Impaired Immune Health in Survivors of Diffuse Large B-Cell Lymphoma

Abstract

Purpose: Therapeutic advances for diffuse large B-cell lymphoma (DLBCL) have led to an increasing number of survivors. Both DLBCL and its treatments perturb the immune system, yet little is known about immune health during extended survivorship.

Methods: In this retrospective cohort study, we compared 21,690 survivors of DLBCL from the California Cancer Registry (CCR) to survivors of breast, prostate, head and neck, and melanoma cancers. We linked their CCR records to a statewide database documenting hospital, emergency room, and ambulatory surgery visits and investigated the incidence of autoimmune conditions, immune deficiencies, and infections 1-10 years after cancer diagnosis.

Results: We found elevated incidence rate ratios (IRRs) for many immune-related conditions in survivors of DLBCL compared with other cancer survivors, including significantly and consistently elevated IRRs for viral and fungal pneumonias (up to 10.8-fold), meningitis (up to 5.3-fold), as well as humoral deficiency (up to 17.6-fold) and autoimmune cytopenias (up to 12-fold). IRRs for most conditions remained high even in the late survivorship period (5-10 years after cancer diagnosis). The elevated risks could not be explained by exposure to chemotherapy, stem-cell transplantation, or rituximab, except for IRRs for humoral deficiency, which were consistently higher after the incorporation of rituximab into DLBCL treatments.

Conclusion: To our knowledge, this is the largest cohort study with extended follow-up to demonstrate impaired immune health in survivors of DLBCL. The observed persistent, elevated risks for autoimmune diseases, immune deficiencies, and infectious conditions may reflect persistent immune dysregulation caused by lymphoma or treatment and may lead to excess morbidity and mortality during survivorship. Improved understanding of these risks could meaningfully improve long-term care of patients with DLBCL.

FIG 1.

Estimated incidence rate ratios of autoimmune, immune deficiency, and infectious conditions among survivors of B-cell lymphoma (DLBCL) compared with survivors of (A) female breast cancer and (B) prostate cancer in the 1-10 years after cancer diagnosis, California, 1991-2012. Models were adjusted for age, race, and year of cancer diagnosis. Error bars show 95% CIs. Only diagnoses for which the comparison yielded a P value ≤ .0002 are shown. Log2 scale is used for the x-axis. RA, rheumatoid arthritis.

FIG 2.

Estimated incidence rate ratios of autoimmune, immune deficiency, and infectious conditions among survivors of B-cell lymphoma (DLBCL) compared with survivors of (A) head and neck cancer and (B) melanoma in the 1-10 years after cancer diagnosis, California, 1991-2012. Models were adjusted for age, race, sex, and year of cancer diagnosis. Error bars show 95% CIs. Only diagnoses for which the comparison yielded a P value ≤ .0002 are shown. Log2 scale is used for the x-axis. RA, rheumatoid arthritis.

FIG 3.

Cumulative incidence curves for (A) humoral deficiency, (B) sicca syndrome, (C) autoimmune hemolytic anemia (AIHA), (D) immune thrombocytopenia, (E) viral pneumonia, (F) fungal pneumonia, (G) influenza, (H) meningitis, (I) enteral infections, and (J) bacterial pneumonia for each cohort of cancer survivors as indicated in the graph legends, from 1-10 years after initial cancer diagnosis. DLBCL, diffuse large B-cell lymphoma.

FIG 4.

Estimated incidence rate ratios (IRRs) among survivors of diffuse large B-cell lymphoma compared with each of the four comparison cohorts during years 5-10 after cancer diagnosis. Error bars show 95% CIs. Only diagnoses for which the IRR had a P value ≤ .0002 are shown. Log2 scale is used for the x-axis.

FIG 5.

Incident rate ratios for 18 categories of immune-related conditions (autoimmune diseases, immune deficiencies, and infectious diseases) among survivors of diffuse large B-cell lymphoma (DLBCL) compared with survivors of breast cancer, prostate cancer, head and neck cancer, or melanoma. Results reaching a P value ≤ .0002 are shown, and the remainder are grayed. Color scale is logarithmic (log¹⁰), with pink and red shading for categories of diagnoses found more commonly in survivors of DLBCL and green shading for categories of diagnoses found more commonly in the comparator cohort.