Antiviral Therapy 2016; 21: (doi: /IMP3011)

Transcription

1 Antiviral Therapy 2016; 21: (doi: /IMP3011) Original article Assessment of etravirine resistance in HIV-1-infected paediatric patients using population and deep sequencing: final results of the PIANO study Lotke Tambuyzer 1 *, Kim Thys 1, Annemie Hoogstoel 1, Steven Nijs 1, Frank Tomaka 2, Magda Opsomer 1, Sandra De Meyer 1, Johan Vingerhoets 1 1 Janssen Infectious Diseases BVBA, Beerse, Belgium 2 Janssen Research & Development, LLC, Titusville, NJ, USA *Corresponding author ltambuyz@its.jnj.com Background: We assessed etravirine resistance in treatmentexperienced, HIV-1-infected children (n=41)/adolescents (n=60) who received twice-daily etravirine 5.2 mg/kg and a background regimen (boosted protease inhibitor plus nucleoside/nucleotide reverse transcriptase inhibitors, optional enfuvirtide/raltegravir) in a Phase II, open-label, multicentre trial (PIANO). Methods: In addition to phenotypes, viral genotypes were assessed by population and deep sequencing (PS and DS) in virological failures (VFs; baseline and end point) and responders (baseline). Minority resistance-associated mutations (RAMs) were defined as those with frequencies above 1% and not detected with PS. Results: By week 48, 41/101 (40.6%) patients experienced VF; 17/41 (41.5%) VFs and 22/54 (40.8%) responders had 1 baseline etravirine RAM by PS, mainly A98G, K101E, V106I and G190A. Baseline minority etravirine RAMs (n) were detected in 8/40 VFs (V90I [2], A98G [1], L100I [1], V106I [1], E138G [1] and Y181C [2]) and 5/38 responders (V90I [3], A98G [1], V106I [1] and E138G [1]). The most frequent emerging non-nucleoside reverse transcriptase inhibitor RAMs detected by PS ( 3 VFs; n) were the etravirine RAMs Y181C (8), V90I (3), L100I (3) and E138A (3). In 15 of 29 (51.7%) VFs with baseline DS/PS and end point PS data, 1 emerging etravirine RAM was detected by PS, which was not detected at baseline by DS in most cases (12/15 [80.0%]). In 10/26 (38.5%) VFs with baseline/end point DS data, 1 additional emerging minority etravirine RAM was detected. Conclusions: Patterns of etravirine resistance in adults, adolescents and children experiencing VF are similar. The presence of minority etravirine RAMs at baseline was not consistently associated with treatment failure. ClinicalTrials.gov: NCT Introduction For HIV-1-infected paediatric patients the number of approved antiretroviral (ARV) agents, although increasing, is limited compared with available treatment options for adults [1]. A key challenge in treating HIV-1 infection in both children and adults is the presence and/or development of resistance-associated mutations (RAMs), which reduce ARV effectiveness thereby limiting treatment options [2,3]. Non-nucleoside reverse transcriptase inhibitors (NNRTIs) are a key component of combination ARV regimens for treatment-naive, HIV-1-infected children and adolescents [1,4]. However, the prevalence of NNRTI RAMs in paediatric patients failing firstline therapy is increasing (for example, from 6% in to 42% in in Spain [5]), and can be very high in resource-poor settings (for example, 88% [6] to 98% [7]). Etravirine 200 mg twice daily is the only NNRTI approved for treating HIV-1 infection in ARV treatment-experienced adults. In the EU, etravirine is combined with a boosted protease inhibitor (PI) and other ARVs [8]; in the USA, etravirine is approved for patients who have viral replication and HIV-1 strains resistant to an NNRTI and other ARVs [9]. In patients who have experienced virological failure (VF) on an NNRTI-containing regimen, etravirine cannot be used with nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs) only. Initial approval of etravirine in adults was based on the efficacy and safety of etravirine plus a background 2016 International Medical Press (print) (online) 317

2 L Tambuyzer et al. regimen of darunavir/ritonavir and NRTIs in the Phase III DUET trials [10 13]. The indication of etravirine was expanded to ARV treatment-experienced paediatric patients aged 6 years and older [8,9,14]. Initially, a Phase I trial in treatment-experienced children and adolescents enabled the selection of a weight-based dose (5.2 mg/kg twice daily) of etravirine for further assessment [15]. Subsequently, the pivotal 48-week Phase II (Pediatric study of Intelence As an NNRTI Option) PIANO study evaluated the safety, efficacy and pharmacokinetics of etravirine with an optimized background regimen (OBR) in treatment-experienced, HIV-1-infected children ( 6 <12 years) and adolescents ( 12 <18 years). At 48 weeks, 53.5% of patients in PIANO had a virological response (HIV-1 RNA plasma viral load <50 copies/ml; time-to-loss of virological response [TLOVR]) to etravirine with an OBR and no new safety issues were identified [14]. A secondary objective of PIANO was to assess the evolution of the viral genotype and phenotype up to 48 weeks, which is presented here. Methods Study