Treatment-Naive

The manufacturer conducted a cost-utility analysis comparing DTG with commonly used regimens (Atripla, raltegravir [RAL], darunavir plus ritonavir [DRV/r], and atazanavir plus ritonavir [ATZ/r]) or alternative regimens (Complera, Stribild, lopinavir plus ritonavir [LPV/r]).² The comparators for the evaluation of DTG with efficacy derived from head-to-head trials are: Atripla (SINGLE study), RAL (SPRING-2), and DRV/r (FLAMINGO). Both SINGLE and SPRING-2 were phase III, randomized, double-blinded studies, while FLAMINGO was a phase IIIb, randomized, open-label study. The comparators with efficacy and safety derived from the network meta-analysis (NMA) are: Complera, Stribild, ATZ/r, and LPV/r. Primary efficacy outcome was less than 50 copies/mL of HIV ribonucleic acid (RNA) at week 48. Viral suppression rate, monthly CD4+ cell count and monthly rate of viral rebound (late-failure rate) from the trial were also used to establish efficacy. The time horizon for the reference case was lifetime with monthly cycle, and used the Canadian public-payer perspective.

A micro-simulation approach was used in the economic model where patients passed through the model one at a time. This allowed individual histories of accumulating events to influence the probability of disease progression, including factors such as CD4+ cell count, viral load, opportunistic infections (OI) prophylaxis status, age, gender, and Framingham score. Simulated patients transition through mutually exclusive health states defined in terms of HIV (with or without [OIs]), combined with cardiovascular disease (CVD) health state. As patients pass through the model, they experienced the natural progression of HIV infection. Up to six successive antiretroviral therapies (ARTs) were followed in the model and these depended on treatment history and resistance status. The rates of adverse events (AE) were derived from the SINGLE, SPRING-2, and FLAMINGO trials.^3–5 Patients could switch treatment after an acute AE, or when a treatment was failing. The development of resistance following virologic failure after initial treatment was also derived from SINGLE, SPRING-2, and FLAMINGO trials.

HIV utilities were derived from a Canadian study that examined the relationship between HUI3-derived health preference score and HIV health status measured by CD4+ cell count.⁶ Utility decrements associated with the CVD were derived from a US study.⁷ The costs for ART and OI prophylaxis treatment were obtained from the RAMQ List of Medications. Health care resource utilization costs (costs of HIV, OI, CVD, and death) were based on Canadian studies.^8–11

Treatment-Experienced

The manufacturer conducted a cost-utility analysis comparing DTG relative to RAL with optimized background therapy (OBT) in integrase inhibitor(INI)–naive TE patients.² Efficacy data were obtained from the SAILING clinical trial, which was a phase III, randomized, double-blinded, active-controlled, non-inferiority study.¹² DRV/r + tenofovir was assumed to be OBT based on the SAILING baseline population. Viral suppression rate, monthly CD4+ cell count, and monthly rate of viral rebound (late-failure rate) from the trial were used to establish efficacy. The reference-case time horizon was lifetime with monthly cycle, and used the Canadian public-payer perspective.

The same micro-simulation approach was used in the economic model where patients passed through the model one at a time with individual characteristics and accumulating events, and same health states as the treatment-naive model. Up to four successive ART regimens were followed in the model and these depended on treatment history and INI-resistance status. Patients could switch treatment after an acute AE, or when a treatment was failing. AEs were considered only through treatment discontinuation due to AEs, as no difference was observed between RAL and DTG in the SAILING trial. The prevalence of resistance was based on data from the SAILING trial. NRTI and non-nucleoside reverse transcriptase inhibitors (NNRTI) plus NRTI resistance were considered at the beginning of the model, and the potential development of INI resistance was considered after failure on the first treatment.

HIV utilities were derived from the same Canadian study that examined the relationship between HUI3-derived health preference score and HIV health status measured by CD4+ cell count.⁷ Utility decrements associated with CVD were derived from a US study.¹³ The costs for ART and OI prophylaxis treatment were obtained from the RAMQ List of Medications. Health care resource utilization costs (costs of HIV, OI, CVD, and death) were based on Canadian studies.^8,14–16

Treatment-Naive

The manufacturer reported DTG being the dominant strategy (less cost and more effective) when compared with Atripla, RAL, DRV/r, and other indirect comparators (Complera, Stribild, ATZ/r, LPV/r).

Treatment-Experienced

The manufacturer reported DTG as the dominant strategy when compared with RAL.

Use of Surrogate End Points

The model uses surrogate end points of viral suppression and CD4+ cell count to predict clinical outcomes, including OI, resistance, and mortality; however, these are well-accepted indicators of clinical outcomes and an accepted standard to adjudicate relative efficacy.

Magnitude of Benefit

The incremental QALYs with DTG compared with other relevant treatment strategies are small: an additional 15 to 48 days of perfect life for treatment-naive and an additional 81 days of perfect life for treatment-experienced; however, this does not alter conclusions, as DTG is less costly in most analyses.

Uncertainty in Relative Efficacy

While RCTs of some relevant comparators are available, the FLAMINGO trial is an open-label study, and there are no head-to-head trials for DTG versus Complera, Stribild, ATZ/r and LPV/r. An NMA was conducted in appropriate manner to estimate relative efficacy. In CDR reanalysis assuming no difference in viral suppression at 48 weeks between comparators did not alter overall conclusions. Even if no additional QALYs occur with DTG, it would still remain attractive given lower net costs.

Drug Cost

The cost of antiretroviral therapy is the key driver of costs (comprising approximately 87% of total costs). ART costs are lower for DTG, driven by either lower drug-acquisition costs of DTG (in some but not all comparators), as well as lower likelihood of treatment failure or resistance (associated with use of second- through sixth-line therapies, which are more costly). The lower likelihood of progressing down the treatment algorithm has face validity (according to the clinical expert) based on clinical estimates of improved efficacy in most outcomes. DTG drug costs are greater compared with generic efavirenz (EFV) plus Truvada (instead of Atripla), leading to an ICUR of $44,604 for DTG; however, the clinical expert indicates that in the current era of therapy, the majority of patients would be receiving Atripla and not generic EFV + Truvada. If there is no difference in progression to second- through sixth-line therapies, DTG would remain less costly than five to seven of the eight comparators considered, depending on the NRTI backbone used.

Integrase Inhibitor–Resistant Patients

The recommended dose of DTG in patients who are INI resistant is 50 mg twice daily. No economic information was provided for this patient population.

Treatment-Naive – Efficacy for Virologic Suppression

When equivalent virological suppression is assumed for DTG versus Complera, Stribild, ATZ/r, and LPV/r, DTG remains the dominant strategy (less costly and more effective).

Issues for Consideration

According to the clinical expert, the level of adherence is a major predictor of the effectiveness of treatment. As such, the once-daily DTG might have an advantage to improve adherence over other regimens with more frequent administration, which may have an impact on real-world effectiveness.