Background

The life expectancy of HIV-infected individuals is approaching that of the general population due to the effectiveness and good tolerability of new antiretroviral drugs.

The main concerns over antiretroviral therapy (ART) are no longer efficacy and tolerance, but antiretroviral drug-related toxicity [1], especially over the long term. This toxicity tends to be subclinical and cumulative and has the potential to interact with comorbidities, aging, and other processes related to immune activation and inflammation in HIV infection. It is mainly associated with Nucleoside Analogue.

Reverse Transcriptase Inhibitors (NRTI) [2,3,4,5,6,7,8], whose frequent use in pairs in classic Triple Therapy (TT) exerts a synergic antiviral effect but also increases the potential toxicity.

For this reason, nuke-sparing regimens (NSRs), that do not include one or both NRTIs, have been developed to reduce and prevent ART-related toxicity [9]. The development of novel and highly effective antiretroviral (ARV) drugs with a high genetic barrier against the development of drug resistance, such as protease inhibitors (PI) [10, 11], allows treatments to be simplified, with the administration of fewer drugs. Permanent TT may no longer be necessary and safe simplification strategies are now available to simplify ART [12].

Numerous studies (clinical trials, observational studies, and meta-analyses) have found that the efficacy of PI monotherapy (MT) [13,14,15,16,17,18,19,20,21,22,23,24,25,26,27] and dual therapy (DT) [28,29,30,31] with lamivudine (3TC) and boosted PIs (generally in patients without advanced HIV infection who had received suppressive ART for at least 6–12 months and/or had no history of virologic failure [VF]) was not inferior to that of TT [15, 17, 18, 20, 23, 24, 28,29,30,31,32,33,34]. Both of these simplification strategies maintain viral suppression while avoiding the development of resistance mutations [13,14,15,16,17,18,19,20,21,22,23,24,25,26,27, 35] and controlling immune activation and chronic inflammation [36,37,38,39,40,41], even in HIV sanctuaries and biological reservoirs [42,43,44,45,46,47,48], similarly to TT and with a good cost-effectiveness ratio [12, 49,50,51,52,53,54].

Fewer data are available on the application of DT in patients with a less favorable profile (advanced HIV infection, long exposure to ART, history of virological failures, long-term toxicity), although they have generally proven high efficacy and a reduction of possible toxicity [55,56,57,58,59,60,61,62].

DT with rilpivirine and boosted darunavir (RPV + bDRV) is an attractive NSR that appears to combine both a high efficacy and genetic barrier with a lower pill burden, good tolerance and toxicity profile. However, despite its utilization in clinical practice, there has been little research on the outcomes.

With this background, we performed a retrospective multi-center investigation to analyze the profile of patients prescribed with this combination and study its efficacy and safety. Confirmation that these NSRs are safe and effective will help to consolidate them as an optimized alternative for ART that could improve the long-term prognosis of HIV-infected patients and reduce treatment costs.

Methods

Study design

An observational, retrospective, multi-center study was conducted in HIV-infected patients who switched to DT with rilpivirine (RPV) (25 mg, once daily) and boosted darunavir (bDRV) (800 mg, once daily) (with either ritonavir or cobicistat).

Objectives

The main objective of the study was to analyze the effectiveness of the DT by calculating the proportion of patients with virologic success (defined as 24 weeks of follow up without VF, considering VF as two consecutive RNA HIV-1 > 50 copies/mL).

Secondary objectives were to establish: a) proportion of patients with RNA HIV-1 < 50 copies/mL at week 24 of treatment; b) stratification of all viral load determinations obtained during exposure to the DT; c) Incidence of new adverse events; d) Impact of the DT on lymphocyte subpopulations, lipid profile, and liver and kidney function; e) Analysis of possible differences in efficacy and potential toxicity between darunavir/ritonavir (DRV/r) and darunvir/cobicistat (DRV/c).

