Access barriers to anti-CD19⁺ CART therapy for NHL across a community transplant and cellular therapy network

Anti-CD19⁺ chimeric antigen receptor T-cell (CART) therapy is a breakthrough oncologic advance because of remarkable effectiveness and durable responses.¹ Utilization has accelerated in the United States after the initial US Food and Drug Administration (FDA) approval in 2017 for several indications in B-cell malignancies. The current generation of CART therapies is complicated by high cost, a lengthy and often unpredictable manufacturing process, and acute as well as delayed toxicities of treatment.^2,3 These challenges have required administration at specialized centers that have implemented the necessary infrastructure.

Obstacles to therapy have included disease-related factors, complicated financial considerations, pharmaceutical capacity, patient aspects (age, fitness, and distance), as well as center-related factors.⁴ Logistic bottlenecks include the referral process to tertiary centers, cell therapy fitness evaluation, financial clearance, apheresis appointments, and cell product manufacturing time. During this time, aggressive disease may advance to a point that renders patients ineligible.

Initial FDA approval for anti-CD19 CART for non-Hodgkin lymphoma (NHL) was based upon a smaller population who had received ≥2 lines of therapy. However, CART therapy is advancing to earlier phases of disease activity. Two products have now been approved as second-line therapy for large B-cell lymphoma refractory to first-line chemoimmunotherapy or relapsing within 12 months of treatment: axicabtagene ciloleucel (Axi-cel) based upon the phase 3 ZUMA-7 trial⁵ and lisocabtagene maraleucel (Liso-cel) based upon the phase 3 TRANSFORM trial.⁶ Although access to a tertiary center with CART expertise may have been a limiting factor initially, there are now ∼130 tertiary centers in the United States accredited by the Foundation for the Accreditation of Cellular Therapy to accommodate the growing demand for CART therapy.⁷ 

Despite the growing availability of CART therapies, there remains a discrepancy between patients who meet the criteria for treatment and the actual number of patients who receive it. Community oncologists may struggle with applying the benefits of breakthrough technology such as CARTs to their patients.^8,9 Access barriers only exacerbate the issues faced in the real world, in which patients may have advanced age, impaired performance status, socioeconomic disadvantage, or belong to racial/ethnic minority groups that are typically underrepresented in registrational clinical trials.¹⁰ Real-world evidence (RWE) has demonstrated the efficacy of CART therapy for NHL outside clinical trials^11,12 and provided valuable insights into access barriers.^13,14 

Even when CART therapy is ultimately delivered, the timeliness of administration can become a significant consideration. Delayed access could result in ineligibility/attrition due to disease progression, diminished CART efficacy because of disease progression, or increased toxicities triggered by disease burden.

Providing quality CART outcomes to the community remains a major challenge. The Sarah Cannon Transplant and Cellular Therapy Network (SCTCTN), as a part of HCA Healthcare, is a non–university-based community-centric organization that offers CART therapies at 5 transplant and cellular therapy programs in 3 states. SCTCTN implemented a coordinated approach to streamline the utilization of CART therapies across our network with standardized referral processes, operating procedures, eligibility criteria in compliance with the FDA label, financial approval process, treatment guidelines, inpatient and outpatient toxicity management, and nursing care. Using this foundation, in which there is limited heterogeneity among our programs in the care for CART recipients, we conducted a retrospective review to identify the barriers to delivery of FDA-approved anti-CD19⁺ CARTs for NHL from the time of referral to infusion.

This study was supported by Sarah Cannon, the cancer institute of HCA Healthcare, and conducted in accordance with the US regulations, applicable International Council for Harmonisation of technical requirements for pharmaceuticals for human use E6 international standards of Good Clinical Practice, and institutional research policies and procedures. All authors testify to the accuracy and completeness of the data with acknowledgement that there are limitations to RWE. This research was approved under expedited review as exempt with full waiver of consent and Health Insurance Portability and Accountability Act authorization by an external governing institutional review board. The design, analysis, and data interpretations were conducted independently by the investigators.

