Outcomes in older versus younger patients over 96 weeks in HIV-1- infected patients treated with rilpivirine or efavirenz in ECHO and THRIVE.

OBJECTIVES: Increasing life expectancy of HIV-1-infected patients raises interest in how trial results apply to older patients. This post-hoc analysis evaluated potential differences in efficacy and safety in older (≥50 years) versus younger (<50 years) patients in the ECHO and THRIVE trials over 96 weeks.
METHODS: HIV-infected, treatment-naïve adults were randomized to receive rilpivirine (RPV) or efavirenz (EFV), plus a background regimen. Virologic response rates (FDA snapshot analysis; HIV-1 RNA <50 copies/mL) were assessed at Week 96. Total-body bone mineral density was evaluated at baseline and Week 96 by dual-energy X-ray absorptiometry scans. Serum concentrations of 25-hydroxy vitamin D (ECHO trial only) were also measured at baseline, Week 24 and Week 48.
RESULTS: 1368 patients were treated. At Week 96, virologic response rates were similar between older (77%) and younger (76%) RPV-treated patients and numerically higher in older (84%) versus younger (76%) EFV-treated patients. No clinically relevant age-related differences were observed in immunologic responses. Small differences were noted in older versus younger patients in adverse events (higher rates of depression, insomnia, and rash in older EFV-treated patients), laboratory abnormalities (increased low-density lipoprotein cholesterol and hyperglycemia in older EFV-treated patients and increased amylase in older patients across treatments), bone mineral density (larger decreases in older patients across treatments), and progression to severe vitamin D deficiency (greater in older versus younger EFV-treated patients).
CONCLUSION: Efficacy and safety outcomes were generally similar in older versus younger patients in the ECHO and THRIVE trials.

RCT Entities:   Population Interventions Outcomes

INTRODUCTION

As the life expectancy of HIV-1–infected patients increases, and HIV-1 prevalence among older individuals rises, the applicability of overall trial results to older patients with HIV-1 becomes more relevant for clinical practice [1]. It is estimated that by 2015, at least 50% of HIV-infected individuals in the United States will be ≥50 years of age [2]. Older patients differ from younger patients in response to therapy, rate of HIV disease progression and level of complications caused by comorbidities [2]. Combination antiretroviral therapy (ART) provides many benefits, but older patients are more susceptible to adverse events (AEs), use more concomitant medications, are at greater risk of drug-drug interactions, and have lower rates of immune recovery and higher mortality rates. Despite these trends, older patients have higher virologic response rates than younger patients [3-6].

ECHO (TMC278-C209, NCT00540449) and THRIVE (TMC278-C215, NCT00543725) were two global, randomized, double-blind, double-dummy, phase III trials in treatment-naïve, HIV-1–infected adults [7, 8]. In the Week 48 primary analyses, the non-nucleoside reverse transcriptase inhibitor (NNRTI) rilpivirine (RPV) 25 mg once daily (qd) demonstrated non-inferior efficacy compared with the NNRTI efavirenz (EFV) 600 mg qd (primary objective) in both trials [7, 8]. In the pooled 96-week analysis, RPV demonstrated sustained overall efficacy that was similar to EFV, with better tolerability [9].

We performed a post-hoc analysis to evaluate potential differences in efficacy and safety outcome by age at Week 96 in patients enrolled in ECHO and THRIVE. Treatment groups were divided into patients <50 years of age (“younger”) and patients ≥50 years of age (“older”).

MATERIALS AND METHODS

Study Design

The ECHO and THRIVE trials have been described previously [7, 8]. Both were conducted in accordance with the Declaration of Helsinki. The primary objective was to demonstrate non-inferiority (12% margin) of RPV to EFV in confirmed response (HIV-1 RNA <50 copies/mL, intent-to-treat time-to-loss-of-virologic-response [ITT-OVR] algo-rithm) at week 48. Secondary objectives included efficacy and safety/tolerability over 96 weeks. Main inclusion criteria were HIV-1 RNA ≥5000 copies/mL, no NNRTI resistance-associated mutations (RAMs; from a list of 39 NNRTI RAMs) and phenotypic sensitivity to the nucleoside/ nucleotide reverse transcriptase inhibitors (determined using virco®TYPE) [10]. HIV-infected, treatment-naïve adults aged ≥18 years were randomized to receive RPV 25 mg qd or EFV 600 mg qd (1:1), plus either tenofovir disoproxil fumarate (TDF)/emtricitabine (FTC) in ECHO or TDF/FTC, zidovudine/lamivudine (3TC) or abacavir/3TC in THRIVE.