design, participants and treatment Details of this Phase II, open-label, multicentre study (PIANO; TMC125-TiDP35-C213; Clinical Trials.gov: NCT ) have been described previously [14]. Key inclusion criteria were: treatment-experienced (stable ARV for 8 weeks), HIV-1-infected children ( 6 <12 years) and adolescents ( 12 <18 years), weighing 16 kg, currently failing virologically (confirmed plasma viral load of >500 copies/ml). Key exclusion criteria were: etravirine-resistant HIV-1 virus at screening (using virco TM TYPE HIV-1; Janssen Diagnostics BVBA, Beerse, Belgium) and any active clinically significant disease. Patients received etravirine at 5.2 mg/kg (maximum dose 200 mg) twice daily for 48 weeks, and an investigator-selected OBR of at least two ARVs (a ritonavir-boosted PI with NRTIs and optional enfuvirtide and/or raltegravir). Resistance analyses VFs and responders were defined using the TLOVR algorithm (plasma viral load <50 copies/ml, non-vf censored) for patients who were still in PIANO at week 12 and therefore had the chance to respond. VFs were categorized as rebounders or non-responders. Rebounders were those who had achieved a confirmed viral load <50 copies/ml followed by either a confirmed viral load 50 copies/ml or discontinued after week 12 with a last viral load 50 copies/ml; non-responders had never achieved a confirmed viral load <50 copies/ml. Viral genotype (population sequencing [PS]) and phenotype evaluations (Antivirogram ; Janssen Diagnostics BVBA) were conducted by Janssen Diagnostics BVBA. RAMs (etravirine, NNRTI, NRTI, primary PI and PI) were identified using published data [16 19]. The etravirine weighted genotypic score (WGS) was derived by adding the individual weight factors for each etravirine RAM (V90I, A98G, L100I, K101E/H/P, V106I, E138A/ G/K/Q, V179D/F/T, Y181C/I/V, G190A/S, M230L) present in a patient s sample [18,19]. A cumulative score of 0 2, and 4 is associated with high, intermediate and reduced etravirine response, respectively. This WGS was reported to reliably predict etravirine sensitivity and was in good correlation with other scoring systems such as those developed by Monogram, Stanford HIVdb, ANRS and Rega [20]. In the phenotype assay, a drug was considered active if the fold change (FC) in effective concentrations inhibiting 50% of viral activity (EC 50 ) was below or equal to the clinical or biological cutoff, that is, 3 and 13 for etravirine [19], 3.3 for nevirapine and 6 for efavirenz. Patients with etravirine FC values 3, 3 13 or >13 were defined as sensitive, intermediate or resistant to etravirine, respectively. Deep sequencing (DS) by Illumina technology [21,22] (of an amplicon encompassing HIV-RT position 1 400) was used to detect minority RAMs, defined as those with frequencies >1% (threshold chosen based on error estimation and plasmid testing) and not detected with PS. Median coverage was 58,713 reads per position (so 1% represented approximately 500 reads). Analyses included the 20 known etravirine RAMs plus 30 other NNRTI RAMs (K101Q, K103H/ N/S/T, V106A/M, V108I, E138R, V179E/G/I/L, Y188C/ H/L, V189I, G190C/E/Q/T, H221Y, P225H, F227C/L, M230I, P236L, K238N/T, Y318F) [16,17]. Data analyses Descriptive genotypic and phenotypic parameters at baseline (for example, type and number of RAMs, median FC values) were summarized for VFs and responders, and for children and adolescents. Changes in these parameters from baseline to end point were described for VFs. An overview of genotypic (PS and DS) and phenotypic data available at baseline and/or end point for VFs and/or responders is shown in Table 1. Matched genotypic (PS) and phenotypic data at both baseline and end point were available for 23 of the 41 VFs. Statistical comparisons were performed using Fisher s exact test (n/total n versus n/total n), ANOVA test (mean values) or Kruskal Wallis test (median numbers). Results Patient demographics and baseline disease characteristics A total of 41 children and 60 adolescents were enrolled and received etravirine; 76 (75.2%) completed 48 weeks of treatment. Demographic and baseline characteristics International Medical Press

3 Etravirine resistance in HIV-1-infected children and adolescents Table 1. Overview of genotypic (PS or DS) and phenotypic data available at baseline and/or end point for patients in PIANO PIANO population VFs (n=41) Responders (n=54) (TLOVR non-vf censored) Children Adolescents Children Adolescents Total (n=95) Baseline PS genotype 11 a 30 a 28 b 26 b 95 DS genotype 11 c 29 c 19 d 19 d 78 Phenotype 9 e 28 e 26 f 26 f 89 End point PS genotype 7 23 NA NA 30 g DS genotype 7 19 NA NA 26 Phenotype 6 18 NA NA 24 Children: 6 <12 years, adolescents: 12 <18 years. a Total equals 41. b Total equals 54. c Total equals 40. d Total equals 38. e Total equals 37. f Total equals 52. g For 29 of these 30 virological failures (VFs), baseline deep sequencing (DS) genotype data are available. NA, not available; PS, population