• In a large public safety-net health system, OS in patients with LBCL is comparable to Surveillance, Epidemiology, and End Results estimates.

  • Factors associated with OS included R-IPI, National Cancer Institute Comorbidity Index, hemoglobin, and International Normalized Ratio.

Abstract

Real-world outcome data for patients with large B-cell lymphomas (LBCLs) who are uninsured or have socioeconomic barriers to care are limited. We performed a retrospective cohort study of patients with newly diagnosed LBCL treated in a large safety-net hospital system. Between January 2011 and June 2022, 496 patients aged >18 years were diagnosed with LBCL at Harris Health System, Houston, Texas. The median age was 53 years, 75% were uninsured, and 81% were in the most disadvantaged Area Deprivation Index national quartiles. Most (69%) had stage III/IV disease, 44% had poor-risk disease by the Revised International Prognostic Index (R-IPI), and 17% had a history of HIV infection. The median diagnosis-to-treatment interval was 17 days. The median follow-up time was 53.5 months. Among 464 evaluable patients, 66% achieved a complete response, and 11% had a partial response. Of 48 patients, 26 (54%) eligible for cell therapies received them. At 5 years, event-free and overall survival (OS) rates were 57% and 68%, respectively. Factors that affected OS included Hispanic ethnicity (hazard ratio [HR], 0.70; P = .027), R-IPI (HR, 4.67 for poor vs very good risk; P < .001), National Cancer Institute Comorbidity Index (HR, 1.53 per unit increment; P = .003), hemoglobin (HR, 0.89 per unit increment; P = .002), and International Normalized Ratio (HR, 2.17 per unit increment; P = .007). Insurance status was not associated with differences in OS. In our safety-net health system with robust financial assistance programs and limited access to cell therapies, uninsured status was not associated with inferior outcomes. Addressing barriers to care may improve outcomes in other settings.

Diffuse large B-cell lymphoma (DLBCL) is the most common type of non-Hodgkin lymphoma (NHL) and accounts for 40% of NHL cases.1 Multiagent chemotherapy remains the cornerstone of treatment for DLBCL and other large B-cell lymphomas (LBCLs). In the early 2000s, rituximab transformed the landscape of lymphoma therapeutics and dramatically increased survival in patients with LBCL.2 The advent of newer targeted agents, such as polatuzumab vedotin (pola), tafasitamab (tafa), ibrutinib, and loncastuximab tesirine, coupled with advances in hematopoietic stem cell transplantation (HSCT) and development of chimeric antigen receptor T-cell (CAR-T) therapy, has further improved clinical outcomes.3-12 Surveillance, Epidemiology, and End Results (SEER) Program data from 2013 to 2019 estimate a 5-year overall survival (OS) of 64.7% among individuals diagnosed with DLBCL.13 Nevertheless, socioeconomic disparities in LBCL outcomes persist in the modern treatment era.14-17 A retrospective cohort study of 33 032 patients diagnosed with DLBCL in California between 1988 and 2009 identified race/ethnicity, neighborhood socioeconomic status (SES), and insurance status as factors that affected lymphoma-specific survival.16 Another retrospective analysis of 3858 patients showed inferior survival in uninsured and Medicaid-insured patients compared with privately insured patients. In both studies, the observed differences were attenuated after adjustment for neighborhood SES and other potential mediators such as age and stage.17 

Although advances in diagnostics and therapeutics have enhanced survival in patients with LBCL, there remains a paucity of real-world data on outcomes in patients who are uninsured or underinsured or who may have socioeconomic barriers to care. As the intent of therapy is curative for most patients diagnosed with LBCL, minute disparities in access to high-quality care may translate to significant differences in survival outcomes. With a population of 4.8 million, Harris County is the third most populous county in the United States and encompasses the city of Houston, Texas.18 The Harris Health System (HHS) is an integrated academic health care system with a primarily indigent patient population.19 HHS consists of 2 inpatient facilities, Ben Taub General Hospital and Lyndon B. Johnson General Hospital, and 18 community health centers.19 Hematology and oncology clinics associated with the 2 hospitals are staffed by physicians from Baylor College of Medicine and The University of Texas MD Anderson Cancer Center, respectively.19 Patient characteristics at these locations are similar, with the primary difference being the proximity of patients’ residence from the clinical sites. Forty-six percent of HHS patients are uninsured, and an additional 33% rely on public insurance for medical care.19 Financial assistance is available to Harris County residents with a household income <150% of the federal poverty level.19 HHS patients have access to standard-of-care therapeutics through the health system’s formulary and pharmaceutical assistance programs. However, barriers to clinical trials and cell therapies including HSCT and CAR-T therapy exist. This study aims to elucidate contemporary practice patterns and clinical outcomes of patients with newly diagnosed LBCL in this large urban safety-net hospital system in the United States.

Patients with newly diagnosed LBCL at HHS between 1 January 2011 and 30 June 2022, aged ≥18 years, were included in the analysis. The following histologies were considered LBCL: DLBCL, not otherwise specified; DLBCL/high-grade B-cell lymphoma (HGBCL) with MYC and BCL2 rearrangements; plasmablastic lymphoma (PBL); primary mediastinal LBCL; T-cell/histiocyte-rich LBCL; primary effusion lymphoma (PEL); mediastinal gray zone lymphoma; intravascular LBCL; and primary cutaneous DLBCL. Primary central nervous system lymphoma was excluded because of significant differences in prognosis and management. The protocol was approved by the institutional review board of Baylor College of Medicine in accordance with the Declaration of Helsinki.

Demographic, disease, treatment, response, and follow-up data were abstracted from the electronic medical record, and data from the HHS Cancer Registry were used to supplement survival data. US Census Block Group Federal Information Processing System codes derived from geocoded home addresses at the time of diagnosis were used to compute the Area Deprivation Index (ADI).20-23 The diagnosis of LBCL required histologic confirmation by biopsy of lymph node, bone marrow, and/or other involved tissue. Immunophenotyping was performed using multiparameter flow cytometry in a Clinical Laboratory Improvement Amendments certified laboratory. Interphase fluorescence in situ hybridization was used to determine BCL2, BCL6, and MYC rearrangement status of each patient’s lymphoma. The National Cancer Institute Comorbidity Index (NCI-CI) was calculated using each patient’s medical diagnoses.24,25 Patients with histologic diagnoses other than PEL were risk-stratified using the Revised International Prognostic Index (R-IPI) according to their age, stage, Eastern Cooperative Oncology Group (ECOG) performance status (PS), serum lactate dehydrogenase level, and number of extranodal sites of involvement at the time of LBCL diagnosis.26 In addition, the Prognostic Nutritional Index (PNI) was computed using the absolute lymphocyte count and serum albumin level at the time of diagnosis.27 Treatment response was assessed using the Lugano Classification lymphoma response criteria.28 Response in patients who underwent positron emission tomography/computed tomography (PET/CT) imaging at the end of frontline therapy was determined using PET/CT; response in the remaining patients was assessed via contrast computed tomography imaging. Eligibility for cell therapies for relapsed/refractory (R/R) disease was verified by a trained clinician abstractor (J.Y.J.) on the basis of patients’ age, PS, comorbidities, and clinical response to salvage chemotherapy at the time of evaluation.

