Dr. Ameet Dravid presentation at International AIDS conference

1. Tenofovir nephrotoxicity in resource limited setting
of Western India : Higher rate of renal function
decline, acute kidney injury and progression to
chronic kidney disease compared to Western data
A.Dravid
1
,A.Sadre
2
,S.Dhande
1
, A.Borkar
1
,M.Kulkarni
1
,M.Dravid
3
1
Ruby hall clinic, Department of HIV Medicine, Pune, India
2
Ruby hall clinic, Department of Nephrology, Pune, India
3
Infectious disease clinic, Department of HIV Medicine, Dhule, India

2. Introduction
•India has the second highest population of HIV positive patients in
the world which stands at 2.5 million out of which 600,000 patients
are on antiretroviral therapy.
•Tenofovir based antiretroviral therapy is increasingly used for
treatment naïve and treatment experienced patients in India over
the last 5 years as per recommendation by national guidelines
•It has coincided with availability of generic fixed dose combinations
of Tenofovir/emtricitabine(TE),
Tenofovir/Emtricitabine/Efavirenz(TEE) and
Tenofovir/Lamivudine/Efavirenz(TLE)
•Tenofovir nephrotoxicityis characterized by proximal tubular cell
dysfunction that may be associated with Fanconi’ssyndrome, acute
kidney injury or progression to chronic kidney disease.
•Tenofovir nephrotoxicitydevelops in 1-2% patients

3. Introduction
•Majority of the data on Tenofovir nephrotoxicity comes from
either randomized controlled trials or from observational
studies conducted in Western,resource rich settings.
•Clinical trials have strict inclusion criteria and tend to exclude
patients with comorbidities which hampers their
generalisability to real world settings.
•Data from resource limited settings like India is sparse and is
plagued by missing data, limited covariates for analysis, high
incidence of lost to fo llow up and short follow up.

4. Objective
•Primary objective of this study was to determine annual
decline in estimated glomerular filtration rate(eGFR) by MDRD
equation in patients taking Tenofovir based ART and
comparing it with patients on Tenofovir sparing regimens
•Secondary objective was to determine incidence of acute
kidney injury(AKI) in Tenofovir exposed population which led
to Tenofovir withdrawal.
•Recovery of renal function on Tenofovir cessation was studied
and proportion of patients who progress to Stage 3-5 Chronic
kidney disease despite Tenofovir withdrawal was estimated.

5. Methods
•The study was carried out at Ruby Hall Clinic,Pune which is a tertiary
centre for HIV/AIDS clinical care in Western India.
•Electronic medical records of patients are stored in a central database
from which demographic, clinical and laboratory data of all patients was
extracted.
•Patients who were initiated on or switched to Tenofovir based
antiretroviral therapy from 1
st
March 2009 to 1
st
March 2013 were
included in this retrospective observational cohort study.
•Patients having atleast 1 follow up serum creatinine and creatinine
clearance values were included.
•Patients already on Tenofovir based ART prior to 2009 were also included
provided they had regular baseline and follow up data available.
•Patients started on Tenofovir sparing regimens (i.e. Zidovudine, Stavudine
and Abacavir based regimens) during the said period were taken as control
population

6. Methods
•Age,Sex,CD4 count, HBsAgstatus, serum creatinine, Baseline
WHO stage 3-4 infection, Use of concomitant antiretroviral
drugs along with Tenofovir( Non nucleoside reverse
transcriptase inhibitors (NNRTI) versus protease inhibitors (PI)
were the demographic variables studied.
•Serum creatininewas measured at baseline and 6 monthly
follow up by Von Jaffe method.
•Serum creatinineestimation was made traceable to Isotope
dilution mass spectrophotometry(IDMS) in 2012.
•Glomerularfiltration rate (eGFR) was measured at baseline
and at every 6 monthly follow up by Cockcroft Gault(CG)
formula and Modification of diet in renal disease(MDRD)
equation.
•Total duration of exposure to Tenofovirin months was
calculated for each patient.

7. Methods
•Tenofovirwas initiated in patients with Creatinine Clearance (Cr Cl) > 50 ml/min.
•Patients started on Tenofovir at baseline Cr Cl < 50 ml/min were excluded.
•Patients having missing baseline and follow up serum creatinine values were also
excluded
•Annual decline in GFR was calculated by CG formula and MDRD equation for
Tenofovircontaining and Tenofovir sparing regimens.
•Acute Kidney Injury was defined as Serum creatinine> 2 mg/dl, Cr Cldecrease to
< 50 ml/min or GFR decrease > 50% of baseline (Rifle criteria 2002).
•Patients with GFR value < 60 ml/min(MDRD equation), 6 months after Tenofovir
discontinuation were classified as having Chronic kidney disease ( CKD).
•Presence of co morbidities which increase incidence of renal toxicity like diabetes
mellitus, hypertension, use of concomitant nephrotoxic drugs, obstructive
uropathy and urinary tract infecton were recorded.
•Angiotensin converting enzyme inhibitors(ACEI), Non steroidal anti-inflammatory
drugs(NSAID’s), Amino glycosides and Amphotericin B were the nephrotoxic drugs
studied.
•Obstructive uropathy included conditions like renal calculus disease, urethral
stricture and benign prostatic hypertrophy.

