This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Assess the Prognostic Usefulness of Flutemetamol (18F) Injection for Identifying Subjects With Amnestic Mild Cognitive Impairment Who Will Convert to Clinically Probable Alzheimer's Disease (NCT01028053), a na phase 3 study. The primary outcome is Hazard Ratio (HR) by PET Scan Readers for Conversion to Probable Alzheimer's Disease Based on Visual Image Interpretation..
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01028053 |
| Phase | Phase 3 |
| Condition | Mild Cognitive Impairment, Alzheimer's Disease |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 365 |
| Arms in the primary outcome | n/a |
| Lead sponsor | GE Healthcare |
| Results posted | Yes |
2. Clinical Question
Population
The study concerned people with amnestic mild cognitive impairment (aMCI) being assessed for later conversion to clinically probable Alzheimer’s disease. The registry lists enrollment of 365. For the primary endpoint, the record says that 8 subjects who withdrew before the first Clinical Adjudication Committee evaluation were excluded and that 224 subjects were included from the stated 232-subject analysis population.
Intervention
Participants received Flutemetamol (18F) Injection, a PET tracer. The prognostic assessment used visual interpretation of PET scans by independent readers, with interpretations described in terms of normal or abnormal amyloid beta status.
Comparator
This was not a randomized comparison between assigned treatment arms. The relevant contrast was between participants classified as normal and abnormal by PET interpretation. Those categories arise from scan readings; they are not like-for-like control groups created by random allocation.
Primary question
The primary endpoint was the hazard ratio for conversion to probable Alzheimer’s disease based on PET scan readers’ visual interpretations, followed for up to 36 months after flutemetamol administration. The registry defines the comparison as the hazard rates in the normal and abnormal groups. It does not report a numerical hazard ratio, confidence interval, or p-value, so the magnitude and precision of the prognostic association cannot be assessed from the registry record.
3. Trial Design
This was a phase 3, completed, open-label study of Flutemetamol (18F) Injection. The registry identifies the study as single-group, with no masking and no randomized allocation. Its listed enrollment is 365. The registry dates the study start to 2009-12 and primary completion to 2014-01.
The study question was prognostic rather than a comparison of treatment effects. After tracer administration, PET images were interpreted visually by independent readers. The endpoint description distinguishes normal scans, negative for amyloid beta, from abnormal scans, positive for amyloid beta, and relates those readings to later conversion to probable Alzheimer’s disease. The comparison therefore concerns outcomes among groups defined by imaging interpretation, not outcomes between groups assigned different therapies.
The registry describes the primary outcome as a hazard ratio for conversion, with follow-up up to 36 months after administration. It also lists a secondary outcome concerning the number of normal and abnormal subjects who convert to probable Alzheimer’s disease within the follow-up period. The outcome description does not give the numerical primary result, event counts by scan category, or a detailed account of the secondary outcome results. The registry lists 2 outcome measures and 0 analyses.
The primary endpoint population text states that 8 subjects withdrew before the first Clinical Adjudication Committee evaluation and were not included in the analysis; it reports 224 subjects included from a stated 232-subject group. The registry does not explain how this analysis population relates to all enrolled participants beyond those statements. Other details of the study’s statistical design, including a full analysis plan and procedures for handling follow-up, are not reported in the registry record.
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
| Stage | Flutemetamol (18F) Injection |
|---|---|
| STARTED | 365 |
| COMPLETED | 232 |
| NOT COMPLETED | 133 |
Reasons for not completing this period:
| Reason | Flutemetamol (18F) Injection |
|---|---|
| Screen Failure | 113 |
| Patient Request | 11 |
| Protocol Violation | 1 |
| Physician Decision | 2 |
| Technical Problem | 1 |
| Terminated by Sponsor | 1 |
| Various reasons | 4 |
Analysis population for the primary outcome, as defined in the registry: Eight Subjects who withdrew prior to the first Clinical Adjudication Committee (CAC) evaluation are not included in the analysis. (232 - 8 = 224 Subjects included).