Inclusion criteria and variables

Patients from 15 Spanish hospitals were evaluated, with the following inclusion criteria: infection with HIV-1, age > 18 years, and receipt of antiretroviral treatment with RPV + bDRV before December 31st 2015 with a minimum follow-up period of 24 weeks. They were also required to have a baseline viremia at switch (bVL) < 1000 copies/mL and to have signed informed consent to retrieve data from their medical records.

A standardized electronic database was use to collect the following variables: age, gender, date of HIV diagnosis, VL at diagnosis, CD4 lymphocyte cell count at diagnosis, CD4 nadir, HIV stage (CDC), date of first ART, number of previous ART combinations, previous ART, previous VF (and ART received and genotype mutations at time of VF), reason for switch to study combination (toxicity, intolerance, VF, simplification/optimization or other reasons), booster (ritonavir or cobicistat), weeks of exposure to ritonavir and cobicistat, and HIV- 1 RNA, CD4 cell count, and bloodwork at baseline, 4–8 weeks, 9–23 weeks, and ≥ 24 weeks.

New adverse events during exposure to DT and reasons for not completing 24 weeks of follow-up were also collected (toxicity/intolerance, VF, voluntary drop-out, or others).

Statistical analysis

For the first and secondary objectives of virologic efficacy, an intention-to-treat (ITT) analysis was carried out (considering losses as failures) as well as an on-treatment (OT) analysis, excluding patients who discontinued treatment with the DT for any reason other than VF (voluntary discontinuation of treatment, toxicity/intolerance, medical decision, patient’s decision).

Means with standard deviation, medians, and interquartile ranges were calculated for quantitative variables and absolute and relative frequencies for qualitative variables. After applying the Kolgomorov-Smirnov test to check the distribution of variables, effects of the DT on change in analytical parameters (lipids, liver and kidney function) were analyzed using the paired samples t-test when the distribution was normal or the signed-rank Wilcoxon test when it was not. For bivariate analyses, the Student’s t-test was applied when the distribution was normal and the Mann-Whitney test when it was not. SPSS v20.0 (IBM Corp, Armonk, NY) was used for data analyses, and the significance level was 0.05 for all tests.

Ethical aspects

The study was approved by the regional ethics committee of Andalusia (CCEIBA) on April 25, 2016 and participants provided written informed consent for data from their medical records to be collected.

Results

General description of study population

One hundred and eighty-nine patients switched to RPV + bDRV: 28 did not meet inclusion criteria (20 with baseline VL > 1000 copies/mL and 8 with follow-up of < 24 weeks). Therefore, the final study sample comprised 161 patients with median age of 49 years (IQR 44–53), median time since HIV diagnosis of 17 years (IQR 10–23), receipt of ART for median of 14 years (IQR 6–18), and median of 5 previous ART combinations (IQR 3–7); 29.3% had a history of AIDS (stage C or 3 according to CDC), and the median CD4 lymphocyte nadir was 170 cells/μL (IQR 73–290) [Table 1]. ART regimens before the switch to DT included NRTI in 59% of patients, non-nucleoside reverse transcriptase inhibitors (NNRTIs) in 56.5%, protease inhibitors (PI) in 79.5%, and integrase inhibitors (II) in 9.9%; TT had been received by 57.1% of the patients, DT by 32.9%, and MT by 9.9%. Of the patients on MT, the reason for switch was optimization in 43.8% patients and VF in 43.8%; bVL was < 50 copies/mL in 6.3%.

Table 1 General description of study population

At baseline, 25.5% did not have suppressed HIV viremia (baseline viral load of 50–1000 copies/mL), although the baseline CD4 lymphocyte cell count was 618 cells/μL (IQR 370–861) [Table 1]; 36.6% (59/161) had previous VF and data were available on the drug therapy at the time of the VF in 51 of these, being PI in 90.2% (46/51) and 9.3% of patients had previous virological failure while receiving NNRTI. In 58.8% (30/51) of cases, the PI was DRV. Hence, one-third of the patients in this study had experienced previous VF while receiving PI and one-fifth (21.5%) while receiving DRV.