Study Assessments

In a pooled week 96 analysis of ECHO and THRIVE, efficacy and safety outcomes were evaluated for each age group within each treatment arm; post-hoc age subgroup cutoffs were <50 years (“younger”) and ≥50 years (“older”) at baseline. The age cutoff of 50 was selected both for its common use in HIV studies and because cutoffs at higher ages resulted in n’s that were too small to make meaningful comparisons. Total body bone mineral density (BMD) was evaluated by dual-energy X-ray absorptiometry (DEXA) scans at baseline and week 96. Serum concentration measurements of 25-hydroxy vitamin D at baseline, Week 24 and week 48 (collected only in the ECHO trial to Week 48) were also assessed. Variables were summarized using descriptive statistics. Statistical analyses were performed in a pooled week 96 post-hoc analysis for response rates (ITT-TLOVR, HIV-1 RNA <50 copies/mL) and rates of AEs (grade ≥2 at least possibly related, grade ≥3 all causality, and serious AEs [SAEs]) using the Mantel-Haenszel chi-square test. Change in CD4+ cell count from baseline to week 96 was assessed using nonparametric statistics (Wilcoxon rank sum test).

RESULTS

A total of 1368 patients were treated (1242 younger; 126 older). Demographics and baseline disease characteristics were generally similar between age and treatment groups (Table ). Most patients were male (76% younger; 70% older) and white (60% younger; 64% older). Hepatitis C virus (HCV) co-infection was noted in 8% of younger patients and 11% of older patients. Mean baseline viral load was 5.1log10 copies/mL in older patients and 4.9 log10 copies/mL in younger patients, and median CD4+ cell count was 233 cells/mm3 in older patients and 258 cells/mm3 in younger patients.

Disposition was generally similar between younger and older patients for both the RPV and EFV treatment arms (Table ). In the EFV treatment arm, 88% of older patients completed treatment through week 96, compared with 79% of younger patients. In the RPV treatment arm, the same proportion (81%) of older and younger patients completed week 96. Overall, a higher number of younger patients discontinued due to “other” reasons (9% younger; 2% older); “other” reasons included patient lost to follow-up, withdrew consent, noncompliant, ineligible to continue trial, and sponsor’s decision.

Week 96 virologic response rates (FDA snapshot analysis; HIV-1 RNA <50 copies/mL) were similar in older (77%) and younger (76%) patients treated with RPV (Table ). Numerically higher response rates were observed in older (84%) patients compared with younger (76%) patients in the EFV arm, mainly explained by a higher attrition rate due to “other reasons” rather than efficacy or safety/tolerability reasons in the younger group. Week 96 virologic response rates determined by TLOVR (HIV-1 RNA <50 copies/mL) were 75% in older patients and 78% in younger patients treated with RPV and 82% in older patients and 77% in younger patients treated with EFV (Table ). When evaluating the EFV and RPV arms combined, no statistically significant difference in week 96 virologic response rates was noted in older (79%) versus younger (78%) patients (p=0.775). Median changes in CD4+ cell counts at Week 96 (observed case, no imputation) were similar between older and younger patients for RPV (older, 237 versus younger, 263 cells/mm3) and for EFV (older, 278 versus younger, 251 cells/mm3; Table ). Almost all of the 1098 older and younger patients with both baseline and week 96 values available saw an increase in CD4+ cell count at Week 96. Among patients in the RPV arm with <200 cells/mm3 at baseline, 85% of older patients and 92% of younger patients experienced an increase in CD4+ cell counts to ≥200 cells/mm3 at week 96; in the EFV arm improvements were 100% and 91%, respectively. Among patients in the RPV arm with 200-349 cells/mm3 at baseline, 90% of older patients and 94% of younger patients experienced an increase in CD4+ cell counts to ≥350 cells/mm3 at week 96; in the EFV arm improvements were 91% and 96%, respectively. When evaluating the EFV and RPV arms combined, no statistically significant difference in week 96 CD4+ cell count change from baseline was noted in older versus younger patients (p=0.303).