The primary outcomes evaluated were event-free survival (EFS; time from diagnosis to primary refractory disease, progression, relapse, initiation of subsequent therapy for reasons other than intolerance, or death from any cause) and OS (time from diagnosis to death from any cause). The distributions of continuous variables were characterized by medians and interquartile ranges; the distributions of categorical variables were summarized by frequencies and percentages. Baseline characteristics were compared using the Wilcoxon rank sum test, Pearson χ2 test, and Fisher exact test, as appropriate. To minimize immortal time bias, initiation of systemic therapy, completion of first-line therapy, and response to first-line therapy were treated as time-varying covariates. Time-to-event analysis was performed using the Kaplan-Meier method, and Cox proportional hazards regression was used to examine factors associated with OS. Bidirectional stepwise regression was used to generate a multivariable Cox model incorporating pretreatment characteristics after locking R-IPI as a mandatory covariate in the model. All statistical analysis was completed using R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).

Demographics and baseline characteristics of the study cohort

Among 496 patients with newly diagnosed LBCL (Table 1), the median age was 53 years (range, 18-93), and 60% identified as Hispanic. Seventy-five percent were uninsured at the time of diagnosis and required county funds for care; 81% were in the 2 most disadvantaged ADI national quartiles. Most had advanced-stage (III or IV) disease and had an ECOG PS of 0 or 1 at the time of diagnosis. By R-IPI, 42% had good-risk and 44% had poor-risk disease.

Table 1.

Baseline characteristics of study cohort

N = 496
Age at diagnosis, median (IQR) 53 (44-62) 
Male sex, n (%) 289 (58%) 
Race/ethnicity, n (%)  
Non-Hispanic White 68 (14%) 
Non-Hispanic Black 97 (20%) 
Hispanic 297 (60%) 
Asian and Pacific Islander 33 (6.7%) 
Preferred language, n (%)  
English 253 (51%) 
Spanish 223 (45%) 
Other/unknown 20 (4.0%) 
Insurance status, n (%)  
Private 26 (5.2%) 
Medicare 24 (4.8%) 
Medicaid 72 (15%) 
Uninsured 373 (75%) 
ADI national quartile, n (%)  
34 (6.9%) 
60 (12%) 
180 (37%) 
217 (44%) 
Histology, n (%)  
DLBCL, NOS, germinal center 184 (37%) 
DLBCL, NOS, nongerminal center 172 (35%) 
DLBCL, NOS, unclassifiable 57 (11%) 
PBL 27 (5.4%) 
DLBCL/HGBCL with MYC/BCL2 rearrangements 16 (3.2%) 
Primary mediastinal LBCL 15 (3.0%) 
T-cell/histiocyte-rich LBCL 10 (2.0%) 
PEL 6 (11.2%) 
Mediastinal gray zone lymphoma 5 (1.0%) 
Intravascular LBCL 3 (0.6%) 
Primary cutaneous DLBCL 1 (0.2%) 
Stage, n (%)  
Early stage (I-II) 156 (31%) 
Advanced stage (III-IV) 340 (69%) 
ECOG PS, n (%)  
0-1 232 (47%) 
2-4 126 (25%) 
Unknown 138 (28%) 
R-IPI risk group, n (%)  
Very good 68 (14%) 
Good 206 (42%) 
Poor 215 (44%) 
History of HIV infection, n (%) 82 (17%) 
NCI-CI, median (IQR) 0.00 (0.00-0.58) 
Hemoglobin, median (IQR), g/dL 11.80 (9.88-13.20) 
Serum creatinine, median (IQR), mg/dL 0.80 (0.70-1.00) 
Serum albumin, median (IQR), g/dL 3.40 (2.90-3.80) 
INR, median (IQR) 1.10 (1.00-1.20) 
Serum LDH (× ULN), median (IQR) 1.22 (0.82-2.16) 
PNI, median (IQR) 40 (33-47) 
N = 496
Age at diagnosis, median (IQR) 53 (44-62) 
Male sex, n (%) 289 (58%) 
Race/ethnicity, n (%)  
Non-Hispanic White 68 (14%) 
Non-Hispanic Black 97 (20%) 
Hispanic 297 (60%) 
Asian and Pacific Islander 33 (6.7%) 
Preferred language, n (%)  
English 253 (51%) 
Spanish 223 (45%) 
Other/unknown 20 (4.0%) 
Insurance status, n (%)  
Private 26 (5.2%) 
Medicare 24 (4.8%) 
Medicaid 72 (15%) 
Uninsured 373 (75%) 
ADI national quartile, n (%)  
34 (6.9%) 
60 (12%) 
180 (37%) 
217 (44%) 
Histology, n (%)  
DLBCL, NOS, germinal center 184 (37%) 
DLBCL, NOS, nongerminal center 172 (35%) 
DLBCL, NOS, unclassifiable 57 (11%) 
PBL 27 (5.4%) 
DLBCL/HGBCL with MYC/BCL2 rearrangements 16 (3.2%) 
Primary mediastinal LBCL 15 (3.0%) 
T-cell/histiocyte-rich LBCL 10 (2.0%) 
PEL 6 (11.2%) 
Mediastinal gray zone lymphoma 5 (1.0%) 
Intravascular LBCL 3 (0.6%) 
Primary cutaneous DLBCL 1 (0.2%) 
Stage, n (%)  
Early stage (I-II) 156 (31%) 
Advanced stage (III-IV) 340 (69%) 
ECOG PS, n (%)  
0-1 232 (47%) 
2-4 126 (25%) 
Unknown 138 (28%) 
R-IPI risk group, n (%)  
Very good 68 (14%) 
Good 206 (42%) 
Poor 215 (44%) 
History of HIV infection, n (%) 82 (17%) 
NCI-CI, median (IQR) 0.00 (0.00-0.58) 
Hemoglobin, median (IQR), g/dL 11.80 (9.88-13.20) 
Serum creatinine, median (IQR), mg/dL 0.80 (0.70-1.00) 
Serum albumin, median (IQR), g/dL 3.40 (2.90-3.80) 
INR, median (IQR) 1.10 (1.00-1.20) 
Serum LDH (× ULN), median (IQR) 1.22 (0.82-2.16) 
PNI, median (IQR) 40 (33-47) 

IQR, interquartile range; LDH: lactate dehydrogenase; NOS: not otherwise specified; ULN: upper limit of normal.

Notably, 17% of patients had a history of HIV infection at the time of diagnosis. The median HIV viral load was 69 400 (range, undetected to 4 530 000), and the median CD4+ T-cell count was 115 (range, 1-523); 29% were on highly active antiretroviral therapy (HAART) at the time of diagnosis. Patients who were HIV-seropositive were more likely to be younger (median age, 47 vs 56 years; P < .0001), male (82% vs 54%; odds ratio [OR], 3.85; 95% confidence interval [CI], 2.09-7.51; P < .0001), Black (40% vs 15%; OR, 3.66; 95% CI, 2.11-6.32; P < .0001), and insured (39% vs 22%; OR, 2.27; 95% CI, 1.32-3.85; P = .002), and to have a lower hemoglobin level (10.2 vs 12.1 g/dL; P < .0001), serum albumin level (3.1 vs 3.5 g/dL; P < .0001), and PNI (36 vs 41; P < .0001) than those who were HIV-seronegative. People with HIV were overrepresented in the PBL (20/27 [74%]; OR, 18.6; 95% CI, 7.18-54.1; P < .0001) and PEL groups (4/6 [67%]; OR, 10.5; 95% CI, 1.47-118; P = .008).

