1. Managing renal and bone toxicity Andrea Antinori INMI L. Spallanzani IRCCS, Roma Simposio “INI in the Long-term HIV Treatment” Catania 9 Novembre, 2015

2. Andrea Antinori Disclosure statement Personal fees for consultancy and lectures from Abbvie, Bristol Myers Squibb, Gilead, Janssen, Merck, ViiV. Travel grants from Abbvie, ViiV. Research grants from Bristol Myers Squibb, Gilead, Janssen, ViiV.

3. Risk and Age at Diagnosis of End-Stage Renal Disease in HIV-Infected Versus Uninfected Adults Althoff KN, Clin Infect Dis, 2015 A total of 98,687 (31% HIV-infected and 69% uninfected) adults contributed >450,000 person-years and 1,135 ESRD incident diagnoses.

4. ARVAlteration of renal function (generally not treatment-limiting) Treatment-limiting renal disease TDF Tubular dysfunction Rapid eGFR decline Proteinuria (non-glomerular) Incident CKD AKI (rare) Tubulo-interstitial nephritis (rare) Tubular disease/FS (uncommon) CKD with progressive eGFR decline ATV/r Inhibition of creatinine secretion Tubular dysfunction Rapid eGFR decline Incident CKD AKI (rare) Tubulo-interstitial nephritis (rare) Nephrolithiasis (uncommon) CKD with progressive eGFR decline LPV/r Incident CKD/CKD progression None reported COBI/EVG Inhibition of creatinine secretion AKI (rare) Tubular disease/FS (uncommon) COBI/ATV Inhibition of creatinine secretion AKI (rare) Tubular disease/FS (uncommon) DTG Inhibition of creatinine secretion None reported RPV Inhibition of creatinine secretion None reported ARV’s and the kidney Post F, 2013

5. D:A:D Non-ARV Predictors of Advanced CKD/ESRD Gender female vs male Ethnicity African vs Caucasian HIV transmission IDU vs MSM Prior AIDS yes vs no HBV pos vs neg HCV pos vs neg Hypertension yes vs no Diabetes yes vs no Prior CVE yes vs no Smoking non vs current eGFR per 10 ml/min lower Age per 10 years higher CD4 per doubling CD4 Nadir per 100 cells/mm 3 higher VL per log 10 higher UnivariateMultivariate 0.250.51248 Advanced CKD/ESRD IRR (95%CI) Poisson regression model adjusted for gender, ethnicity, HIV transmission group, enrolment cohort, Prior AIDS, HBV status*, HCV status*, smoking status*, hypertension*, diabetes*, cardiovascular events (CVE)*, baseline year, eGFR, age, current CD4 count*, CD4 Nadir, HIV-1 viral load (VL)*, and use of TDF, ATV/r, ATV, LPV/r, other PI/r and IND. Incidence rate-ratio, IRR. *time-updated. Ryom L, et al. AIDS, 2014

6. Age Family history ARTDiabetes HIV Hypertension Hepatitis C Ethnicity KD Risk = Modifiable = Non-modifiable Risk factors for kidney disease in the HIV-positive population Gupta et al. Clin Infect Dis 2005;40:1559–85. Decreased CD4 cell count Increased viral load

7. DHHS 2015 Antiretroviral Regimen Considerations as Initial Therapy based on Specific Clinical Scenarios - 1 Patient or Regimen Characteristics Clinical Scenario Consideration(s)Rationale/Comments Presence of other conditions Chronic kidney disease (defined as eGFR<60 mL/min) Consider avoiding TDF. If eGFR is <70 mL/min, Do Not Use: EVG/c/TDF/FTC, or ATV/c with TDF, or DRV/c with TDF Options for CKD Patients Use ABC/3TC if HLA-B*5701 Negative: If HIV RNA >100,000 copies/mL, do not use ABC/3TC with EFV or ATV/r. If CrCl <50 mL/min, do not use coformulated ABC/3TC because 3TC requires dose adjustment. Other Options (See Text for Discussion): DRV/r plus RAL (if HIV 200/mm3 ), or LPV/r plus 3TC, or Modify TDF dose TDF has been associated with renal tubulopathy. OsteoporosisConsider avoiding TDF. Use ABC/3TC if HLA-B*5701 negative If HIV RNA >100,000 copies/mL, do not use ABC/3TC plus (EFV or ATV/r) TDF is associated with greater decrease in bone mineral density along with renal tubulopathy, urine phosphate wasting, and osteomalacia.

