Antiretroviral Treatment Outcomes Among Late HIV Presenters Initiating Treatment With Integrase Inhibitors or Protease Inhibitors

G Schuettfort; L Boekenkamp; A Cabello; AG Cotter; P De Leuw; J Doctor; M Górgolas; L Hamzah; E Herrmann; G Kann; P Khaykin; PW Mallon; A Mena; M Del Palacio Tamarit; CA Sabin; C Stephan; T Wolf; AE Haberl


HIV Medicine. 2021;22(1):47-53. 

In This Article


To our knowledge, this study is the first to describe discontinuation and virological response in those presenting late for HIV diagnosis and starting treatment with regimens including either an INI or a PI. There were no significant differences in discontinuation proportions at either 12 or 48 weeks between the two groups, and both demonstrated comparably high rates of viral suppression after 48 weeks. Although a smaller proportion of those on PI-based regimens had an undetectable viral load at 12 weeks, this is likely to reflect well-described rapid viral load decay with INI-based therapy.[19]

Data on optimal ART regimens and their outcomes for this patient group are highly relevant, as a high percentage of people newly diagnosed with HIV infection in the USA and Europe continue to present late.[17,18] There were high proportions of discontinuation of initial ART regimens at weeks 12 and 48, but this did not differ by ART group. At 12 weeks, the majority of discontinuations were driven by changes to the NRTI backbone (73.9% and 35.7% in the INI and PI groups, respectively). At 48 weeks, change of NRTI backbone remained the commonest reason for discontinuation in the INI group (28.5%), while in the PI group, side effects were the most frequent reason for switch (18.2%).

At the beginning of the evaluation period, 78% of the participants received TDF/FTC as their NRTI backbone. Approval of co-formulated TAF/FTC/elvitegravir/cobicistat in 2015 and TAF/FTC in 2016[20,21] occurred during the observation period, and it is likely that switches to the NRTI backbone were in part driven by the availability of a single-tablet regime as well as the assumed improved side effect profile of the newly approved NRTIs.

Both INIs and PIs were well tolerated, with the frequency and type of side effects being comparable in the two groups to week 48. This may in part be attributed to DRV being the most frequently used PI, which is associated with significantly milder gastrointestinal side effects than other PIs.[22,23] By week 12, more patients in the PI group reported diarrhoea, but the rates of this event were not significantly different and it did not generally result in discontinuation of therapy. Other common side effects including dizziness, joint pain, itching, and sleep disturbances did not differ in frequency between groups. This reflects similar findings from studies in individuals with CD4 cell counts > 350 cells/μL or without AIDS-defining conditions.[24,25]

The two groups had similar absolute CD4 cell counts and CD4:CD8 ratios at the start of ART, and changes to the CD4 count over the first 48 weeks of therapy did not differ significantly, with median CD4 count values at 48 weeks being comparable to those reported in previous studies.[26,27] Even with the current INIs, faster recovery of the immune system compared with PIs did not seem to occur, consistent with previous findings for RAL[28] and DTG.[29]

No significant difference was noted in mortality between the INI and PI groups during the first 48 weeks of treatment. The overall low mortality reported in this study compared to published data may reflect the high efficacy and excellent tolerability of more recent HIV therapy, combined with greater experience in the treatment of late presenters and opportunistic infections.[30] Data on IRIS were not collected in this study, as evaluation of retrospective data cannot provide reliable results and the definition of IRIS remains complex and encompasses not only infection, but also autoimmune disease and malignancy.[31] This is reflected in wide variation of reported IRIS rates of between 5 and 30%.[32] Furthermore, the occurrence of IRIS in most instances does not prompt a switch of ART, as the prognosis is good and mortality is not reported to be increased.[33] We did not perform a gender-specific analysis because of the small number of women included in this study.

Other limitations include the retrospective nature of the study and the potential for unmeasured confounding. We obtained data on mortality but not IRIS and we did not examine factors that led to the choice of third agent and therefore could not assess whether this choice was influenced by, for example, the possibility of drug interactions, prior documented resistance or the need for immediate treatment (limiting the ability to wait for results of resistance tests). Further limitations may include a prescribing bias for sociodemographic reasons, loss to follow-up or missing data because of the retrospective design of the study.

Overall, in this European cohort of late presenters starting first-line therapy with either an INI- or a PI-based ART regimen, there were no significant difference in discontinuation proportions and virological response after 48 weeks. Further research with a prospective study design is required to determine if an INI- or PI-based regimen is preferable in late presenters and if there are any gender-specific differences. However, our findings support the conclusion that the choice of third agent can and should be made on an individual basis, with both INI- and PI-based regimens performing well in terms of virological response and mortality.