sequencing; TLOVR, time-to-loss of virological response. have been described [14]. The majority of enrolled patients were female (63.4%), and 49.5%, 30.3% and 20.2% were White, Black/African-American and Asian, respectively. Most patients were from Thailand (19.8%), Argentina (14.9%), the USA (14.9%) and South Africa (9.9%). Median baseline log 10 viral load was 3.6 (children) and 4.0 (adolescents) log 10 copies/ml. Most patients harboured HIV-1 group M clade B (47.5%); other observed clades were CRF01_AE (19.8%), C (11.9%), CRF12_BF (9.9%), CRF02_AG (4.0%), F1 (3.0%), CRF14_BG (2.0%), CRF06_CPX (1.0%) and CRF11_CPX (1.0%). All patients had previously used at least two ARVs; 2 NRTIs 100%; 1 NNRTI 75.2% (children 65.9%; adolescents 81.7%); 1 PI 80.2% (children 80.5%; adolescents 80.0%). Three adolescents had previously used enfuvirtide. During PIANO, all patients received a boosted PI, most frequently darunavir/ritonavir (51.2% children; 51.7% adolescents) or lopinavir/ritonavir (43.9% children; 35.0% adolescents). Most patients (63.4% children; 56.7% adolescents) also received two NRTIs; the most frequently used were lamivudine (51.2%) or zidovudine (39.0%) in children and tenofovir disoproxil fumarate (51.7%) or lamivudine (35.0%) in adolescents. Enfuvirtide and raltegravir were used by two (2%) and 11 (10.9%) patients, respectively. Overall baseline resistance At baseline, 41/101 (40.6%: 17 children; 24 adolescents) patients had 1 etravirine RAM by PS. The median number of etravirine, NNRTI, NRTI and PI RAMs at baseline by PS are shown in the Additional file 1. One adolescent had three etravirine RAMs (K101E, V179T, G190A), with a WGS of 3. Overall, the etravirine RAMs (n) observed were G190A (13), K101E (9), A98G (8), V106I (8), Y181C (4), E138A (3), V90I (3), K101H (2), G190S (2), E138K (1), E138Q (1), L100I (1), V179D (1) and V179T (1). In 91.1% (92/101; 38 children; 54 adolescents) of patients, an etravirine WGS of 0-2 was noted. In 8.9% (9 of 101; 3 children; 6 adolescents) of patients, the etravirine WGS was None of the patients had a score 4. At baseline, 80.0% adolescents and 65.9% of children had 1 NNRTI RAM, 85.0% and 75.6% 1 NRTI RAM and 53.3% and 34.1% 1 primary PI mutation, respectively. When summing up mutations detected by PS and DS, 90.0% of adolescents (43/48) and 77.0% of children (23/30) had 1 NNRTI RAM at baseline. The frequency of baseline NNRTI RAMs using PS was not significantly different in adolescents versus children (P=0.1637), and remained as such when adding DS data (P=0.1957). The most frequent minority NNRTI RAMs (n) at baseline were V179I (13), V108I (6), V90I (5) and V189I (5) observed among VFs and responders (Table 2). The frequency of baseline minority etravirine RAMs (n) was not significantly different (P=0.5469) between VFs (8/40, 20%) and responders (5/38, 13%); V90I (2), A98G (1), L100I (1), V106I (1), E138G (1) and Y181C (2) were detected in VFs, and V90I (3), A98G (1), V106I (1) and E138G (1) in responders (Tables 2 and 3). Overall, the prevalence of minority etravirine RAMs at baseline ranged from 1.3% to 6.4% (Table 2). The high-impact (etravirine WGS 2.5) etravirine RAMs L100I and Y181C were detected as minority RAMs at baseline in three of eight VFs and not in any of the five responders (Table 3). After accounting for the minority etravirine RAMs at baseline, the etravirine WGS changed from high (based on PS data only) to intermediate or reduced response (PS + DS) in three of eight VFs, but also in three of five responders. The minority NNRTI RAMs V179E, Y188L and F227L were detected at baseline in VFs and not in responders. All three minority NNRTI RAMs were observed together with at least one etravirine RAM. In all patients, etravirine was an active ARV at screening (based on virco TM TYPE HIV-1) as per protocol. At baseline, 89/101 patients had phenotypic data Antiviral Therapy

4 L Tambuyzer et al. Table 2. Minority NNRTI RAMs (by DS) at baseline Minority NNRTI RAMs at baseline VFs (n=40) Responders (n=38) All patients (n=78) V90I 2 (5.0) 3 (7.9) 5 (6.4) A98G 1 (2.5) 1 (2.6) 2 (2.6) L100I 1 (2.5) 0 1 (1.3) K101Q 2 (5.0) 1 (2.6) 3 (3.8) K103S 1 (2.5) 2 (5.3) 3 (3.8) V106I 1 (2.5) 1 (2.6) 2 (2.6) V108I 3 (7.5) 3 (7.9) 6 (7.7) E138G 1 (2.5) 1 (2.6) 2 (2.6) V179E 1 (2.5) 0 1 (1.3) V179I 5 (12.5) 8 (21.1) 13 (16.7) Y181C 2 (5.0) 0 2 (2.6) Y188C 0 1 (2.6) 1 (1.3) Y188L 1 (2.5) 0 1 (1.3) V189I 3 (7.5) 2 (5.3) 5 (6.4) G190E 0 1 (2.6) 1 (1.3) H221Y 2 (5.0) 1 (2.6) 3 (3.8) F227L 1 (2.5) 0 1 (1.3) M230I 0 1 (2.6) 1 (1.3) K238T 2 (5.0) 1 (2.6) 3 (3.8) Y318F 1 (2.5) 1 (2.6) 2 (2.6) Data are n (%). Etravirine resistance-associated mutations (RAMs) are shown in bold. DS, deep sequencing; NNRTI, non-nucleoside reverse transcriptase inhibitor; VFs, virological failures. (Antivirogram ), and 9 (4 children, 5 adolescents) of those 89 patients (10.1%) did not have active etravirine. Eight patients had an etravirine FC 3-13 and one patient had an etravirine FC>13. Efavirenz and nevirapine were active in 21/35 (60.0%) and 18/35 (51.4%) children, and 18/54 (33.3%) and 17/54 (31.5%) adolescents, respectively. Most patients had at least one active PI (97.8%), and at least one active NRTI (95.5%). 