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.
| Characteristic | Unit | Flutemetamol (18F) Injection |
|---|---|---|
| Participants analysed | 232 | |
| Age, Continuous | years | 71.1 (SD 8.62) |
| Age, Categorical: <=18 years | Participants | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 63 |
| Age, Categorical: >=65 years | Participants | 169 |
| Sex: Female, Male: Female | Participants | 118 |
| Sex: Female, Male: Male | Participants | 114 |
| Region of Enrollment: United States | participants | 109 |
| Region of Enrollment: Belgium | participants | 13 |
| Region of Enrollment: Denmark | participants | 8 |
| Region of Enrollment: Finland | participants | 15 |
| Region of Enrollment: Sweden | participants | 7 |
| Region of Enrollment: United Kingdom | participants | 80 |
6. Primary Endpoint
The registry lists the primary outcome as Hazard Ratio (HR) by PET Scan Readers for Conversion to Probable Alzheimer's Disease Based on Visual Image Interpretation., measured over Up to 36 months post flutemetamol administration.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | Ratio of visual interpretations |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is a hazard ratio based on visual PET interpretation by independent readers. It compares the hazard rates of conversion to probable Alzheimer’s disease between the group classified as normal, meaning negative for amyloid beta, and the group classified as abnormal, meaning positive for amyloid beta. The stated timeframe is up to 36 months after flutemetamol administration.
A hazard ratio expresses the relative event rates over follow-up among people who have not yet experienced the event. It is not itself the proportion of participants who converted, nor does it directly state when an individual will convert. The registry describes a hazard ratio of 1 as the null value for equal rates. It also states that values increasing above 1 correspond to increased chances of probable Alzheimer’s disease. The endpoint is intended to assess whether scan interpretation distinguishes groups with different subsequent conversion rates.
For interpretation, the estimate needs to be considered alongside its uncertainty and the definitions of the groups and event. However, the registry record does not report a numerical hazard ratio, confidence interval, p-value, or group-specific event counts. Consequently, the endpoint’s direction, magnitude, and precision cannot be determined from the registry record. The presence of an endpoint definition alone does not establish how strongly scan interpretation was associated with conversion in the observed data.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record establishes what the primary endpoint was intended to measure, but it does not provide the numerical hazard-ratio result. It therefore does not permit an assessment of the size or direction of the observed association between PET classification and conversion to probable Alzheimer’s disease. No confidence interval or p-value for the endpoint is reported in the registry record. Without an estimate and its uncertainty, it is not possible to judge how precisely the study estimated the association or whether the observed data were compatible with a range of clinically different associations.
The endpoint compares scan-defined groups: normal and abnormal visual interpretations. That is a prognostic comparison, not a randomized treatment contrast. A difference in conversion rates between such groups, if reported, would describe an association between PET classification and subsequent outcome in the study population. It would not by itself show that the tracer or scan interpretation caused a change in disease progression. Participants were not allocated to positive or negative scan results, and the registry records no randomized allocation.
The endpoint description gives a timeframe of up to 36 months after administration. The registry does not report the number of conversions in each scan category, the timing or distribution of events, or the handling of participants with incomplete follow-up. Those details matter because a hazard ratio uses event timing and the time people remain under observation, rather than only a simple count of conversions. The posted information does not allow a reader to check how the stated endpoint was estimated.
The primary endpoint text reports that 8 subjects who withdrew before the first Clinical Adjudication Committee evaluation were excluded, and that 224 subjects were included from the stated 232-subject group. These registry entries describe the analysis population as recorded, but they do not explain why those withdrawals occurred or whether other participants had incomplete outcome information. A statistical interpretation should not assume that excluded or unobserved participants had the same outcomes as those included.
The registry lists 2 outcome measures and 0 analyses. It does not report the primary estimate, secondary outcome results, reader-specific estimates, or any analysis plan. Thus there is no basis in the registry record for comparing readers, assessing consistency across readings, or determining how the secondary outcome relates to the primary endpoint. The appropriate conclusion is limited: the registry records a planned prognostic endpoint and an analysis population statement, while the numerical outcome evidence needed to evaluate the association is not reported in the registry record.
Any planning calculations based on an observed effect should also be interpreted separately from the trial’s observed findings. Such calculations are design illustrations under stated assumptions, not evidence of the strength or reliability of the observed association. They are not observed power and cannot replace the missing estimate and uncertainty for the primary endpoint.
