Cysteamine Eye Drops to Treat Corneal Crystals in Cystinosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cysteamine Eye Drops to Treat Corneal Crystals in Cystinosis (NCT00001213), a na phase 2 study. The primary outcome is Number of Participants With Serious and Non-Serious Adverse Events.

2. Clinical Question

3. Trial Design

The registry describes a Phase 2, completed, single-group study sponsored by the National Eye Institute. Enrollment was 328. The listed condition is cystinosis, and the intervention is cysteamine topical solution administered hourly while awake in both eyes. Masking is listed as none, and allocation is listed as not applicable. These features mean the posted results are not from a randomized comparison between treatment groups.

The study began in 1986-04, and the registry lists primary completion in 2013-07. The record describes an initial aim of testing whether cysteamine eye drops could prevent or reduce corneal cystine crystals. It later states that, after efficacy had been established and a New Drug Application filed, the primary outcome measure was changed post hoc to evaluating safety information. The registry’s brief summary also describes a longer-term purpose of evaluating safety and effectiveness until the drops became available for commercial purchase.

The primary safety outcome is titled as the number of participants with serious and non-serious adverse events. Its timeframe is recorded as any time point up to 27 years. The accompanying description says there was no specified timeframe and that safety data were collected until the drug became available for commercial purchase in May 2013. These statements characterize the registry’s safety collection; they do not give participant-level follow-up durations or a uniform observation period.

Further protocol details, including eligibility criteria, the planned statistical analysis, and any sample-size justification, are not reported in the registry record. The absence of a comparator also means the study design cannot separate changes associated with treatment from changes over time or other influences using a concurrent control group.

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.

Overall Study

Reasons for not completing this period:

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.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Serious and Non-Serious Adverse Events, measured over Any Time Point up to 27 Years.

The posted primary outcome counts participants recorded as having serious and non-serious adverse events. The registry describes it as a post-hoc primary measure: after ophthalmic cysteamine efficacy had been established and a New Drug Application filed, the focus of this outcome was safety. The timeframe is labeled any time point up to 27 years, while the outcome description says there was no specified timeframe and that safety information was collected until commercial availability in May 2013.

A participant count is a descriptive summary of how many people were recorded with an event category. It is not the same as a count of every event, because one participant may experience more than one event. The posted results list 11 of 328 participants at risk with serious adverse events and 257 of 328 at risk with non-serious adverse events. The record does not specify whether participants could appear in both categories, so these counts should not be added together to create a total.

Safety outcomes are used to describe events observed during study follow-up, including events that may warrant attention because of their seriousness. A count alone does not show whether an event was caused by the intervention. Interpretation also depends on how events were defined and collected, how long participants were observed, and whether exposure and follow-up were comparable. Those details are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

For a binary safety outcome, each participant can be classified according to whether the event category was recorded or not. A simple count describes the number classified with the event; its interpretation depends on the denominator and on the period over which events could be detected. In this study, the registry gives 328 participants at risk for both the serious and non-serious categories, with 11 and 257 participants affected, respectively. Those counts are descriptive. They do not by themselves describe the number of distinct event episodes, severity within a category, or timing.

Participant counts and event counts answer different questions. If one participant has repeated episodes, a count of participants with at least one event does not count every episode. The primary outcome title specifies participants with serious and non-serious adverse events, so the measure is at the participant level. The registry does not report event-level totals or enough detail to determine whether the serious and non-serious groups overlap. Adding the two participant counts would therefore not be justified by the record.

A denominator is essential for interpreting a count. The registry identifies 328 participants at risk for each category. In a study with an observation period that differs among participants, a simple proportion does not account for time under observation. A rate based on person-time would require participant-level information about follow-up and event timing, which is not reported in the registry record. The timeframe wording also needs care: the outcome is labeled any time point up to 27 years, while its description says no specified timeframe was used and safety collection continued until commercial purchase in May 2013. The posted summary does not reconcile those descriptions.

Confidence intervals are a general statistical tool for expressing uncertainty around an estimate. For a participant proportion, an interval can convey that the observed count is only one estimate of the underlying frequency in a broader population. Its width depends partly on the amount of information available and on the method used. No confidence intervals or interval-estimation method are reported for this outcome, so uncertainty should not be inferred from the counts alone. A confidence interval would not, by itself, resolve the absence of a control group or show that an event was caused by treatment.

In a randomized comparative trial, an analysis might estimate a difference between groups, with uncertainty around that contrast. Here, the registry identifies one treatment group and no like-for-like control comparison. There is no treatment contrast to estimate from the posted safety outcome. Descriptive counts remain useful for conveying what was recorded, but they cannot distinguish treatment effects from background events, the underlying condition, other care, or other influences over follow-up. The record does not report an adjusted model or another strategy intended to address those issues.

