ONLINE SUPPLEMENTARY MATERIAL

Efficacy and Safety of Switching to the 2-Drug Regimen Dolutegravir/Lamivudine Versus

Continuing a 3- or 4-Drug Regimen for Maintaining Virologic Suppression in Adults

Living With HIV-1: Week 48 Results From the Phase 3, Non-inferiority SALSA

Randomized Trial

Josep M. Llibre,1 Carlos Brites,2 Chien-Yu Cheng,3,4 Olayemi Osiyemi,5 Carlos Galera,6 Laurent

Hocqueloux,7 Franco Maggiolo,8 Olaf Degen,9 Stephen Taylor,10,11 Elizabeth Blair,12 Choy

Man,12 Brian Wynne,12 James Oyee,13 Mark Underwood,12 Lloyd Curtis,13 Gilda Bontempo,12

Jean van Wyk14

1Hospital Universitari Germans Trias i Pujol, Barcelona, Spain; 2Universidade Federal da Bahia,

Salvador, Brazil; 3Department of Infectious Diseases, Taoyuan General Hospital, Ministry of

Health and Welfare, Taoyuan, Taiwan; 4Institute of Public Health, School of Medicine, National

Yang-Ming Chiao Tung University, Taipei, Taiwan; 5Triple O Research Institute PA, West Palm

Beach, FL, USA; 6Hospital Clínico Universitario Virgen de la Arrixaca, Murcia, Spain; 7Centre

Hospitalier Régional d’Orléans, Orléans, France; 8ASST Papa Giovanni XXIII, Bergamo, Italy;
9Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany; 10Birmingham Heartlands

Hospital, Birmingham, UK; 11University of Birmingham, Birmingham, UK; 12ViiV Healthcare,

Research Triangle Park, NC, USA; 13GlaxoSmithKline, Brentford, UK; 14ViiV Healthcare,

Brentford, UK

Page 1
Table of Contents

Supplementary Graphical Abstract ............................................................................................... 3

Supplementary Methods ............................................................................................................... 4
Supplementary Figure 1. ............................................................................................................... 5
Supplementary Figure 2. ............................................................................................................... 6
Supplementary Figure 3. ............................................................................................................... 8
Supplementary Figure 4. ............................................................................................................... 9
Supplementary Figure 5. ............................................................................................................. 11

Page 2
Supplementary Graphical Abstract

Page 3
Supplementary Methods

Procedures

Samples were analyzed using the Abbott RealTime HIV-1 assay (Abbott Molecular Inc,

Des Plaines, IL; lower detection limit: 40 copies/mL). Criteria for confirmed virologic withdrawal

(CVW) were defined as HIV-1 RNA ≥50 copies/mL followed by a second consecutive HIV-1

RNA ≥200 copies/mL. If CVW criteria were met, the participant was withdrawn from the study

and plasma samples from Day 1 and the initial elevated viral load at suspected virologic

withdrawal were tested for HIV-1 protease, reverse transcriptase, and integrase genotype and

phenotype (Monogram Biosciences, South San Francisco, CA).

Adverse events (AEs) and weight were recorded at each study visit through the last

participant’s Week 52 visit. Clinical chemistry, hematology, and CD4+ and CD8+ cell counts

were assessed at all visits through Week 48. Renal, bone, and inflammatory biomarkers;

homeostasis model of assessment–insulin resistance (HOMA-IR; fasting serum glucose and

insulin); fasting lipids; and urinalysis were evaluated at baseline, Week 24, and Week 48.

Page 4
Supplementary Figure 1. Change from baseline in HOMA-IR in safety population. CAR,

current antiretroviral regimen; DTG/3TC, dolutegravir/lamivudine; HOMA-IR, homeostasis model

of assessment–insulin resistance. aEstimated adjusted ratio was calculated using mixed-model

repeated measures applied to change from baseline in loge-transformed data adjusting for

treatment, visit, baseline third agent class, CD4+ cell count (continuous), age (continuous), sex,

race, body mass index (continuous), presence of hypertension, loge-transformed baseline

HOMA-IR value (continuous), treatment-by-visit interaction, and loge-transformed baseline

value-by-visit interaction, with visit as the repeated factor. n indicates number of participants

with non-missing data at baseline, Week 24, and Week 48.