Rates of grade ≥2 treatment-related AEs, grade ≥3 AEs regardless of relationship and SAEs were numerically higher in older versus younger patients for both treatment arms, except for grade ≥2 treatment-related AEs, for which rates were higher in younger RPV-treated patients compared with older RPV-treated patients (Table ). However, the most frequent (≥5% of patients) grade ≥2 treatment-related AEs had similar rates between younger and older patients in the RPV treatment arm (Table ). Higher rates of depression (7%, n=4), insomnia (7%, n=4), and rash (9%, n=5) were observed in older EFV-treated patients compared with younger EFV-treated patients (1% [n=8], 2% [n=13], and 5% [n=31], respectively; Table ). However, because of the small number of patients in this subgroup, these data should be interpreted with caution. When evaluating the EFV and RPV arms combined, no statistically significant difference in rates of grade ≥2 treatment-related AEs from baseline was noted in older (25%) versus younger (25%) patients (p=0.914); no statistically significant difference in rates of SAEs (13% older, 10% younger; p=0.162) or grade ≥3 AEs regardless of relationship was observed (24% older, 18% younger; p=0.143).

No clinically relevant differences in grade ≥3 metabolic laboratory abnormalities occurred between age groups, except abnormal low-density lipoprotein cholesterol (LDL-C) levels in 18% (n=10) of older compared with 5% (n=30) of younger EFV-treated patients (Table ) and hypergly-cemia, observed in 4% (n=2) of older compared with 0.3% (n=2) of younger EFV-treated patients. Thirteen patients (1%) entered the trials on lipid-lowering medication; 28 patients initiated lipid-lowering agents during treatment (23 in EFV arm), and no differences were observed in younger (2%) versus older (2%) patients overall with regard to initiation of lipid therapy. Grade ≥3 amylase levels were observed in older (12%) compared with younger patients overall (4%; Table ). Reported non-metabolic laboratory abnormalities (grade ≥3) were similar between age and treatment groups, including levels of evaluated hepatic transaminases, electrolytes and hematologic parameters (Table ). It should be noted that the number of patients who experienced grade ≥3 metabolic or non-metabolic laboratory abnormalities was low overall, so comparisons between groups should be made with caution.

Estimated glomerular filtration rate (eGFR), based on serum creatinine, decreased from baseline to Week 96 regardless of age or treatment group; the magnitude of change was similar between age categories within the two treatment groups. The median change from baseline at Week 96 for the RPV treatment arm was –13.6 mL/min/1.73 m2 in younger and –11.0 mL/min/1.73 m2 in older patients, and for the EFV treatment arm, –3.9 mL/min/1.73 m2 in younger and –9.2 mL/min/1.73 m2 in older patients. The majority of eGFR changes in patients taking RPV occurred during the first 2 weeks of exposure and remained fairly stable thereafter. Bone mineral density, as measured by DEXA, decreased from baseline to week 96 regardless of age or treatment group. Larger median decreases were observed in older (–0.035 g/cm2, RPV arm; –0.049 g/cm2, EFV arm) compared with younger patients (–0.014 g/cm2, RPV arm; –0.014 g/cm2, EFV arm).

Median 25-hydroxy vitamin D changes from baseline were greater in older (–3.2 ng/mL; n=29) versus younger (–1.6 ng/mL; n=261) patients in the EFV arm. Levels remained relatively unchanged for both older (0.8 ng/mL; n=31) and younger (–0.8 ng/mL; n=261) patients in the RPV arm. Risk of progression from insufficient (50-74 nmol/L) or deficient (25-49 nmol/L) to severely deficient (<25 nmol/L) 25-hydroxy vitamin D levels from baseline to Week 48 occurred in 0% (n=0) of older and 2% (n=4) of younger RPV-treated patients. Among EFV-treated patients, risk of progression was 13% (n=2) for older and 8% (n=13) for younger patients.

DISCUSSION

Several challenges exist for the aging HIV-1–infected population, including age-related physiologic changes in renal and hepatic function and attenuated immune competence, independent of suppression of HIV replication. Comorbid conditions requiring drug treatments that may complicate ART also need to be considered [11]. Many HIV clinical trials that examine the effect of novel antiretrovirals (ARVs) exclude or underrepresent older patients, as well as those with comorbid conditions, and many do not compare outcomes in older versus younger patients [12, 13].

