A Study of Encapsulated Cell Technology (ECT) Implant for Participants With Early Stage Retinitis Pigmentosa: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Encapsulated Cell Technology (ECT) Implant for Participants With Early Stage Retinitis Pigmentosa (NCT00447980), a randomized phase 2 study. The primary outcome is Change in Humphrey Visual Fields - Total Sensitivity.

2. Clinical Question

The clinical question is framed around a continuous measure of visual-field sensitivity rather than a count of events or a time-to-event outcome. Change from baseline places the endpoint on a scale of change over the stated follow-up interval. The study record lists the endpoint as primary and also lists two secondary outcomes concerning mean Humphrey visual-field sensitivity.

3. Trial Design

The study, NCT00447980, is titled “A Study of Encapsulated Cell Technology (ECT) Implant for Participants With Early Stage Retinitis Pigmentosa.” The official title describes a Phase II/III study of encapsulated human NTC-201 cell implants releasing CNTF, with visual-field sensitivity as the primary outcome. The registry lists the phase as PHASE2, allocation as randomized, masking as double, and the model as parallel.

The record describes two dose groups. In each, an NT-501 implant was placed in one eye and sham surgery was performed in the fellow eye. The low-dose summaries include 20 participants for the implant and 20 for sham; the high-dose summaries include 47 for each. These are eye-arm summaries within the described implant-and-fellow-eye design, so they should not be read as unrelated groups of people. The primary endpoint population is identified as MITT, with one high-dose participant excluded because they had no visits after surgery.

The registry lists enrollment as 73, the study status as completed, and the sponsor as Neurotech Pharmaceuticals. The recorded start date is 2007-01-22 and primary completion date is 2010-04-19. The registry describes the purpose as examining safety and effectiveness of CNTF implants on vision in people with retinal degenerations. It does not report the trial’s original sample-size justification or detailed design assumptions, so those should not be inferred from the later planning illustrations.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: A total of 90 participants were screened and 73 were randomized to dose groups. 17 of 90 patients were screen failures and not associated (randomized) to either arm. Of the 73 randomized participants, 5 participants, 2 in the high dose and 3 in the low dose arms were randomized but did not receive surgery and did not take part in the study. 68 participants had surgery. For participants randomized to treatment dose, one eye was randomized to surgery and the fellow eye was randomized to sham.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: MITT Population excluding 1 participant from the high dose arm who did not complete any visits subsequent to surgery.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: The baseline analysis population includes all 73 participants who were randomized to a treatment arm including the 3 participants from low dose arm and the 2 participants from the high dose arm who discontinued prior to surgery, and study start.

6. Primary Endpoint

The registry lists the primary outcome as Change in Humphrey Visual Fields - Total Sensitivity, measured over 12 months.

The primary endpoint is change in Humphrey Visual Fields total sensitivity from baseline to month 12. The registry describes the assessment as the HVF 30-2 test, comprising 76 points, with the measure calculated as the sum of actual thresholds at all 76 locations. The unit is dB, and the reported summary is a mean with standard deviation.

A continuous endpoint retains information about the magnitude and direction of change rather than reducing results to a binary category. In general, a change score expresses how the measured value at follow-up differs from its baseline value. For this endpoint, a negative mean change indicates a lower summed sensitivity at follow-up relative to baseline; the clinical meaning of a particular change is not established by the summary statistics alone.

Visual-field sensitivity is used in the study because the stated research question concerns vision in people with retinitis pigmentosa. The registry identifies visual-field sensitivity as the primary outcome and lists mean Humphrey visual-field sensitivity outcomes as secondary measures. It does not report a threshold for a clinically important change, nor does the posted endpoint description explain how missing month-12 measurements were handled in the trial’s own analysis.

7. Results Posted in the Registry

• Low Dose - Implant: mean 1.40 versus 16.80 on control; difference -15.40 (95% CI -120.80 to 90.00), standardized effect d = -0.09; z-test p = 0.775.
• High Dose - Implant: mean -164.80 versus 16.80 on control; difference -181.60 (95% CI -261.20 to -102.00), standardized effect d = -1.38; z-test p = <0.001.
• High Dose - Sham: mean -66.70 versus 16.80 on control; difference -83.50 (95% CI -161.77 to -5.23), standardized effect d = -0.67; z-test p = 0.037.

8. Statistical Interpretation

9. Statistical Methods Explained

The endpoint is continuous: it is a mean change in a numerical visual-field measure. A mean gives the average value within an arm, while the standard deviation describes variability among measurements around that mean. In general, a large standard deviation relative to a mean difference signals that individual measurements vary considerably, making the arm-level contrast less precise. Here, all four posted standard deviations are large compared with the low-dose contrast, and the confidence interval for that contrast is correspondingly broad.