Treatment regimens, deviations from treatment plan, and treatment response

A total of 478 patients (96%) received lymphoma-directed therapy (Table 2). The remainder received no treatment because of poor PS (n = 13), loss to follow-up (n = 4), or definitive evidence of lymphoma resolution on follow-up evaluation (n = 1). The median diagnosis-to-treatment interval (DTI; time from diagnosis to treatment initiation) was 17 days (range, 0-432). The most common frontline treatment regimens were rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP. 48%); rituximab, etoposide, prednisone, vincristine, cyclosphophamide, and doxorubicin (R-EPOCH, 40%); and rituximab, cyclosphosphamide, vincristine, and prednisone (R-CVP. 3.8%). Other regimens included rituximab, cyclophosphamide, etoposide, vincristine, and prednisone as well as rituximab with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with methotrexiate and cytarabine (supplemental Table 1). Rituximab was omitted in cases of CD20 disease. Central nervous system prophylaxis was administered as clinically indicated in accordance with standard practice. R-EPOCH was more likely to be selected over R-CHOP in patients with PBL (OR, not estimable; 95% CI, 5.78-∞; P < .0001), primary mediastinal LBCL (OR, 18.2; 95% CI, 2.72-774; P < .0001), DLBCL/HGBCL with MYC and BCL2 rearrangements (OR, 8.37; 95% CI, 1.86-77.3; P = .001), HIV infection (OR, 5.51; 95% CI, 2.87-11.2; P < .0001), and R-IPI ≥2 (OR, 2.11; 95% CI, 1.36-3.30; P = .0006).

Table 2.

Treatment regimens and clinical response

N = 496
Treatment status, n (%)  
Treatment 478 (96%) 
No treatment 18 (3.6%) 
Time to treatment (d), median (IQR) 17 (8, 37) 
Initial regimen, n (%)  
R-CHOP 231 (48%) 
R-EPOCH 190 (40%) 
R-CVP 18 (3.8%) 
Other 39 (8.2%) 
Cycles received, n (%)  
1-2 47 (10%) 
3-5 77 (16%) 
289 (61%) 
≥7 60 (13%) 
Treatment completed, n (%) 363 (76%) 
End-of-treatment PET/CT, n (%) 298 (64%) 
Initial response, n (%)  
CR 307 (64%) 
PR 49 (10%) 
SD 11 (2.3%) 
PD 97 (20%) 
Unknown 14 (2.9%) 
Evidence of R/R disease  
During follow-up period, n (%) 189 (38%) 
Receipt of subsequent therapy, n (%) 138 (28%) 
Response to subsequent therapy, n (%) 69 (14%) 
Eligibility for cell therapy, n (%) 48 (9.7%) 
Receipt of cell therapy, n (%) 26 (12%) 
Type of cell therapy, n (%)  
HSCT only 14 (54%) 
CAR-T only 7 (27%) 
HSCT and CAR-T 5 (19%) 
Vital status, n (%)  
Alive 325 (66%) 
Deceased 171 (34%) 
N = 496
Treatment status, n (%)  
Treatment 478 (96%) 
No treatment 18 (3.6%) 
Time to treatment (d), median (IQR) 17 (8, 37) 
Initial regimen, n (%)  
R-CHOP 231 (48%) 
R-EPOCH 190 (40%) 
R-CVP 18 (3.8%) 
Other 39 (8.2%) 
Cycles received, n (%)  
1-2 47 (10%) 
3-5 77 (16%) 
289 (61%) 
≥7 60 (13%) 
Treatment completed, n (%) 363 (76%) 
End-of-treatment PET/CT, n (%) 298 (64%) 
Initial response, n (%)  
CR 307 (64%) 
PR 49 (10%) 
SD 11 (2.3%) 
PD 97 (20%) 
Unknown 14 (2.9%) 
Evidence of R/R disease  
During follow-up period, n (%) 189 (38%) 
Receipt of subsequent therapy, n (%) 138 (28%) 
Response to subsequent therapy, n (%) 69 (14%) 
Eligibility for cell therapy, n (%) 48 (9.7%) 
Receipt of cell therapy, n (%) 26 (12%) 
Type of cell therapy, n (%)  
HSCT only 14 (54%) 
CAR-T only 7 (27%) 
HSCT and CAR-T 5 (19%) 
Vital status, n (%)  
Alive 325 (66%) 
Deceased 171 (34%) 

PD, progressive disease; SD, stable disease.

A total of 363 patients (76%) received all planned cycles of initial therapy. The median number of cycles received was 6 (range, 1-9) for patients with both early- and advanced-stage disease. Reasons for discontinuation were progression (n = 54, 47%), complication (n = 25, 22%), patient factors (n = 23, 20%), loss to follow-up (n = 7, 6.1%), loss of medical coverage (n = 5, 4.3%), and incarceration (n = 1, 0.8%). Specific patient factors that led to discontinuation included missed appointments (n = 9), refusal of further therapy (n = 7), lack of transportation (n = 2), substance use (n = 2), psychiatric illness (n = 1), housing instability (n = 1), and work obligations (n = 1). Patients were more likely to complete first-line therapy (supplemental Table 2) if they had no insurance at the time of diagnosis (OR, 1.93; 95% CI, 1.18-3.14; P = .006), a lower ECOG PS (OR, 3.66 for 0-1 vs 2; 95% CI, 2.14-6.33; P < .0001), a more favorable R-IPI (OR, 2.73; very good- or good- vs poor-risk disease; 95% CI, 1.73-4.35; P < .0001), no history of HIV infection (OR, 3.12; 95% CI, 1.80-5.36; P < .0001), and a lower NCI-CI (median, 0 vs 0.29; P < .0001). Of 447 patients who received ≥2 cycles of first-line therapy, 238 (53%) had at least 1 cycle of treatment delayed by >7 days. The most common causes for delays were complication (n = 120, 50%; including infection in 53 and cytopenia in 29), bed or appointment availability (n = 47, 20%), patient factors (n = 39, 16%), lapse in medical coverage (n = 26, 11%), severe weather event (n = 5, 2.1%), and incarceration (n = 1, 0.4%). Specific patient factors that resulted in treatment delays included missed appointments (n = 21), personal commitments (n = 13), lack of transportation (n = 2), substance use (n = 1), psychiatric illness (n = 1), and work obligations (n = 1).

The median follow-up time for the cohort was 53.5 months (range, 0.2-152 months). Fourteen treated patients (2.9%) did not have adequate follow-up or subsequent imaging and could not be evaluated for response. Among 464 evaluable patients, 64% underwent metabolic assessment by PET/CT. A total of 307 patients (66%) achieved an initial complete response (CR) and 49 (11%) achieved a partial response (PR), whereas 108 (23%) had refractory disease (stable disease or progressive disease). Of those with initial CR or PR, 79 (22%) had a relapse. There were 138 patients (73%) who received subsequent-line therapy (supplemental Table 3), which included chemoimmunotherapy (n = 133, 96%), pola (n = 16, 12%), ibrutinib (n = 9, 6.5%), tafa (n = 7, 5.1%), brentuximab vedotin (n = 3, 2.2%), bispecific antibodies (n = 1, 0.1%), and radiation therapy (n = 33, 24%). Nineteen (14%) of these patients, including 12 patients who were uninsured at the time of diagnosis, were referred to other centers and enrolled in clinical trials; all patients who were uninsured at the time of diagnosis had to have acquired health insurance before clinical trial evaluation at other centers. The median number of subsequent lines of treatment received was 1 (range, 1-9). Many uninsured patients were able to receive novel therapies such as pola (10/16 [63%] of patients who received the drug were uninsured at the time of diagnosis), ibrutinib (7/9 [78%]), tafa (5/7 [57%]), and brentuximab vedotin (2/3 [67%]) by means of pharmaceutical assistance programs. Half achieved CR or PR with salvage therapy.

