Viral load and immunophenotype of cells obtained from lymph nodes by fine needle aspiration as compared with peripheral blood cells in HIV-infected patients.

We wished to establish the feasibility of fine needle aspiration of lymph nodes as a noninvasive method for measuring subsets of immune cells and viral load in HIV-infected patients. Twenty-five patients (CD4+ T cell range 4-760/microl, median 362) were selected. Lymph node aspiration was attempted in 21 patients. Lymph node cells (LNC), ranging from 6 x 10(3) to 2 x 10(6) (median 6 x 10(5)) were obtained in 17 subjects, and compared with peripheral blood mononuclear cells (PBMC) obtained simultaneously. Immunophenotype could be determined by flow cytometry in 9 patients. Mean percent of CD4+ CD3+ T cells in LNC and PBMC and 23.2 and 14.6. Mean percent of CD8+ CD3+ T cells in LNC and PBMC was 23.1 and 45.0, respectively. Therefore, CD4+/CD8+ ratios were much higher in LNC (mean +/- SD: 1.06 +/- 0.31) than in PBMC (0.35 +/- 0.13). The amount of HIV DNA (11 patients) and RNA (8 patients) was determined in the plasma, LNC, and PBMC by competitive reverse transcriptase-polymerase chain reaction (RT-PCR). The number of copies of viral DNA/10(5) cells was higher in LNC than in PBMC (LNC/PBMC ratio ranger: 0.54-25, median 3.4). The number of copies of unspliced viral RNA/10(5) cells was much higher in LNC than in PBMC (LNC/PBMC ratio range 65-1,159, median 435). The plasma RNA copy number, a measure of circulating cell-free virus, was correlated with the RNA copy number in PBMC, but not in LNC. A RT-PCR system specific for spliced transcripts was also used to assess the level of transcripts independent of genomic RNA. This assay also detected more signal in LNC than in PBMC. The level of spliced transcripts in LNC and PBMC correlated with the amount of full-length RNA detected by competitive PCR. A semiquantitative coculture assay with lymphoblasts from healthy donors was used to assess the infectivity of LNC as compared with PBMC in 14 patients. The minimum number of LNC necessary to cause a positive coculture ranged from 10(3) to > 10(5) (median 10(4)); the corresponding number for PBMC ranged from 10(3) to > 10(6) (median 5 x 10(5)). In most patients selected for palpable lymph nodes, LNC could be obtained by fine needle aspiration, thus allowing noninvasive monitoring of viral burden in lymphoid tissue. The present study also suggests that both T-cell subsets and viral load differ in the blood and lymphoid tissue, which raises the question of whether the study of the lymphoid tissue would yield better prognostic markers of disease course.