What does a mean difference represent?

The listed mean differences are unadjusted recalculations using the posted group means. For the low-dose implant comparison, the calculation contrasts 1.4 dB with the low-dose sham mean of 16.8 dB. For the high-dose implant comparison, it contrasts -164.8 dB with that same low-dose sham mean. The high-dose sham comparison similarly contrasts -66.7 dB with low-dose sham. These comparisons are arithmetic summaries of the posted arm means, not evidence that the original trial analysis used these exact contrasts or methods.

The study design has an important paired feature: implant and sham were assigned to fellow eyes of a participant. Measurements from two eyes in the same person may be related. A paired analysis uses that within-person relationship; a calculation based only on separate arm means and standard deviations does not. The registry record does not report the relationship between fellow-eye measurements or an original paired analysis, so the recalculated intervals and p-values are best read as simple, unadjusted illustrations rather than as the trial’s definitive inferential analysis.

How should a confidence interval be read?

In general, a confidence interval expresses uncertainty around an estimated contrast under the assumptions of the method used. It shows a range of values compatible with the observed data and calculation procedure; it is not a guarantee that the true value lies within that particular interval. A narrow interval indicates greater precision than a wide interval, while an interval spanning zero includes values representing no difference as well as values in either direction.

The low-dose interval runs from -120.8 to 90.0, so it covers zero and a wide range on both sides. The high-dose implant interval, from -261.2 to -102.0, lies below zero. The high-dose sham interval, from -161.8 to -5.23, also lies below zero. Those interval positions describe the unadjusted recalculations; they do not account for possible multiplicity, the paired structure, or an analysis model that is not reported in the registry record.

What do the p-values say?

In general, a p-value measures how unusual data at least as extreme as those observed would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The recalculated low-dose comparison has p = 0.775; the high-dose implant comparison has p < 0.001; and the high-dose sham comparison has p = 0.037. Because these are unadjusted recalculations, they should not be interpreted as confirmatory trial results or as accounting for a family of comparisons.

Unadjusted means that the calculations do not incorporate adjustments for other variables or multiple testing. The record does not report covariate-adjusted estimates, a multiplicity procedure, or the sponsor’s statistical analysis plan. The primary endpoint population is specified as MITT, excluding one high-dose participant with no visits after surgery, but the registry record does not describe the analysis method for missing endpoint measurements beyond that statement.

Why can over-dispersion matter?

General explanation: over-dispersion refers to variability greater than a model expects. It is a concern for outcomes analyzed with models that assume a particular variance structure, such as some count-data models. The primary endpoint is continuous and is summarized with means and standard deviations. The posted standard deviations show substantial spread, but the registry record does not describe a count model or an over-dispersion adjustment. For this endpoint, the relevant general point is that high variability can reduce precision and that the analysis needs a variance approach appropriate to the data and design.

Finally, the registry lists three outcome measures and two secondary outcomes, but does not report the full analysis plan or how the outcomes were prioritized in statistical testing. That missing information limits conclusions about the intended inferential hierarchy. The recalculated statistics are therefore useful for understanding the posted summaries, while the trial’s own statistical methods and estimates are not reported in the registry record.

10. Confidence Intervals and Precision

Precision is reflected in how narrowly an estimate is bounded by its confidence interval. The low-dose implant contrast is -15.4 dB, with a 95% confidence interval from -120.8 to 90.0. The interval is wide and crosses zero, so the recalculation is compatible with a range of negative and positive contrasts. The estimate alone should not be read as a precise measurement of the difference.

The high-dose implant contrast is -181.6 dB, with a 95% confidence interval from -261.2 to -102.0. The high-dose sham contrast is -83.5 dB, with a 95% confidence interval from -161.8 to -5.23. Both intervals lie below zero, although they also span a range of possible values. Their interpretation remains subject to the unadjusted calculation and the unaccounted-for fellow-eye pairing.

In general, interval width depends on the observed variability, sample information, and assumptions used in the calculation. The posted standard deviations are 174.6 and 165.4 for the low-dose implant and sham groups, and 115.0 and 103.2 for the high-dose implant and sham groups, respectively. The registry does not provide confidence intervals for the four arm means themselves. It also does not report the trial’s own adjusted interval estimates or a model that could incorporate the within-participant structure.

Confidence intervals convey statistical uncertainty, not clinical importance. The registry record does not report a clinically meaningful threshold for this endpoint. Consequently, even an interval lying on one side of zero cannot by itself establish the practical significance of a contrast. Precision must be considered alongside the design, endpoint definition, variability, and analysis limitations.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total participants across the study. They are planning illustrations calculated from the observed effect, not the trial’s original sample-size justification; that original justification is not reported in the registry record. The illustrations assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario uses the stated fraction of the observed effect as the effect to detect.