This post-hoc analysis of the pooled ECHO and THRIVE trials showed that when treatment arms were combined, older (aged ≥50 years) patients generally responded to treatment as well as younger (aged <50 years) patients, both immunologically and virologically. There were slightly increased response rates among older versus younger patients treated with EFV. No evaluation of differences by NRTI backbone could be made as 80% of patients were receiving TDF.

Older age is associated with reduced hepatic and renal function, potentially affecting the ability to metabolize drugs [14]. This may result in elevated serum concentrations of ARVs, and associated drug toxicity. Comorbidities requiring drug treatment also complicate ART; drug-drug interactions potentially alter tolerability or efficacy [2]. Similarly, older patients are more likely to change therapy due to AEs [15]. In this study, rates of depression, insomnia and rash were slightly higher in older versus younger EFV-treated patients. Generally, older patients did not experience higher levels of neuropsychiatric, hepatic or renal treatment-related AEs compared with younger patients. Rates of grade ≥3 metabolic laboratory abnormalities were similar in older and younger patients in both RPV and EFV treatment arms, except for LDL-C and glucose, which were greater in older EFV-treated patients. The need to introduce lipid-lowering therapy was not greater in older versus younger patients; generally, few patients started such therapy, regardless of age. Because of the smaller number of patients aged ≥50 and the even smaller number of patients who experienced AEs in our study, comparisons between age and treatment group should be interpreted with caution.

Decreases in eGFR from baseline to Week 96 were modest and did not differ considerably between age groups. eGFR changes during the first 2 weeks of RPV treatment were likely related to inhibition of tubular secretion of creatinine. Increases in eGFR, estimated using cystatin C (indicator of renal filtration), following treatment with RPV (THRIVE trial) suggest that RPV was not associated with decreases in glomerular function [7].

In this analysis, BMD decreased from baseline in both age groups with the largest median decrease found in older patients. Cigarette use (self-reported), known to worsen BMD, was infrequent. Decreased BMD is a feature of osteoporosis, which is common in the older HIV-infected population. HIV disease-related chronic immune activation and treatment with ARV regimens reduce BMD, therefore increasing the risk of osteoporosis and subsequent risk of fractures [16-19].

Vitamin D insufficiency and deficiency are common among HIV-infected individuals, and EFV treatment is associated with reduction in 25-hydroxy vitamin D levels [20]. Vitamin D analysis at Week 48 (ECHO trial) demonstrated that older RPV-treated patients showed no increased risk of lower 25-hydroxy vitamin D levels or progression to severe deficiency compared with younger RPV-treated patients. Older EFV-treated patients had greater decreases in 25-hydroxy vitamin D levels than younger EFV-treated patients and greater risk of progression to severe vitamin D deficiency.

Older HIV-1–infected patients face additional challenges compared with younger patients. Thus, patient age needs to be considered when choosing an optimal treatment plan. This substudy is informative in that similar data on age-related HIV drug treatment outcomes are not well represented in published literature. Limitations of this analysis are the small number of older compared with younger patients and its post-hoc nature; an age-matched HIV-negative group for comparison was not included in ECHO/THRIVE; therefore, an evaluation of whether changes observed between younger and older patients were simply age-related is not possible.

In this analysis, safety outcomes were similar between older and younger patients. Although rates of grade ≥3 AEs and SAEs were higher in older patients, discontinuations due to AEs were low and similar between older and younger patients. Changes in eGFR and rates of laboratory abnormalities were similar between age groups with few exceptions. Higher rates of LDL-C and glucose elevations in older patients were driven by higher rates of these abnormalities in older EFV-treated patients, and amylase elevations were noted in older patients in both treatment arms. Furthermore, differences in bone metabolism markers between older and younger patients warrant further investigation.

In summary, older and younger patients had similar efficacy outcomes in both RPV and EFV treatment arms, with some small differences noted in safety/tolerability. On the basis of these data, there were no clinically relevant differences in outcomes between older and younger patients in the phase III ECHO and THRIVE trials.

Table 1.

