Evaluation of Immune Recovery Following Autologous Hematopoietic Cell Transplantation in HIV-Related Lymphoma: Results of the BMT CTN 0803/AMC 071 Trial

Background: Clinical outcomes for patients with HIV-related lymphomas who have undergone autologous hematopoietic cell transplantation (AHCT) are similar to HIV-negative patients (Alvarnas et al., Blood 2016). Here we report a detailed, longitudinal immunophenotypic and functional evaluation of immune recovery of patients enrolled on the BMT CTN 0803/AMC 071 multicenter phase II study (clinicaltrials.gov NCT01141712).

Methods: Comprehensive analysis of cellular immunome was performed using 5 color flow cytometry. Acquisition and analysis was performed via FC500 cytometry analyzers with CXP software and prism plot. Comparisons were made between HIV+ and HIV- cohorts of peripheral blood mononuclear cell (PBMC) subsets at 56, 180, and 360 days post AHCT. The HIV- cohort was collected from 30 multiple myeloma patients enrolled in a longitudinal immune recovery study after AHCT (median age 52.5 years (18-71); 57% male, no post AHCT exposure to IMID or other treatment). Control samples were collected from 72 healthy volunteers (median age of 49 (21-68); 53%, M). A Wilcoxon rank sum test was utilized to compare the HIV+ and HIV- groups to controls and to each other at each time point for 18 immune cell subsets common to all three panels. An unsupervised analysis was performed utilizing a principal component analysis (PCA) to look for overall differences in the cohorts. Similar methodologies were used to compare HIV+ to controls that analyzed 100 PBMC subsets. Functional immune recovery was evaluated by IFNg Elispot assay where 2x10⁵ PBMC collected at each time point were pulsed with control, EBV (BZLF1) or HIV (GAG) pepmix preparations. As a control for TCR responsiveness, anti CD3/CD28 antibody-beads were used to immobilize TCR in ELISPOT assay. T cell responses from PBMCs of each of the three time points of HIV+ patients on trial were compared to PBMCs from HIV- donors (n=6).

Results: Wilcoxon Rank sum tests show significant differences between transplant patients and controls and between HIV+ and HIV- patients at all visits. There are fewer cell subsets significantly different at day 365 compared to day 56 or 120 in all comparisons. The PCA showed group differences between HIV+, HIV- and control subjects. CD3+/HLA-DR+ (late activation), CD8+/CD25- (cytotoxic T cells) and CD3+/CD314+ (T cells with activating NKG2D) were found to be more prevalent in HIV+ transplant patients. These findings may be consistent with expanded populations of chronically activated cytotoxic T lymphocytes in HIV+ transplant patients. Subsets of NK, Th1 and Th2 cells showed statistically significant differences between HIV+ (low), HIV- (higher) and controls (higher). When the principal components are plotted by visit there is a pattern of both HIV+ and HIV- transplant patients clustering closer to controls as patients recover from AHCT. The PCA was also utilized to compare the HIV+ cohort to controls which had the same panel of cell subsets tested and allowed for the use of 100 cell subsets in the analysis. This analysis showed a similar group separation and pattern of clustering closer to controls in later visits. These findings demonstrate complex interactions between T and NK cell subsets. Functional assessment of antigen-specific T cell responsiveness was evaluated in Elispot assays with EBV (BZLF1) and HIV (GAG) recall antigens and anti-CD3/CD28 controls. Of 30 evaluable patients, 28 HIV+ patients demonstrated measurable IFNg production in response to GAG (spots/2x10⁵ PBMC, range: 8-615), 21 showed measurable response to BZLF1 pepmix (range 12-450); and all patients demonstrated responsiveness to anti CD3/CD28 stimulation. Magnitude of IFNg production from HIV+ samples was generally higher than that observed healthy, HIV- controls. Assessment of NK cell responsiveness is currently underway.

Conclusions: While clinical outcomes following AHCT between HIV+ and HIV- patients is comparable, clear distinctions were observed with immune recovery of specific PBMC subsets during the first year following AHCT with differences diminishing as patients recover post transplant. Longitudinal immune responsiveness of PBMC from HIV+ patients to EBV and HIV recall antigens and TCR stimulation generally showed more robust IFNg production compared to PBMCs from HIV- volunteer controls. These data provide further justification supporting AHCT as an option for HIV+ patients provided they meet standard transplant criteria.