As previously reported 28
, the distribution of CD3+
T-cells and CD4+
T-cell subsets within CD3+
T-cells was similar across the three study groups: HPV negative, HPV persistence, and HPV clearance groups (– and data not shown). When immunophenotypic subsets of CD3+
cells were evaluated and the frequency distributions compared between the HPV persistence group and the HPV negative controls (), statistically significant differences in the distributions of the respective marker were observed with increases in the median levels among the persistence group for the following activation CD3+
T-cell subsets: CD69+
. In addition, a statistically significant difference in the distribution was observed for the CD45RO+
subset, a marker of undifferentiated memory cells with a decrease in the median level among the persistence group. Significant differences were also observed when the HPV persistence group was compared to the HPV clearance group for the same markers with median levels similarly increased or decreased as when compared to the HPV negative group (data not shown). No significant differences for any of the markers analyzed were observed when the HPV clearance group was compared to HPV negative controls (data not shown).
Next, we compared immunophenotypic subsets of CD3+CD8+ cells (). The only statistically significant difference noted between the HPV persistence and HPV negative group was for the mean rank distribution of the CD45RO+CD27−CD8+ cell population, a marker of differentiated memory cells with an increase in the median level for the persistent HPV positive group. A higher median value of CD45RO+CD27−CD8+ was also observed for the persistent HPV positive group when compared with the HPV clearance group with a significantly different quartile distribution (data not shown). No significant differences were observed when the HPV clearance group was compared to HPV negative controls (data not shown).
Associations between risk of HPV persistence and increasing percentages of specific subsets within CD3+CD4+ and CD3+CD8+ are presented in and , respectively. These analyses were restricted to the subsets whose population’s mean rank distribution differed significantly between the HPV persistence and the HPV negative control groups. Although the distribution difference was significant for CD38+HLADR+CD3+CD4+ and CD38−HLADR+CD3+CD4+, these two subsets were not evaluated further since their observed effects were driven by the main effect seen for HLADR+CD3+CD4+ (correlation between these two subsets and HLADR+CD3+CD4+ was 0.53 and 0.96, respectively).
Odds ratios for HPV persistence by proportion of CD4+ lymphocyte immunophenotypic subsets
Odds ratios for HPV persistence by proportion of CD8+ lymphocyte immunophenotypic subsets
Consistent with findings based on the T-cell markers population distributions, a statistically significant trend in risk for HPV persistence was observed among women with increasing quartile categories when compared with women in the lowest quartile for the following markers of immune activation and/or differentiation after adjusting for age CD4+CD69+ (p-trend < 0.001; 4th quartile OR = 5.4, 95% CI 2.2 – 13.3), HLADR+CD3+CD4+ (p-trend = 0.04; 4th quartile OR = 2.6, 95% CI = 1.2 – 5.9), and CD45RO+CD27−CD8+ (p-trend = 0.01; 4th quartile OR = 2.3, 95% CI 1.1 – 4.7;). Conversely, a decreased risk for HPV persistence was observed among women in the highest quartiles for the following marker of undifferentiated memory cells CD45RO+CD27+CD4+ (p-trend = 0.006; 4th quartile OR = 0.36, 95% CI 0.17 – 0.76). Effects of stronger magnitude were observed in analyses that restricted the case group to long term persistors ( and ). Similar effects were observed in analyses that evaluated persistence with carcinogenic HPV types (data not shown). We observed no evidence that the effects differed by age (45–54, 55–64, 65–74) (data not shown).
Next, we evaluated whether our previously reported association between proliferative responses and risk of HPV persistence could be explained by differences in the activation/differentiation state of CD3+CD4+ or CD3+CD8+ subsets (). Adjustment of the effects seen for proliferative responses against PHA and HPV-16 VLP for levels of the T-cell markers found to be associated with persistence (CD69+CD4+, HLADR+CD3+CD4+, CD45RO+CD27+CD4+, and CD45RO+CD27−CD8+) only slightly attenuated effects (data not shown). A modest attenuation in the odds ratios was observed when we adjusted proliferative responses for the joint marker variable that combined across the four subsets significantly associated with risk of persistence (). The age-adjusted OR comparing low to high responders to PHA was reduced from 2.7 (95% CI = 1.4 – 5.1) to 1.9 (95% CI = 0.94 – 3.7). Similarly, the age-adjusted OR comparing low to high responders to HPV-16 VLP was reduced from 2.7 (95% CI = 1.4 – 5.3) to 1.9 (95% CI = 0.90 – 3.8). Adjustment of the immunophenotype effects for proliferative responses to PHA or HPV-16 VLP had a minor effect on the magnitude of the immunophenotype associatiosn with HPV persistence (). Finally, although limited by study size, we observed no evidence that the proliferation effects differed by levels of specific immunophenotypic markers or vice versa (i.e., no evidence for effect modification) (data not shown).
Odds ratios and confidence intervals for persistent HPV positive group by PHA and VLP proliferation assays