Of the 48 patients who were eligible for cell therapies after diagnosis of R/R LBCL, 26 (54%) received them; 19 patients (73%) received an autologous or haploidentical allogeneic HSCT, 12 (46%) received CAR-T therapy, and 5 (19%) received both modalities during their treatment course. Twenty-one patients could not receive cell therapies because of lack of insurance or residency status; 1 declined to be evaluated for cell therapies. The median lines of salvage therapy were 1.5 (range, 1-5) and 2 (range, 1-9) for eligible patients who received cell therapy and those who could not receive cell therapy, respectively. Five-year OS from the time of progression was 59% (95% CI, 41%-84%) vs 42% (95% CI, 22%-78%; P = .6).

Survival outcomes and prognostic factors

At the time of analysis, 38.1% of patients (n = 189) had developed R/R disease, and 34.5% (n = 171) had died. Cause of death was related to progression or therapy in 77.2% of patients (n = 132), due to other comorbidities (including other malignancies) in 4.7% (n = 8), and unknown in 18.1% because of loss to follow-up (n = 31). There were 27 early deaths (5.4%), defined as deaths that occurred within 90 days of diagnosis. At 2 and 5 years, EFS rates were 63% and 57% and OS rates were 76% and 68% for the cohort and for patients with classic DLBCL (DLBCL, not otherwise specified, and DLBCL/HGBCL with MYC and BCL2 rearrangements; Figure 1; supplemental Table 4). Compared with patients with germinal center B-cell–like DLBCL, those with non–germinal center B-cell-like DLBCL (hazard ratio [HR], 1.63; 95% CI, 1.13-2.35; P = .009), DLBCL/HGBCL with MYC and BCL2 rearrangements (HR, 2.16; 95% CI, 1.02-4.55; P = .044), and PBL had a greater risk of death (HR, 3.47; 95% CI, 2.02-5.94; P < .001). In univariable Cox regression models, other factors associated with OS included age, Hispanic ethnicity, stage, ECOG PS, R-IPI, NCI-CI, hemoglobin, serum creatinine, serum albumin, International Normalized Ratio (INR), serum lactate dehydrogenase, and PNI (Table 3). Patients with HIV infection also showed a trend toward shorter OS (OR, 1.45; 95% CI, 1.00-2.10; P = .050). Among patients with HIV-associated LBCL, HIV viral load, CD4+ T-cell count, and receipt of HAART at the time of diagnosis did not affect OS (P = .3, .5, and .2).

Figure 1.

Kaplan-Meier estimates of EFS and OS by R-IPI. ∗Classic DLBCL includes DLBCL, not otherwise specified, and DLBCL/HGBCL with MYC/BCL2 rearrangements.

Figure 1.

Kaplan-Meier estimates of EFS and OS by R-IPI. ∗Classic DLBCL includes DLBCL, not otherwise specified, and DLBCL/HGBCL with MYC/BCL2 rearrangements.

Close modal
Table 3.

Effect of baseline characteristics and DTI on OS

Univariable modelMultivariable model
HR95% CIP value∗HR95% CIP value∗
Age at diagnosis 1.03 1.02-1.04 <.001    
Male sex 1.10 0.81-1.49 .5    
Race/ethnicity       
Non-Hispanic White Ref Ref     
Non-Hispanic Black 1.20 0.75-1.93 .4    
Hispanic 0.69 0.45-1.06 .091    
Asian and Pacific Islander 0.83 0.41-1.67 .6    
Hispanic ethnicity 0.65 0.48-0.88 .005 0.70 0.52-0.96 .027 
Preferred language       
English Ref Ref     
Spanish 0.74 0.54-1.02 .063    
Other/unknown 1.53 0.80-2.93 .2    
Insurance status       
Private Ref Ref     
Medicare 1.57 0.71-3.45 .3    
Medicaid 0.95 0.48-1.89 .9    
Uninsured 0.56 0.30-1.04 .067    
ADI national quartile       
Ref Ref     
0.93 0.49-1.77 .8    
0.76 0.43-1.34 .3    
0.67 0.38-1.17 .2    
Histology       
DLBCL, NOS, germinal center Ref Ref  Ref Ref  
DLBCL, NOS, nongerminal center 1.63 1.13-2.35 .009 1.29 0.89-1.87 .2 
DLBCL, NOS, unclassifiable 0.93 0.54-1.62 .8 0.94 0.54-1.65 .8 
PBL 3.47 2.02-5.94 <.001 3.82 2.09-6.99 <.001 
DLBCL/HGBCL with MYC/BCL2 rearrangements 2.16 1.02-4.55 .044 1.35 0.63-2.87 .4 
Primary mediastinal LBCL 0.39 0.10-1.61 .2 0.35 0.08-1.50 .2 
T-cell/histiocyte-rich LBCL 1.10 0.34-3.54 .9 0.75 0.23-3.48 .6 
PEL 1.15 0.28-4.74 .3 0.80 0.18-3.45 .8 
Mediastinal gray zone lymphoma 2.05 0.50-8.44 .3 2.03 0.49-8.43 .3 
Intravascular LBCL 0.00 NE >.9 0.00 NE >.9 
Primary cutaneous DLBCL 0.00 NE >.9 0.00 NE >.9 
Stage       
Early stage (I-II) Ref Ref     
Advanced stage (III-IV) 2.91 1.94-4.37 <.001    
ECOG PS       
0-1 Ref Ref     
2-4 3.68 2.58-5.25 <.001    
Unknown 1.52 1.03-2.26 .037    
R-IPI risk group       
Very good Ref Ref  Ref Ref  
Good 2.27 1.08-4.80 .031 1.95 0.91-4.16 .084 
Poor 6.71 3.27-13.8 <.001 4.67 2.23-9.78 <.001 
History of HIV infection 1.45 1.00-2.10 .050 0.70 0.44-1.11 .13 
NCI-CI 1.94 1.51-2.50 <.001 1.53 1.15-2.04 .003 
Hemoglobin, g/dL 0.81 0.76-0.87 <.001 0.89 0.83-0.96 .002 
Serum creatinine, mg/dL 1.29 1.14-1.46 <.001    
Serum albumin, g/dL 0.54 0.44-0.66 <.001    
INR 2.40 1.57-3.65 <.001 2.17 1.24, 3.79 .007 
Serum LDH, ×ULN 1.10 1.07-1.14 <.001    
PNI 0.94 0.92-0.96 <.001    
Time to treatment       
≤6 d Ref Ref     
7-13 d 1.04 0.64-1.70 .9    
14-20 d 1.24 0.75-2.08 .4    
21-27 d 1.20 0.67-2.15 .5    
28-34 d 0.57 0.27-1.24 .2    
≥35 d 0.09 0.62-1.55 .9    
Univariable modelMultivariable model
HR95% CIP value∗HR95% CIP value∗
Age at diagnosis 1.03 1.02-1.04 <.001    
Male sex 1.10 0.81-1.49 .5    
Race/ethnicity       
Non-Hispanic White Ref Ref     
Non-Hispanic Black 1.20 0.75-1.93 .4    
Hispanic 0.69 0.45-1.06 .091    
Asian and Pacific Islander 0.83 0.41-1.67 .6    
Hispanic ethnicity 0.65 0.48-0.88 .005 0.70 0.52-0.96 .027 
Preferred language       
English Ref Ref     
Spanish 0.74 0.54-1.02 .063    
Other/unknown 1.53 0.80-2.93 .2    
Insurance status       
Private Ref Ref     
Medicare 1.57 0.71-3.45 .3    
Medicaid 0.95 0.48-1.89 .9    
Uninsured 0.56 0.30-1.04 .067    
ADI national quartile       
Ref Ref     
0.93 0.49-1.77 .8    
0.76 0.43-1.34 .3    
0.67 0.38-1.17 .2    
Histology       
DLBCL, NOS, germinal center Ref Ref  Ref Ref  
DLBCL, NOS, nongerminal center 1.63 1.13-2.35 .009 1.29 0.89-1.87 .2 
DLBCL, NOS, unclassifiable 0.93 0.54-1.62 .8 0.94 0.54-1.65 .8 
PBL 3.47 2.02-5.94 <.001 3.82 2.09-6.99 <.001 
DLBCL/HGBCL with MYC/BCL2 rearrangements 2.16 1.02-4.55 .044 1.35 0.63-2.87 .4 
Primary mediastinal LBCL 0.39 0.10-1.61 .2 0.35 0.08-1.50 .2 
T-cell/histiocyte-rich LBCL 1.10 0.34-3.54 .9 0.75 0.23-3.48 .6 
PEL 1.15 0.28-4.74 .3 0.80 0.18-3.45 .8 
Mediastinal gray zone lymphoma 2.05 0.50-8.44 .3 2.03 0.49-8.43 .3 
Intravascular LBCL 0.00 NE >.9 0.00 NE >.9 
Primary cutaneous DLBCL 0.00 NE >.9 0.00 NE >.9 
Stage       
Early stage (I-II) Ref Ref     
Advanced stage (III-IV) 2.91 1.94-4.37 <.001    
ECOG PS       
0-1 Ref Ref     
2-4 3.68 2.58-5.25 <.001    
Unknown 1.52 1.03-2.26 .037    
R-IPI risk group       
Very good Ref Ref  Ref Ref  
Good 2.27 1.08-4.80 .031 1.95 0.91-4.16 .084 
Poor 6.71 3.27-13.8 <.001 4.67 2.23-9.78 <.001 
History of HIV infection 1.45 1.00-2.10 .050 0.70 0.44-1.11 .13 
NCI-CI 1.94 1.51-2.50 <.001 1.53 1.15-2.04 .003 
Hemoglobin, g/dL 0.81 0.76-0.87 <.001 0.89 0.83-0.96 .002 
Serum creatinine, mg/dL 1.29 1.14-1.46 <.001    
Serum albumin, g/dL 0.54 0.44-0.66 <.001    
INR 2.40 1.57-3.65 <.001 2.17 1.24, 3.79 .007 
Serum LDH, ×ULN 1.10 1.07-1.14 <.001    
PNI 0.94 0.92-0.96 <.001    
Time to treatment       
≤6 d Ref Ref     
7-13 d 1.04 0.64-1.70 .9    
14-20 d 1.24 0.75-2.08 .4    
21-27 d 1.20 0.67-2.15 .5    
28-34 d 0.57 0.27-1.24 .2    
≥35 d 0.09 0.62-1.55 .9    

