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 P03277 Dose Finding Study in Central Nervous System (CNS) Magnetic Resonance Imaging (MRI) (NCT02633501), a randomized phase 2 study. The primary outcome is Contrast to Noise Ratio (CNR) Difference.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02633501 |
| Phase | Phase 2 |
| Condition | CNS Lesion, Blood Brain Barrier Defect |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Enrollment | 312 |
| Arms in the primary outcome | 12 |
| Lead sponsor | Guerbet |
| Results posted | Yes |
2. Clinical Question
Population
Male and female participants with known or highly suspected focal disruption of the blood-brain barrier, including at least one expected enhancing lesion of minimum 5 mm, who were scheduled for routine contrast-enhanced CNS MRI. The registry reports an enrollment of 312.
Intervention
P03277-enhanced MRI at doses of 0.025, 0.05, 0.1, or 0.2 mmol/kg, depending on the study subset.
Comparator
Gadobenate dimeglumine-enhanced MRI at 0.1 mmol/kg. Each participant's two MRI examinations were performed in an assigned order, with P03277 and gadobenate dimeglumine given in sequence.
Primary question
How did the contrast-to-noise ratio (CNR) with each P03277 dose compare with CNR after gadobenate dimeglumine, as assessed by independent readers? The primary measure was the difference, calculated as P03277 CNR minus gadobenate dimeglumine CNR.
This was a phase 2 dose-finding study. Its central statistical question was whether CNR changed across the tested P03277 doses relative to the active comparator, and whether that pattern was consistent across readers. The endpoint was designed to quantify image contrast rather than to assess a clinical outcome. A positive difference indicates a higher estimated CNR for P03277 than for gadobenate dimeglumine; a negative difference indicates a lower estimated CNR.
The registry describes two study subsets. One subset tested P03277 doses of 0.05 or 0.1 mmol/kg, while the other tested 0.025, 0.05, 0.1, or 0.2 mmol/kg. Each subset included both possible examination sequences: P03277 followed by gadobenate dimeglumine, or gadobenate dimeglumine followed by P03277. The results are reported separately by dose and reader, rather than as one pooled estimate.
3. Trial Design
The registry identifies this as a randomized, quadruple-masked, parallel-group study with a crossover comparator procedure. It was a prospective, multicenter, international phase 2 trial sponsored by Guerbet. The study status is completed; the registry lists a start in 2016-06 and primary completion in 2018-01.
Participants were assigned to a sequence of two contrast-enhanced MRI examinations. One examination used P03277 at a dose determined by the study subset, and the other used gadobenate dimeglumine at 0.1 mmol/kg. The order varied: some participants received P03277 first, while others received the comparator first. Thus, the contrast-agent comparison was within-participant, although the registry describes the overall model as parallel and presents separate dose groups. Randomization and masking are stated in the registry; details of the randomization procedure and the roles included in quadruple masking are not reported in the registry record.
The study included participants with CNS lesions or blood-brain-barrier defects, as described in its listed conditions and eligibility summary. The primary outcome was a continuous imaging measure assessed during a 1 day procedure. Three independent blinded off-site readers evaluated lesion signal measurements. The registry specifies that the analysis used lesions detected by both MRIs after lesion tracking and that the maximum of 3 enhanced lesions was assessed.
The primary analysis population was the Per Protocol Set, described as including participants without major protocol deviations. A participant was analyzed once there was a matching CNR evaluation at both MRIs by at least 1 off-site reader. The full details of the protocol, dose allocation within subsets, and any additional design safeguards are not reported in the registry record. The reported enrollment is 312, but the analysis counts differ by dose and reader; the registry outcome record therefore does not present one common analysis sample size for all estimates.
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: Among the 312 screened subjects, 32 subjects were screening failure. 280 subjects were randomized: 28 in subset 1 and 252 in subset 2. Images from the subset 1 were used for training of off-site readers and validation of imaging protocol. Consequently, these subjects were not included in the efficacy analysis but only in the safety assessment.
