Characterization of virologic failure patients on darunavir/ritonavir in treatment-experienced patients in TITAN
10 September 2009
De Meyer, Sandra; Lathouwers, Erkki; Dierynck, Inge; De Paepe, Els; Van Baelen, Ben; Vangeneugden, Tony; Spinosa-Guzman, Sabrina; Lefebvre, Eric; Picchio, Gaston; de Bethune, Marie-Pierre
aTibotec BVBA, Mechelen, Belgium
bJanssen-Cilag BV, Amsterdam, The Netherlands
cTibotec Inc., Yardley, Pennsylvania, USA.
Objective: Characterization of resistance development in virologic failure patients on the protease inhibitor darunavir administered with low-dose ritonavir (DRV/r) in the 48-week analysis of TMC114/r In Treatment-experienced pAtients Naive to lopinavir (TITAN).
Design: TITAN is a randomized, controlled, open-label, phase III, noninferiority trial comparing the efficacy and safety of DRV/r with that of lopinavir/ritonavir (LPV/r) in HIV-1-infected, treatment-experienced, LPV-naive patients. The primary endpoint was the proportion of patients with HIV-1 RNA less than 400 copies/ml at week 48.
Methods: Patients received DRV/r 600/100 mg twice daily (n = 298) or LPV/r 400/100 mg twice daily (n = 297), and an optimized background regimen. Patients who lost or never achieved HIV-1 RNA less than 400 copies/ml after week 16 were considered virologic failure patients. Genotyping and phenotyping were performed.
Results: The virologic failure rate in the DRV/r arm (10%, n = 31) was lower than in the LPV/r arm (22%, n = 65). Furthermore, fewer virologic failure patients in the DRV/r arm than in the LPV/r arm developed primary protease inhibitor mutations (6 vs. 20) or nucleoside reverse transcriptase inhibitor resistance-associated mutations (4 vs. 15). In addition, fewer virologic failure patients on DRV/r than on LPV/r lost susceptibility to the protease inhibitor (3 vs. 13) or nucleoside reverse transcriptase inhibitor(s) (3 vs. 14) used in the treatment regimen or to other protease inhibitors. Most DRV/r-treated virologic failure patients retained susceptibility to all protease inhibitors.
Conclusion: In treatment-experienced, LPV-naive patients, the overall virologic failure rate in the DRV/r arm was low and was associated with limited resistance development. These findings showed that the use of DRV/r in earlier lines of treatment was less likely to lead to cross-resistance to other protease inhibitors compared with LPV/r.
One of the challenges in treating patients with HIV infection is choosing the most appropriate therapeutic regimen. When deciding upon therapy from the available treatments, efficacy, safety and convenience should be considered. Treatment selection can be limited by the resistance to HAART that emerged upon failing prior regimens or the concern that failure on the regimen being selected may preclude future treatment options. These factors continue to present a significant challenge in managing HIV infection [1-3]. Therefore, given its potential to limit future therapy options for antiretroviral-experienced patients, characterizing the development of resistance to new antiretroviral drugs may help clinicians improve treatment decisions in earlier lines of therapy.
The protease inhibitors remain a widely used class for HIV therapy. These drugs have shown efficacy in treatment-naive and treatment-experienced patients [4-8]. Pharmacokinetic boosting of most protease inhibitors with low-dose ritonavir has increased their bioavailability [9-12], which in turn has helped patients overcome and prevent resistance, allowed less frequent dosing and encouraged treatment adherence .
Protease inhibitors have an important advantage over other antiretroviral drugs; numerous mutations are often necessary for the virus to develop clinically significant resistance to them . Pharmacokinetic boosting of protease inhibitors has increased their virologic potency by increasing plasma trough concentrations and in turn improving the barrier against the development of resistance [15-17]. However, the emergence of HIV strains that are cross-resistant to different protease inhibitors has been observed and is a major challenge in treatment . Therefore, protease inhibitors with activity against protease inhibitor-resistant HIV strains and a good tolerability profile are still needed.