9. Statistical Methods Explained
Several statistical ideas help clarify what the endpoint would mean if its estimate and supporting results were reported. The registry’s primary measure is a hazard ratio, derived from time-to-event information. The event is conversion to probable Alzheimer’s disease, and the comparison is between groups defined by visual PET interpretation as normal or abnormal. Follow-up is described as extending up to 36 months after flutemetamol administration.
What does a hazard ratio describe?
In general, a hazard is the rate at which an event occurs at a particular point in follow-up among people who have not yet experienced that event. A hazard ratio compares those rates between groups over the period being analyzed. It uses information about both whether an event occurred and when it occurred. It should not be read as a ratio of the overall proportions who experienced the event, and it does not give an individual’s probability or timing of conversion.
The registry describes the null value as a hazard ratio of 1, corresponding to equal hazard rates in the two groups. It also says that an increase above 1 corresponds to increased chances of probable Alzheimer’s disease. The record does not specify the precise coding order for the groups in the numerical ratio, so the direction should be interpreted using the registry’s stated convention rather than inferred from an absent estimate. No numerical hazard ratio is reported in the registry record.
Why does uncertainty matter?
In general, a confidence interval expresses the range of effect estimates that are reasonably compatible with the data and the statistical model at a stated confidence level. A narrow interval indicates greater precision than a wide interval, while an interval spanning effects in different directions or of different practical sizes may leave substantial uncertainty. A p-value addresses compatibility with a specified null model under assumptions; it does not describe the magnitude of an association or the probability that a hypothesis is true. The registry record does not report a confidence interval or p-value for the primary endpoint, so neither precision nor statistical compatibility with the null can be evaluated.
For a time-to-event result, the number and timing of events, length of observation, and censoring of follow-up all affect the estimate and its uncertainty. Censoring occurs when a participant’s event time is not observed through the full period, but some follow-up information is still available up to the last known observation. In general, time-to-event methods require assumptions about how censoring relates to event risk. The registry does not report event counts by PET classification, censoring details, or the assumptions and model used for the primary analysis.
What does over-dispersion mean for count outcomes?
In general, count data may vary more than a simple statistical model expects. This is called over-dispersion, and it can make standard errors too small if the extra variation is ignored. A count outcome can also depend on the amount of time participants are observed and on whether repeated events are possible. The registry lists a secondary outcome about the number of normal and abnormal subjects who convert within follow-up, but it does not report the counts, model, or analysis details. The presence of a count outcome does not establish that a particular count model was used.
How should the planning illustrations be read?
The planning assumptions describe calculations using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. For time-to-event calculations, the assumptions specify total events using the Schoenfeld approach; for other calculations, results are expressed as participants per arm. These are planning illustrations based on an effect, not observed power and not a record of the study’s actual design.
This distinction is particularly important because the trial is identified as single-group with no randomized allocation. The normal and abnormal categories are based on PET readings, not assigned arms. Equal allocation in a planning calculation is a hypothetical assumption and should not be mistaken for the study’s enrollment pattern or an actual randomized comparison. The registry record does not report the trial’s own sample-size justification or statistical plan, nor does it provide numerical planning outputs in the registry information described for the endpoint.
Interpretation of the primary endpoint would also depend on how the reader assessments were incorporated. The registry says visual interpretations were made by independent readers but does not report the number of readers, how disagreements were resolved, whether readings were combined, or whether estimates were calculated separately for each reader. These choices can affect the definition and reproducibility of the scan categories. The record does not state whether a proportional-hazards model was used or whether that assumption was assessed. In general, a hazard ratio may be difficult to summarize with one number if relative hazards change materially over follow-up.
These general methods explain what would be needed for a fuller statistical assessment; they do not supply missing results. The posted registry information does not report a numerical primary estimate, its uncertainty, the model specification, or the diagnostic checks. Those omissions limit interpretation even though the endpoint and some population details are described.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 2 outcome measures and 0 analyses. It identifies a primary hazard-ratio outcome and a secondary outcome concerning the number of normal and abnormal subjects who convert during follow-up. However, it does not report a statistical analysis plan, a hierarchy among secondary outcomes, or any procedures for controlling the risk of false-positive findings across multiple outcomes or reader assessments. Whether multiplicity was addressed is not reported in the registry record.