Overall Study
| Stage | Subset 1 - P03277 0.05 mmol/kg/Gadobenate Dimeglumine | Subset 1 - P03277 0.1 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.025 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.05 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.1 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.2 mmol/kg/Gadobenate Dimeglumine |
|---|---|---|---|---|---|---|
| STARTED | 14 | 14 | 62 | 63 | 64 | 63 |
| COMPLETED | 12 | 11 | 55 | 59 | 53 | 48 |
| NOT COMPLETED | 2 | 3 | 7 | 4 | 11 | 15 |
Reasons for not completing this period:
| Reason | Subset 1 - P03277 0.05 mmol/kg/Gadobenate Dimeglumine | Subset 1 - P03277 0.1 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.025 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.05 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.1 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.2 mmol/kg/Gadobenate Dimeglumine |
|---|---|---|---|---|---|---|
| Adverse Event | 0 | 0 | 1 | 0 | 1 | 1 |
| Withdrawal by Subject | 2 | 1 | 2 | 2 | 1 | 3 |
| Various other reasons | 0 | 2 | 4 | 1 | 7 | 11 |
| Lost to Follow-up | 0 | 0 | 0 | 1 | 2 | 0 |
Analysis population for the primary outcome, as defined in the registry: The analysis was done for each of the 3 independent off-site readers using the Per Protocol Set that included all subjects without major protocol deviations. A subject was analyzed as soon as he had a matching evaluation of the CNR at both MRIs by at least 1 off-site reader. Therefore, the numbers of subjects analyzed by reader could be different.
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: Subjects of subset 1 were only included in the safety assessment. The Full Analysis Set included 240 subjects from subset 2 who had available results for CNR from both unenhanced and enhanced measurements derivable from one MRI by at least one off-site reader. Analyses using the FAS were based on the contrast agent allocated by the randomization.
Only the first 6 of 7 groups are shown.
| Characteristic | Unit | Subset 2 - P03277 0.025 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.05 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.1 mmol/kg/Gadobenate Dimeglumine | Subset 2 - P03277 0.2 mmol/kg/Gadobenate Dimeglumine | Subset 1 - P03277 0.05 mmol/kg/Gadobenate Dimeglumine | Subset 1 - P03277 0.1 mmol/kg/Gadobenate Dimeglumine |
|---|---|---|---|---|---|---|---|
| Participants analysed | 57 | 62 | 61 | 60 | 14 | 14 | |
| Age, Continuous | years | 54.8 (SD 10.9) | 54.0 (SD 14.8) | 54.2 (SD 14.4) | 52.1 (SD 13.7) | 54.9 (SD 13.3) | 56.1 (SD 16) |
| Sex: Female, Male: Female | Participants | 34 | 39 | 30 | 39 | 10 | 5 |
| Sex: Female, Male: Male | Participants | 23 | 23 | 31 | 21 | 4 | 9 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 1 | 0 | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 8 | 5 | 8 | 4 | 2 | 1 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 1 | 1 | 4 | 2 | 2 | 2 |
| Race (NIH/OMB): White | Participants | 47 | 55 | 49 | 54 | 10 | 11 |
| Race (NIH/OMB): More than one race | Participants | 0 | 1 | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 | 0 | 0 | 0 |
6. Primary Endpoint
The registry lists the primary outcome as Contrast to Noise Ratio (CNR) Difference, measured over 1 day procedure.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | Ratio |
| Endpoint classification | Continuous Endpoint |
The primary endpoint was the difference in contrast-to-noise ratio (CNR) between P03277 and gadobenate dimeglumine. CNR is a continuous imaging measure based on signal intensity and noise. The registry describes it as a quantitative parameter directly related to contrast-agent efficacy and says it was selected to support a precise dose determination. The difference was defined as CNR with P03277 minus CNR with gadobenate dimeglumine, so positive values favor P03277 on this imaging measure and negative values favor the comparator.
Signal intensity was measured in enhanced lesions by 3 independent blinded readers. The endpoint included only lesions detected by both MRI examinations after lesion tracking, with a maximum of 3 enhanced lesions considered. The registry reports least-squares means and 95% confidence intervals for each dose-reader combination. A confidence interval describes the uncertainty around an estimated difference under the analysis assumptions; it is not a range containing all participant-level results.
