimen-level observations. With a larger number of CAR-T datasets, these comparisons could change. Therefore, these inferential results were considered exploratory, and descriptive median and fold-difference comparisons were emphasized in the main analysis. A further limitation concerns the scope of cost estimation. The present analysis focused primarily on drug acquisition and administration costs and did not comprehensively itemize supportive care, long-term follow-up, hospitalization, adverse event management, or procedure-related expenditures. This is particularly relevant for CAR-T therapy, where apheresis, bridging therapy, lymphodepleting chemotherapy, manufacturing logistics, inpatient monitoring, management of cytokine release syndrome or immune effector cell-associated neurotoxicity syndrome, intensive care requirements, and prolonged follow-up may contribute substantially to total cost of care. Although this restricted cost scope was applied uniformly across all regimens to preserve internal comparability, the reported values should be interpreted as approximations of direct pharmacological expenditure rather than comprehensive total cost of care. Similarly, the favorable economic profile projected for CAR-T within this analytical framework reflects modeled long-term therapeutic value, not real-world affordability or accessibility. In practice, CAR-T access remains constrained by high upfront costs, specialized infrastructure, manufacturing capacity, apheresis availability, trained multidisciplinary teams, post-infusion monitoring, and reimbursement variability, particularly in low- and middle-income settings. Therefore, the present analysis should be interpreted as a comparative framework for estimating therapeutic value, not as a direct guide to health-system affordability or patient-level access. Finally, bispecific antibodies were not included in the quantitative cost-effectiveness calculations. This exclusion was intentional and reflects the predefined analytical scope and data-maturity requirements of the model. The present analysis relied on mature and comparable OS/PFS-based inputs, whereas bispecific antibodies entered routine clinical practice relatively recently, with initial approvals occurring in 2022–2023 and additional agents, indications, and combinations continuing to emerge through 2025–2026. Their inclusion would therefore have required extrapolation from relatively immature, heterogeneous, and often non-randomized or single-arm evidence, potentially increasing uncertainty and reducing methodological comparability with more established regimens. Accordingly, bispecific antibodies were discussed as an important and rapidly evolving therapeutic class in relapsed/refractory multiple myeloma, but were not incorporated into the present OS/PFS-based economic model in order to preserve internal consistency and fairness of comparison. 8. Conclusion This study establishes a comprehensive and transparent model-based comparative framework demonstrating that CAR-T cell therapy delivers superior survival and quality-adjusted clinical outcomes for heavily pretreated multiple myeloma patients compared with conventional stepwise regimens. Although the upfront acquisition cost of cellular therapy remains substantial, its single-administration structure effectively mitigates the continuous, compounding financial burden frequently observed with multi-agent later-line conventional combinations. Consequently, CAR-T therapy demonstrates a highly favorable incremental cost-effectiveness profile in relapsed/refractory settings, presenting a compelling clinical and economic rationale for its integration into earlier lines of treatment. However, translating these robust quality-adjusted survival gains into global clinical practice requires overcoming significant structural and ethical barriers. Discrepancies in specialized manufacturing infrastructure, complex regional reimbursement frameworks, and socio-economic inequities in patient access must be systematically addressed. Healthcare authorities, researchers, and clinicians must collaboratively utilize standardized utility models to optimize resource allocation, ensure equitable access, and maximize the long-term value of advanced therapeutic modalities in multiple myeloma care. Supplemental material Supplemental material - From toxicity to immunity: Evaluating CAR-T and conventional therapies in multiple myeloma through quality-adjusted life