The record’s efficacy history and its posted primary safety outcome are also distinct. The registry says the safety measure became primary after efficacy had been established and a New Drug Application filed. This is a post-hoc change described in the outcome text; the record does not provide the original statistical plan, dates or details of the change, or analyses supporting the earlier efficacy statement. The safety counts should not be treated as a test of efficacy, and the brief summary’s efficacy language is not a substitute for an outcome analysis.

The statistical planning assumptions associated with the sample-size illustration are general planning assumptions, not a description of the study’s actual design. The illustration uses a two-sided significance level of 0.05, equal allocation between two hypothetical arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. Such assumptions describe a hypothetical planning exercise. They do not transform this single-group record into a two-arm comparison, and they do not provide observed power. Any sample-size figure generated under those assumptions would be a planning illustration rather than an observed trial result or evidence that the study was adequately powered for its actual safety outcome.

Over-dispersion is a general concern when observations vary more than a statistical model assumes. For participant-level binary counts, the central issue is not a reported over-dispersion adjustment; no model is reported. More broadly, repeated events within the same participant can be correlated, so treating every episode as independent would require methods that account for clustering. The registry outcome is framed as the number of participants affected, not the number of episodes, and the record does not provide analyses of repeated events.

Finally, the registry does not report a missing-data method, an analysis population definition beyond the displayed at-risk count, or a sensitivity analysis. It is therefore not possible to determine from the record whether participants with incomplete follow-up were handled differently, whether all enrolled participants contributed throughout safety collection, or whether any assumptions were tested. These omissions limit assessment of the completeness and precision of the reported safety summary.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cysteamine Topical Solution")
n      <- c(328)
events <- c(257)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cysteamine Topical Solution  328  257  78.4  73.5  82.7  73.6  82.5  73.6  82.5

11. Multiplicity, Missing Data and Analysis Populations

Multiplicity concerns arise when a study assesses multiple outcomes, comparisons, time points, or subgroups and then makes several statistical tests. Without a prespecified approach, the chance of finding at least one apparently notable result can increase as the number of tests grows. The registry lists two outcome measures, but does not report a multiplicity strategy, a hierarchy of outcomes, or an adjustment for multiple testing. It also reports no analyses in the registry record. It is not possible to determine how many analyses were conducted or how any multiplicity was handled.

The primary safety outcome is described as post hoc, following the establishment of efficacy and the filing of a New Drug Application. The registry does not report the original outcome plan, the timing or process of the change, or whether the safety outcome was selected before or after examination of safety data. The description establishes that the measure was post hoc; it does not establish whether the change affected analysis choices or conclusions.

Missing-data handling is not reported in the registry record. There is no account of missing event information, incomplete follow-up, exclusions, or methods for addressing missing observations. The displayed at-risk group size does not by itself establish that every participant had complete safety information throughout the full collection period.

Analysis populations are also not defined beyond the participant counts at risk. The record does not state whether the counts include every enrolled participant, how discontinuation or incomplete observation was handled, or whether any subgroup analyses were planned. Without those details, readers cannot assess the consequences of missingness or determine whether the reported counts reflect a prespecified population. The absence of these details is a reporting limitation, not evidence that missing data or multiplicity did or did not affect the findings.

12. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: benign intracranial hypertension (2.4%), renal failure (0.6%), optic disc disorder (0.3%), blindness (0.3%), gastrointestinal infection (0.3%).

The registry reports 11 of 328 participants at risk with serious adverse events and 257 of 328 with non-serious adverse events in the cysteamine topical solution group. These figures count participants, not every event episode. Because the record does not say whether someone could be included in both categories, the counts should not be combined into a single total.

The serious-event terms listed are benign intracranial hypertension (2.4%), renal failure (0.6%), optic disc disorder (0.3%), blindness (0.3%), gastrointestinal infection (0.3%), and intestinal obstruction (0.3%). The registry does not specify the denominator or precise calculation underlying these percentages. They should therefore be read as registry-listed term percentages, not assumed to be directly comparable participant-level risks.

These are observations recorded during safety collection, not proof that cysteamine caused the events. The study has no control group, and the registry does not report individual exposure or follow-up durations, event timing, or a causality assessment. The timeframe is described as any time point up to 27 years, while the outcome description says data collection continued until commercial purchase in May 2013 without a specified timeframe. Those limitations prevent a more detailed comparison of event frequency over time.