The protease inhibitor darunavir (DRV, TMC114) has shown significant, broad-spectrum in-vitro antiretroviral activity against both wild-type and multidrug-resistant HIV-1 strains . DRV has a high genetic barrier to the development of resistance and allows the retention of antiviral activity despite the presence of multiple protease mutations . In clinical trials, DRV with low-dose ritonavir (DRV/r) has shown significant antiretroviral activity and tolerability in a wide range of adult patients, from treatment-naive to highly treatment-experienced patients [20-22]. In the Performance Of TMC114/r When evaluated in treatment-Experienced patients with protease inhibitor Resistance (POWER) 1 and 2 trials, a significantly higher proportion of patients treated with DRV/r and an optimized background regimen (OBR) achieved HIV-1 RNA less than 50 copies/ml at week 48 than those treated with investigator-selected protease inhibitors and OBR (45 vs. 10%, P < 0.0001) . In the 48-week primary analysis of the phase III ARTEMIS (AntiRetroviral Therapy with TMC114 ExaMined In naive Subjects) trial (TMC114-C211), the efficacy and tolerability of once-daily DRV/r (800/100 mg) has been recently demonstrated in treatment-naive patients . In ARTEMIS trial, no primary protease inhibitor mutations developed in DRV/r virologic failure patients, which is consistent with other trials performed with boosted protease inhibitors as first-line therapy . However, in the recently published 48-week results of the CASTLE study, one virologic failure patient taking atazanavir (ATV)/ritonavir developed primary protease inhibitor mutations .
DRV/r was also studied in treatment-experienced, lopinavir (LPV)-naive patients in the phase III TITAN (TMC114/r In Treatment-experienced pAtients Naive to lopinavir) trial (TMC114-C214). In the 48-week primary analysis of this trial, treatment with DRV/r 600/100 mg twice daily (b.i.d.) was not only noninferior to LPV/r 400/100 mg b.i.d. in the proportion of patients with confirmed HIV-1 RNA less than 400 copies/ml [77 vs. 68%; estimated least square mean (LSM) difference 9%, 95% confidence interval (CI) 2-16; P < 0.0001, per-protocol, time-to-loss of virologic response (TLOVR)] but also DRV/r treatment led to a significantly higher proportion of patients with a broad range of treatment experience achieving this primary endpoint (77 vs. 67% for LPV/r; estimated LSM difference 10%, 95% CI 2-17; P = 0.008, intent-to-treat (ITT)-TLOVR] . Moreover, statistically significantly more patients in the DRV/r group achieved full virologic suppression (HIV-1 RNA < 50 copies/ml) through 48 weeks than in the LPV/r group (71 vs. 60%; estimated LSM difference 11%, 95% CI 3-19; P = 0.005, ITT-TLOVR). Furthermore, in patients who had a baseline LPV fold-change in 50% effective concentration (FC50) 10 or less, the virologic response (HIV-1 RNA < 50 copies/ml) at week 48 in the DRV/r group was noninferior to that of the LPV/r group (70 vs. 63%; estimated LSM difference 8%, 95% CI -1 to 16; P < 0.0001, ITT-TLOVR). Both treatment regimens were generally well tolerated.
The present analysis studies the influence of the number of baseline mutations on virologic response and characterizes the development of resistance in virologic failure patients in TITAN.
In the TITAN trial, 595 patients were randomized and treated (DRV/r, 298; LPV/r, 297). Baseline characteristics and treatment history were similar between the two groups (Table 1) . The mean baseline viral load was 4.30 log10 copies/ml and the median CD4 cell count was 232 cells/µl; 31% patients were protease inhibitor naive and 31% had used more than one protease inhibitor. Treatment groups were generally similar with respect to phenotypic and genotypic susceptibility at baseline; 98% of patients were fully susceptible to DRV (FC50 ≤ 10) and 90% were fully susceptible to LPV (FC50 ≤ 10). The median phenotypic sensitivity score of the OBR was two.
Influence of the number of baseline mutations on virologic response
At week 48, the percentage of patients with HIV-1 RNA less than 400 copies/ml was 90% in the DRV/r group vs. 79% in the LPV/r group (ITT-TLOVR nonvirologic failure-censored analysis). The virologic RR (HIV-1 RNA < 400 copies/ml) to DRV/r at week 48 was diminished (60%) when at least three DRV RAMs were present at baseline (Fig. 1a). In this analysis, a diminished RR was defined as a RR below 75% of the overall RR for the DRV/r group (90%), that is, below 68%. The subgroup of patients (n = 10) with at least three DRV RAMs had a median number of 13 protease inhibitor RAMs . Regardless of the number of DRV RAMs at baseline, the virologic RR at week 48 was numerically higher in the DRV/r group than in the LPV/r group (91 vs. 83%, 97 vs. 74%, 89 vs. 36% and 60 vs. 46% in the subgroups with 0, 1, 2 or ≥3 DRV RAMs, respectively). Similar results were obtained when studying the proportion of patients achieving HIV-1 RNA less than 50 copies/ml at week 48 (Fig. 1b).