19 patients had only 0 or 1 active ARV in their OBR; 4 were children and 15 were adolescents. Virological failures At week 48, 41/101 (40.6%) of patients experienced VF (11 children; 30 adolescents). 29 VFs were nonresponders (of which 21 were adolescents) and 12 were rebounders (of which 9 were adolescents). In 8 of 12 rebounders, the viral load did not increase above 400 copies/ml. Sixteen VFs harboured HIV-1 group M clade B; other observed clades in VFs (n) were CRF01_AE (11), C (5), CRF02_AG (4) CRF12_ BF (3) and F1 (2). No significant differences (P>0.05) were observed in baseline disease characteristics between VFs (n=41) and non-vfs (n=60) [23]. Mean (range) viral load was 4.07 ( ) and 3.85 ( ) log 10 copies/ml, the median (range) number of NNRTI RAMs [17] was 2.0 (0 5.0) and 1.0 (0 4.0), and the median (range) number of NRTI RAMs [16] was 2.0 (0 6.0) and 4.0 (0 7.0) in VFs versus non-vfs, respectively. Of the 41 VFs, 23 (56.1%) were NNRTI-experienced, and had previously received either efavirenz (n=9), nevirapine (n=10) or both (n=4). Emerging resistance in VFs In VFs, the median (range) number of etravirine RAMs (by PS) increased from baseline to end point (0 [0-3] to 1 [0-5]); the median (range) etravirine WGS also increased from 0 (0-3.0) at baseline to 1.75 (0-8.5) at end point. The most frequently emerging NNRTI RAMs (n) by PS, occurring in 3 VFs overall, were Y181C (8), V90I (3), L100I (3) and E138A (3; Figure 1). Those most frequently emerging by DS in addition to those emerging by PS ( 3 VFs) were V189I (6), V179I (4), Y181C (4), H221Y (4), E138K (3), F227L (3) and K238T (3; Figure 1). In five of eight VFs with emerging Y181C, one or more other etravirine RAMs emerged (in three instances in combination with a mutation at RT position 138). 15 of 23 (65.2%) adolescents and 3 of 7 (42.9%) children developed NNRTI RAMs as detected by PS. When combining NNRTI RAMs by PS and DS at baseline and at end point, 15 of 19 (78.9%) adolescents and 4 of 7 (57.1%) children had at least one emerging NNRTI RAM. The frequency of emerging NNRTI RAMs by PS was not significantly different between children and adolescents (P=0.3915) and remained as such when adding DS data (P=0.3401). In 15/29 (51.7%) VFs with baseline DS/PS and end point PS data, one or more emerging etravirine RAMs International Medical Press

5 Etravirine resistance in HIV-1-infected children and adolescents Table 3. Overview of VFs and responders with minority etravirine RAMs by DS at baseline End point Baseline and/or screening Emerging Etravirine etravirine RAMs: DS Etravirine Emerging Emerging RAMs: DS Etravirine Etravirine frequency, Etravirine WGS b : Etravirine etravirine etravirine frequency, Etravirine Etravirine Patient RAMs: PS RAMs: DS a % WGS b : PS PS+DS FC c RAMs: PS d RAMs: DS e % WGS b : PS f FC c VFs Children (n=2) Pt 1 None V106I NA NA NA NA NA Pt 2 None V90I NA None None 0 NA Adolescents (n=6) Pt 3 None E138G NA NA NA NA Pt 4 None A98G NA NA NA NA Pt 5 None V90I V90I A98G, 3.6, V106I, 2.2, E138A 1.5 Pt 6 None L100I L100I None Pt 7 V179D, Y181C NA Y181C None 2.5 NA G190A Pt 8 V90I Y181C None None g Responders Children (n=2) Pt 9 G190A V106I Pt 10 K101E, V90I, 1.4, G190A E138G 7.8 Adolescents (n=3) Pt 11 V106I V90I Pt 12 None A98G Pt 13 None V90I a Additional resistance-associated mutations (RAMs) compared with RAMs observed by population sequencing (PS). b Normal number denotes high response to etravirine; italic number denotes intermediate response to etravirine; bold number denotes reduced response to etravirine. c Normal number denotes sensitive to etravirine; italic number denotes intermediate sensitivity to etravirine; bold number denotes resistant to etravirine. d Emergence of RAMs compared with baseline PS. e Emergence of additional RAMs compared with the emerging RAMs observed by PS and compared with baseline deep sequencing (DS). f Etravirine weighted genotypic score (WGS) by PS+DS is not shown at end point. g Y181C observed (1.7%) but not emerging, present by DS at baseline. FC, fold change; NA, not available; VFs, virological failures. were detected by PS. Only a small number of these VFs (3 of 15 [20.0%]) had minority etravirine RAMs at baseline. In 10/26 (38.5%) VFs with baseline and end point DS data, emergence of 1 additional minority etravirine RAM was detected (Tables 3 and 4). Some of the baseline minority etravirine RAMs detected in VFs were also observed at end point by PS or DS (V90I, L100I and Y181C). Y181C was the most frequent emerging etravirine RAM at end point by DS (n=4). For 1 patient, Y181C was observed by DS only at baseline as well as at end point (see Table 3). Phenotypic data were available for 37 (baseline) and 24 (end point) VFs. The