The measure addresses image contrast, not diagnostic accuracy, patient outcomes, or long-term safety. The relationship to contrast-agent efficacy is the rationale stated in the registry, but the endpoint alone does not establish broader clinical benefit. The primary endpoint timeframe is listed as a 1 day procedure.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The estimated CNR differences follow a dose-related pattern across the posted results: all three readers reported negative estimates at 0.025 mmol/kg, estimates close to zero at 0.05 mmol/kg, and positive estimates at 0.1 and 0.2 mmol/kg. For Reader 1, the estimates were -10.59, 2.18, 8.66, and 14.45 across those doses. For Reader 2, they were -14.14, 1.91, 19.38, and 38.37. For Reader 3, they were -23.96, 2.47, 29.23, and 51.96. Each value is the estimated difference in CNR between P03277 and gadobenate dimeglumine, not a raw CNR measurement.
The confidence intervals convey how precisely those differences were estimated. At 0.025 mmol/kg, every interval lies below zero: for example, Reader 1's interval is -14.59 to -6.58, and Reader 3's is -33.78 to -14.13. At 0.05 mmol/kg, each interval includes zero, with bounds of -2.67 to 7.03 for Reader 1, -7.28 to 11.1 for Reader 2, and -8.39 to 13.32 for Reader 3. At 0.1 mmol/kg and 0.2 mmol/kg, each reader's interval lies above zero. The intervals are wider for some reader-dose estimates than others, indicating that their precision varies.
The size of the estimates also varies across readers, particularly at the higher doses. At 0.1 mmol/kg, estimates range from 8.66 for Reader 1 to 29.23 for Reader 3. At 0.2 mmol/kg, they range from 14.45 for Reader 1 to 51.96 for Reader 3. This variation matters: the direction of the estimated difference is consistent among readers at each dose, but the estimated magnitude is not identical. The registry does not report a single pooled reader estimate.
The reported statistical tests are from mixed models. For the 0.05 mmol/kg results, the reported p-values are 0.186, 0.338, and 0.325 across the three readers. The tests reported for the 0.1 and 0.2 mmol/kg results have smaller p-values; the 0.025 mmol/kg entries have no p-value listed. These tests and the confidence intervals are evidence about estimated CNR differences under the analysis, not proof of a particular clinical consequence. The registry does not report an adjustment for multiple testing, so the collection of dose- and reader-specific tests should not be treated as if it were one prespecified, multiplicity-controlled conclusion.
Interpretation is limited to the analyzed participants and the imaging endpoint. The analysis used the Per Protocol Set and required matching evaluations from both MRI examinations. The registry's text describing the analysis population ends before the complete count information, and the number analyzed varies by dose and reader. The results therefore support a description of the posted imaging comparisons, but do not by themselves establish effects on clinical outcomes or characterize all randomized participants.
9. Statistical Methods Explained
The posted primary estimates are least-squares means from mixed models. A least-squares mean is a model-based estimate of the average outcome for a group or comparison, adjusted according to the terms included in the model. The registry names the mixed-model method but does not report the model formula, covariates, covariance structure, or other implementation details. Those details are important for reproducing the analysis and understanding exactly how the estimates account for the study's paired MRI examinations and other design features.
What does the difference estimate represent?
The contrast is defined as P03277 CNR minus gadobenate dimeglumine CNR. A value above zero therefore corresponds to higher estimated CNR with P03277, while a value below zero corresponds to lower estimated CNR. Because the endpoint is continuous, the estimate is expressed as a difference on the CNR scale. It is not a relative change, a probability, or a count of lesions. No new transformation or unit conversion is needed to interpret the sign.
The values are reported separately for each dose and reader. That structure reflects that three independent readers assessed the images; it also means that each displayed estimate answers a reader-specific comparison rather than representing an average across readers. Differences in estimates between readers may reflect variation in reader assessments, but the registry record does not provide enough detail to attribute that variation to a specific source.
How should the confidence intervals be read?