8. D:A:D Study ARV Exposure (Per Year) and Incidence Rate Ratios of ceGFR ≤70 and CKD From eGFR >90 Cumulative TDF (aIRR 1.18 [1.12-1.25] per year) and ATV/r (1.19 [1.09-1.32]) use were independently associated with increased rates of ceGFR ≤70 from eGFR >90, but not with CKD (eGFR <60), whereas lopinavir/r (LPV/r) use was associated with both endpoints (1.11 [1.05-1.17]) and 1.22 [1.16-1.28] respectively. Inconsistent trends were seen with abacavir use. Ryom L, et al. J Infect Dis, 2013

9. D:A:D: ARV Exposure and Risk of CKD Retrospective analysis of pts with BL eGFR > 90/mL/min (N = 23,560) – Evaluated cumulative exposure to TDF, ABC, ATV/RTV, LPV/RTV, other PIs and risk of CKD – 210 pts developed CKD Multivariate analysis: exposure to TDF, ATV/RTV, and LPV/RTV significantly associated with CKD development – Risk  greatly over 5 yrs Association with TDF or LPV/RTV and CKD remains when excluding those who stopped drugs during or before study entry When TDF exposure censored, CKD risk per yr of ATV/RTV or LPV/RTV exposure increased substantially CKD risk  with time after stopping TDF Mocroft A, et al. CROI 2015. Abstract 142. Relationship Between Increasing Exposure to ARVS and CKD 1.80 1.60 1.40 1.20 1.00 0.00 ATV/RTVLPV/RTV TDF IRR (95% CI) Univariate Multivariate On treatment TDF censored

10. Association of cumulative ARV exposure (per year) with risk of kidney disease outcomes *** p<.0001, ** p<.001, * p<.01, + p<.05 Scherzer R, et al. AIDS, 2012

11. Studies of renal uptake transporters showed that RAL inhibited organic anion transporters 1 and 3 (OAT1 and OAT3) with 50% inhibitory concentrations (IC50s) (108 Mand 18.8 M, respectively) well above the maximum concentration of drug in plasma (Cmax) at the clinical 400-mg dose and did not inhibit organic cation transporter 2 (OCT2). Raltegravir Has a Low Propensity To Cause Clinical Drug Interactions through Inhibition of Major Drug Transporters Rizk ML, et al. AAC, 2014

12. Low risk of TDF discontinuation for renal adverse events in treatment-naive HIV-1 patients Meta-analysis of Randomized Clinical Studies Twenty-one clinical studies met the selection criteria. Treatment duration ranged from 48 to 288 weeks. Renal AEs led to study drug discontinuation in 44 of 10,129 patients exposed to TDF (0.43%; 95% CI, 0.32%-0.58%) and 2 of 2,013 patients exposed to non-TDF-containing regimens (0.10%; 95% CI, 0.01%-0.36%). In 5 randomized, controlled studies that included a non-TDF comparator, the estimated risk difference between the treatment groups (TDF vs non-TDF) was 0.50% (95% CI, 0.13%-0.86%; P =.007). In clinical studies using TDF-containing regimens, the rate of discontinuations due to renal AEs was low, but was slightly higher than in studies using non-TDF comparators. Winston J, et al. HIV Clin Trials, 2014

13. Mean (± SD) Change From Baseline To Week 48 In Creatinine (µMol/L) By NRTIs Curtis LD, et al. IAS 2013. Poster TUPE282 DTG 50 mg + TDF/FTC QDRAL 400 mg + TDF/FTC BID DTG 50 mg + ABC/3TC QDRAL 400 mg + ABC/3TC BID Subjects receiving each NRTI background, n (%) DTGRAL TDF/FTC242 (59)247 (60) ABC/3TC169 (41)164 (40)

14. GS 104 & 111 pooled analysis Changes in quantitative proteinuria at week 48 Sax PE, et al. Lancet, 2015 At 48 weeks, quantitative proteinuria (total urinary protein, albumin, retinol binding protein and β2-microglobulin to urine creatinine ratios) increased from baseline in the E/C/F/TDF. Reductions or significantly smaller increases in these urinary proteins were noted in the E/C/F/TAF group. Other measures of proximal renal tubular function (fractional excretion of phosphate and uric acid) showed significantly less change in patients receiving E/C/F/TAF compared with the E/C/F/TDF.