Patient Baseline Demographics and Disease Characteristics

	Treatment Arm, Age Group, Years				Both Treatment Arms, Age Group, Years
	RPV		EFV
	<50 N=617	≥50 N=69	<50 N=625	≥50 N=57	<50 N=1242	≥50 N=126
Median age, years (range)	35 (18, 49)	54 (50, 78)	34 (19, 49)	54 (50, 69)	35 (18, 49)	54 (50, 78)
Sex, n (%)
Female	148 (24)	20 (29)	145 (23)	18 (32)	293 (24)	38 (30)
Male	469 (76)	49 (71)	480 (77)	39 (68)	949 (76)	88 (70)
Race, n (%)
White	375 (61)	45 (65)	375 (60)	35 (61)	750 (60)	80 (64)
Black or African American	145 (24)	20 (29)	140 (22)	16 (28)	285 (23)	36 (29)
Asian	74 (12)	4 (6)	94 (15)	3 (5)	168 (14)	7 (6)
Othera	7 (1)	0	4 (1)	3 (5)	11 (1)	3 (2)
Smoker	227 (37)	19 (28)	210 (34)	13 (23)	437 (35)	32 (25)
HCV,b n (%)	44 (7)	9 (13)	50 (8)	5 (9)	94 (8)	14 (11)
Clinical stage of HIV, n (%)
A	442 (72)	43 (62)	436 (70)	38 (67)	878 (71)	81 (64)
B	146 (24)	20 (29)	154 (25)	15 (26)	300 (24)	35 (28)
C	29 (5)	6 (9)	35 (6)	4 (7)	64 (5)	10 (8)
Median (range) duration of HIV disease at screening, years	1 (0, 23)	2 (0, 24)	1 (0, 25)	1 (0, 28)	1 (0, 25)	2 (0, 28)
N(t)RTI use, n (%)	617 (100)	69 (100)	625 (100)	57 (100)	1242 (100)	126 (100)
Abacavir sulfate with lamivudine	38 (6.2)	6 (8.7)	34 (5.4)	2 (3.5)	72 (6)	8 (6)
Emtricitabine and tenofovir disoproxil fumarate	493 (79.9)	59 (85.5)	498 (79.7)	49 (86.0)	991 (80)	108 (86)
Zidovudine with lamivudine	101 (16.4)	6 (8.7)	98 (15.7)	6 (10.5)	199 (16)	12 (10)
Median (range) CD4+ cell count at baseline, cells/mm3	253c (1, 888)	220 (5, 879)	261 (1, 919)	257 (2, 1137)	258d (1, 919)	233 (2, 1137)
Mean (SD) HIV-1 RNA at baseline, log10 copies/mL	4.9 (0.62)	5.0 (0.64)	4.9 (0.63)	5.1 (0.56)	4.9 (0.63)	5.1 (0.61)

RPV, rilpivirine; EFV, efavirenz; HCV, hepatitis C virus; N(t)RTI, nucleoside/nucleotide reverse transcriptase inhibitor; SD, standard deviation.

aIncludes “not allowed to ask”; bHCV antibody positive or reactive; cN=616; dN=1241.

Table 2.

Patient Disposition and Week 96 Virologic Response Rates

	Treatment Arm, Age Group, Years				Both Treatment Arms,  Age Group, Years
	RPV		EFV
	<50 N=617	≥50 N=69	<50 N=625	≥50 N=57	<50 N=1242	≥50 N=126
Disposition, n (%)
Completed Week 96	498 (81)	56 (81)	495 (79)	50 (88)	993 (80)	106 (84)
Discontinuation due to adverse events	24 (4)	4 (6)	53 (9)	6 (11)	77 (6)	10 (8)
Discontinuation due to virologic failure	46 (8)	7 (10)	19 (3)	0 (0)	65 (5)	7 (6)
Discontinuation due to other reasonsa	49 (8)	2 (3)	58 (11)	1 (2)	107 (9)	3 (2)
Efficacy Parameters
HIV-1 RNA <50 copies/mL, FDA snapshot analysis, n (%)	471 (76)	53 (77)	474 (76)	48 (84)	945 (76)	101 (80)
HIV-1 RNA <50 copies/mL, TLOVR, n (%)	480 (78)	52 (75)	482 (77)	47 (82)	962 (77)	99 (79)
Median (range) change in CD4+ cell count from baseline, cells/mm3	263b (–108, 763)	237c (37, 815)	251d (–126, 1216)	278e (–328, 763)	257f (–126,1216)	246g (–328, 815)

RPV, rilpivirine; EFV, efavirenz; FDA, Food and Drug Administration; TLOVR, time to loss of virologic response.

aOther reasons included lost to follow-up, withdrew consent, noncompliant, ineligible to continue trial, and sponsor’s decision; bN=497; cN=58;dN=494; eN=49; fN=991;gN=107.