The reported intervals are 95% confidence intervals. In general, a confidence interval gives a range of values compatible with the data and the statistical model at its stated confidence level. A narrower interval indicates greater precision for the estimate, while a wider interval indicates more uncertainty. If an interval includes zero, the data are compatible with both a positive and a negative difference at that interval level. If the entire interval is above zero, the range is consistent with a positive difference; if it is entirely below zero, the range is consistent with a negative difference. These statements describe the statistical estimate and do not determine clinical importance.
For example, all three intervals for 0.05 mmol/kg include zero, while the intervals for 0.1 and 0.2 mmol/kg are wholly positive. The 0.025 mmol/kg intervals are wholly negative. The width of the interval is not the spread of individual participant responses. It describes uncertainty about the estimated comparison, conditional on the analysis.
What do the p-values add?
The registry reports p-values for most dose-reader comparisons under a mixed-model superiority analysis. A p-value measures how incompatible the observed data are with a specified null hypothesis, under the statistical model and its assumptions. It does not measure the probability that the null hypothesis is true, the size of the CNR difference, or the clinical importance of the result. The estimates and confidence intervals are therefore essential for interpreting direction, magnitude, and precision alongside the tests.
The registry does not state the exact null hypothesis and testing details beyond identifying the analyses as superiority analyses. A p-value is absent for the listed 0.025 mmol/kg reader results. Missing p-values should not be interpreted as evidence that no difference exists; they are simply not listed for those entries.
Why does the crossover procedure matter?
Each participant underwent MRI with both agents, in one of two sequences. Such paired measurements can make the within-participant comparison informative because each participant contributes assessments under both agents. At the same time, sequence, period, carryover, and within-participant correlation can matter in crossover analyses. The registry record does not report how these features were handled in the mixed model, so no specific adjustment or assumption about them can be asserted.
The study is also described as having a parallel model and dose groups. These descriptions refer to different aspects of the design: P03277 dose assignment is organized into groups, while the comparator MRI is part of a paired sequence. The registry does not give a more detailed design account sufficient to resolve all implementation details.
Why does over-dispersion matter in general?
Over-dispersion is a general concern when observed variation is greater than a statistical model expects. It can lead to standard errors that are too small if the model does not account for the extra variability. This endpoint is continuous, and the registry does not report an over-dispersion assessment or identify over-dispersion as an issue. The concept is relevant generally, but the posted record does not support a claim that it affected these results.
The outcome description specifies the Per Protocol Set, excluding participants with major protocol deviations, and a matching CNR evaluation at both MRIs by at least 1 reader. This selection defines who contributes to an estimate and may differ from the complete randomized population. The registry does not report a complete participant flow for this endpoint, a missing-data method, or sensitivity analyses. These omissions constrain assessment of how incomplete imaging, protocol deviations, or other missing evaluations could affect the estimates.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Mixed Models Analysis | Least Squares Mean -10.59 | not reported | not reported |
| Mixed Models Analysis | Least Squares Mean 2.18 | not reported | 0.186 |
| Mixed Models Analysis | Least Squares Mean 8.66 | not reported | < 0.001 |
| Mixed Models Analysis | Least Squares Mean 14.45 | not reported | < 0.001 |
| Mixed Models Analysis | Least Squares Mean -14.14 | not reported | not reported |
| Mixed Models Analysis | Least Squares Mean 1.91 | not reported | 0.338 |
The registry reports the primary results as least-squares means from mixed-model analyses, with reader-specific estimates and associated confidence intervals in the outcome record. It also lists mixed-model superiority analyses and p-values for most dose-reader comparisons. These are the sponsor-reported analyses and are the appropriate formal inferential results to distinguish from any simple re-calculation based only on posted summary values.
No discrepancy between the reported results and the simple estimates is flagged in the record. The displayed estimates and reported least-squares means are consistent in value. The posted confidence intervals provide information about precision, while the listed p-values provide the tests reported by the registry; neither should be replaced by an unadjusted calculation that omits the model's structure.
Important details of the sponsor's model are not reported in the registry record, including the full model specification and how the paired examination sequence was represented. Consequently, the mixed-model results can be described as reported, but cannot be independently reconstructed in full from the registry information alone. No additional sponsor interpretation or dose-selection conclusion is reported in the registry record.