All patients referred to SCTCTN specifically for the FDA-approved anti-CD19⁺ CART therapy between 2018 and 2022 were tracked in our prospective registry. These patients were sent by referring physicians, most often external to SCTCTN, who thought their patient would be a candidate. The registry did not include patients with NHL who were seen in our network but not specifically referred or considered for CART. We identified 357 referrals, which were distributed across 5 SCTCTN programs (3 in Texas, 1 in Colorado, and 1 in Tennessee; all are Foundation for the Accreditation of Cellular Therapy–accredited for allogeneic and autologous transplantation and for immune effector cell therapy). The SCTCTN adopted eligibility criteria for standard-of-care CART that were consistent with the FDA label and were prospectively reviewed for eligibility (supplemental Table 1). Patient eligibility conditions also changed over time as Axi-cel and Liso-cel received a second-line indication from the FDA in 2022, leading to a higher proportion of healthier patients, along with earlier referrals from referring oncologists, and greater confidence among network physicians to treat patients with impaired performance status. The choice of CART therapy was determined by individual treating physicians in consultation with their patient. However, the pretreatment evaluation, lymphodepletion, postinfusion monitoring, and toxicity management followed standardized procedures across all centers. For the first 30 days after infusion, all patients were required to have a full-time adult caregiver, whereas outpatient and those who lived >1 hour from the center were required to relocate. Patients were not managed at satellite clinics during the initial 30-day period.

These data represent real-world observational data that were gathered during the course of delivering standard-of-care treatment. Data were extracted from the SCTCTN registry hosted by Genospace, Inc, which integrates information from the electronic medical records of HCA Healthcare (MEDITECH, Canton, MA) and SCTCTN’s transplant and cell therapy clinical workflow and data information technology solution (STAFA Cellular Therapy, Fremont, CA). These efforts were supplemented by manual data extraction for reasons for ineligibility.

The selection for the final sample data is summarized in a filtering flowchart (supplemental Figure 1). Of the 357 patients with NHL referred for CART therapy during the study time period, 51 were “on track” and were waiting to receive their infusion (eg, undergoing workup or cells were undergoing manufacturing) at the time of analysis and were excluded. The remaining 306 patients were categorized into those who were able to receive the infusion (CART infused; n = 182) and those who did not (CART ineligible; n = 124). All CART-infused patients were available for analysis, whereas those who did not receive infusion were not followed beyond the decision not to infuse.

The area deprivation index (ADI) is a composite measure of 17 census variables designed to describe socioeconomic disadvantage based upon neighborhood income, education, employment, household characteristics, and housing, with higher ADI scores indicative of greater social disadvantage.^15-17 State- and national-level ADI were derived from the postal ZIP code of patient residence.

The distance of the patient’s residence to the treatment center was approximated by acquiring the geographical coordinates for the centroid of the patient’s residential ZIP code and the ZIP code of the treatment center. Distance was calculated using Vincenty formula for oblate spheroids in the R package “geosphere.”

Standard descriptive analyses, Fisher exact test for categorical variables, Mann-Whitney U test for continuous variables, 1-way analysis of variance with test for trend, and multivariable logistic regression were performed using GraphPad Prism version 9 for Windows/MacOS (GraphPad Software, Boston, MA; www.graphpad.com). Data were assessed for missingness and generally found adequate for all variables described; no imputation was performed.

Table 1 shows the patient and disease characteristics of the cohorts representing the CART-infused vs the CART-ineligible groups. Among all patients referred for CART therapy, the median age at referral was 61 years (range, 23-87). We observed that White (76%) and non-Hispanic patients (69%) constituted the majority of those referred for CART. However, this is similar to the state census data normalized by the proportion of patients contributed by each center.¹⁸ Diffuse large B-cell lymphoma was the most common indication for referral (69%), followed by follicular NHL (13%), follicular/large B-cell NHL (7%), mantle cell lymphoma (4%), and other B-cell lymphomas (7%). As expected, the number of patients referred increased over time. The primary health coverage for the majority of patients (61%) was private insurance, whereas 31% of patients had Medicare or Medicaid coverage. For those receiving CART, the products used were Axi-cel in 111 (62%), Tisa-cel in 30 (16%), brexucabtagene autoleucel in 24 (13%), and Liso-cel in 17 (9%). The number of lines of prior therapy and the choice of prebridging or bridging therapy, if any, was not available. However, all patients received the minimum lines of prior therapy to be consistent with the FDA label. A significant decline in the proportion of patients deemed ineligible for CART has occurred in the more recent era. The number of ineligible patients in 2018 to 2020 vs 2021 to 2022 was 58 (53.2%) and 66 (33.5%), whereas the number of patients receiving infusion was 51 (46.8%) and 131 (66.5%), respectively (P = .001). Eleven patients (9%) in the group who did not receive CARTs were lost to follow-up, as shown in Table 2.

