Transcript Outcome research

Journal Club
“Implementing a Tenofovir-Base First-Line
Regimen in Rural Lesotho: Clinical Outcomes
and Toxicities After Two Years”
JAIDS 2011
Bygrave et al
Background
-Stavudine (d4T) was the most commonly used
backbone drug in resource limited settings
(RLS)
-Based on RCT’s by Gallant et al in JAMA 2004
(TDF vs d4T) and in NEJM 2006 (TDF/FTC + EFZ
vs AZT/3TC + EFZ) the WHO guidelines of 2009
recommended TDF or AZT in favor of d4T
-Lesotho adopted these recommendations
before the release of these guidelines
Background
-Pop. ~ 2 million, 40% below poverty line, GDP
around 6 bi $ (wiki)
-Almost ¼ of the population is HIV + (UNAIDS)
-HIV co-infections: 35.5 % (28.9-42.6) HbCAg +,
5.5% (2.8-9.6%) HbSAg +, 0.5% (0-2.8%) HCV +
(Rebenau et al Sex Transm Dis 2010)
-In US: MMWR 2011 survey on Armed Forces>
HIV/HBV: 1.6% HIV/HCV: 1.4%
-MSF established a RN run clinic at the Scott Health
Service Area in rural Lesotho in 2006
Study Outline
-Goal: evaluate real life application of WHO guidelines
-Design: Retrospective cohort outcomes study
-Setting: RN run rural clinic
-Participants: Adults starting HAART from Jan 1 2008 to
Dec 31 2008, follow up until Dec 31 2009. Exclusion
criteria: GFR < 30 ml/min or no Cr or CD4 at baseline
-Data collection: extraction from files by a team of 3
clinicians
-End points: (1) death (2) loss for follow up {defined by
missing an appointment for > 90 days} (3) 1st toxicity
driven switch
Study Outline
-TDF prescribed to all non pregnant adults with
GFR> 50ml/mim
-AZT is pregnant or GFR < 50 ml/min
-d4T: excluded from d4T and Hb < 8 g/dl
-All on triple Rx with 3TC + nevirapine or efavirenz
-HAART started when CD4 < 350 or WHO stage 3 or
4
Study Outline
-Nelson Aalen cumulative hazards estimative
method to calculate cumulative hazards for
regimen change using Cox regression for the
following confounders: Age, gender, TB at
HAART initiation, pregnancy at HAART
initiation, CD4 baseline, and GFR
-Nelson Aalen is used to estimate cumulative
hazards (vs survival Kaplan-Meier) (wikipedia)
Results
Results
-Mortality: 6.5 per 100 patient-years (5.3-7.9)
-Non significant trend for higher mortality on
non-TDF regimens: TDF CI (3.8-7.0), AZT CI
(5.0-11.1), and d4T CI (5.8-11.7)
-Trend was sustained on multivariate analysis
-Difference of loss for follow up non statistically
significant
Results
Switch rates
• TDF (2.0-4.5)
“renal toxicity”
• AZT (5.4-12.1)
“severe anemia”
• d4T (14.8-24.1)
“severe neuropathy”
“lipodystrophy”
-This was sustained in the
multivariate analysis
Strengths
-Feasible study design in a setting the is very
different from where RCT’s were done
-No blinding: switches are not protocol driven.
Allows us to understand why providers or
patients elect to switch between regimens
when they are aware of the medication used.
-Nurse managed: realistic to RSL settings
Weaknesses
-Retrospective
-No validation system for data collection
mentioned
-Death in Lesotho
-Side effects vs loss for follow up
-Anemia not included in multivariate analysis:
bias against d4T?
Weaknesses
-HBV status? This data is not included in the
pivotal TDF trial, but given HBV as much more
prevalent in Lesotho than in US, this could
have introduced bias.
-Definition of toxicity?
Weaknesses
-How often can they measure Cr? Can they
check proteinuria? What about phosphate?
-Is Lesotho ready to deal with CKD?
-LDL? To they care about dyslipidemia as much
as we care?