To evaluate the performance of E6 HPV16/18 Viral load, (E6-E2) HPV16/18 Viral load (viral load "integrated") and E2/E6 ratio (integration coefficient to detect cervical high-grade lesions (CIN2 +)) in a cohort of African women infected with HIV enrolled in the HARP (HPV Research Partnership in Africa) project. A total of 1238 women were enrolled in the HARP study. At baseline and at 18 months of follow-up (endline) the following screening tests were performed: HPV test, Cytology, VIA / VILI, colposcopy and biopsy if one test positive. Histology was systematically confirmed by the End Point Committee. In addition, HPV detection and genotyping was performed using the INNO-LiPA HPV genotyping Extra assay. 245 women co-infected with HPV16 and/or HPV18 were included in the present study. Their median (range) age was 35 (25-49) years, CD4 + cell count was 417 (7-830) cells/µL, and 158 (66.4%) women were on ART; 122 positive for HPV 16 and 78 for HPV18 at inclusion; 103 positive for HPV 16 and 66 for HPV18 at 18 months. 25 co-infected at baseline and 11 at endline.
Quantitative real-time PCR targeting the E6 and E2 genes were performed using serial dilutions of HPV16 and HPV18 plasmids as standard curves. Total cellular DNA was measured by real-time PCR of the GAPDH gene, and the results were expressed as the number of copies of E6 and E2 per 1,000 cells. SiHa, Caski and Hela cell lines were used as controls.
It was observed a very significant (p<0.0001) increase in HPV16 E6 viral load, and HPV16 (E6-E2) viral load and a decrease in E2/E6 ratio as the grade of cervical lesions increased. It was also observed an increase (p=0.029) in viral loads with the lesion grade. There was a strong correlation between viral loads E6 and E2-E6 HPV16 and HPV18 (Rs = 0.899 and 0.962 respectively). E6-E2 and E2/E6 are not independent of E6. E6 viral loads of 4 log copies/1,000 cells for HPV16 and of 2 log copies/1,000 cells were associated with high grade lesions. Sensitivity and specificity of these viral load levels were 83% and 75% for HPV16, and 50% and 70% for HPV18, respectively.
A high HPV16 viral load (> 4 log E6 DNA copies / 1000 cells) or, to a lesser extent, for HPV18 (> 2 log E6 DNA copies / 1000 cells) is associated with cervical high-grade lesions. Among women without high-grade lesion at baseline, a high HPV16 viral load is associated with progression to high-grade lesion at 18 months, this was not observed for HPV18. E6 Viral load and E2-E2 ratio are not independent markers of viral load E6. HPV16 viral load might constitute a triage test for HPV16-positive women. HPV18 viral load is of more limited interest given the low sensitivity / specificity ratio.