∗ Bold values denote statistical significance at the P <0.05 level.

BCL2, B-cell leukemia/lymphoma 2 gene; MYC, myelocytomatosis gene; NE, not estimable; NOS, not otherwise specified; Ref, reference.

In a multivariable Cox model, factors associated with OS included Hispanic ethnicity (HR, 0.70; P = .027), R-IPI (HR, 4.67 for poor- vs very good-risk disease; P < .001), NCI-CI (HR, 1.53 per unit increment; P = .003), hemoglobin (HR, 0.89 per unit increment; P = .002), and INR (HR, 2.17 per unit increment; P = .007). Preferred language and insurance status were not associated with differences in OS. DTI and delays between cycles of first-line chemotherapy also did not have an appreciable effect on OS before or after adjusting for baseline characteristics. After adjustment for histology, HIV status, and R-IPI, R-CHOP and R-EPOCH were associated with a similar OS (HR, 1.04 vs R-CHOP; P = .8), whereas other regimens were associated with inferior OS (HR, 3.07 vs R-CHOP; P < .0001) (supplemental Table 5). Receiving fewer than planned cycles of frontline therapy was also associated with higher mortality (HR, 4.78; 95% CI, 3.44-6.64; P < .0001), as was not attaining CR at the end of therapy (HR, 6.43; 95% CI, 4.40-9.40; P < .0001; time-varying covariates not included in baseline models).

To our knowledge, our study represents one of the largest real-world studies to date examining the clinical outcomes of underserved, primarily uninsured, and mostly Hispanic patients with newly diagnosed LBCL in a safety-net hospital system in the United States. Overall, our patients had shorter EFS and OS than highly selected patients in contemporary clinical trials.29-31 However, the observed 5-year OS of patients with classic DLBCL was comparable with SEER estimates and to OS rates in other real-world retrospective analyses, including the R-IPI derivation and validation cohorts.13,26,32,33 

This study provides valuable insights into the delivery of care for uninsured patients diagnosed with LBCL. Notably, three-quarters of patients in our cohort were uninsured. This figure far exceeds the uninsured segment in the general HHS patient population (46%), Harris County (24%), and other similar studies (10%-20%).16-18 Unlike in previous population-based studies,16,17 uninsured status was not associated with inferior outcomes in our cohort, and uninsured patients were also paradoxically more likely to complete first-line therapy than insured patients. This may be related in part to the fact that many HHS patients are migrants from outside the Unites States, primarily Latin America. Although many of these patients do have barriers to care associated with social determinants of health (SDOH), they tend to be younger, have fewer comorbidities, and have better PS in an instance of the “healthy migrant effect.”34 Our study also recapitulates the finding that Hispanic patients with LBCL may have longer survival.35,36 Further studies are needed to clarify whether there are biological differences underlying the survival difference.34 

Additionally, as a safety-net health system, HHS has accrued strategies to reduce barriers to care related to lack of insurance. Upon patients’ presentation to the HHS, a financial support team is engaged to facilitate procurement of insurance for eligible patients and enrollment in the HHS financial assistance program for others with household incomes <150% of the federal poverty level. To minimize treatment delays, although patients await verification of eligibility, lymphoma-directed therapy is frequently commenced in the inpatient setting. As a result, DTI in our cohort was comparable to that reported in population-based studies.15 In contradistinction to previous studies, a shorter DTI was not associated with adverse clinical features and decreased survival in our cohort.15 Because the impact of DTI on survival in patients with LBCL is likely mediated by disease aggressiveness, we hypothesize that the association between DTI and survival may be less pronounced in populations with prominent SDOH-related access issues.