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

The posted results describe safety events recorded among participants in a single-group study of topical cysteamine. They report 11 participants with serious adverse events and 257 with non-serious adverse events among 328 at risk. The serious-event list includes several terms and associated percentages. These figures convey the registry’s recorded safety information; they do not establish that the intervention caused any particular event.

The study has no comparator group, so its results cannot show how the observed counts differ from those under another treatment or no treatment. The counts also do not account for each participant’s amount of exposure or observation time. A participant observed for a different duration may have had a different opportunity for an event to be detected, but participant-level durations are not reported. The serious and non-serious counts should not be added because overlap is not described.

The registry’s summary states that the study initially examined whether drops could prevent or reduce corneal cystine crystals. It also states that efficacy had been established and that the primary outcome was subsequently changed to safety after a New Drug Application was filed. Those are statements in the registry record, not efficacy results that can be independently evaluated from the posted safety counts. No efficacy analysis is reported in the registry record.

The record can support a narrow descriptive reading: safety events were recorded in this treatment group over the study’s stated collection period. It cannot, from the information posted, quantify a causal treatment effect, establish comparative safety, or determine how the event profile would differ in another population. Missing details about event definitions, follow-up, analysis methods, and missing observations further limit the precision of interpretation. The results are best considered in the context of those design and reporting constraints.

14. Why This Trial Matters Statistically

This record illustrates why a trial’s purpose and its posted primary outcome need to be read together. The registry describes an initial focus on corneal cystine crystals and later identifies safety evaluation as the post-hoc primary outcome after efficacy had been established and a New Drug Application filed. A safety count is not an efficacy estimate, so the two kinds of claims should not be conflated.

It also highlights the limits of descriptive safety data without a comparator. Counts can show how many participants in a study group were recorded with an event category, but they cannot by themselves distinguish intervention-related events from events that would occur for other reasons. The lack of participant-level follow-up information also matters when the collection period is long and the outcome description does not specify one uniform timeframe.

The study’s single-group design is important when interpreting any comparative language. Sample-size planning illustrations based on hypothetical two-arm assumptions may help describe what a different study design would require, but they do not alter the actual design or create a treatment contrast. Such planning figures are not observed power and are not evidence that the posted safety counts demonstrate a treatment effect.

Finally, registry records are most informative when outcome definitions, analytic populations, missing-data procedures, and statistical methods are documented. Those details are not reported for this outcome. Recognizing what the registry does and does not report helps separate a useful description of accumulated safety observations from claims that require comparative or causal evidence.

15. 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 single-group design is the main limitation for interpreting the safety counts. With no concurrent control group, the registry cannot support a like-for-like comparison or distinguish events associated with treatment from events related to cystinosis, other care, or background occurrence. The record lists masking as none and allocation as not applicable, so the posted findings do not have the protections of a masked randomized comparison.

The timeframe is not fully clear. The outcome is labeled any time point up to 27 years, while its description says no specified timeframe was used and that safety collection continued until the drug became available for commercial purchase in May 2013. Participant-specific follow-up and exposure durations are not reported. Neither rates accounting for time nor a common follow-up window can be evaluated from these aggregate counts.

The registry does not report detailed adverse-event definitions, event dates, severity criteria, relatedness assessments, or whether individual participants could contribute to both serious and non-serious categories. It also does not explain the denominator used for the percentages attached to the listed serious-event terms. These gaps constrain interpretation of the event categories and term-level figures.

Further limitations include absent analysis methods, confidence intervals, missing-data procedures, analysis-population definitions, and sensitivity analyses. The record also does not report detailed participant characteristics or eligibility criteria, which limits assessment of who the results may describe. The primary safety measure is identified as post hoc, but the record does not document the original analysis plan or the process behind that change. These limitations restrict the conclusions available from the registry summary; they do not establish that the underlying study collected no further information.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Cysteamine Eye Drops to Treat Corneal Crystals in Cystinosis — NCT00001213.
• Registry record last updated 2014-07-22; status last verified 2014-06.
• Results first posted 2014-07-22.
• 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.

19. Record Summary

The registry record describes a completed Phase 2, single-group study sponsored by the National Eye Institute, with 328 participants enrolled and cystinosis listed as the condition. Cysteamine topical solution was administered hourly while awake in both eyes. The study began in 1986-04, with primary completion listed as 2013-07. Masking is listed as none, and allocation as not applicable.

The post-hoc primary outcome counts participants with serious and non-serious adverse events. The registry reports 11 of 328 at risk with serious events and 257 of 328 with non-serious events. The timeframe is described as any time point up to 27 years, with safety collection continuing until commercial purchase in May 2013. No analyses are reported in the registry record. Comparative effects, event attribution, analysis methods, missing-data handling, and participant-level follow-up are not reported in the registry record.