proportion of VFs who had active etravirine (FC 3) at baseline was 33/37 (89.2%), whereas 4 (10.8%) and 0 had an etravirine FC 3-13 and >13, respectively. The corresponding proportions for non-vfs at baseline were 47/52 (90.4%), 4 of 52 (7.7%) and 1 of 52 (1.9%), respectively. At end point, the proportion of VFs with an etravirine FC 3 had decreased to 45.8% (11/24), and 33.3% (8 of 24) were resistant to etravirine (FC >13). In the 23 VFs with phenotypic and genotypic data at both baseline and end point, the median (range) etravirine FC was 0.9 ( ) at baseline and 3.6 (0.8->1,599) at end point. In 10 of 23 (43.5%) VFs with matched data, both genotypic (increase in etravirine WGS from 0-2 to 2.5 or to 4 by PS only) and phenotypic (increase to etravirine FC >3) development of etravirine resistance was observed, whereas this was not observed in 10 other VFs (43.5%). In 3 of 23 (13.0%) VFs only Antiviral Therapy

6 L Tambuyzer et al. Figure 1. Emerging NNRTI RAMs (ETR RAMs) by PS and DS NNRTI RAM V90I A98G L100I K101E K101P K101Q K103N K103S V106A V106I V108I E138A E138Q E138K E138G V179D V179E V179F V179I Y181C Y181I Y181V V189I G190A G190E H221Y F227L K238T Y318F Emerging NNRTI RAM observed at end point by PS (n=30) Minority NNRTI RAM observed at baseline by DS (n=29) Emerging minority NNRTI RAM observed at end point by DS (on top of PS; n=26) Number of VFs with emerging NNRTI RAMs (ETR RAMs) at end point by PS and DS Underline represents etravirine (ETR) resistance-associated mutations (RAMs). DS, deep sequencing; NNRTI, non-nucleoside reverse transcriptase inhibitor; PS, population sequencing; VF, virological failure. phenotypic resistance was seen that could be attributed to the single etravirine RAMs (K101E, E138G and Y181C) detected by DS. Discussion We studied the evolution of HIV-1 viral genotype and phenotype in treatment-experienced, HIV-1-infected children and adolescents virologically failing on a regimen containing etravirine in the PIANO study. Genotypic and phenotypic resistance to etravirine developed in 9 of 23 VFs with matched data. The most frequent emerging NNRTI RAMs observed by PS were Y181C, L100I, E138A and V90I. These are all etravirine RAMs which were defined based on the data from the DUET trials [18,19,24]. Overall, the etravirine resistance profile in paediatric patients was similar to that in adults. Management of HIV-1 infection in children and adolescents is challenging for many reasons [1,4,25]. A key challenge is the presence or development of RAMs which reduce further treatment options [1,3,4]. The NNRTIs efavirenz and nevirapine are used widely in treatment-naive, HIV-1-infected paediatric patients [26,27], despite their low genetic barrier to resistance. A single mutation can confer cross-resistance between these drugs [1,4,16,28]. In contrast, etravirine resistance requires multiple etravirine RAMs to confer resistance [16,24,29], making this NNRTI a valuable option for treatment-experienced patients [1,4,8,9]. The overall virological response rate to etravirine with an OBR in PIANO was similar to that achieved with other ARV combinations in treatment-experienced paediatric patients [1]. Causes of VF with ARVs are complex and multifactorial, for example, prior treatment failure, high baseline viral load, poor adherence, suboptimal exposure and the presence of baseline RAMs [1]. Low adherence to etravirine was reported in children and adolescents in PIANO [14]. VFs tended to have lower adherence to etravirine versus non-vfs; adherence in VFs and non- VFs was 53.7% versus 72.2%, respectively (adherence questionnaire; P=0.0836), and 25.7% versus 48.1% (pill count method) [14]. Although population pharmacokinetics indicated that the exposure to etravirine International Medical Press

7 Etravirine resistance in HIV-1-infected children and adolescents Table 4. Overview of VFs and responders without minority etravirine RAMs by DS at baseline End point Baseline and/or screening Emerging Etravirine Etravirine Etravirine Emerging Emerging etravirine RAMs: Etravirine Patient RAMs: PS RAMs: DS a WGS b : PS Etravirine FC c etravirine RAMs: PS d etravirine RAMs: DS e DS frequency, % WGS a : PS f Etravirine FC c VFs Children (n=9) Pt 14 A98G None E138A/Q V90I, E138K, Y181C 6.1, 2.4, Pt 15 None None None None Pt 16 None None None None Pt 17 V106I None None None Pt 18 None None Y181I V90I, Y181C 1.1, Pt 19 None None V90I, E138A, V179F, Y181C None 6.5 >1,598.8 Pt 20 Y181C None NA NA NA NA Pt 21 Y181C None NA NA NA NA Pt 22 None None NA NA NA NA Adolescents (n=23) Pt 23 None None None None Pt 24 K101E, G190A None None E138G Pt 25 None None E138Q, Y181C K101E, E138K, Y181I 2.2, 31.2, Pt 26 None None None None Pt 27 None None None None Pt 28 None None None Y181C Pt 29 None None Y181C None Pt 30 Y181C None V179F E138A Pt 31 None None E138A, Y181C/V V106I, E138K 14.9, Pt 32 K101E, G190S None K101P, Y181C None 6.5 NA Pt 33 None None L100I None 2.5 NA Pt 34 V106I None G190A None 