Number of virologic failure patients
The virologic failure rate in the DRV/r group was lower than in the LPV/r group (10%, 31/298 patients vs. 22%, 65/297 patients, respectively; P < 0.001; TLOVR nonvirologic failure-censored algorithm), with a similar proportion of rebounders and never-suppressed patients present per arm (Fig. 2a). The number of virologic failure patients was also determined in the subgroup with baseline LPV FC50 10 or less (the low clinical cutoff for LPV ) and in the subgroup of patients who previously used one or less protease inhibitor. In the subgroup with baseline LPV FC50 10 or less, 26 out of 263 (10%) DRV/r-treated patients compared with 50 out of 261 (19%) LPV/r-treated patients experienced virologic failure (P = 0.003) (Fig. 2b). In the subgroup that previously used one or less protease inhibitor, 19 out of 202 (9%) DRV/r-treated patients compared with 36 out of 208 (17%) LPV/r-treated patients experienced virologic failure (P = 0.021) (Fig. 2c).
Development of mutations in virologic failure patients
The development of mutations at endpoint as compared with baseline was studied in virologic failure patients with an available genotype at both time points. Baseline and endpoint genotypes were available for 28 out of 31 DRV/r-treated and for 56 out of 65 LPV/r-treated virologic failure patients. Fewer virologic failure patients in the DRV/r group than in the LPV/r group developed primary protease inhibitor mutations (21%, 6/28 virologic failure patients vs. 36%, 20/56 virologic failure patients, respectively) or NRTI RAMs (14%, 4/28 virologic failure patients vs. 27%, 15/56 virologic failure patients, respectively) (Fig. 2d).
In the DRV/r group, six virologic failure patients developed a primary protease inhibitor mutation. Table 2 shows data for these six patients and Table 3 shows data for the additional DRV/r-treated virologic failure patients without developing primary protease inhibitor mutations. Tables 2 and 3 summarize the previous protease inhibitor use and the resistance data, showing the baseline protease inhibitor RAMs (including the primary protease inhibitor mutations), the protease inhibitor RAMs developing at endpoint relative to baseline and the susceptibility to protease inhibitors. In one out of six virologic failure patients with developing primary protease inhibitor mutations (patient #6), the L90M mutation, detected at endpoint but not at baseline, had been detected earlier in the screening sample. In the other five virologic failure patients (patients #1-5), the following primary protease inhibitor mutations (all DRV RAMs) developed: V32I (n = 3), I47V and L76V (n = 2) and I54L (n = 1). These five patients had all previously used at least two protease inhibitors.
In the LPV/r group, 20 virologic failure patients developed a primary protease inhibitor mutation. In one out of 20 virologic failure patients, the primary protease inhibitor mutation M46I, detected at endpoint but not at baseline, had previously been detected in the screening sample. In the other 19 virologic failure patients, the following primary protease inhibitor mutations (all LPV RAMs) developed: M46I (n = 6), M46L and L76V (n = 4), L33F, I47A, I47V, V82A and I84V (n = 2) and V32I, I50V and I54M (n = 1) . Nine of these 19 patients had previously used only one protease inhibitor, whereas the other 10 had previously used at least two protease inhibitors .
Furthermore, the difference between treatment groups in the number of virologic failure patients who developed primary protease inhibitor mutations or NRTI RAMs was maintained or increased in the subgroups who were fully susceptible to LPV at baseline and had previously used one or less protease inhibitor. Among patients fully susceptible to LPV at baseline, only one out of 23 (4%) DRV/r-treated as compared with 14 out of 43 (33%) LPV/r-treated virologic failure patients developed a primary protease inhibitor mutation (Fig. 2e). Among patients who had previously used one or less protease inhibitor, none of the 17 (0%) DRV/r-treated virologic failure patients developed a primary protease inhibitor mutation compared with 10 out of 31 (32%) LPV/r-treated virologic failure patients (Fig. 2f).