11. Multiplicity, Missing Data and Analysis Populations
The primary endpoint is presented across multiple P03277 doses and three independent readers. The registry reports multiple dose-reader estimates and tests, but a multiplicity adjustment or testing hierarchy is not reported in the registry record. When many related statistical tests are conducted, the chance of at least one apparently unusual result can increase. This is a general statistical concern; it does not show that any particular reported result is false. The dose-specific estimates, confidence intervals, and reader consistency should therefore be considered together rather than treating each p-value as an isolated confirmatory finding.
The record identifies the Per Protocol Set as the analysis population. It includes participants without major protocol deviations, and a participant was analyzed when at least one off-site reader had matching CNR evaluations at both MRIs. The registry's population description is incomplete at the point where it begins to state the numbers included. The numbers analyzed differ across dose-reader results, and a full accounting of exclusions and missing evaluations is not reported in the registry record.
The posted record does not describe a missing-data strategy, imputation, sensitivity analysis, or analysis of all randomized participants for this endpoint. It also does not give the detailed prespecified statistical analysis plan, so it is not possible to establish from the record which comparisons were primary for inference or whether the displayed analyses were adjusted for all design features. These limits matter because available paired images and protocol adherence determine which participants contribute to the outcome estimates.
The registry reports 312 enrolled participants, but the endpoint-specific analysis counts are lower and vary by reader and dose. Enrollment and analyzed sample size answer different questions; enrollment alone does not establish how much endpoint information contributed to each estimate. No further reason for the differences in analysis counts is reported in the registry record.
12. 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 |
|---|---|---|---|
| P03277 0.025 mmol/kg | 0 | 56 | 0.0% |
| P03277 0.05 mmol/kg | 0 | 74 | 0.0% |
| P03277 0.1 mmol/kg | 1 | 70 | 1.4% |
| P03277 0.2 mmol/kg | 1 | 56 | 1.8% |
| Gadobenate Dimeglumine | 2 | 256 | 0.8% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| P03277 0.025 mmol/kg | 0 of 56 |
| P03277 0.05 mmol/kg | 0 of 74 |
| P03277 0.1 mmol/kg | 0 of 70 |
| P03277 0.2 mmol/kg | 0 of 56 |
| Gadobenate Dimeglumine | 0 of 256 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| P03277 0.025 mmol/kg | 14 of 56 |
| P03277 0.05 mmol/kg | 12 of 74 |
| P03277 0.1 mmol/kg | 19 of 70 |
| P03277 0.2 mmol/kg | 18 of 56 |
| Gadobenate Dimeglumine | 57 of 256 |
Most frequent serious adverse event terms
| Event term | Organ system | P03277 0.025 mmol/kg | P03277 0.05 mmol/kg | P03277 0.1 mmol/kg | P03277 0.2 mmol/kg | Gadobenate Dimeglumine |
|---|---|---|---|---|---|---|
| Seizure | Nervous system disorders | 0 of 56 | 0 of 74 | 0 of 70 | 0 of 56 | 1 of 256 |
| Meningioma surgery | Surgical and medical procedures | 0 of 56 | 0 of 74 | 0 of 70 | 0 of 56 | 1 of 256 |
| Blood creatinine increased | Investigations | 0 of 56 | 0 of 74 | 1 of 70 | 0 of 56 | 0 of 256 |
| Femoral neck fracture | Injury, poisoning and procedural complications | 0 of 56 | 0 of 74 | 0 of 70 | 1 of 56 | 0 of 256 |
Most frequent non-serious adverse event terms
| Event term | Organ system | P03277 0.025 mmol/kg | P03277 0.05 mmol/kg | P03277 0.1 mmol/kg | P03277 0.2 mmol/kg | Gadobenate Dimeglumine |
|---|---|---|---|---|---|---|
| Injection site pain | General disorders | 0 of 56 | 2 of 74 | 5 of 70 | 7 of 56 | 9 of 256 |
| Headache | Nervous system disorders | 4 of 56 | 3 of 74 | 1 of 70 | 0 of 56 | 7 of 256 |
| Nausea | Gastrointestinal disorders | 0 of 56 | 0 of 74 | 0 of 70 | 1 of 56 | 9 of 256 |