At HHS, expert physicians, advanced practice providers, and hematology/oncology fellows from major academic medical centers provide care for patients with lymphoma. Uninsured patients who qualify for financial assistance receive standard-of-care chemoimmunotherapy and novel therapies including pola and tafa at a low or no cost. Higher-cost therapies are secured through pharmaceutical company assistance programs, whereas the costs of other drugs are subsumed into the HHS operating budget. Response rates to first-line and salvage therapy in our cohort were nearly identical to those in other cohorts.7 Uninsured patients who are US citizens or who have been US residents for a minimum of 5 years can also receive HSCT or CAR-T therapy through cell therapy programs affiliated with HHS. Other uninsured patients do not have access to cell therapy unless they can purchase health insurance. Altogether, only 54% of patients with R/R disease eligible for cell therapy received such therapy, with citizenship and residency status being the major barrier to cell therapy for all but 1 patient. Although some patients from Latin America did return to their home countries to pursue HSCT, most declined out of concerns of no re-entry, persecution, or financial burden. Similarly, uninsured patients generally cannot be evaluated for clinical trials at other institutions without obtaining insurance, as HHS does not have funding to defray the costs of such evaluations.

The patient population served by HHS reflects the demographics of Harris County residents, 44.6% of whom self-identify as Hispanic, 27.1% as non-Hispanic White, and 20.6% as Black.18 Cell therapies prolong survival in patients with R/R NHL,7-12,37 yet studies have shown that patients who are Hispanic, Black, underinsured, and less affluent are less likely to receive cell therapies.38-40 With CAR-T therapy now being used as a second-line treatment, the survival gap related to SDOH may increase further because of differential access to cell therapies. Thus, efforts to reduce barriers to these treatments for marginalized patient populations are urgently needed. Toward this end, the Dan L Duncan Comprehensive Cancer Center at Baylor College of Medicine is piloting the first cell therapy program within a safety-net health system by implementing phase 1 and 2 cell therapy trials at HHS.41 By bringing these opportunities to patient populations traditionally underrepresented in clinical trials and helping HHS achieve Foundation for the Accreditation of Cellular Therapy accreditation, the pilot aims to foster collaboratives with industry partners and other funding sources and to effect policy changes that facilitate expansion of cell therapy programs to other health care settings.41 

Additional opportunities to improve the delivery of care exist. Minority individuals, who may experience more SDOH-related barriers to care, comprise most HHS patients. In our study, a sizable proportion of treatment delays and interruptions were attributable to potentially preventable causes such as schedule conflicts, lack of transportation, lapse in insurance coverage, and missed appointments for other reasons. In one prospective study of 204 patients with aggressive LBCL treated at a comprehensive cancer center, nurse navigation support for minority patients was proposed to have contributed to equal outcomes.42 Implementing patient navigation support at HHS and other similar safety-net institutions may improve treatment adherence. Furthermore, although admissions to initiate lymphoma therapy may reduce DTI on a patient level, this practice risks congesting an already busy health system and decreasing bed availability. Increasing primary care access and expanding the health care workforce may represent more sustainable means of facilitating lymphoma diagnosis and therapy in the long run.

R-IPI retained its prognostic significance in this primarily indigent patient population with a large proportion of Hispanic individuals. Independent of R-IPI, anemia, higher INR, and higher NCI-CI at diagnosis were also associated with increased mortality in our cohort. Anemia has been linked to inferior outcomes in DLBCL in several retrospective studies.43,44 Proposed mechanisms have invoked the role of hemoglobin as a surrogate biomarker for bone marrow involvement, hematopoietic reserve, and inflammation.43,44 INR, a marker of liver synthetic function, has not been well studied as a prognostic factor for lymphoma-related survival and warrants further investigation. Although NCI-CI has not been validated as a prognostic index in LBCL, the Charlson and Hematopoietic Cell Transplantation-Specific Comorbidity indices have been identified as relevant predictors of survival in DLBCL patients.45 Future research is needed to verify the prognostic significance of these parameters in other populations. ADI, a measure of neighborhood SES, was not associated with differences in survival in this cohort. We propose that the relative socioeconomic homogeneity of the HHS patient population may have effaced the survival disparities attributable to neighborhood SES. Further investigations with adequate power to examine the impact of ADI on survival in LBCL patients are necessary.

Finally, HIV infection was more prevalent in our cohort than in historical comparators.46-48 HIV-associated LBCL is characterized by more advanced and aggressive disease. Clinical outcomes have markedly improved in the current antiretroviral therapy era, although this improvement is contingent upon access to HAART and appropriate lymphoma therapy.49,50 Moreover, people with HIV have not benefited from advances in autologous CAR-T strategies, as the clinical trials that led to the approval of commercially available CAR-T products systematically excluded people with HIV infection.8-12,49 In our study, HIV-seropositive patients had a trend toward shorter survival. HIV control and receipt of HAART at the time of lymphoma diagnosis did not have a clear impact on survival in our cohort. One possible explanation for the parity is that patients are nearly uniformly initiated on HAART when they are diagnosed with LBCL, an acquired immunodeficiency syndrome-defining illness. Longitudinal studies exploring HIV viral load and CD4+ T-cell count kinetics are needed to shed light on their effects on outcomes.

There are several limitations to this retrospective study. First, our cohort had a high proportion of underinsured and lower-income individuals, with a median age at diagnosis much lower than that quoted by SEER estimates (66 years). As a result, our findings may not be fully generalizable to other health care systems and patient populations. Nonetheless, to our knowledge, our study is among the first to examine clinical outcomes in this understudied population and will inform optimal management of these patients. Second, SDOH-associated delays in initial presentation might have served as a source of survivorship bias, as patients with more aggressive disease or worse PS at diagnosis might not have been equally captured. We adjusted for differences in baseline covariates such as disease characteristics and medical comorbidities using multivariate Cox regression. We also attempted to identify SDOH-related drivers of treatment delays and interruptions in our study. Further studies are needed to identify other barriers to care in this patient population. Lastly, a small number of patients were lost to follow-up during the study period, and some were missing baseline characteristics including ECOG PS and PNI. Overall, the attrition rate was relatively low (<5%). We categorized unknown ECOG PS as a separate categorical variable, and sensitivity analysis of missing continuous variables comparing complete-case analysis and multiple imputation by chained equations did not alter our findings.

In summary, in this public safety-net health system with robust financial assistance programs and access to cell therapies for some patients, survival in patients with newly diagnosed LBCL is comparable to SEER estimates. Importantly, uninsured status at the time of diagnosis was not associated with inferior outcomes. Addressing extant barriers to care, particularly to clinical trials and cell therapies, may improve outcomes for uninsured patients in other settings.

J.Y.J. was supported by the Cancer Prevention and Research Institute of Texas (CPRIT; RP210027) via the Baylor College of Medicine Comprehensive Cancer Training Program. M.P.M. was supported by CPRIT (RP210143). A.L., a CPRIT scholar in cancer research, was supported by CPRIT (RR190104); the National Heart, Lung, and Blood Institute (K23 HL159271); and the National Institutes of Health Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity (AIM-AHEAD 3OT2-OD032581). C.R.F., a CPRIT scholar in cancer research, was supported by CPRIT (RR190079), the Burroughs Wellcome Fund, Eastern Cooperative Oncology Group, National Cancer Institute, and V Foundation.

Contribution: J.Y.J., C.N., and A.L. conceptualized and designed the study; J.Y.J. conducted abstraction and adjudication of clinical data, such as histopathology and fluorescence in situ hybridization results, performance status, treatment response, and cell therapy eligibility; D.G. handled cohort derivation; A.O.O. and O.R. performed linkage and computation of Area Deprivation Index data; R.B. conducted bulk extraction of data from the electronic health record; J.Y.J., D.G., R.K., and A.L. carried out data analysis; J.Y.J., C.N., D.G., A.D., P.S.T., A.L., and C.R.F. contributed to data interpretation; J.Y.J. and A.L. wrote the manuscript; and all authors approved the final version of the manuscript.