year (QALY) outcomes Supplemental material for From toxicity to immunity: Evaluating CAR-T and conventional therapies in multiple myeloma through quality-adjusted life year (QALY) outcomes by Guluzar Gulnur Itez and Asuman Sunguroglu in Therapeutic Advances in Hematology. Acknowledgements The authors thank Ankara University Stem Cell Institute for academic support. Appendix A Table A1.Modern therapies in multiple myeloma: Clinical role.Therapy/ClassYear (Approval/Milestone)Mechanism/TargetCurrent use in MMKey toxicitiesAutologous Stem Cell Transplant (ASCT)1983 (MM application)Hematopoietic rescueStandard in eligible patientsMyelosuppression, infection riskBortezomib (first PI)108,1092003Reversible 20S proteasome inhibitorBackbone of induction (VRd etc.)Peripheral neuropathyCarfilzomib (2nd-gen PI) 109 2012Irreversible PIRelapsed/refractory, combo regimensCardiac toxicityIxazomib (oral PI) 109 2015Oral PIMaintenance, frail patientsGI, mild cytopeniasLenalidomide (IMiD) 107 2005IMiD, cytokine modulation, T-cell activationFrontline & maintenance backboneCytopenias, thrombosisPomalidomide (IMiD) 110 2013Next-gen IMiDRelapsed/refractory settingsCytopenias, infectionsDaratumumab (anti-CD38 mAb)1,1112015ADCC, CDC, ADCP, direct apoptosisWidely used frontline & RRMMNeutropenia, infusion reactionsIsatuximab (anti-CD38 mAb)1,1112020Distinct CD38 epitope, direct apoptosisCombo with Pd in RRMMNeutropenia, infectionsSelinexor (XPO1 inhibitor)1,1112019Nuclear export inhibition, p53 reactivationTriple-class refractoryGI toxicity, cytopeniasVenetoclax (BCL-2 inhibitor)1,111InvestigationBCL-2 inhibitionTargeted subgroup (t(11;14))Tumor lysis, cytopeniasMelflufen (peptide-drug conjugate)1,1112021 (revoked FDA approval)Alkylating payload via peptide conjugateEMA-approved, not FDACytopenias, survival concernHDAC inhibitors (e.g. panobinostat)1,1112015 (panobinostat FDA)Histone deacetylase inhibitionAdjunct in refractory diseaseGI, fatigue, cytopeniasCheckpoint inhibitors (PD-1/PD-L1)1,111Trials (halted in MM)Immune checkpoint blockadeInvestigational onlyImmune toxicitiesBispecific Abs/CAR-T/NK 1,1112020sRedirected T/NK cytotoxicityRapidly emergingCRS, neurotoxicity, cytopenias Modern therapies in multiple myeloma: Clinical role. Appendix B Proposed treatment pathway for multiple myeloma based on the author’s synthesis of recent clinical studies (2018-2025), reflecting risk stratification, treatment eligibility, and therapeutic failure sequences. This algorithm was constructed by integrating data from the clinical trials and publications cited in the main text. It represents an interpretative synthesis rather than an adaptation of any single guideline. All references used to generate this figure are included in the manuscript’s reference list. Author contributions: Gülüzar Gülnur İtez conceptualized the study, designed the QALY model, and drafted the manuscript. Asuman Sunguroğlu supervised the work and critically reviewed the final version. Both authors read and approved the final manuscript. Funding: The authors received no financial support for the research, authorship, and/or publication of this article. The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Supplemental material: Supplemental material for this article is available online. ORCID iDs Guluzar Gulnur Itez https://orcid.org/0009-0007-5367-6687 Asuman Sunguroğlu https://orcid.org/0000-0001-7693-0958 Ethical considerations This study did not involve any new studies with human participants or animals performed by any of the authors. All data were derived from previously published studies. Therefore, ethical approval and informed consent were not required. Data Availability Statement All data used in this study were obtained from publicly available clinical trial publications and databases, which are cited in the manuscript.* References Associated Data Supplementary Materials Supplemental material - From toxicity to immunity: Evaluating CAR-T and conventional therapies in multiple myeloma through quality-adjusted life year (QALY) outcomes Supplemental material for From toxicity to immunity: Evaluating CAR-T and conventional therapies in multiple myeloma through quality-adjusted life year (QALY) outcomes by Guluzar Gulnur Itez and Asuman Sunguroglu in Therapeutic Advances in Hematology. Data Availability Statement All data used in this study were obtained from publicly available clinical trial publications and databases, which are cited in the manuscript.*