2.5 NA Pt 35 V106I None None None 1.5 NA Pt 36 K101E/H, G190A None None Y181C Pt 37 None None None K101E Pt 38 L100I None NA NA NA NA Pt 39 K101E, V179T, G190A None NA NA NA NA Pt 40 A98G, V106I None 2.5 NA V90I, L100I, Y181C NA Pt 41 None None NA NA NA NA Pt 42 None None NA NA NA NA a Additional resistance-associated mutations (RAMs) compared with RAMs observed by population sequencing (PS). b Normal number denotes high response to etravirine; italic number denotes intermediate response to etravirine; bold number denotes reduced response to etravirine. c Normal number denotes sensitive to etravirine; italic number denotes intermediate sensitivity to etravirine; bold number denotes resistant to etravirine. d Emergence of RAMs compared with baseline PS. e Emergence of additional RAMs compared with the emerging RAMs observed by PS and compared with baseline deep sequencing (DS). f Etravirine weighted genotypic score (WGS) by PS+DS is not shown at end point. FC, fold change; NA, not available; VFs, virological failures. Antiviral Therapy

8 L Tambuyzer et al. Table 4. Continued End point Baseline and/or screening Emerging Etravirine Etravirine Etravirine Emerging Emerging etravirine RAMs: Etravirine Patient RAMs: PS RAMs: DS a WGS b : PS Etravirine FC c etravirine RAMs: PS d etravirine RAMs: DS e DS frequency, % WGS a : PS f Etravirine FC c Pt 43 A98G None Y181C NA Pt 44 None None None NA Pt 45 None None NA NA NA NA Responders Children (n=17) Pt 46 V106I None Pt 47 A98G None Pt 48 A98G None Pt 49 K101H, G190S None Pt 50 K101E, G190A None Pt 51 E138A None Pt 52 K101E, G190A None Pts None None Adolescents (n=16) Pt 63 A98G None Pt 64 E138Q None Pt 65 V106I None Pt 66 G190A None Pt 67 V90I, Y181C None Pts None None International Medical Press

9 Etravirine resistance in HIV-1-infected children and adolescents in paediatric patients was, overall, comparable to that in adults [14], the low adherence in paediatric patients, particularly in VFs compared with non-vfs, may explain some of the VFs in PIANO. A likely consequence of low adherence is the development of resistance to one or more of the drugs used, which is further explored for etravirine below. Over 48 weeks in PIANO, etravirine resistance increased in the VFs as shown by an increase in the number of etravirine RAMs and etravirine WGS from baseline. Keeping in mind the small sample sizes as a limitation of this statistical analysis, no significant differences were observed between adolescents and children with regard to previous NNRTI use and the presence or development of NNRTI RAMs detected by PS only, as well as when DS data were added to the analysis. Phenotypic resistance to etravirine (data available for 24 VFs at end point) also developed in PIANO as shown by the increase from baseline in the median (range) etravirine FC. Of note, of the 23 VFs with matched data, a similar percentage of patients (43.5%) did as did not develop etravirine resistance (increase in etravirine WGS as well as in etravirine FC), and 13% developed only phenotypic resistance which could be attributed to the single etravirine RAMs K101E, E138G and Y181C as detected by DS. Deep sequencing analyses of baseline samples from VFs and responders detected minority NNRTI RAMs in both groups, and the frequency of minority etravirine RAMs in VFs (20%) versus responders (13%) was not significantly different (P=0.5469). The high-impact etravirine RAMs L100I and Y181C (etravirine WGS 2.5) were only detected by DS in VFs and not in responders; however, PS also detected these etravirine RAMs in patients responding to the treatment. When accounting for the additional etravirine RAMs found as minority variants, the etravirine WGS changed from sensitive to intermediate or resistant for etravirine in three of eight VFs, but also in three of five responders. Therefore, it cannot be concluded that the presence of minority variants with etravirine RAMs would always result in VF. V179E, Y188L and F227L were the only other minority NNRTI RAMs detected by baseline DS in VFs (in one VF patient each) and not in responders. These three RAMs were observed together with at least one etravirine RAM. Generally, the etravirine RAMs detected by DS at baseline were similar to those observed by PS at baseline, and some of the baseline minority etravirine RAMs were also observed at end point either by PS or DS. Conversely, most of the etravirine RAMs that emerged at end point as observed by PS were not detected at baseline by DS. Finally, there is still uncertainty whether minority variants affect treatment outcome, and whether DS is useful in routine clinical testing [30 40], likely driven by differences in patient population and treatment regimens. The results from PIANO do not indicate that DS should be performed in addition to PS to guide the selection of etravirine for treatment. Further, the presence of minority variants was not consistently associated with treatment failure in this