The main reason for the switch to a DT with RPV + bDRV was simplification/optimization of their ART (49.7%), followed by previous ART toxicity (17.4%), and insufficient effectiveness of previous ART (10.6%).

Initially, the DT boosted with ritonavir (DRV/r) was administered to 131 patients (81.4%), 29 of whom switched to cobicistat (DRV/c) during the first 24 weeks of treatment. A total of 59 patients (36.6%) were exposed to DRV/c during the study period.

Effectiveness of RPV + bDRV at 24 weeks of treatment

Figure 1 depicts the flow chart of patients through the study. The ITT population included all 161 patients, with a median follow up of 33 weeks. At week 24, 87.6% (141/161) of patients continued receiving RPV + bDRV and showed no VF criteria, while 17 patients discontinued this DT: 6 for mild toxicity/intolerance, 4 due to voluntary abandonment of ART, 1 by physician decision (due to one VL > 50 copies/mL), 1 by patient decision to return to previous ART and 8 VF (3 confirmed after week 24).

Fig. 1
figure 1

Flow chart of study patients

Excluding the 12 patients who dropped out of the study for reasons other than VF (population for OT analysis: 149 patients), 94.6% (141/149) completed the study without showing VF criteria.

In 82.6% (133/161) of the study population, the VL was < 50 copies/mL at 24 weeks (ITT analysis). In 89.3% (133/149) of the OT population, the VL was < 50 copies/mL at 24 weeks (OT analysis).

As shown in Table 2, genotyping was performed in three of the total of eight patients with VF and no drug resistance mutations were detected, while two cases were associated with poor adherence to treatment. Resuppression of HIV viremia was achieved in all eight cases with no modification of the DT (3 cases) or after switching to a TT (4 cases) or different DT (1 case).

Table 2 Virologic failures with RPV + bDRV

Out of the 334 VL determinations during exposure to RPV + bDRV, over 105 patients-years, < 50 copies/mL were found in 87.7% (< 20 copies/mL in 79%), 50–1000 copies/mL in 11.1%, and > 1000 copies/mL in 1.2%; 237 of these determinations were obtained during exposure to RPV + DRV/r, with 87.3% being < 50 copies/mL, and 97 during exposure to RPV + DRV/c, with 88.7% being < 50 copies/mL.

The last observed VL determination of the 159 patients with virologic data after the switch was < 50 copies/mL in 86.2% of patients (< 20 copies/mL in 80.5%), between 50 and 1000 copies/mL in 12.6%, and > 1000 copies/mL in 1.3%.

During the 24 weeks of exposure to the DT, the CD4 lymphocyte cell count increased by 34 cells/mm3 from 630 cells/mL to 667 cells/mm3 (p = 0.071) and the CD4/CD8 ratio by 0.04 from 0.75 to 0.79 (p = 0.004).

Patients with previous failure while receiving darunavir were not more likely to present VF with the study DT [13.3% (4/30) vs 5.5% (1/18), p = 0.348] and overall baseline viral load > 50 copies/mL was not a predictor of VF (9.8% vs 3.3%, p = 0.115).

We analyzed the influence of CD4 nadir on VF and found no statistically significant differences between those who failed and those who didn’t (257 vs 156 cells/mm3, p = 0.470). There were also no differences in VF rate between those who received the DT boosted with ritonavir vs cobicistat (7,5% vs 7,1%, p = 1.000).

Safety of RPV + bDRV

Although no severe adverse events were notified, six patients switched therapies due to mild toxicity or intolerance (Table 3).