For all patients referred, the time from initial referral to consultation with a cell therapy specialist was not a significant contributor to delays, with a median of 11 days (interquartile range, 4.5-21). Among the CART-infused cohort, the median time from consultation to infusion was 145 days; and from apheresis to infusion was 32 days. The median duration from consultation to infusion declined steadily over time: the median durations were 207 days in 2019 (and prior), 170 days in 2020, 144 days in 2021, and 108 days in 2022 (P < .0001; Figure 1A) The reduction was driven by a steady reduction in the median time from consultation to collection; 176 days in 2019 (and prior), 137 days in 2020, 98 days in 2021, and 73 days in 2022 (P < .0001; Figure 1B) without any appreciable change in median times from referral to consultation (range, 8-12 days) or between collection and infusion (range, 26-34 days; Figure 1C) There was no significant difference in timings between private and governmental insurance.

The reasons why patients did not proceed to CART are shown in Table 2. A large subgroup of 18 patients were categorized as poor general health. These included patients with general debility (n = 8), death (n = 5), identification of a different malignancy (n = 2), preference for hospice (n = 1), infection (n = 1), or psychosocial ineligibility (n = 1).

Several social determinants of health were explored to compare the CART-infused vs the CART-ineligible groups. Both groups had comparable levels of private insurance and distance from center. CART-ineligible patients had higher median ADI scores at the state level, although the difference was not statistically significant. However, this was significant at the national level (P = .01), suggesting a higher neighborhood social disadvantage. These results are summarized in Table 1.

Patients were similar in both groups with respect to age, sex, race, ethnicity, and disease indication (Table 1). Multivariable logistic regression modeling including age, sex, race/ethnicity, referral year, ADI, financial coverage, and facility distance did not find any variables with statistical significance (supplemental Table 2). National ADI score, which was significant in univariable analysis, was not significant in the multivariable analysis.

We noted that the duration between consultation to infusion decreased steadily over time, and the proportion of patients who received CART also increased over this period. Linear regression analysis confirmed a statistically significant relationship between the mean time to infusion and the proportion of patients receiving CART each year (P = .02).

In this large study of community-based referrals for CART therapies for NHL, we found that 41% of referred patients were not treated. However, over the study period, the proportion of ineligible patients declined from 53.2% (2018-2020) to 33.5% (2021-2022). Ineligible patients and CART-infused patients were similar in terms of demographics and disease characteristics and mostly similar in terms of social determinants of health including private insurance. The primary reasons for ineligibility were attributed to disease progression (34%) and poor health (15%). Notably, CART administration occurred at a median of 145 days from the initial consultation, of which manufacturing time only accounting for a median of 32 days. The time from consultation to infusion declined steadily from 207 days in 2019 to 108 days in 2022, and this was driven by a reduction in time between consultation and collection.

Our findings represent RWE from, to our knowledge, the largest community-focused multicenter network in the United States and highlights several known issues with the currently available CD19-directed CARTs. The complexity of CARTs involves evaluation, authorization, apheresis, and manufacturing. In our study, this practically resulted in a duration of 3 to 4 months until CART infusion. The long duration between consultation and infusion is a unique finding. Despite prompt evaluation at the time of referral, a large proportion could not ultimately benefit from receiving this breakthrough therapy. Disease progression and declining health status while waiting for CART were deemed the primary drivers for ineligibility, whereas demographic, financial, and social factors were less relevant.

There are several implications of our real-world findings. CART response rates and survivals have been generalized from the earlier registrational phase 2 trials of high-performance clinical trial subjects; such studies, analyzed without intent to treat, conceal a substantial selection bias. Analyzing the benefit of CART at a population level or for making broader policy decisions needs to account for patient attrition. Our data suggest that the degree of attrition is significant. The primary driver of ineligibility seemed to be disease progression and health decline, factors that were exacerbated by the long duration between initial consultation, eligibility determination, and when the treatment would ultimately be administered. Newer developments with rapid autologous CART production that reduce vein-to-vein time as well as off-the-shelf allogeneic CART products could fulfill an unmet need.¹⁹ We observed that reducing the duration from consultation to eventual CART infusion is associated with higher rates of CART eligibility. Apart from reducing delays, there are likely other factors favoring increasing eligibility over time, such as earlier referrals, administration at an earlier line in therapy leading to healthier patients, and/or an improving confidence to treat patients with impaired performance status. Apart from an emphasis on overall operational efficiency, we could not identify any specific intervention or policy change that led to improvement.