Loss of susceptibility to study protease inhibitor or nucleoside reverse transcriptase inhibitors in virologic failure patients
The loss of susceptibility at endpoint as compared with baseline was studied in virologic failure patients with available phenotypes at both time points. Baseline and endpoint phenotypes (Antivirogram) were available for 28 out of 31 DRV/r-treated and for 54 out of 65 LPV/r-treated virologic failure patients. Fewer DRV/r-treated than LPV/r-treated virologic failure patients lost susceptibility relative to baseline to the assigned study protease inhibitor or NRTIs used in the OBR (Fig. 2g-i). Among the patients with available data and who were susceptible to the protease inhibitor in the study regimen at baseline, three out of 26 (12%) DRV/r-treated virologic failure patients lost susceptibility to DRV compared with 13 out of 42 (31%) LPV/r-treated virologic failure patients who lost susceptibility to LPV (Fig. 2g). More information on the three DRV/r-treated virologic failure patients who lost susceptibility to DRV (patients #3-5) can be found in Table 2. All three DRV/r-treated virologic failure patients who lost susceptibility to DRV at endpoint had used at least two protease inhibitors and had extensive baseline protease inhibitor resistance. One patient (patient #3) was susceptible to IDV and TPV at baseline, whereas the other two (patients #4 and 5) were only susceptible to DRV and had already a decreased susceptibility to all other protease inhibitors at baseline. These three DRV/r-treated virologic failure patients developed primary protease inhibitor mutation(s) (all DRV RAMs). Furthermore, among the patients with available data and who were susceptible to an NRTI in the study regimen at baseline, three out of 24 (13%) DRV/r-treated compared with 14 out of 43 (33%) LPV/r-treated virologic failure patients were susceptible to fewer NRTI(s) used in their OBR at endpoint (Fig. 2g).
Among virologic failure patients who were phenotypically fully susceptible to LPV at baseline, none of 23 (0%) DRV/r-treated virologic failure patients with baseline susceptibility to DRV lost susceptibility to DRV, whereas 13 out of 42 (31%) LPV/r-treated virologic failure patients with baseline susceptibility to LPV lost susceptibility to LPV (Fig. 2h). Among virologic failure patients who previously used one or less protease inhibitor, none of 17 (0%) DRV/r-treated virologic failure patients with baseline susceptibility to DRV lost susceptibility to DRV, whereas eight out of 29 (28%) LPV/r-treated virologic failure patients with baseline susceptibility to LPV lost susceptibility to LPV (Fig. 2i).
Development of cross-resistance to other protease inhibitors
The maintenance of phenotypic and genotypic susceptibility to other protease inhibitors in virologic failure patients was also evaluated in this analysis. The phenotypic (Antivirogram) and genotypic (ANRS) resistance interpretation calls from the baseline and endpoint samples of all DRV/r-treated virologic failure patients are shown in Tables 2 and 3.
On the basis of Antivirogram data, almost all DRV/r-treated virologic failure patients retained susceptibility to APV (22/22 patients with baseline susceptibility to APV), ATV (22/22), IDV (23/24), LPV (23/23), SQV (22/22) and TPV (24/25) (Fig. 3a). The proportion of LPV/r-treated virologic failure patients retaining phenotypic susceptibility to APV, ATV, IDV, LPV and SQV was lower (Fig. 3a).
Similar results were obtained using the ANRS HIV-1 genotypic drug resistance interpretation system. Again, almost all DRV/r-treated virologic failure patients retained genotypic susceptibility to APV (22/24), ATV (22/22), IDV (19/20), LPV (24/25), SQV (18/20) and TPV (23/23), whereas the proportion of LPV/r-treated virologic failure patients retaining susceptibility to APV, ATV, IDV and LPV was lower (Fig. 3b).
Efficacy, tolerability, convenience and a high barrier to the development of resistance are important factors to consider when choosing an antiretroviral regimen. The efficacy and tolerability of DRV/r in HIV-1-infected patients with limited treatment experience has been demonstrated in the primary 48-week analysis of the TITAN trial . Results from the present analysis showed that DRV/r, compared with LPV/r, was less frequently associated with virologic failure, the development of resistance to the protease inhibitor in the regimen, loss of OBR activity and cross-resistance to other protease inhibitors. These findings contribute to the understanding of the efficacy and resistance profile of DRV/r when used in HIV-1-infected patients with limited treatment experience.
A diminished virologic RR with DRV/r was observed at week 48 when at least three DRV RAMs were present at baseline, confirming previous findings derived from the POWER studies . Nevertheless, the virologic RR at week 48 was higher with DRV/r than with LPV/r, irrespective of the number of DRV RAMs at baseline. A similar study using the IAS-USA list of primary protease inhibitor mutations  has shown that the virologic RR (HIV-1 RNA < 50 copies/ml, ITT-TLOVR nonvirologic failure-censored analysis at week 48) was consistently higher (by at least 10%) in the DRV/r arm than in the LPV/r arm among patients with at least one primary protease inhibitor mutation at baseline . This held true, even in patients in whom both DRV and LPV showed evidence of full susceptibility and thus would be expected to provide similar responses .
