Post on 26-Jul-2020
The controversy about Paclitaxel-coated devices---latest evidence:
A Critical Appraisal of JVIR CLI Paclitaxel Analysis
William A. Gray MD FACC FSCAISystem Chief of Cardiovascular Services
President, Lankenau Heart InstituteWynnewood PA
USA
No finding of paclitaxel effect on death at 6-12 months
No finding of paclitaxel effect on major amputations at 6-12 months
Finding of a paclitaxel effect on amputation-free survival at 6-12
months
Exploratory analysis: dose effect
Inadequate numbers to construct a study-level meta-analysis with significant risk of Type I error (False positive)
Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISM) Flowchart
~1400 subjects
Where have we seen this before?JAHA SFA meta-analysis also fell short
in the relevant time period
Enrollment by duration of trial follow up
4663
4432
2316
863
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
Initial Year 1 Year 2 Year 5
# of Patients
# of Patients
28 RCT
28 RCT
12 RCT
3 RCT
Main Line HealthLankenau Heart Institute
120 subjects
138 subjects
85 subjects
Significant inclusion of non-peer reviewed data
~25% (343/1420)
A Consistent Error:
PTA group is likely not paclitaxel naïve for the entirety of the analysis
• Paclitaxel device approvals in EU and US preceded all of the trial data
• Prior use in proximal, contra-lateral lesions during or prior to the trial is not known
• This was also a problem with the prior JAHA analysis
– Not possible in a study-level analysis---requires patient level data
The math is wrong: IN.PACT Deep
A Consistent Error: The lost-to-follow up (LTFU) and
withdrawals (WD) are not completely/accurately accounted for in the
JAHA analysis
THUNDER trial
After WD and LTFU:mortality at 5 years=
33% (12/36) and 28% (8/29)
RR: 1.17 (not 1.69)
At year 2Control: 12 LTFU DCB: 4 LTFU
At year 5Control: 13 LTFU DCB: 7 LTFU
Dose “analysis” is highly flawed
• No lesion length, number of balloons used, or adjustments for selection bias or cross-trial differences
• Inconsistent with prior Katsanos methodology
Included studies with non-standard follow-up: mixed 6 month to 1 year
The JAHA purported mortality effect was noted at 2 and 5 years,
but not at 1 year
This brings the entire proposed mechanistic explanation–which was already a tortured
one---into even further question
Hoisted by his own petard!
Summary
This “analysis” is very poorly constructed and conducted, and therefore should have no meaningful impact on this high-risk, in-need CLI population, especially given the
marked improvement in patency documented in the same manuscript
Shame on JVIR
• The poor conduct and construct of this analysis is evident to even a cursory review
• Therefore, one can only assume that the editors knowingly allowed the publication of poor science
• Their motives can only be speculated on…
• But their downgrade in standards should be clear to everyone, and should be called out
The controversy about Paclitaxel-coated devices---latest evidence:
A Critical Appraisal of JVIR CLI Paclitaxel Analysis
William A. Gray MD FACC FSCAISystem Chief of Cardiovascular Services
President, Lankenau Heart InstituteWynnewood PA
USA