Page 5
Supplementary Figure 2. Change in serum inflammatory biomarkers from baseline. CAR,

current antiretroviral regimen; DTG/3TC, dolutegravir/lamivudine. Mixed-model repeated

measures analysis was not performed for D-dimer due to high proportion of participants with D-

dimer < lower limit of quantification in both treatment groups. Baseline geometric mean ratio

values (DTG/3TC group; CAR group): C-reactive protein (1.34; 1.27), interleukin-6 (1.73; 1.68),

soluble CD14 (1.55 × 106; 1.46 × 106), and soluble CD163 (538.18; 541.70). aRatio is the

estimated adjusted ratio (Week 48 to baseline) in each group calculated using mixed-model

repeated measures applied to change from baseline in loge-transformed data adjusting for

treatment, visit, baseline third agent class, CD4+ cell count (continuous), age (continuous), sex,

race, body mass index (continuous), smoking status, hepatitis C virus co-infection status, loge-

transformed baseline biomarker value (continuous), treatment-by-visit interaction, and baseline

value-by-visit interaction, with visit as the repeated factor.

Page 6
Page 7
Supplementary Figure 3. Change from baseline in fasting lipids at Week 48. CAR, current

antiretroviral regimen; DTG/3TC, dolutegravir/lamivudine; HDL, high-density lipoprotein; LDL,

low-density lipoprotein. Adjusted mean treatment difference (95% CI) displayed above treatment

groups. aPercent change from baseline based on adjusted ratio (Week 48 to baseline) in each

group calculated from mixed-model repeated measures applied to change from baseline in loge-

transformed data adjusting for treatment, visit, baseline third agent class, CD4+ cell count

(continuous), age (continuous), sex, race, loge-transformed baseline value (continuous),

treatment-by-visit interaction, and loge-transformed baseline value-by-visit interaction, with visit

as the repeated factor. bn/N = number of participants with non-missing fasting lipid data at

baseline and Week 48, excluding those with lipid-modifying agent administered at baseline (lipid

data collected after initiation of a lipid-modifying agent were censored), of total number of

participants in treatment group.

Page 8
Supplementary Figure 4. Change from baseline at Week 48 in (A) plasma/serum and (B) urine

renal biomarkers. CAR, current antiretroviral regimen; CKD-EPI, Chronic Kidney Disease

Epidemiology Collaboration equation; DTG/3TC, dolutegravir/lamivudine; eGFR, estimated

glomerular filtration rate. aEstimated mean change from baseline at Week 48 in each group

calculated from mixed-model repeated measures adjusting for treatment, visit, baseline third

agent class, CD4+ cell count (continuous), age (continuous), sex, race, body mass index

(continuous), presence of diabetes mellitus, presence of hypertension, baseline biomarker value

(continuous), treatment-by-visit interaction, and baseline value-by-visit interaction, with visit as

the repeated factor. bBased on estimated adjusted ratios of Week 48 vs baseline. Based on the

same model as plasma/serum biomarkers except adjusting for loge-transformed baseline

biomarker value. n indicates number of participants with non-missing data at baseline and Week

48.

Page 9
Page 10
Supplementary Figure 5. Change in serum bone biomarkers from baseline to Week 48. CAR,

current antiretroviral regimen; DTG/3TC, dolutegravir/lamivudine. aEstimated mean change from

baseline at Week 48 in each treatment group calculated from mixed-model repeated measures

adjusting for treatment, visit, baseline third agent class, CD4+ cell count (continuous), age

(continuous), sex, race, body mass index (continuous), smoking status, baseline biomarker

value (continuous), treatment-by-visit interaction, and baseline value-by-visit interaction, with

visit as the repeated factor. n indicates number of participants with non-missing data at baseline

and Week 48.

Page 11