Table 3 Adverse events during the study

Impact of RPV + bDRV on analytical parameters (lipid profiles, and kidney and liver functions)

Statistically significant differences between values at baseline and 24 weeks were found in creatinine (0.944 mg/dL vs. 0.977 mg/dL, p = 0.001), total cholesterol (183 mg/dL vs. 196 mg/dL, p < 0.001), HDL-cholesterol (48 mg/dL vs. 52 mg/dL, p = 0.005), LDL-cholesterol (107 mg/dL vs. 124 mg/dL, p = 0.003), GOT (40.9 U/L vs. 30.8 U/L, p = 0.031), GPT (45.9 U/L vs. 30.3 U/L, p = 0.011), GGT (57.5 U/L vs. 59.4 U/L, p = 0.040), and alkaline phosphatase (98 U/L vs. 93 U/L, p = 0.027) but not in total cholesterol/HDLcholesterol ratio (Table 4). A higher increase of LDL-cholesterol was observed with ritonavir (+ 20 mg/dL) than with cobicistat (+ 10 mg/dL), but the difference did not reach statistical significance (p = 0.462).

Table 4 Impact of the dual therapy on analytical parameters

Discussion

Data obtained in this study confirm that nuke-sparing DT with RPV + bDRV may be an acceptable alternative to triple therapy in suppressed and stable HIV-infected patients, as previously suggested in the PROBE CT [63]. However, the present patients generally had a long history of exposure to HIV and ART and included numerous cases of severe previous immunodepression (AIDS stage and/or low CD4 nadir), VF, toxicity associated with antiretroviral drugs, and even a previous non suppressive ART. These conditions have usually been considered as exclusion criteria in studies of MT and DT with 3TC. Nevertheless, the DT under study was found to achieve and maintain viral suppression in > 90% of the present patient population.

Despite the disadvantageous profile of our study population, the proportion of virologic suppression obtained with this DT was similar to that obtained with TT (including stable patients with no history of VF and with suppressed viremia in this switch scenario). Stratification of all viral load determinations in the entire cohort during the study period showed similar rates of blips and VFs, and almost all of the latter could be attributed to poor treatment adherence. No drug resistance mutations against the protease or the inverse transcriptase were observed in any case, and all patients achieved viral resuppression by maintaining the DT or adding a third drug. Although RPV has a low genetic barrier and patients who showed VF could potentially develop resistance to the drug, in this study there were no VF with real exposure to the DT (multiple patients reported poor treatment adherence) and/or high viral loads and few drug resistance tests were performed.

This is a preliminary analysis of a cohort that we are still following, but we believe that a ‘24 weeks Analysis’, considering a threshold of 50 copies/mL for VF, is enough to determine virologic effectiveness for previously suppressed patients. This is the minimum timeframe required to prove ART’s ability to suppress viral replication in naïve patients [64,65,66,67,68] and for rescue strategies in patients with prior virological failure [68]. Twenty-four weeks is also the minimum timeframe required in switch studies to consider a previous HAART stable and effective [69, 70], and we know that the maximum suppression of HIV viremia can be achieved with < 20 weeks of treatment [71], that virological failures with simplification strategies occur during the first months [72] and that this rate does not increase with follow-up time [73].

Despite their long history of ART, the immunological recovery was similar to that reported for TT, with an increase in CD4 lymphocytes and CD4/CD8 ratio.

Tolerance of the combination was generally good, although several patients asked to switch to previous or alternative treatments due to toxicity, which was considered mild in all of these cases.

A slight increase in total cholesterol and LDL-cholesterol levels was observed with the DT under study; however, there was also an increase in HDL-cholesterol levels, with no change in the atherogenic index over the 24-week observation period. There was a significant decrease in transaminase levels, implying a reduction in the potential toxicity of this DT, which supports the idea that NRTIs such as tenofovir could have certain level of hepatotoxicity.

More than one-third of the patients received DRV/c (from the start of the study in half of these cases) and showed no difference in safety and effectiveness outcomes with those receiving DRV/r.

Study limitations include its retrospective, multi-center design, although the necessary data were recovered for almost all patients. Inclusion bias was minimized by recruiting all patients who had been prescribed with the DT under study in the participating hospitals. However, it was not possible to evaluate potential long-term changes in the toxicity of this combination due to the study design and short follow-up period.

Conclusions

Dual therapy with RPV + bDRV in the clinical setting has proven to be effective, even in patients with advanced HIV infection, extended exposure to ART, low CD4 lymphocyte nadir, history of VF, and/or previous non-suppressive ART.