| Leukocyturia | Renal and urinary disorders | 1 of 56 | 2 of 74 | 1 of 70 | 1 of 56 | 5 of 256 |
| Urinary tract infection | Infections and infestations | 1 of 56 | 1 of 74 | 3 of 70 | 1 of 56 | 4 of 256 |
| Haemoglobinuria | Renal and urinary disorders | 1 of 56 | 0 of 74 | 1 of 70 | 0 of 56 | 5 of 256 |
| Proteinuria | Renal and urinary disorders | 1 of 56 | 2 of 74 | 1 of 70 | 1 of 56 | 2 of 256 |
| Injection site coldness | General disorders | 1 of 56 | 0 of 74 | 1 of 70 | 0 of 56 | 4 of 256 |
| Diziness | Nervous system disorders | 0 of 56 | 0 of 74 | 0 of 70 | 2 of 56 | 4 of 256 |
| Bilirubinuria | Renal and urinary disorders | 0 of 56 | 2 of 74 | 0 of 70 | 0 of 56 | 4 of 256 |
Serious events with the highest arm-level rate: femoral neck fracture (1.8%), blood creatinine increased (1.4%), seizure (0.4%), meningioma surgery (0.4%).
The registry lists serious adverse events by treatment group, including the number at risk and the number affected. The reported serious adverse event counts are 0 of 56 for P03277 0.025 mmol/kg, 0 of 74 for P03277 0.05 mmol/kg, 1 of 70 for P03277 0.1 mmol/kg, 1 of 56 for P03277 0.2 mmol/kg, and 2 of 256 for gadobenate dimeglumine. No deaths are listed in any group.
These are descriptive counts. They do not, on their own, establish whether one treatment has a safer profile than another. The groups differ in size, and the registry record does not provide a comparative statistical analysis of serious adverse events. The duration and completeness of safety follow-up are not reported in the registry record.
The record also lists serious-event terms with percentages: femoral neck fracture, 1.8%; blood creatinine increased, 1.4%; seizure, 0.4%; and meningioma surgery, 0.4%. Those terms are not presented as a complete clinical interpretation of all adverse events. Other adverse events are reported as affecting 14, 12, 19, and 18 participants in the four P03277 dose groups, and 57 in the gadobenate dimeglumine group. These counts should be read in their listed groups and not treated as adjusted comparisons.
13. What the Result Does, and Does Not, Mean
The results describe a dose- and reader-specific pattern in an imaging measure: estimates are negative at 0.025 mmol/kg, near zero at 0.05 mmol/kg, and positive at 0.1 and 0.2 mmol/kg for all three readers. The confidence intervals likewise change from wholly negative at the lowest dose, to including zero at 0.05 mmol/kg, to wholly positive at the higher doses. This is evidence about CNR differences in the analyzed MRI examinations under the reported analysis.
That pattern does not, by itself, establish a clinically meaningful improvement in diagnosis, management, symptoms, or health outcomes. The primary endpoint is an imaging measure, and the registry does not report a clinical outcome linked to these CNR differences. Nor does the result establish that any particular dose is the safest or most appropriate dose. The registry describes the study purpose as dose finding, but does not report a final dose-selection conclusion in the posted results.
The results also do not establish that the estimates apply to every participant enrolled. The endpoint analysis used a Per Protocol Set and required matched CNR assessments at both MRI examinations for at least one reader. The number analyzed varies across dose and reader, and the full accounting of missing evaluations is not reported in the registry record. The population-level interpretation should remain tied to the participants and images included in each analysis.
Confidence intervals and p-values answer different statistical questions. The intervals indicate uncertainty and the range of differences compatible with the fitted analysis, while p-values quantify evidence against a null hypothesis under that analysis. Neither substitutes for a judgment about clinical relevance. The statistical plan and multiplicity handling are not reported in the registry record, limiting how strongly the set of dose- and reader-specific tests can be treated as confirmatory.
