Immunologic consequences of chemotherapy for ovarian cancer: impaired responses to the influenza vaccine

Abstract

Objectives: To examine the effect of chemotherapy for ovarian cancer on immunologic function and to define the effect on the serologic response to the influenza vaccine.

Methods: Under IRB approved protocols, patients with ovarian cancer were administered seasonal trivalent killed influenza vaccines. Peripheral blood was collected for immunologic assessments. Serum was analyzed for hemagglutination inhibition (HAI) antibody titers. Peripheral blood mononuclear cells were isolated to characterize T and B cell populations and function.

Results: Thirty-one patients were recruited: 13 in remission receiving a dendritic cell vaccine with or without a single dose of low-dose cyclophosphamide, 3 in remission not receiving treatment, and 15 undergoing standard therapy. Significant effects on T cell and B cell subset distributions were seen. Functional effects were also seen. Few patients were able to mount a 4-fold HAI antibody response. A 4-fold response was observed for H1N1 in 20%, for H3N2 in 26%, and for influenza B in 6%. Pre-existing exposure to influenza was predictive of responders.

Conclusions: Despite CDC recommendations that patients undergoing chemotherapy receive influenza vaccine, there is little evidence to support its serologic effectiveness in this population. Patients with ovarian cancer are almost uniformly unable to mount a meaningful antibody response. These findings have serious implications for future resource allocation for both seasonal and novel pandemic influenza outbreak and understanding the immunologic deficits as a result of chemotherapy may improve patient care.

Keywords: Antibody; Cyclophosphamide; DC vaccine; ELISpot; Immune compromised; Monoclonal B cell lymphocytosis; Treg.

Figure 1

HAI titer responses of subjects. (A) The percent of patients able to demonstrate at least a four-fold rise in titer to any one of the three influenza strains after vaccination. (B) H1N1 titers across all time points do not vary dramatically. (C) The percentage of the entire cohort with baseline titers ≥1:40. OVCA=ovarian cancer patients off protocol, DC-C=dendritic cell vaccine recipients with Cytoxan, DC-NC-dendritic cell vaccine recipients without Cytoxan.

Figure 2

T cell subsets. Flow cytometry was used to define T cell subsets in patients and controls. Mean and standard deviation are shown. Asterisks indicate p<0.05 comparing baseline and controls.

Figures 3

DC-C patients have improved T cell proliferation after cyclophosphamide. Proliferative responses were measured by CFSE. The mean for each population is shown. OVCA=ovarian cancer patients off protocol, DC-C=dendritic cell vaccine recipients with Cytoxan, DC-NC-dendritic cell vaccine recipients without Cytoxan.

Figure 4

B cell subsets. Flow cytometry was used to define B cell subsets in patients and controls. Mean and standard deviation are shown. Asterisks indicate p<0.05 comparing baseline and controls.

Figure 5

Monoclconal B cell expansions. Top panel: CD20+ lymphocytes (B cells) from whole blood are stained with CD27 and CD38 and separated into the following subsets in rough order of maturity, starting with the least mature: Tr (transitional, CD27−, CD38++), MN (mature naive, CD27−, CD38+), MA (mature activated, CD27+, CD38+), PB (plasmablast, CD27++, CD38++), RM (resting memory, CD27+, CD38−), LL (left lower quadrant, CD27−, CD38−). Lower panel: Persistent monoclonal B cell lymphocytosis in a patient with ovarian cancer. Cytometric fingerprinting of B cell subsets was performed as described in the methods section. The y-axis indicates the fold change in the frequency of events vs. the binning model (produced from the aggregated data of the normal subjects) and the x-axis indicates the bins. Black lines = healthy subjects. Red lines = ovarian cancer patients. Data from ovarian cancer patient #1, analyzed at four separate time points (c1-1, c1-2, c1-3 and c1-4) over a 1-year period, reveal the persistence of an expanded population. The cells in the corresponding bins were mapped onto conventional flow cytometry plots and reveal a light chain restricted, class-switched B cell population.

Figure 6

B cell function. Total immunoglobulin producing B cells were enumerated in an ELISPOT as well as influenza-specific IgG producing B cells. Means are shown as a horizontal line. The controls are significantly different than the patient populations for total IgG at all time points but the difference between patients and controls for influenza-specific IgG is not significant.