8. Methods
•Entire data was analysed by the SPSS software(STATA) version
18
•One way Anova test was used to compare GFR decline
amongst subgroups of patients
•Multivariate logistic regression analysis was applied to the
dataset to identify factors significantly associated with
increasing risk of acute kidney injury in Tenofovir exposed
patients

9. Results

10. Baseline characteristics
TDF containing regimensTDF sparing regimens
Total number743340
Age (Mean yrs)43 yrs 39.5 yrs
Sex (M: F)68 : 3262 : 38
Median Baseline CD4 count168 cells/mm
3
121 cells/mm
3
Weight (kg)55.45 kg52.2 kg
Serum Creatinine 0.85 mg/dl0.8 mg/dl
GFR (CG) mean90.19 ml/min84.51 ml/min
GFR (MDRD equation) mean96.04 ml/min100.04 ml/min
Baseline OI36.6%23.2%
Mean duration of F/U21 months33 months

11. Results (TDF exposed cohort)
•Mean weight of cohort was 55.5 kg with 286/743 (38.5 %) patients
having weight <= 50 kg
•437/743(58.81 %) patients had baseline CD4 count <= 200
cells/mm
3
and 258/743 (34.72)% had baseline CD4 count <= 100
cells/mm
3
•30/743 (4 %) were HBsAg positive
•588/743 (79%) patients were exposed to Tenofovir with Non
nucleoside reverse transcriptase inhibitors (NNRTI) and 155/743(21
%) to Tenofovir with Protease inhibitors (PI).
•Mean duration of follow up was 21 months with 553/743(74.42%)
patients having follow-up >=12 months. .
•214/743 (28.8%) patients had baseline creatinine clearance (CG
formula) between 50-70 ml/min
•168/743(22.61 %) patients had baseline GFR (MDRD equation)
between 60-90 ml/min while 13/743(1.7%) had baseline GFR
between 30-60 ml/min

12. Baseline GFR values
GFR by CG formulaNo of patientsGFR by MDRD
equation
No of patients
50-70 ml/min21430-60 ml/min22
71-90 ml/min23361-90 ml/min266
91-120 ml/min21891-120 ml/min328
>120 ml/min78>120 ml/min127

13. TDF containing antiretroviral regimens
247
285
56
138
17
0
50
100
150
200
250
300
No of patients
No of patients

14. eGFR decline on follow up for entire cohort
12 months24 months36 months48 months
Number of patients
completing F/U
55327413256
Mean GFR by CG formula91.04 ml/min91.70ml/min92.36 ml/min92.12 ml/min
Mean GFR by MDRD
equation
89.14 ml/min84.92ml/min82.87 ml/min84.52 ml/min
GFR baseline – GFR F/U by
MDRD equation
6.9 ml/min6.3 ml/min4.37 ml/min-0.19ml/min

15. e
GFR decline on follow up (TDF + NNRTI vsTDF + PI)
12 months24 months36 months48 months
Number of patients4041969038
GFRdecline in TDF +
NNRTI by MDRD equation
(ml/min)
5.5 ml/min6.3 ml/min4.19 ml/min0.9 ml/min
Number of patients9355289
GFR decline in TDF + PI by
MDRD equation (ml/min)
11.75 ml/min7.04 ml/min7.33 ml/min-2.7 ml/min

16. Follow up GFR decline
•Mean decline in GFR in Tenofovir exposed
cohort (MDRD Equation) : 5.29 ml/min/year
•Mean decline in GFR in patients exposed to
TDF + NNRTI only : 4.18 ml/min/year
•Mean decline in GFR in patients exposed to
TDF + PI only : 9.19 ml/min/year
•Mean decline in GFR in patients exposed to
Tenofovir sparing regimens : 1.3 ml/min/year