15. GS-US-292-0109 Renal Safety Results Statistically significant improvements for participants who switched from either E/C/F/TDF or from boosted ATV + FTC/TDF  Serum creatinine (p <0.001); eGFR (p <0.001)  Fractional excretion of phosphate, FEPO4 (p=0.05); fractional excretion of uric acid, FEUA (p <0.001) Changes began by Week 2 and persisted to Week 48 UPCR: urine protein: creatinine ratio; UACR: urine albumin: creatinine ratio; RBP, retinol-binding protein; β-2-m:Cr, beta-2 microglobulin. E/C/F/TAF TDF-Based Regimen RBP:Crβ-2-m:CrUPCRUACR Tubular Proteinuria Each difference between treatment arms was statistically significant (p <0.001). Mills A, et al. IAS 2015. Vancouver, CA; #TUAB0102.

16. Long-term Bone Mineral Density Changes in Antiretroviral- treated HIV-Infected Individuals Change in LS BMD by Serostatus In HIV, BMD decline significantly slows after initial steep decline but continues through 7.5 years of follow-up During early period, HIV-related factors associated with BMD loss (lower CD4, higher VL, TDF use) During late period, total lean body mass, but not HIV-related factors, associated with BMD loss Comparing A5318 study results with two HIV-uninfected cohorts (BACH/Bone and WIHS)

17. HIV infection affects bone formation HIV seroconversion was associated with decreases in osteoblast activity We did not observe an effect of HIV infection on P1NP and CTX levels ART initiation was associated with decreases in sclerostin levels, suggesting that ART initiation may affect osteoblast regulation. Both HIV infection and ART have an impact on bone metabolism Slama L, et al. EACS, 2015

18. START Substudy Greater Declines in Bone Mineral Density With Immediate vs Deferred ART Over 3 Years Significantly greater declines in BMD at spine and hip with immediate vs deferred initiation of ART over mean follow-up of 2.2 years Declines in total spine BMD greatest after 12 months, then appeared to plateau – At 12 and 24 months, decline in total spine BMD approximately 2.0% with immediate ART vs < 0.5% for deferred ART (P <.001 at both time points) – At 36 months, decline in total spine BMD approximately 2.5% with immediate ART vs < 0.5% for deferred ART (P =.001) – Estimated mean difference in total spine BMD from baseline with immediate vs deferred ART groups -1.6% (95% CI: -2.2 to - 1.0; P <.001) Declines in total hip BMD appeared to remain fairly steady over 36 months – At 12 months, decline in total hip BMD approximately 2.0% with immediate vs approximately 0.5% for deferred ART (P <.001) – At 12 months, decline in total hip BMD approximately 3.0% with immediate vs < 1.5% for deferred ART (P <.001) – At 36 months, decline in total spine BMD approximately 3.5% with immediate vs approximately 2.0% for deferred ART (P =.06) – Estimated mean difference in total hip BMD from baseline with immediate vs deferred ART groups: -1.5% (95% CI: -2.3% to - 0.8%; P <.001) BMD substudy included participants from 33 sites in 11 countries on 6 continents 424 patients randomized n = 206 to immediate ART; n = 218 to deferred ART 397 with follow-up data n = 193 to immediate ART; n = 204 to deferred ART Osteoporosis in 13% of bone substudy patients at baseline Hoy JF, et al. EACS, 2015

19. Renal and bone disease in HIV 1. Reynes J. et al., AIDS 2013; 2. Schouten J. et al., CID 2014; 3. Mocroft A. et al., AIDS 2007; 4. Brown TT, Qaqish RB, AIDS 2006; 5. Cotter AG. et al., AIDS 2014

20. Effect of renal tubular markers on BMD Estimate of effect (B) and 95% confidence interval (95%CI) from multivariate linear regression models adjusted by demographic factors and BMI (model 1), laboratory variables (PTH and ALP – model 2) and BTMs (model 3). Tubular proteinuria, measured by RBP/Cr, but not total proteinuria, is associated with reduced BMD at the 3 anatomical sites. This association persists even correcting for by HIV infection and BTMs. Alvarez E, et al. EACS, 2015. Abst.#