Conflict-of-interest disclosure: A.D. is a consultant to Incyte. C.R.F. is a consultant to AbbVie, Bayer, BeiGene, Celgene, Denovo Biopharma, Foresight Diagnostics, Genentech, Genmab, Gilead, Janssen Pharmaceuticals, Karyopharm Therapeutics, N-Power Medicine, Pharmacyclics, Seagen, and Spectrum; receives research funding from 4D, AbbVie, Acerta, Adaptimmune, Allogene, Amgen, Bayer, Celgene, Cellectis, Guardant, Genentech, Gilead, Iovance Biotherapeutics, Janssen Pharmaceuticals, Kite, MorphoSys, Nektar, Novartis, Pfizer, Pharmacyclics, Sanofi, Takeda, TG Therapeutics, Xencor, and Ziopharm; and owns stock options in Foresight Diagnostics and N-Power Medicine. The remaining authors declare no competing financial interests.

Correspondence: Ang Li, Baylor College of Medicine, One Baylor Plaza, 011DF, Houston, TX 77030; email: ang.li2@bcm.edu.

1.
Teras
LR
,
DeSantis
CE
,
Cerhan
JR
, et al
.
2016 US lymphoid malignancy statistics by World Health Organization subtypes
.
CA Cancer J Clin
.
2016
;
66
(
6
):
443
-
445
.
2.
Pierpont
TM
,
Limper
CB
,
Richards
KL
.
Past, present, and future of rituximab—the world’s first oncology monoclonal antibody therapy
.
Front Oncol
.
2018
;
8
:
163
.
3.
Tilly
H
,
Morschhauser
F
,
Sehn
LH
, et al
.
Polatuzumab vedotin in previously untreated diffuse large B-cell lymphoma
.
N Engl J Med
.
2022
;
386
(
4
):
351
-
363
.
4.
Salles
G
,
Duell
J
,
Barca
EG
, et al
.
Tafasitamab plus lenalidomide in relapsed or refractory diffuse large B-cell lymphoma (L-MIND): a multicentre, perspective, single-arm, phase 2 study
.
Lancet Oncol
.
2020
;
21
(
7
):
978
-
988
.
5.
Wilson
WH
,
Young
RM
,
Schmitz
R
, et al
.
Targeting B cell receptor signaling with ibrutinib in diffuse large B cell lymphoma
.
Nat Med
.
2015
;
21
(
8
):
922
-
926
.
6.
Farooq
U
,
Maurer
MJ
,
Thompson
CA
, et al
.
Clinical heterogeneity of diffuse large B cell lymphoma following failure of front-line immunochemotherapy
.
Br J Haematol
.
2017
;
179
(
1
):
50
-
60
.
7.
Crump
M
,
Neelapu
SS
,
Farooq
U
, et al
.
Outcomes in refractory diffuse large B-cell lymphoma: results from the international SCHOLAR-1 study
.
Blood
.
2017
;
130
(
16
):
1800
-
1808
.
8.
Locke
FL
,
Ghobadi
A
,
Jacobson
CA
, et al
.
Long-term safety and activity of axicabtagene ciloleucel in refractory large B-cell lymphoma (ZUMA-1): a single-arm, multicentre, phase 1-2 trial
.
Lancet Oncol
.
2019
;
20
(
1
):
31
-
42
.
9.
Schuster
SJ
,
Bishop
MR
,
Tam
CS
, et al
.
Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma
.
N Engl J Med
.
2019
;
380
(
1
):
45
-
56
.
10.
Abramson
JS
,
Palomba
ML
,
Gordon
LI
, et al
.
Lisocabtagene maraleucel for patients with relapsed or refractory large B-cell lymphomas (TRANSCEND NHL 001): a multicentre seamless design study
.
Lancet
.
2020
;
396
(
10254
):
839
-
852
.
11.
Locke
FL
,
Miklos
DB
,
Jacobson
CA
, et al
.
Axicabtagene ciloleucel as second-line therapy for large B-cell lymphoma
.
N Engl J Med
.
2022
;
386
(
7
):
640
-
654
.
12.
Abramson
JS
,
Solomon
SR
,
Arnason
J
, et al
.
Lisocabtagene maraleucel as second-line therapy for large B-cell lymphoma: primary analysis of the phase 3 TRANSFORM study
.
Blood
.
2023
;
141
(
14
):
1675
-
1684
.
13.
National Cancer Institute. Surveillance, Epidemiology, and End Results Program. All cancer sites combined. Recent trends in SEER age-adjusted incidence rates, 2000-2021
. Accessed 20 February 2023. https://seer.cancer.gov/statistics-network/explorer.
14.
Warnecke
RB
,
Oh
A
,
Breen
N
, et al
.
Approaching health disparities from a population perspective: the National Institutes of Health Centers for Population Health and Health Disparities
.
Am J Public Health
.
2008
;
98
(
9
):
1608
-
1615
.
15.
Maurer
MJ
,
Ghesquières
H
,
Link
BK
, et al
.
Diagnosis-to-treatment interval is an important clinical factor in newly diagnosed diffuse large B-cell lymphoma and has implication for bias in clinical trials
.
J Clin Oncol
.
2018
;
36
(
16
):
1603
-
1610
.
16.
Tao
L
,
Foran
JM
,
Clarke
CA
, et al
.
Socioeconomic disparities in mortality after diffuse large B-cell lymphoma in the modern treatment era
.
Blood
.
2014
;
123
(
23
):
3553
-
3562
.
17.
Han
X
,
Jemal
A
,
Flowers
CR
, et al
.
Insurance status is related to diffuse large B-cell lymphoma survival
.
Cancer
.
2014
;
120
(
8
):
1220
-
1227
.
18.
US Census Bureau. QuickFacts
.
Harris County, Texas
. Accessed 20 February 2023. https://www.census.gov/quickfacts/fact/table/harriscountytexas/POP010220.
19.
Harris Health System
.
Facts and figures
. Accessed 20 February 2023. https://www.harrishealth.org/about-us-hh/who-we-are/Pages/statistics.aspx.
20.
da Costa
WL
,
Guffey
D
,
Oluyomi
A
, et al
.
Patterns of venous thromboembolism risk, treatment, and outcomes among patients with cancer from uninsured and vulnerable populations
.
Am J Hematol
.
2022
;
97
(
8
):
1044
-
1054
.
21.
Kind
AJH
,
Buckingham
W
.
Making neighborhood disadvantage metrics accessible: the neighborhood atlas
.
N Engl J Med
.
2018
;
378
(
26
):
2456
-
2458
.
22.
Area Deprivation Index v3.1. University of Wisconsin School of Medicine and Public Health
. 2015. Accessed 20 February 2023. https://www.neighborhoodatlas.medicine.wisc.edu/.
23.
Area Deprivation Index v3.2. University of Wisconsin School of Medicine and Public Health
. 2020. Accessed 20 February 2023. https://www.neighborhoodatlas.medicine.wisc.edu/.
24.
Klabunde
CN
,
Potosky
AL
,
Legler
JM
,
Warren
JL
.
Development of a comorbidity index using physician claims data
.
J Clin Epidemiol
.
2000
;
53
(
12
):
1258
-
1267
.
25.
Klabunde
CN
,
Legler
JM
,
Warren
JL
, et al
.
A refined comorbidity measurement algorithm for claims-based studies of breast, prostate, colorectal, and lung cancer patients