The higher RR in the DRV/r group as compared with the LPV/r group was not driven by the presence of baseline mutations that hypersensitize HIV to DRV. In previous publications [34,35], the mutations D30N, I50L and N88S were described as associated with an increased susceptibility to DRV when present as the only primary protease inhibitor mutation, and mutations E35D and V82A were identified as having a positive impact on virologic response to DRV/r. In the TITAN trial, no statistical difference in response was found in patients in the DRV/r group with or without one of those mutations at baseline (data not shown).
The higher virologic failure rate observed in the week 48 analysis with LPV/r treatment was consistently observed, even after excluding patients with baseline LPV resistance (LPV FC50 > 10) or those who previously used at least two protease inhibitors.
Both genotypic and phenotypic analyses showed that DRV/r treatment was less likely to lead to resistance development compared with LPV/r among virologic failure patients in the trial. Fewer DRV/r-treated than LPV/r-treated virologic failure patients developed primary protease inhibitor mutations or NRTI RAMs. This outcome was observed even when patients with baseline LPV resistance or those who previously used at least two protease inhibitors were excluded from the analysis. Fewer DRV/r-treated than LPV/r-treated virologic failure patients lost susceptibility to the regimen protease inhibitor. In the group that was susceptible to LPV at baseline and in those that used less than two protease inhibitors at baseline, none of the DRV/r-treated virologic failure patients lost susceptibility to DRV. Consistent with these observations, fewer DRV/r-treated than LPV/r-treated virologic failure patients lost genotypic or phenotypic susceptibility to any protease inhibitor, which showed that using DRV/r in the earlier lines of antiretroviral therapy was less likely to confer cross-resistance to other protease inhibitors than using LPV/r. Analyses on the development of resistance were exploratory and the sample size of the subgroups was small. Therefore, further investigation may be needed to confirm the current findings.
In summary, these analyses from the TITAN study provide important information about virologic failure patients on DRV/r and LPV/r that may have important implications for their future treatment options. First, virologic failure was less common with DRV/r than with LPV/r in treatment-experienced patients. Second, development of resistance to the protease inhibitor or NRTIs in the regimen was less common with DRV/r than LPV/r. Finally, cross-resistance to other protease inhibitors occurred less frequently in patients failing on DRV/r than in patients failing on LPV/r, even in patients susceptible to LPV at baseline. Results from the ARTEMIS trial, examining the use of once-daily DRV/r in treatment-naive patients, also suggest that protease inhibitor resistance is not likely to emerge upon virologic failure with DRV/r treatment . In conclusion, these findings help demonstrate some of the potential clinical benefits that can be provided by use of DRV/r in less treatment-experienced, HIV-infected populations.
Study design and patient population
TITAN is an international, randomized, controlled, open-label, 96-week, phase III noninferiority trial designed to assess the efficacy and safety of DRV/r compared with LPV/r in LPV-naive, treatment-experienced, HIV-1-infected patients . Patients with HIV-1 RNA more than 1000 copies/ml and on HAART for at least 12 weeks were randomized to receive either DRV/r 600/100 mg b.i.d. or LPV/r 400/100 mg b.i.d., and an OBR. The OBR [at least two nucleoside reverse transcriptase inhibitors (NRTIs), nonnucleoside reverse transcriptase inhibitors (NNRTIs) or both] was selected on the basis of treatment history and screening resistance data. Use of enfuvirtide (ENF) was disallowed in the OBR. For cases in which an NNRTI was included in the OBR, the dosage of LPV/r was increased as recommended by the prescribing information to 533/133 mg b.i.d. for the capsule  and to 600/150 mg b.i.d. for the tablet formulation . Patients on a structured treatment interruption lasting at least 4 weeks were permitted to enroll. Patients with previous or current use of LPV, DRV, tipranavir (TPV) or ENF and current use of investigational antiretroviral drugs were excluded.
The study protocol was reviewed and approved by independent ethics committees and appropriate health authorities and was conducted in accordance with the principles of Good Clinical Practices and the Declaration of Helsinki. All patients gave their written informed consent.
Analyses were performed on the dataset from the primary efficacy analysis of the TITAN trial, when all patients had reached week 48 of treatment or discontinued earlier (ITT population).