The primary endpoint population statement says that 8 subjects withdrew before the first Clinical Adjudication Committee evaluation and were excluded, and that 224 subjects were included from a stated 232-subject population. The record does not explain the reasons for these withdrawals, describe any further exclusions, or state whether other participants lacked follow-up or outcome measurements. It also does not report a method for handling missing data, incomplete follow-up, or censoring. No assumptions about the outcomes of excluded participants should be made from the information recorded.
The registry does not identify an intention-to-treat, per-protocol, or other named analysis population for this prognostic study. The information about the primary analysis population is limited to the withdrawal and inclusion statement. How readers’ interpretations were combined, whether all readers’ assessments contributed to the endpoint, and how disagreements were handled are not reported in the registry record.
Interim analyses and stopping rules are also not reported in the registry record. The absence of these details does not establish that a procedure was or was not used; it means the registry information does not describe one. These gaps leave uncertainty about how the reported endpoint was selected and analyzed, in addition to the lack of a numerical result.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Flutemetamol (18F) Injection | 1 | 232 | 0.4% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Flutemetamol (18F) Injection | 0 of 232 |
Most frequent serious adverse event terms
| Event term | Organ system | Flutemetamol (18F) Injection |
|---|---|---|
| Anaphylactoid Reaction | Immune system disorders | 1 of 232 |
Serious events with the highest arm-level rate: anaphylactoid reaction (0.4%).
The serious adverse event figures describe events recorded for the Flutemetamol (18F) Injection group. The registry lists 232 subjects at risk for serious adverse events and 1 subject affected. The listed top serious term is Anaphylactoid Reaction, with 0.4% recorded. These are safety counts and a term frequency; they do not provide a comparison with a randomized control group.
The registry also lists 232 subjects at risk for other adverse events and 0 subjects affected in that category. The death count is not reported in the registry record. The serious event count should not be combined with the other-event count, and the other-event category should not be treated as an alternative estimate of the serious-event frequency.
These figures do not establish whether an event was caused by the tracer, nor do they characterize event timing, severity beyond the serious-event classification, duration, or outcomes. The record does not provide further details about the event or a comparative safety analysis. Because the study has no randomized allocation and no like-for-like control arm recorded, these figures should be read as reported events in the study group, not as evidence of a relative risk against a control treatment.
12. What the Result Does, and Does Not, Mean
The study was designed to assess whether visual interpretation of flutemetamol PET scans could provide prognostic information about conversion from amnestic mild cognitive impairment to probable Alzheimer’s disease. Its primary endpoint compares time-to-conversion hazards between people classified as having normal and abnormal scans. This framing concerns the relationship between an imaging classification and a subsequent clinical outcome.
The registry record does not report the numerical hazard ratio, a confidence interval, a p-value, or event counts by scan category. It therefore does not show how large the association was, how precisely it was estimated, or whether the observed data were consistent with equal hazards. A stated endpoint and a completed study are not substitutes for a numerical outcome result.
Even if a hazard-ratio estimate were available, a prognostic association would not by itself establish causation. The scan categories were not assigned at random, and the study is recorded as single-group with no randomized allocation. A difference in outcomes between scan-defined groups could indicate that the classification is associated with later conversion in the population studied, but it would not show that the tracer or the scan result changed the underlying course of disease.
The registry identifies visual interpretation by independent readers, but does not report how reader results were combined or how classification disagreements were handled. It also does not report a model specification or the uncertainty around the endpoint. These omissions prevent assessment of reproducibility and statistical precision from the registry information.
The analysis population statement records exclusions for withdrawal before the first Clinical Adjudication Committee evaluation and gives the included population size. It does not explain the broader pattern of missing follow-up or establish how excluded participants would have been classified or observed. As a result, the registry information does not support conclusions about whether the analysis population represents all enrolled participants.
In short, the record documents the study question, design features, endpoint definition, selected population information, and adverse-event figures. It does not report the primary numerical result needed to determine what the observed data show about prognostic performance. No treatment recommendation or conclusion about clinical effectiveness follows from the endpoint description alone.