ARV-experienced population with the majority harbouring 1 or more NNRTI RAMs. Overall, the present etravirine resistance analyses of PIANO along with the efficacy data previously reported [14] confirm the value of etravirine as an NNRTI treatment option for HIV-1-infected children and adolescents aged 6-<18 years who used efavirenz and/or nevirapine previously [8,9]. Furthermore, the patterns of etravirine resistance observed in adults, adolescents and children experiencing VF are similar. Acknowledgements The authors are grateful to the patients and their families for their participation and support during the study; the PIANO Janssen study team; the study centre staff and principal investigators; and members of the Janssen etravirine team, in particular, Fien Blancke (Janssen Infectious Diseases, BVBA, Beerse, Belgium), Anick Vandingenen (Janssen Infectious Diseases, BVBA, Beerse, Belgium) and David Anderson (Janssen R&D, Titusville, NJ, USA) for their input and Elizabeth Van Rossem (Janssen Infectious Diseases, BVBA, Beerse, Belgium) for running the deep sequencing of the samples in the laboratory. This study was sponsored by Janssen Research and Development. Medical writing support was provided by Jackie Phillipson and assistance in incorporating author comments by Ian Woolveridge, both of Zoetic Science, an Ashfield company, Macclesfield, UK; this support was funded by Janssen. Data contained in this article were presented at the 14th European AIDS Conference, October 2013, Brussels, Belgium. Abstract PE9/23. All authors substantially contributed to the study s conception, design and performance, and all had significant involvement in the data analyses. All authors were involved in the development of the primary manuscript, interpretation of data, have read and approved the final version, and have met the criteria for authorship as established by the ICMJE. Disclosure statement All authors are Janssen employees and may be Johnson & Johnson stockholders. Additional file Additional file 1: Table of etravirine, NNRTI, NRTI and PI RAMs (by PS) at baseline can be found at Tambuyzer_Addfile1.pdf Antiviral Therapy

10 L Tambuyzer et al. References 1. Panel on Antiretroviral Therapy and Medical Management of HIV-Infected Children. Guidelines for the use of antiretroviral agents in pediatric HIV infection. Department of Health and Human Services. (Updated 5 March Accessed 2 July 2015.) Available from gov/guidelines/html/2/pediatric-arv-guidelines/0/ 2. Cortez KJ, Maldarelli F. Clinical management of HIV drug resistance. Viruses 2011; 3: Tang MW, Shafer RW. HIV-1 antiretroviral resistance: scientific principles and clinical applications. Drugs 2012; 72:e1 e Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Duman Services. (Updated 8 April Accessed 2 July 2015.) Available from gov/guidelines/html/1/adult-and-adolescent-arv-guidelines/0 5. de Mulder M, Yebra G, Navas A, et al. Trends in drug resistance prevalence in HIV-1-infected children in Madrid: 1993 to 2010 analysis. Pediatr Infect Dis J 2012; 31:e213 e Sigaloff KC, Calis JC, Geelen SP, van Vugt M, de Wit TF. HIV-1-resistance-associated mutations after failure of firstline antiretroviral treatment among children in resourcepoor regions: a systematic review. Lancet Infect Dis 2011; 11: Puthanakit T, Jourdain G, Hongsiriwon S, et al. HIV-1 drug resistance mutations in children after failure of firstline nonnucleoside reverse transcriptase inhibitor-based antiretroviral therapy. HIV Med 2010; 11: Etravirine EMA summary of product characteristics (Updated 22 May Accessed 2 July 2015.) Available from 9. Etravirine US prescribing information (Updated August Accessed 2 July 2015) Available from Katlama C, Haubrich R, Lalezari J, et al. Efficacy and safety of etravirine in treatment-experienced, HIV-1 patients: pooled 48 week analysis of two randomised, controlled trials. AIDS 2009; 23: Katlama C, Clotet B, Mills A, et al. Efficacy and safety of etravirine at week 96 in treatment-experienced HIV type-1- infected patients in the DUET-1 and DUET-2 trials. Antivir Ther 2010; 15: Lazzarin A, Campbell T, Clotet B, et al. Efficacy and safety of TMC125 (etravirine) in treatment-experienced HIV- 1-infected patients in DUET-2: 24 week results from a randomised, double-blind, placebo-controlled trial. Lancet 2007; 370: Madruga JV, Cahn P, Grinsztejn B, et al. Efficacy and safety of TMC125 (etravirine) in treatment-experienced HIV-1-infected patients in DUET-1: 24 week results from a randomised, double-blind, placebo-controlled trial. Lancet 2007; 370: Tudor-Williams G, Cahn P, Chokephaibulkit K, et al. Safety, efficacy and resistance analysis of etravirine in treatmentexperienced, HIV-1-infected children and adolescents: 48-week results from the phase II PIANO study. HIV Med 2014; 15: Königs C, Feiterna-Sperling C, Esposito S, et al. Pharmacokinetics and short-term safety and tolerability of etravirine in treatment-experienced HIV-1-infected children and