In short, the posted findings support a description of reader-specific CNR differences across tested doses relative to gadobenate dimeglumine. They do not independently establish effects beyond that imaging endpoint, resolve every design and analysis uncertainty, or provide medical advice.
14. Why This Trial Matters Statistically
This trial illustrates a common statistical challenge in dose-finding research: the outcome is quantitative, multiple doses are compared with an active comparator, and the measurement is repeated across independent readers. Looking only at one estimate would hide the dose pattern and the reader-to-reader variation. Reading estimates with their confidence intervals makes both direction and precision visible.
The within-participant MRI sequence offers a paired comparison of contrast agents, while dose assignment is organized into groups and subsets. That combination makes the analysis dependent on appropriate handling of repeated measurements and examination sequence. The registry identifies a mixed-model analysis, but does not report the full model specification. This is a useful reminder that naming a model class is not the same as publishing enough detail to reproduce an analysis.
Multiplicity is also important. The outcome record contains comparisons for several doses across three readers. A collection of nominal tests can give a different inferential picture from a prespecified testing strategy that accounts for multiple comparisons. Since the registry does not report multiplicity handling, the dose-response pattern, estimated magnitudes, and intervals deserve attention alongside individual tests.
Finally, the endpoint demonstrates the distinction between a biomarker or imaging measure and a clinical outcome. CNR was selected as a quantitative measure related to contrast-agent efficacy, but the posted endpoint only measures the imaging comparison. Sound interpretation keeps the conclusions close to that measure and recognizes that precision, reader variation, analysis-population selection, and missing information all shape what can be learned from the record.
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 registry gives limited detail for independently evaluating the primary model. Although it identifies a mixed-model analysis and reports least-squares means, the formula, covariates, covariance structure, and handling of examination order are not reported in the registry record. Those details are particularly relevant because participants received two MRI examinations and the comparison is paired.
The results are reported separately for three readers, and the estimated magnitudes vary across readers, especially at the higher doses. The record does not report a pooled reader estimate or a formal analysis of reader agreement. The reader-specific results are informative, but they do not establish how reproducible the measurements would be across other readers or settings.
The endpoint analysis uses the Per Protocol Set rather than presenting a complete analysis of all randomized participants. A matching evaluation at both MRIs from at least one reader was required, and the analysis counts vary by dose and reader. The full numbers excluded, reasons for missing evaluations, and sensitivity analyses are not reported in the registry record.
The registry does not report multiplicity control, a complete statistical analysis plan, or a sample-size justification. It also does not state the clinical importance threshold for a CNR difference. Consequently, statistical evidence of a difference cannot be equated with evidence that the difference is clinically meaningful. The outcome is an imaging measure observed during a 1 day procedure; longer-term clinical effects and safety follow-up are not reported in the registry record.
Finally, while the study is described as dose finding, no dose-selection decision is reported in the outcome information. The results should therefore be interpreted as the reported dose- and reader-specific CNR comparisons rather than as a complete account of the clinical or regulatory decision-making process.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: P03277 Dose Finding Study in Central Nervous System (CNS) Magnetic Resonance Imaging (MRI) — NCT02633501.
- Registry record last updated 2021-10-25; status last verified 2021-10.
- Results first posted 2019-05-08.
- 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.
19. Record Summary
NCT02633501 is a completed phase 2 randomized, quadruple-masked study sponsored by Guerbet, with 312 participants enrolled. It compared P03277-enhanced MRI at doses ranging from 0.025 to 0.2 mmol/kg, according to subset, with gadobenate dimeglumine at 0.1 mmol/kg.
The primary outcome was the CNR difference, calculated as P03277 minus gadobenate dimeglumine, and assessed separately by three independent blinded readers. Reported estimates were negative at 0.025 mmol/kg, close to zero at 0.05 mmol/kg, and positive at 0.1 and 0.2 mmol/kg for all readers. Analysis used the Per Protocol Set and mixed models; important model details, multiplicity handling, and complete missing-data accounting are not reported in the registry record.
The endpoint is an imaging measure, so the findings describe CNR comparisons rather than clinical outcomes. Serious adverse event counts and selected event terms are also listed, but no comparative safety analysis is reported.