13. Why This Trial Matters Statistically
This study is statistically notable because it asks a prognostic question rather than testing a randomized treatment contrast. The principal comparison is based on normal or abnormal PET interpretations, followed by observation of conversion to probable Alzheimer’s disease. That distinction affects what can be concluded: the endpoint concerns association and prediction, not the causal effect of assigning a therapy.
The use of a hazard ratio makes event timing relevant. A time-to-event measure can use information from participants with different lengths of follow-up, but its interpretation depends on event timing, censoring, and the model used. The registry specifies follow-up up to 36 months but does not report event counts, censoring information, model assumptions, or the numerical estimate. The endpoint can therefore be described, but its statistical behavior and precision cannot be checked from the registry record.
Independent visual readings also make classification procedures important. Reader variation, agreement, and the handling of discordant interpretations can influence how participants are grouped and how stable a prognostic association appears. The registry records independent readers but does not describe the number of readers, their agreement, or how their interpretations were used in the analysis. These are relevant statistical details for an imaging-based prognostic study.
The record also illustrates why reporting an endpoint definition is different from reporting a result. Without an estimate and uncertainty interval, a reader cannot distinguish a strong, precise association from a weak or imprecise one. Listing an outcome and an analysis population does not resolve that uncertainty. For this study, the registry’s account supports understanding the intended question and some features of the design, while leaving the main quantitative conclusion unavailable.
Finally, planning calculations based on an observed effect need to be kept separate from evidence about the study’s findings. Assumptions such as equal allocation and a chosen fraction of an effect are hypothetical inputs for planning, not descriptions of this single-group study’s actual allocation or observed power. This separation is essential when interpreting prognostic studies whose categories are defined by measurement rather than random assignment.
14. 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 main limitation for interpreting the study statistically is that the registry record does not report a numerical primary hazard ratio or its uncertainty. It also does not report group-specific conversion counts, so the primary endpoint cannot be reconstructed or checked from the registry information. The secondary outcome is listed, but its numerical results are not reported in the registry record.
The endpoint is based on PET classifications made through visual interpretation by independent readers. The record does not describe the number of readers, their agreement, how disagreements were resolved, or how multiple readings were incorporated. Without those details, the reproducibility of the scan categories and their relationship to the endpoint cannot be assessed from the registry entry.
The registry does not report the primary statistical model, its assumptions, diagnostics, or the handling of censoring. Although the timeframe is specified as up to 36 months, the record does not describe individual follow-up patterns or how incomplete outcome information was treated. The analysis population statement identifies withdrawals before the first Clinical Adjudication Committee evaluation, but gives no further account of missing data or the representativeness of the included population.
Because this is a single-group study without randomized allocation, the normal and abnormal PET groups are not randomized treatment arms. Differences between them, if reported, would not by themselves establish a causal effect of flutemetamol or a treatment effect. Finally, the registry does not report its own sample-size justification, a statistical plan, interim analyses, or multiplicity procedures. These details are not reported in the registry record and should not be inferred from the planning assumptions.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Assess the Prognostic Usefulness of Flutemetamol (18F) Injection for Identifying Subjects With Amnestic Mild Cognitive Impairment Who Will Convert to Clinically Probable Alzheimer's Disease — NCT01028053.
- Registry record last updated 2014-09-11; status last verified 2014-09.
- Results first posted 2014-09-11.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry identifies NCT01028053 as a completed, phase 3, open-label, single-group study of Flutemetamol (18F) Injection, with no masking and no randomized allocation. Its listed enrollment is 365. The study start is recorded as 2009-12 and primary completion as 2014-01.
The primary outcome is a hazard ratio for conversion to probable Alzheimer’s disease based on visual PET interpretation, with follow-up up to 36 months after administration. The endpoint compares normal and abnormal scan classifications. The registry states that 8 subjects withdrew before the first Clinical Adjudication Committee evaluation and were excluded, and that 224 subjects were included from a stated 232-subject population.
The record lists 2 outcome measures and 0 analyses. It does not report a numerical primary hazard ratio, confidence interval, p-value, or conversion counts by scan classification. For safety, it records 1 subject affected by a serious adverse event among 232 at risk, with Anaphylactoid Reaction listed at 0.4%; deaths are not reported in the registry record. The statistical plan and missing-data procedures are also not reported.