Table 3.

Adverse Events, Serious Adverse Events and Grade ≥3 Laboratory Abnormalities

n (%)	Treatment Arm, Age Group, Years					Both Treatment Arms, Age Group, Years
	RPV		EFV
	<50 N=617a	≥50 N=69	<50 N=625b	≥50 N=57c		<50 N=1242d	≥50 N=126e
Grade ≥2 treatment-relatedf AEs	107 (17)	9 (13)	204 (33)	22 (39)		311 (25)	31 (25)
Grade ≥3 AEs regardless of relationship	104 (17)	16 (23)	125 (20)	14 (25)		229 (18)	30 (24)
SAEs	55 (9)	10 (15)	64 (10)	7 (12)		119 (10)	17 (13)
Grade ≥2 Treatment-Relatedf AEs Occurring in ≥5% of Patients
Nervous system disorders	22 (4)	1 (1)	64 (10)	5 (9)		86 (7)	6 (5)
Dizziness	4 (1)	0 (0)	42 (7)	2 (4)		46 (4)	2 (2)
Psychiatric disorders	39 (6)	2 (3)	57 (9)	8 (14)		96 (8)	10 (8)
Depression	10 (2)	1 (1)	8 (1)	4 (7)		18 (2)	5 (4)
Insomnia	12 (2)	1 (1)	13 (2)	4 (7)		25 (2)	5 (4)
Skin and subcutaneous tissue disorders	10 (2)	1 (1)	51 (8)	9 (16)		61 (5)	10 (8)
Rash	3 (1)	1 (1)	31 (5)	5 (9)		34 (3)	6 (5)
Grade ≥3 Metabolic Laboratory Abnormalities Regardless of Relationship, Occurring in Any Patient
Lipids (fasted)
Total cholesterol	1 (0.2)	0 (0)	19 (3)	3 (5)		20 (2)	3 (2)
Triglycerides	3 (1)	1 (1)	22 (4)	1 (2)		25 (2)	2 (2)
LDL-C	10 (2)	1 (1)	30 (5)	10 (18)		40 (3)	11 (9)
Glucose (fasted)
Glucose increased	2 (0.3)	1 (1)	2 (0.3)	2 (4)		4 (3)	3 (2)
Grade ≥3 Non-Metabolic Laboratory Abnormalities Regardless of Relationship, Occurring in Any Patient
Amylase	22 (4)	8 (12)	30 (5)	7 (13)		52 (4)	15 (12)
Lipase	4 (1)	2 (3)	8 (1)	2 (4)		12 (1)	4 (3)
Phosphorous	7 (1)	0 (0)	7 (1)	2 (4)		14 (1)	2 (2)
AST	13 (2)	2 (3)	23 (4)	1 (2)		36 (3)	3 (2)
ALT	9 (2)	1 (1)	25 (4)	2 (4)		34 (3)	3 (2)
White blood cell count decreased	7 (1)	1 (1)	5 (1)	0 (0)		12 (1)	1 (1)

RPV, rilpivirine; EFV, efavirenz; AE, adverse event; SAE, serious adverse event; LDL-C, low-density lipoprotein cholesterol; AST, aspartate aminotransferase; ALT, alanine aminotransferase.

The number of observations for each parameter does not always equal the total number of patients in each subgroup due to unavailable test results for a small number of patients:

aN=616 for metabolic laboratory abnormalities; bN=613 for total cholesterol and triglycerides, N=611 for LDL-derived, N=612 for glucose, N=617 for phosphorous and AST, N=623 for white blood cell count; cN=56 for metabolic laboratory abnormalities; dN=1241 for metabolic laboratory abnormalities, N=1230 for total cholesterol and triglycerides, N=1228 for LDL-derived, N=1229 for glucose, N=1234 for phosphorous and AST, N=1240 for white blood cell count; eN=125 for metabolic laboratory abnormalities; fDeemed at least possibly related by investigator.