17. GFR decline in presence of risk factors known to
increase Tenofovir renal toxicity
GFR decline/year
P value
Age >= 50 yrsYes (n=99)
5.46 ml/min/year
0.057
No (n=451)
5.58 ml/min/year
SexMale (n=388)
6.3 ml/min/year
0.367
Female (n=165)
3.74 ml/min/year
CD4 count =< 100
cells/mm3
Yes (n= 188)
6.4 ml/min/year
0.036
No (n=365)
5.19 ml/min/year
Diabetes mellitusYes (n=35)
11.43 ml/min/year
0.014
No (n = 518)
5.37 ml/min/year
HypertensionYes (n=57)
12.24 ml/min/year
0.765
No (n=496)
4.99 ml/min/year
Nephrotoxic drugsYes (n=50)
6.88 ml/min/year
0.399
No (n =477)
5.13 ml/min/year
Obstructive uropathyYes (n=28)
17.25 ml/min/year
0.770
No (n=525)
5.07 ml/min/year

18. Tenofovir and Acute kidney injury
•Number of patients who developed AKI : 36/743
(4.8%)
•Time to developing AKI
•< 6 months : 16
•6 –12 months : 5
•12 -24 months : 8
•> 24 months : 7
•Median time to developing AKI : 8.5 months.
•Number of patients requiring haemodialysis : 3/36
(8.33 %)
•Number of patients who died : 4/36 (11.11 %)

19. Analysis of risk factors which increase risk of AKI in Tenofovir
exposed population by multiple logistic regression
Risk factorNumber of patientsNumber of patients
developing AKI
P value by multiple
logistic regression
Creatinine clearance
50-70 ml/min
21422P = 0.01
CD 4 count < 100
cells/mm
3
25822P = 0.002
Tenofovir with PI15517P = 0.042
Diabetes mellitus489P = 0.69
Hypertension655P = 0.105
Concomitant
nephrotoxic drugs
5012P = 0.031
Obstructive uropathy3711P = 0.001

20. Recovery of renal function post Tenofovir withdrawal
and progression to grade 3-5 chronic kidney disease
•Out of 36 patients who developed AKI, 18 patients completed
6 months of follow up post Tenofovir cessation.
•4/36 patients died,2/36 were lost to follow up and 1 patient
had to take Tenofovir alternate day for his Hepatitis B
coinfection.
•7/18 patients continued to have eGFR < 60 ml/min 6 months
post TDF cessation
•Only 1 patient had Serum creatinine >2 mg/dl, 6 months after
Tenofovir withdrawal.
•None of the patients have required long term renal
replacement therapy as of now.

21. Renal function recovery on TDF cessation
Scenarios of eGFR recovery after TDF cessation Number of patients
Complete recovery of renal function : eGFR on recovery
equal to or better than eGFR baseline
4/18
Incomplete recovery of renal function : eGFR after 6
months of TDF cessation > 60 ml/min but not reaching
baseline
5/18
Incomplete recovery of renal function and progression to
stage 3 CKD : eGFR after 6 months of TDF cessation
between 30-60 ml/min (Stage 3 CKD)
6/18
Incomplete recovery of renal function and progression to
stage 4 CKD : eGFR after 6 months of TDF cessation
between 15 -30 ml/min (Stage 4 CKD)
1/18
eGFR baseline < 60 ml/min and eGFR after TDF cessation
showed incomplete recovery and remained < 60 ml/min
(Stage 3 CKD)
2/18

22. Discussion
•Renal function decline in patients on Tenofovir based ART
was much higher than patients taking Tenofovir sparing ART
(mean 5.29 ml/min/year vs 1.3 ml/min/year)
•Renal function decline was higher in patients taking Tenofovir
with PI’s than in patients taking Tenofovir with NNRTI (9.19
ml/min/year vs 4.14 ml/min/year).
•Normal age related GFR decline in HIV negative population is
1 ml/min/year.
•So in effect, eGFR decline seen in patients taking Tenofovir
based ART in our cohort is similar to that seen in patients
suffering from diabetic nephropathy.
•The decline seems to be progressive over a period of 3 years.

23. Discussion and conclusions
•We also found higher incidence of acute kidney injury
amongst our Tenofovir exposed population compared to that
seen in Western resource rich settings (4.8% vs 1 %)
•This could be attributable to lower baseline creatinine
clearance, lower eGFR, lower baseline CD4 count and higher
incidence of co-morbidities in our cohort.
•Recovery of eGFR after withdrawal of Tenofovir is incomplete
in significant proportion of patients in our cohort. These
patients are at risk of progression to stage 3-5 Chronic kidney
disease.
•Finally, management of Tenofovir nephrotoxicity in resource
limited settings like India is tough due to limited access to
routine laboratory monitoring, renal replacement therapy and
alternate antiretroviral drugs like Abacavir.

24. CONCLUSIONS
Drawbacks of our study :
•retrospective observational cohort design
•CKD EPI equation and AKIN criteria could not be applied to
our dataset
•Routine urine examination data was not available at baseline
and follow up for all the patients. It could have helped us to
identify nephrotoxicity earlier.
•Although we tried to identify all factors which increase risk of
renal toxicity by Tenofovir, there could be unknown
confounding factor which could have been missed.