Disease burden is an important predictor of response to and toxicity of CART therapy.²⁰ Bridging therapy will be necessary in the majority of these patients with NHL.^6,21 Although “bridging therapy” is often defined as an antineoplastic therapy during the period of CART manufacture (ie, between apheresis and infusion), this period accounted for less than one-fourth of the duration between consultation and infusion. Careful attention, therefore, needs to be given to the continuum of therapy starting with “salvage therapy” (ie, prebridging therapy) before apheresis, in which it is critical to use regimens that are not lymphodepleting, followed by “bridging therapy” after apheresis. Although the choice of salvage/bridging therapy, if any, needs to be individualized, polatuzumab plus rituximab is a popular regimen. Patients who require bridging therapy have inferior outcomes, suggesting higher disease risk.²² The overall efficacy of disease-controlling strategies in our population was suboptimal, with 34% being deemed ineligible for disease progression and a further 15% for a decline in health, largely attributable to underlying malignancy. As noted, however, within our network, the disease-controlling strategies did become less of a concern over time because patients proceeded to CART therapy sooner. Regardless, this remains a source for optimization to improve outcomes and lower the toxicity of the therapy.

This study explored social determinants of health in CART access for adults with NHL. One important patient-related access barrier is the burden of traveling a long distance for consultation, apheresis, as well as treatment. In a study from the Vizient database between 2018 and 2020, one-third of CART therapy patients lived >2 hours away from the tertiary center, and the majority of patients belonged to a higher socioeconomic stratum.¹³ In our community-based study, distance did not appear to significantly contribute to patient ineligibility. Unfortunately, 10% of patients were deemed ineligible due to a deficiency in financial coverage, however, we found no significant impact of the type of primary insurance (private vs government) between ineligible patients and patients who received infusion. The ADI at the national level, but not the state level, affected eligibility in univariate analysis but not in multivariable analysis. The state ADI and national ADI of patients who received infusion was substantially more affluent than the state/national averages of 5 and 50, respectively, which is concerning for disparities in infusion relative to the general population, and there could have been a sample-size effect. Furthermore, ADI may not be reliable for multistate analysis.²³ Taken together, these data suggest that treatment delay and disease progression were more consequential than social determinants. However, a limitation from our analysis is the referral bias for the treatment itself; our analysis only included patients who were referred to our centers and does not address potential CART candidates who were not referred to begin with.

A limitation of this study is the lack of granular data regarding the number of lines of prior therapy and the choice of prebridging and/or bridging therapy; disparities in these are highly relevant to access.²⁴ Nevertheless, because disease progression was the single most important factor for nonreceipt of CART, attention to timeliness and effectiveness of interim therapy is critical. We do not present survival outcomes of CART-infused patients in our analysis, but this has been previously described in the real-world setting and found similar to clinical trial data.^11,22,25,26 In addition, our data are not applicable to the pediatric population nor to diseases that are not NHL. Even though the demographics of the referred patients corresponded closely to that of state census data, we cannot exclude the possibility of a minority population being disproportionately excluded from referral for CART because we did not track data on the overall SCTCTN heme malignancy population. Our study does not comprehensively address social determinants of health, by virtue of being limited to the measures studied and by using neighborhood measures rather than actual self-reported longitudinal tracking.

In conclusion, disease-related factors, likely exacerbated by delays in administration, were the primary barriers to CART ineligibility for NHL in the community. Age, disease subtype, and social determinants of health were less relevant in our patient population. Time to administration as well as patient ineligibility have both improved over time. Implications of our findings include the necessity for reducing the time to financial approval, manufacturing time, education efforts to support early referral, and careful attention to disease-controlling strategies while waiting for treatment.

The authors acknowledge members of the HCA Healthcare Research Institute for contributions to project management and manuscript upload and Mary Alice Keller for document preparation. This study was funded by Sarah Cannon Transplant and Cellular Therapy Network and analysis was performed by teams at Sarah Cannon, Genospace, Inc, and HCA Healthcare.

This research was supported, in whole, by HCA Healthcare and/or an affiliated entity of HCA Healthcare.

The views expressed in this publication represent those of the author(s) and do not necessarily represent the official views of HCA Healthcare or any of its affiliated entities.

Contribution: M.B. and N.S.M. conceptualized the study design and contributed to all aspects of the study, including the development of the statistical analysis plan and interpretation of results; B.B., R.B., and C.M. contributed to the data acquisition process and analysis; and all authors contributed critically to manuscript writing and review.

Conflict-of-interest disclosure: All authors report employment with HCA Healthcare.

Correspondence: Minoo Battiwalla, Sarah Cannon Transplant and Cellular Therapy Program at Tristar Centennial Medical Center, HCA Healthcare, 2300 Patterson St, 5th Floor, Nashville, TN 37203; email: minoo.battiwalla@hcahealthcare.com.