Viral isolates and laboratory assays
Plasma samples were collected at screening, baseline and regular intervals from patients participating in the study. Viral phenotypic (Antivirogram; VIRCO BVBA, Mechelen, Belgium) and genotypic determinations were performed by Virco BVBA (Mechelen, Belgium). The samples taken at screening, baseline, week 24, week 48 or withdrawal visit were analyzed when received. Samples taken at other time points were analyzed upon request of the Tibotec protocol virologist. Only samples with HIV-1 RNA more than 1000 copies/ml were analyzed. Plasma HIV-1 RNA levels were measured using the Roche Amplicor HIV-1 Monitor assay version 1.5 or the Roche UltraSensitive method version 1.5 (Roche Molecular Systems, Branchburg, New Jersey, USA). Analyses of primary protease inhibitor mutations, DRV resistance-associated mutations (RAMs), LPV RAMs and protease inhibitor RAMs were based on the International AIDS Society (IAS)-USA list .
The relationship between baseline genotype and the corresponding virologic response at week 48 was studied using a TLOVR nonvirologic failure-censored algorithm. This means that no viral load imputations were performed at time points after the last on-treatment viral load determination for patients who discontinued for reasons other than virologic failure. As a result, higher response rates (RR) were observed in the nonvirologic failure-censored analyses compared to the TLOVR analyses.
Virologic failure patients were defined as either 'rebounders' (initially achieved a confirmed virologic response, i.e., HIV-1 RNA < 400 copies/ml, followed by two consecutive measurements of HIV-1 RNA at least 400 copies/ml or discontinuation of treatment after one measurement of HIV-1 RNA at least 400 copies/ml) or 'never suppressed' (never achieved a confirmed virologic response, i.e., HIV-1 RNA < 400 copies/ml). Patients who discontinued before week 16 and did not have the chance to respond were not taken into account in this analysis. Virologic response was determined using the TLOVR nonvirologic failure-censored method, which means that no viral load imputations were performed at time points after the last on-treatment viral load determination for patients who discontinued for reasons other than virologic failure.
The development of a mutation was defined as the detection of a mutation by population sequencing at endpoint (i.e., the last available time point with a genotype during the treatment period) but not at baseline.
The loss of phenotypic susceptibility to an antiretroviral was defined as having an FC50 value above the biological/clinical cutoff of the Antivirogram at endpoint (i.e., the last available time point with a phenotype during the treatment period) but not at baseline. A clinical cutoff value of 10 was used for DRV  and LPV  and 3 for TPV . Biological cutoff values of 2.2, 2.4, 2.4 and 1.8 were used for amprenavir (APV), ATV, indinavir (IDV) and saquinavir (SQV), respectively. Genotypic interpretation of protease inhibitor susceptibility was performed using the Agence Nationale de Recherches sur le Sida (ANRS) algorithm of 2007 . Interpretation using the ANRS system results in three possible calls: susceptible (S), intermediate (I) and resistant (R). The loss of genotypic susceptibility to an antiretroviral was defined as having an intermediate or resistant call at endpoint, while having a susceptible call at baseline. Cross-resistance was studied by measuring the loss of susceptibility to other protease inhibitors at endpoint.
An exact chi-squared test was used to study whether the difference between the two treatment groups in number of virologic failure patients was statistically significant. No formal statistical tests were used to compare the differences in the development of resistance between the treatment groups, as the study was not powered to achieve this and also because the sample sizes were too small in these subgroups.
Financial assistance to support this service was provided by Tibotec BVBA, Mechelen, Belgium. The authors thank the patients and their families for their participation and support during this study, and the TITAN study team, including the investigators for their contributions. The authors also acknowledge Kerry Padilla-Dumlao (Medical Writer, Gardiner-Caldwell Communications, Macclesfield, UK) for assistance in preparing the outline, editing the manuscript and collating author contributions.
All authors are current employees of Tibotec and contributed to the design and analyses of the TMC114-C214 (TITAN) trial. Furthermore, all authors were involved in the development of the manuscript and have reviewed and approved the final version.
Some of the data in this manuscript have been presented at the Fourth IAS Conference on HIV Pathogenesis, Treatment and Prevention (22-25 July 2007; Sydney, Australia), the 47th Interscience Conference on Antimicrobial Agents and Chemotherapy (17-20 September 2007; Chicago, Illinois, USA), the 11th European AIDS Conference (24-27 October 2007; Madrid, Spain) and the 15th Conference on Retroviruses and Opportunistic Infections (3-6 February 2008; Boston, Massachusetts, USA).