.
Ann Epidemiol
.
2007
;
17
(
8
):
584
-
590
.
26.
Sehn
LH
,
Berry
B
,
Chhanabhai
M
, et al
.
The revised International Prognostic Index (R-IPI) is a better predictor of outcome than the standard IPI for patients with diffuse large B-cell lymphoma treated with R-CHOP
.
Blood
.
2007
;
109
(
5
):
1857
-
1861
.
27.
Luan
C
,
Wang
F
,
Wei
N
,
Chen
B
.
Prognostic nutritional index and the prognosis of diffuse large B-cell lymphoma: a meta-analysis
.
Cancer Cell Int
.
2020
;
20
:
455
.
28.
Cheson
BD
,
Fisher
RI
,
Barrington
SF
, et al
.
Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification
.
J Clin Oncol
.
2014
;
32
(
27
):
3059
-
3068
.
29.
Cunningham
D
,
Smith
P
,
Mouncey
P
, et al
.
R-CHOP14 versus R-CHOP21: Result of a randomized phase III trial for the treatment of patients with newly diagnosed diffuse large B-cell non-Hodgkin lymphoma
.
J Clin Oncol
.
2011
;
29
(
suppl 15
):
8000
.
30.
Delarue
R
,
Tilly
H
,
Mounier
N
, et al
.
Dose-dense rituximab-CHOP compared with standard rituximab-CHOP in elderly patients with diffuse large B-cell lymphoma (the LNH03-6B study): a andomized phase 3 trial
.
Lancet Oncol
.
2013
;
14
(
6
):
525
-
533
.
31.
Bartlett
NL
,
Wilson
WH
,
Jung
S-H
, et al
.
Dose-adjusted EPOCH-R compared with R-CHOP as frontline therapy for diffuse large B-cell lymphoma: clinical outcomes of the phase III intergroup trial alliance/CALGB 50303
.
J Clin Oncol
.
2019
;
37
(
21
):
1790
-
1799
.
32.
El-Galaly
TC
,
Villa
D
,
Alzahrani
M
, et al
.
Outcome prediction by extranodal involvement, IPI, R-IPI, and NCCN-IPI in the PET/CT and rituximab era: a Danish-Canadian study of 443 patients with diffuse-large B-cell lymphoma
.
Am J Hematol
.
2015
;
90
(
11
):
1041
-
1046
.
33.
Olszewski
AJ
,
Winer
ES
,
Castillo
JJ
.
Validation of clinical prognostic indices for diffuse large B-cell lymphoma in the national cancer data base
.
Cancer Causes Control
.
2015
;
26
(
8
):
1163
-
1172
.
34.
Blansky
D
,
Fazzari
M
,
Mantzaris
I
, et al
.
Racial and ethnic differences in diffuse large B-cell lymphoma survival among an underserved, urban population
.
Leuk Lymphoma
.
2021
;
62
(
3
):
581
-
589
.
35.
Crozier
JA
,
Sher
T
,
Yang
D
, et al
.
Persistent disparities among patients with T-cell non-Hodgkin lymphomas and B-cell diffuse large cell lymphomas over 40 years: a SEER database review
.
Clin Lymphoma Myeloma Leuk
.
2015
;
15
(
10
):
578
-
585
.
36.
Li
Y
,
Wang
Y
,
Wang
Z
, et al
.
Racial differences in three major NHL subtypes: descriptive epidemiology
.
Cancer Epidemiol
.
2015
;
39
(
1
):
8
-
13
.
37.
Philip
T
,
Guglielmi
C
,
Hagenbeek
A
, et al
.
Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive non-Hodgkin's lymphoma
.
N Engl J Med
.
1995
;
333
(
23
):
1540
-
1545
.
38.
Vaughn
JL
,
Soroka
O
,
Epperla
N
, et al
.
Racial and ethnic differences in the utilization of autologous transplantation for lymphoma in the United States
.
Cancer Med
.
2021
;
10
(
20
):
7330
-
7338
.
39.
Landry
I
.
Racial disparities in hematopoietic stem cell transplant: a systematic review of the literature
.
Stem Cell Investig
.
2021
;
8
:
24
.
40.
Ahmed
N
,
Shahzad
M
,
Shippey
E
, et al
.
Socioeconomic and racial disparity in chimeric antigen receptor T cell therapy access
.
Transplant Cell Ther
.
2022
;
28
(
7
):
358
-
364
.
41.
Badr
H
,
Rouce
R
,
Scheurer
ME
,
Lulla
P
,
Mims
M
,
Reddy
P
.
Bringing CAR T cell therapy trials to underserved populations
.
Cancer Cell
.
2023
;
41
(
12
):
2007
-
2010
.
42.
Hu
B
,
Boselli
D
,
Pye
LM
, et al
.
Equal access to care and nurse navigation leads to equitable outcomes for minorities with aggressive large B-cell lymphoma
.
Cancer
.
2021
;
127
(
21
):
3991
-
3997
.
43.
Hong
J
,
Woo
HS
,
Kim
H
, et al
.
Anemia as a useful biomarker in patients with diffuse large B-cell lymphoma treated with R-CHOP immunochemotherapy
.
Cancer Sci
.
2014
;
105
(
12
):
1569
-
1575
.
44.
Clausen
MR
,
Maurer
MJ
,
Ulrichsen
SP
, et al
.
Pretreatment hemoglobin adds prognostic information to the NCCN-IPI in patients with diffuse large B-cell lymphoma treated with anthracycline-containing chemotherapy
.
Clin Epidemiol
.
2019
;
11
:
987
-
996
.
45.
Kocher
F
,
Mian
M
,
Seeber
A
, et al
.
The prognostic impact of comorbidities in patients with de-novo diffuse large B-cell lymphoma treated with R-CHOP immunochemotherapy in curative intent
.
J Clin Med
.
2020
;
9
(
4
):
1005
.
46.
Baptista
MJ
,
Garcia
O
,
Morgades
M
, et al
.
HIV-infection impact on clinical-biological features and outcome of diffuse large B-cell lymphoma treated with R-CHOP in the combination antiretroviral therapy era
.
AIDS
.
2015
;
29
(
7
):
811
-
818
.
47.
Besson
C
,
Lancar
R
,
Prevot
S
, et al
.
Outcomes for HIV-associated diffuse large B-cell lymphoma in the modern combined antiretroviral therapy era
.
AIDS
.
2017
;
31
(
18
):
2493
-
2501
.
48.
Wu
J
,
Miao
Y
,
Qian
C
, et al
.
Clinical characteristics and outcomes in HIV-associated diffuse large B-cell lymphoma in China: a retrospective single-center study
.
J Cancer
.
2021
;
12
(
10
):
2903
-
2911
.
49.
Noy
A
.
Optimizing treatment of HIV-associated lymphoma
.
Blood
.
2019
;
134
(
17
):
1385
-
1394
.
50.
Jiang
JY
,
Reid
EG
.
Impact of insurance status on outcomes in individuals with AIDS-defining cancers
.
AIDS Res Hum Retroviruses
.
2021
;
37
(
11
):
884
-
892
.

Author notes

A.L. and C.R.F. are joint senior authors and contributed equally to this study.

Presented at the 65th American Society of Hematology Annual Meeting and Exposition in San Diego, CA, 11 December 2023.

Original data are available on request from the corresponding author, Ang Li (ang.li2@bcm.edu).

The full-text version of this article contains a data supplement.

Supplemental data