The 80% power column gives the planned participants per arm under those assumptions for a design with an 80% chance of detecting the specified effect if that effect is truly present and the assumptions hold. The 90% power column gives the corresponding per-arm planning figure for a 90% chance. These are design calculations, not counts observed in the trial. They should not be read as evidence that the completed study had a particular probability of finding a result.

The rows describe planning for 100%, 75%, or 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so it generally requires more participants. For the low-dose implant contrast, the 80% power figures are 1,915 participants per arm for 100% of the observed effect, 3,403 for 75%, and 7,657 for 50%. For the high-dose implant contrast, the corresponding figures are 9, 15, and 34. For the high-dose sham contrast, they are 35, 63, and 140. The 90% power figures are listed separately in the planning display. These values are not observed power.

The illustrations inherit limitations of the observed estimates, including their uncertainty and the fact that the recalculations do not account for the fellow-eye pairing. They should be treated as conditional planning aids rather than recommendations or a substitute for a design justified using the intended analysis and clinical assumptions.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -15.400; pooled standard deviation σ = 170.057.
2. n = 2 × (1.96 + 0.842)² × 170.057² / -15.400² = 1,914.2, rounded up to 1,915 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -15.4                      # observed difference in means
sigma <- 170.056820651216                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(447980)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 1,915 (80% power), 2,563 (90% power)
 75% of observed effect: n per arm = 3,403 (80% power), 4,556 (90% power)
 50% of observed effect: n per arm = 7,657 (80% power), 10,251 (90% power)
power.t.test: 1915.1 
Simulated power at n = 1,915 per arm: 0.799 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists the primary outcome as change in Humphrey visual fields total sensitivity at 12 months. It also lists two secondary outcomes: change in mean Humphrey visual-fields sensitivity and mean Humphrey visual-fields sensitivity. The record does not report how the investigators controlled the chance of false-positive findings across multiple endpoints, comparisons, or time points. A multiplicity adjustment or testing hierarchy is therefore not reported in the registry record.

The posted comparisons include three unadjusted recalculations: low-dose implant versus low-dose sham, high-dose implant versus low-dose sham, and high-dose sham versus low-dose sham. Their p-values should be understood individually and without a reported adjustment for multiple comparisons. The registry lists no analysis records, and it does not provide the sponsor’s primary-analysis estimate, adjusted p-value, or statistical analysis plan. The recalculated results do not establish what the trial’s prespecified testing strategy was.

The primary endpoint population is described as the MITT population, excluding one participant in the high-dose arm who did not complete any visits after surgery. That exclusion is stated in the endpoint description. The record does not explain how other missing measurements, incomplete follow-up, or any further departures from the planned population were handled. It also does not report whether endpoint values were imputed or whether a sensitivity analysis assessed missing-data assumptions.

The paired-eye design is another analysis issue. The registry describes implant treatment in one eye and sham surgery in the fellow eye, but does not report the correlation between eyes or the method used to account for it in the sponsor’s analysis. Separate-arm recalculations do not capture that relationship. Without the analysis plan and trial-specific model, the role of pairing, covariate adjustment, and missing-data handling cannot be assessed from the registry record.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 4% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cubital tunnel syndrome (5.0%), acute appendicitis (2.1%), ovarian cyst (2.1%), hypertrophic medial synovial plica with synovitis (2.1%), non-ischemic cardiomyopathy (2.1%).

The registry reports serious adverse event counts by dose group. In the low-dose group, 1 participant was affected among 20 at risk; in the high-dose group, 4 were affected among 48 at risk. No deaths are reported in either group. These are group-level safety figures and should not be confused with the visual-field efficacy endpoint or the recalculated comparisons.

The record also reports other adverse events for 20 of 20 at-risk participants in the low-dose group and 48 of 48 at-risk participants in the high-dose group. The serious-event terms listed include cubital tunnel syndrome (5.0%), acute appendicitis (2.1%), ovarian cyst (2.1%), hypertrophic medial synovial plica with synovitis (2.1%), and non-ischemic cardiomyopathy (2.1%). The registry excerpt does not clarify the denominator or event-counting convention for these listed percentages, so they should be read as reported rather than reinterpreted.

In general, adverse-event counts describe events reported in the groups and do not by themselves establish causation. The record does not provide a detailed account of event timing, severity beyond serious-event classification, attribution, or comparative statistical testing. These safety summaries are descriptive and should be interpreted within those limits.