adolescents. AIDS 2012; 26: Johnson VA, Clavez V, Günthard HF, et al update of the drug resistance mutations in HIV-1. Top Antivir Med 2011; 19: Tambuyzer L, Azijn H, Rimsky LT, et al. Compilation and prevalence of mutations associated with resistance to nonnucleoside reverse transcriptase inhibitors. Antivir Ther 2009; 14: Tambuyzer L, Nijs S, Daems B, Picchio G, Vingerhoets J. Effect of mutations at position E138 on HIV-1 reverse transcriptase on phenotypic susceptibility and virologic response to etravirine. J Acquir Immune Defic Syndr 2011; 58: Vingerhoets J, Tambuyzer L, Azijn H, et al. Resistance profile of etravirine: combined analysis of baseline genotypic and phenotypic data from the randomized, controlled Phase III clinical studies. AIDS 2010; 24: Vingerhoets J, Nijs S, Tambuyzer L, Hoogstoel A, Anderson D, Picchio G. Similar predictions of etravirine sensitivity regardless of genotypic testing method used: comparison of available scoring systems. Antivir Ther 2012; 17: Thys K, Van Hove C, Van Rossem E, et al. Assessing the use of Illumina technology for deep sequencing in virology application. Antivir Ther 2012; 17 Suppl 1:A Bennett S. Solexa Ltd. Pharmacogenomics 2004; 5: Tambuyzer L, Nijs S, Tomaka F, Picchio G, Vingerhoets J. Baseline and emerging resistance to etravirine in HIV-1- infected children and adolescents: final results from the Phase II PIANO study. Antivir Ther 2012; 17 Suppl 1:A Tambuyzer L, Vingerhoets J, Azijn H, et al. Characterization of genotypic and phenotypic changes in HIV-1-infected patients with virologic failure on an etravirine-containing regimen in the DUET-1 and DUET-2 clinical studies. AIDS Res Hum Retroviruses 2010; 26: Phelps BR, Rakhmanina N. Antiretroviral drugs in pediatric HIV-infected patients: pharmacokinetic and practical challenges. Paediatr Drugs 2011; 13: Funk MB, Notheis G, Schuster T, et al. Effect of first line therapy including efavirenz and two nucleoside reverse transcriptase inhibitors in HIV-infected children. Eur J Med Res 2005; 10: Luzuriaga K, McManus M, Mofenson L, et al. A trial of three antiretroviral regimens in HIV-1-infected children. N Engl J Med 2004; 350: Delaugerre C, Rohban R, Simon A, et al. Resistance profile and cross-resistance of HIV-1 among patients failing a non-nucleoside reverse transcriptase inhibitor-containing regimen. J Med Virol 2001; 65: Vingerhoets J, Azijn H, Fransen E, et al. TMC125 displays a high genetic barrier to the development of resistance: evidence from in vitro selection experiments. J Virol 2005; 79: Johnson JA, Li JF, Wei X, et al. Minority HIV-1 drug resistance mutations are present in antiretroviral treatmentnaive populations and associate with reduced treatment efficacy. PLoS Med 2008; 5:e Paredes R, Lalama CM, Ribaudo HJ, et al. Pre-existing minority drug-resistant HIV-1 variants, adherence, and risk of antiretroviral treatment failure. J Infect Dis 2010; 201: Rowley CF, Boutwell CL, Lee EJ, et al. Ultrasensitive detection of minor drug-resistant variants for HIV after nevirapine exposure using allele-specific PCR: clinical significance. AIDS Res Hum Retroviruses 2010; 26: Li JZ, Paredes R, Ribaudo HJ, et al. Low-frequency HIV-1 drug resistance mutations and risk of NNRTI-based antiretroviral treatment failure: a systematic review and pooled analysis. JAMA 2011; 305: Goodman DD, Zhou Y, Margot NA, et al. Low level of the K103N HIV-1 above a threshold is associated with virological failure in treatment-naive individuals undergoing efavirenz-containing therapy. AIDS 2011; 25: Gianella S, Delport W, Pacold ME, et al. Detection of minority resistance during early HIV-1 infection: natural variation and spurious detection rather than transmission and evolution of multiple viral variants. J Virol 2011; 85: Metzner KJ, Rauch P, Braun P, et al. Prevalence of key resistance mutations K65R, K103N, and M184V as minority HIV-1 variants in chronically HIV-1 infected, treatment-naive patients. J Clin Virol 2011; 50: International Medical Press

11 Etravirine resistance in HIV-1-infected children and adolescents 37. Li JZ, Paredes R, Ribaudo HJ, et al. Relationship between minority nonnucleoside reverse transcriptase inhibitor resistance mutations, adherence, and the risk of virologic failure. AIDS 2012; 26: Messiaen P, Verhofstede C, Vandenbroucke I, et al. Ultradeep sequencing of HIV-1 reverse transcriptase before start of an NNRTI-based regimen in treatment-naive patients. Virology 2012; 426: Lehman DA, Wamalwa DC, McCoy CO, et al. Lowfrequency nevirapine resistance at multiple sites may predict treatment failure in infants on nevirapine-based treatment. J Acquir Immune Defic Syndr 2012; 60: Snedecor SJ, Khachatryan A, Nedrow K, et al. The prevalence of transmitted resistance to first-generation non-nucleoside reverse transcriptase inhibitors and its potential economic impact in HIV-infected patients. PLoS ONE 2013; 8:e Accepted 1 November 2015; published online 13 November 2015 Antiviral Therapy