Clinical Outcomes and Patterns of Care Among Patients with Newly Diagnosed Hodgkin Lymphoma in a Safety-Net Health System

Introduction: Patients with Hodgkin lymphoma (HL) have excellent outcomes, with a 5-year survival of 88.9% per Surveillance, Epidemiology, and End Results (SEER) Program estimates. Nevertheless, few studies have explored the patterns of care for uninsured patients, who have inferior survival (Parikh et al., Cancer 2015). We conducted a retrospective cohort study to examine the outcomes for patients with HL treated in a large safety-net health system in the US.

Methods: Patients aged ≥18 years diagnosed with HL between 1/2011-12/2021 at Harris Health (HH), Houston, TX, were included. We abstracted baseline, treatment, and follow-up data from the electronic medical record. FIPS codes derived from home addresses were used to compute the Area Deprivation Index (ADI). We performed time-to-event analysis using the Kaplan-Meier method and examined factors associated with overall survival (OS) using Cox regression.

Results: Of 240 patients with HL, the median age was 39 years (range, 19-88), 65% were men, 82% were uninsured, and 84% lived in a neighborhood with an ADI below the national median. Nearly half (48%) identified as Hispanic and 34% as Black. Most (96%) had classical HL; 72% had stage III/IV disease (median International Prognostic Score [IPS] 4; range, 1-6), and 28% had stage I/II disease (20% unfavorable-risk). Most (63%) had an ECOG performance status (PS) of 0-1. Forty-five (19%) had a history of HIV infection.

Most (97%) received therapy, with a median diagnosis-to-treatment interval (DTI) of 25 days (range, 0-898). The most common first-line (1L) treatments were ABVD (71%), AVD (13%), and BV+AVD (5%); 37 (16%) received radiation therapy (RT; 4 with RT alone). Of treated patients, 80% received all planned 1L cycles. The predominant reasons for treatment discontinuation were progression (6.5%), therapy complication (6.0%), and loss to follow-up (5.6%). Between-cycle delays of >7 days were experienced by 60% of patients because of complications (37%), missed appointments (16%), and appointment unavailability (7.5%). Of 225 evaluable patients, 153 (68%) attained a complete response (CR) and 18 (8%) a partial response (PR); of these, 22 (13%) relapsed. Of 74 with relapsed/refractory HL, 57 (77%) received subsequent therapy, with which 43 (58%) achieved ≥PR. Twenty-two of 40 patients eligible for hematopoietic stem cell transplantation (HSCT) underwent autologous HSCT (n=18), allogeneic HSCT (3), or both (1).

The median follow-up was 72.8 months (range, 0.8-160), with a 5-year OS of 80.4% (95% CI 75.4%-85.8%); 5-year OS rates were 95.5% (95% CI 90.7%-100%) for stage I-II and 74.1% (95% CI 67.6%-81.3%) for stage III-IV disease. The cause of death was lymphoma in 27/53 patients (51%), comorbidity in 12 (23%), and treatment complication in 6 (11%). Factors associated with OS in the multivariable model included age (HR 1.38 per 10-year increase; p=0.003), ECOG PS (HR 2.28 for 2-4 vs. 0-1; p=0.029), lymphocyte-depleted subtype (HR 6.28; p=0.006), albumin (HR 0.56 per 1 g/dL increase; p=0.002), and total bilirubin (HR 1.12 per 1 mg/dL increase; p=0.029). HIV infection (HR 2.08; p=0.0145), ≥3 nodal regions involved (HR 2.51; p=0.004), and IPS (HR 1.37 per point for stage III-IV; p=0.008) were associated with shorter OS only in the univariable model. Sex, race, ethnicity, insurance, ADI, ESR, and DTI were not associated with differences in OS. Adjusted for baseline factors, ABVD, AVD, and BV+AVD were associated with similar OS. Not receiving all 1L cycles (HR 6.78; p<0.0001) and not attaining CR (HR 4.84; p<0.0001) were associated with greater mortality.

Conclusions: This is one of the largest real-world studies examining outcomes and patterns of care of socioeconomically disadvantaged, largely uninsured patients with HL treated in a safety-net health system. Owing to the robust financial assistance programs available at HH, Harris County residents have access to standard-of-care therapeutics regardless of insurance status. A greater proportion of advanced-stage, higher-risk, and HIV-associated HL, frequent treatment delays/interruptions, and limited access to HSCT may have resulted in a slightly shorter OS in this cohort than national SEER estimates. However, we did not observe significant racial/ethnic disparity. Future efforts to expand access to care, enhance treatment adherence, and reduce barriers to follow-up may improve outcomes for uninsured patients in similar settings.