15. What the Result Does, and Does Not, Mean

The posted data describe changes in Humphrey visual-field total sensitivity and simple contrasts among the listed groups. The low-dose implant mean change was 1.4 dB, compared with 16.8 dB for low-dose sham. The unadjusted recalculated difference was -15.4 dB, with a broad interval that includes zero. The high-dose implant and high-dose sham mean changes were -164.8 dB and -66.7 dB, respectively; their listed unadjusted contrasts against low-dose sham were -181.6 dB and -83.5 dB.

These estimates do not establish that an implant caused the observed changes. The recalculations are not reported as the sponsor’s own analysis, do not account for the fellow-eye pairing, and have no reported multiplicity adjustment. The registry does not give the trial’s analysis plan, adjusted endpoint estimates, or missing-data methods. Those omissions make it inappropriate to treat the simple calculations as definitive evidence about treatment effects.

Statistical significance is not the same as clinical importance. The registry does not report a threshold for a meaningful change in dB, so the clinical importance of the estimates cannot be determined from these figures alone. Confidence intervals describe uncertainty under the calculation assumptions, while p-values do not quantify the probability that a treatment works or the size of a clinically meaningful benefit.

The safety figures report serious adverse events and other adverse events by dose group, with no deaths listed. They are descriptive counts, not proof of cause or a complete characterization of safety. Overall, the registry record supplies a study framework and outcome summaries, but does not report enough of the original analysis to make the recalculated comparisons a substitute for a full trial report.

16. Why This Trial Matters Statistically

This study illustrates why an endpoint’s design matters as much as its summary statistics. The primary outcome is a continuous change in summed visual-field sensitivity, and the trial compares an implanted eye with a sham-operated fellow eye. That paired-eye structure creates a natural within-participant relationship that a simple comparison of arm means does not represent. It is therefore important to distinguish descriptive arm summaries from analyses that account for the design.

The posted variability and intervals also show why a point estimate is not a complete account of evidence. The low-dose contrast is modest relative to its uncertainty, while the high-dose contrasts have intervals below zero in the unadjusted recalculations. Yet those calculations remain constrained by absent information about the sponsor’s analysis, multiplicity, and handling of missing observations. Statistical interpretation requires both the observed estimates and clarity about how they were derived.

The planning illustrations further demonstrate that sample-size calculations depend on the effect assumed for planning. The figures change across the listed fractions of observed effects, and smaller assumed effects require more participants per arm. They are not observed power and do not reconstruct the study’s original design rationale.

For clinical researchers, the central statistical lesson is to separate what the registry reports from what can only be established by a full analysis. Endpoint definitions, paired measurements, uncertainty, missing data, multiplicity, and safety denominators all shape interpretation. When those details are not reported, that absence is part of the evidence assessment rather than a reason to fill gaps with assumptions.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record has important reporting limitations for interpreting efficacy. It does not report the sponsor’s own primary analysis, the statistical analysis plan, a multiplicity procedure, or an analysis model for the fellow-eye design. The recalculated contrasts use posted summary statistics and should not be read as adjusted or paired trial estimates. The registry also does not report the relationship between fellow-eye measurements, which is needed to assess the precision of a paired comparison.

The endpoint description identifies the MITT population and notes the exclusion of one high-dose participant who did not complete any visits after surgery. It does not describe how other missing endpoint measurements were handled, whether values were imputed, or whether sensitivity analyses were conducted. The posted arm summaries and analysis-population description may not provide enough detail to reconstruct the trial’s full participant flow.

The record lists secondary outcomes but does not provide their numerical results in the registry record for this analysis. It also does not report a clinically important threshold for the dB endpoint. Therefore, the magnitude of the estimates cannot be translated into a threshold-based assessment of clinical relevance. The serious adverse event listing provides counts and selected terms, but not a complete account of timing, attribution, or event-level detail.

Finally, the sample-size figures are planning illustrations based on observed effects and stated assumptions, not the study’s original design justification. They do not resolve the uncertainty in the observed effects or replace a plan based on the trial’s intended paired analysis.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Study of Encapsulated Cell Technology (ECT) Implant for Participants With Early Stage Retinitis Pigmentosa — NCT00447980.
• Registry record last updated 2025-05-01; status last verified 2025-04.
• Results first posted 2025-05-01.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT00447980 was a completed randomized, double-masked, parallel Phase II study of NT-501 encapsulated cell implants releasing CNTF for people with retinitis pigmentosa. The record lists enrollment of 73, a start date of 2007-01-22, and primary completion on 2010-04-19. Each dose group included an implant in one eye and sham surgery in the fellow eye.

The primary endpoint was change in Humphrey visual-fields total sensitivity from baseline to 12 months, measured as the sum of thresholds at 76 locations. The registry lists group means and standard deviations, along with summary-based unadjusted recalculations. It does not report the trial’s own analysis plan or estimates in the record. Safety reporting lists serious and other adverse events by dose group and no deaths.