Impact of HIV infection on TCR repertoire diversity in a fixed volume of whole blood. In this figure, as well as those that follow, circles represent HIV− subjects; and squares, HIV+ subjects. (A) Whole-blood TCR repertoire diversity compared for HIV-infected and -uninfected subjects. Diversity is expressed in relative Cot units.17 Solid symbols represent subjects meeting prespecified criteria for comparison; and open symbols, subjects that did not meet these prespecified criteria. Diversity values for HIV-infected subjects were significantly lower (P = .0005; 2-tailed Mann-Whitney test for prespecified comparison groups). (B) Absolute T-cell count did not correlate with whole-blood TCR repertoire diversity (Spearman r = −0.07; P = .79 for HIV−; Spearman r = −0.26; P = .27 for HIV+; Spearman r = 0.25; P = .14 for all subjects). (C) CD4+ T-cell percentage and whole-blood TCR repertoire diversity measured for HIV-infected subjects. Solid symbols represent subjects with viral loads > 1000 copies/mL; and open symbols, subjects with viral loads < 1000 copies/mL. CD4% and sum TCR repertoire diversity were significantly correlated for all HIV-infected subjects (Spearman r = 0.55; P = .01); the strength of correlation increases if the 2 subjects with viral loads < 1000 copies/mL are excluded (Spearman r = 0.72; P = .001). (D) CD4+ T-cell absolute counts and whole-blood TCR repertoire diversity measured for HIV-infected subjects. Solid symbols represent subjects with viral loads > 1000 copies/mL; and open circles, subjects with viral loads < 1000 copies/mL. Absolute CD4 count and whole-blood TCR repertoire diversity were significantly correlated (Spearman r = 0.46; P = .048), but the strength of this correlation improved if the 2 subjects with viral loads < 1000 copies/mL are excluded (Spearman r = 0.61; P = .01). (E) Naive CD4+ T-cell absolute counts and whole-blood TCR repertoire diversity measured for HIV-uninfected subjects (Spearman r = 0.61; P = .008). (F) HIV viral load and blood TCR diversity for HIV-infected subjects. The correlation was not significant (Spearman r = −0.42, P = .07). (G) Absolute memory CD8+ cell count and whole-blood TCR diversity. Memory cells were defined as in Table 3 (Spearman r = −0.21, P = .41 for HIV−; Spearman r = −0.27, P = .29 for HIV+). (H) Percentage CD8 effector/memory cells of all CD8 cells and whole-blood TCR diversity in HIV+ subjects. Open symbols represent subjects with viral loads < 1000 copies/mL. Memory and effector cells were defined as in Table 3 (Spearman r = −0.52, P = .03).