The OPHTEC ARTISANTM Myopia lens: endothelial cell counts

1
The OPHTEC ARTISANTM Myopia lens endothelial
cell counts

• Gerry Gray, Ph.D.
• Cardiovascular and Ophthalmic Team Leader,
• Division of Biostatistics
• Office of Surveillance and Biometrics
• Center for Devices and Radiological Health
• February 5, 2004

2
Study design

• Purpose investigate effects of device on
  endothelial cells through time.
• Endothelial cell counts measurements
• Specular microscope photographs
• Multiple images per eye (fall 2000 later)
• Counts at Baseline, 6 months, 1, 2, 3 years.
• Due to large variability, images were re-read
• 662 eyes originally enrolled counted
• Variability was quite high
• 353 available photos from reliable machines were
  re-counted at one reading center
• NO CONTROL GROUP AVAILABLE

3
ECD recount eye accountability
No baseline visit 170
one subsequent 9
two or more subsequent 161
Baseline, 0,1 subsequent 22
Baseline, 2,3 subsequent 104
All visits 57
two or more visits after baseline 322
TOTAL 353
All eyes accountability based on ECD recount
dataset submitted on 12/10/2003. Includes 215
first eyes and 138 fellow eyes
4
Available recount data
Green indicates an ECD count is present.
Individual patients can be traced
horizontally. Numbers of patients with ECD counts
indicated for selected regions.
5
Available recount data
Green indicates an ECD count is present.
Individual patients can be traced
horizontally. Numbers of patients with ECD counts
indicated for selected regions.
6
ECD results for all eyes and visits
Blue line connects mean counts at each time.
7
Mean cell counts for various cohorts
COHORT n BL 6 mo 1 yr 2 yr 3 yr
All eyes 353 2761 2768 2720 2662 2578
w/o BL 170 NA 2779 2672 2638 2552
w/ BL 183 2761 2764 2758 2688 2612
2FU 322 2767 2772 2714 2663 2583
All visits 57 2816 2811 2766 2758 2691
2FU patients with two or more post-op visits.
8
Estimates of cell loss over duration of the study

• Baseline to 3 years 4.75 - 5 loss
• Using 107 BL3yr patients, sponsor estimates
  4.75 baseline to 3 year loss,
  95 CI 3.3, 6.2.
• Summing sponsors change by period
  5.05 baseline to 3 year loss,
  95 CI 3.1, 7.0
• Includes both initial operational loss (if any)
  and normal loss due to ageing.

9
Steady state long term loss

• What is the long term rate of change in
  endothelial cell density that we can expect to
  see?
• Estimate depends mostly on
• If we believe there is any initial operative loss
• If we estimate using
• Difference between baseline 3 years, or
• All visits (regression)
• Cohort that is used for estimation
• Whether we use all eyes, 2FU, eyes w/all visits,
  etc.

10
What are the standards?

• Want reasonable assurance that endothelial cell
  density is preserved.
• Normal ageing about 0.6 per year
• Consensus (?) points for concern
• 1500 cells/mm2, 800 cells/mm2
• ALL of the standards are written in terms of
  the upper confidence limit on the rate of loss.
• FDA draft guidance ( Ophthalmic panel)
• annual rate from 3 mo. 3 yr., upper 90 CI 1.5
• ISO, ANSI (not standards for acceptable rate of
  loss)
• calculate sample size using 2.0 upper 90 CI

11
Steady state long term loss

• Percent change from baseline (NOT equivalent to
  steady state)
• 107 patients with baseline 3 year counts
• Annual loss 1.59, 95 CI 1.1, 2.1, 2.0
  one-sided limit
• 57 patients with counts at all time points
  (consistent cohort)
• Annual loss 1.27, 95 CI 0.4, 2.1, 2.0
  one-sided limit
• Regression-based analyses using all time points
  produce a loss of 1.7 -2.0 per year
• Sponsor 183 eyes w/ baseline count, including
  baseline, GEE
• Annual loss 1.72, 95 CI 1.23, 2.22
• Alternate 322 eyes w/2FU, 6mo through 3yrs, not
  including baseline
• Annual loss 1.9, 95 CI 1.5, 2.3
• Annual loss 2.0, 95 CI 1.6, 2.5
  depending on GEE model
• Additional 4-year data on 27 patients 1.63 loss
  between 3 and 4 years

12
Test for change in slope

• For eyes with baseline measurements, virtually no
  change between baseline and 6 months
• Piecewise linear model
• Baseline to 6 months, 6 months to 3 years
• Various methods of testing do not show a
  significant change in slope
• Data do not indicate any peri-operative period
  with increased rate of endothelial cell loss
• Conflicts with expected outcomes in all guidance
  documents

13
Long term loss

• Percent change from baseline is not equivalent to
  the steady state long term rate
• Regression-based methods use intermediate time
  points and should generally be considered
  superior to simple differences between start
  end times
• Estimate depends somewhat on whether baseline
  count is included in regression
• 1.7 loss including baseline
• 1.9 - 2.0 not including baseline
• Upper CI limits 2.2 - 2.5
• All regressions use GEE methods for
  within-patient correlation

14
Results from three different fits
15
Extrapolation from different fits
?
16
Extrapolation caveats

• ALWAYS a questionable exercise to extrapolate
  beyond the range of available data, especially to
  the degree we want here.
• Highly dependent on the model we use the
  assumptions we make.
• ALL of the previous extrapolations assume that
  loss will continue linearly.
• Probably much more important to think about
• If its necessary to obtain good long-term data.
• If so, how to go about it.

17
How do individual patients fare?

• Perhaps more important than average cell loss
  through time.
• What proportion of the patients will show cell
  loss greater than some critical amount?
• What proportions of patients will have cell
  densities less than 1500, 800 cells/mm2 in 10,
  20, 30 years?
• Difficult to answer this with any confidence
• Extrapolation of percentiles

18
Estimated outcomes for individual patients
Random effect regression, 6 months subsequent
on 2FU patients.
19
Summary of estimates from individual patient
modeling
Percentiles Percentiles Percentiles Percentiles
5 ile 10 ile 25 ile 50 ile
Annual rate after 6 months -2.3 -2.2 -2.1 -1.9
Percent change in cells/mm2 from preop. mean
2767 cells/mm2
20
Predictors of endothelial cell loss
21
Question for panel

• Do the endothelial cell data presented provide
  reasonable assurance of safety of the Artisan
  myopia lens?

22
(No Transcript)