21. Mean changes in bone mineral density during TDF-containing trials (HIV+ and HIV-) Trials of HIV-infected subjects: GS-903, ASSERT, A5224s. Trials of HIV-uninfected subjects (HIV PrEP): iPrEx, MSM PrEP 1. Gallant JE et al. JAMA 2004;292:191–201; 2. Moyle G et al., LIPO 2010; London. Oral #23; 3. McComsey G et al. J Infect Dis 2011;203:1791–1801; 4. Mulligan K et al. CROI 2011. #94LB; 5. Liu A et al. CROI 2011. #93 *This slide depicts data from multiple studies published from 2004–2011. Not all regimens have been compared head-to-head in a clinical trial Change in spine BMD, % 1 0 -2 -3 -4 -5 1 0 -2 -3 -4 -5 Change in hip BMD, % 024487296 TDF use, weeks GS-903 1 iPrEx 4 ASSERT 2 MSM PrEP 5 A5224s 3 Spine BMD Hip BMD

22. Trends in TBS over time according to ART groups Naive (n=26) TeRiPi (n=14) RiPi (n=10) Te (n=27) NoTeRiPi (n=13) TeRiPi < Te = RiPi < NoTeRiPi  Time (yrs) 90 patients, who remained naïve or had 1 st ART unchanged and had  2 bone DXA assessments Martinez E, et al. 17th International Workshop on Co-morbidities and Adverse Drug Reactions in HIV, Barcelona, 2015

23. A5257 BMD losses were less pronounced with RAL compared to the Pis, both combined with TDF/FTC Brown TT, et al. CROI 2014, Boston (MA). Abst.#779LB Brown TT, et al. J Infect Dis, 2015

24. Intragroup differences from baseline: (A) lumbar spine at week 48, standard group, p<0.0001; NtRTI-sparing group, p=0.05; week 96 standard group, p<0.0001; NtRTI-sparing group, p=0.44; (B) total hip at week 48, standard group, p<0.0001; NtRTI- sparing group, p=0.17; week 96 standard group, p<0.0001; NtRTI-sparing group, p=0.016; (C) femoral neck at week 48, standard group, p<0.0001; NtRTI-sparing group, p=0.0005; week 96 standard group, p=0.02;

25. Studies 104/111 Significantly Smaller Decline in Hip and Spine BMD With TAF Significantly smaller decline in hip and spine BMD with TAF 02448 0 24 48 -8 -6 -4 -2 0 2 Wk -8 -6 -4 -2 0 2 = 845 = 850 797 816 784 773 836 848 789 815 780 767 -1.30 -2.86 -0.66 -2.95 P <.001 Mean % Change From BL Sax PE, et al. CROI 2015. Abstract 143. Sax PE, et al. Lancet, 2015 n TAF/FTC/EVG/COBI (n = 866) TDF/FTC/EVG/COBI (n = 867) Change in Spine BMDChange in Hip BMD

26. GS-US-292-0109 DXA Scan Results: Spine BMD  Regardless of prior treatment regimen, differences between arms were statistically significant  More than 2% difference between the arms at Week 48 26 Median % Change in BMD (Q1, Q3) E/C/F/TAF TDF-Based Regimen Change From Baseline to Week 48 All Participants (N=1,369) p <0.001 Mills A, et al. IAS 2015. Vancouver, CA; #TUAB0102.

27. Summary Renal impairment as frequent in HIV disease due to multifactorial pathway. Among ARV drugs, TDF and boosted PIs mainly affect renal function. Association of TDF with drugs characterized by lower risk of kidney damage and TDF substitution (ABV, RAL, TAF) as possible strategies of harm reduction. BMD changes more frequent in HIV population. Bone injury by HIV from time of seroconversion. Greater BMD decline observed with early ARV treatment. Bone and renal disease in a common pathway driven by tubular proteinuria. TDF and boosted PIs as main factors of BMD changes. INSTI (RAL)-including (even NRTI-sparing) and future TAF-including regimens as possible strategies to minimize ART impact on BMD decline.