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 Evaluation of the Safety, Tolerability, Pharmacokinetics (PK) and Effects on Liver Iron Concentration of ICL670 Relative to Deferoxamine(DFO). (NCT01090323), a non-randomized phase 2 study. The primary outcome is Number of Participants With Adverse Events After Start of ICL670.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01090323 |
| Phase | Phase 2 |
| Condition | Sickle Cell Disease |
| Allocation | Non-Randomized |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 185 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Novartis Pharmaceuticals |
| Results posted | Yes |
2. Clinical Question
Population
People with sickle cell disease and transfusional hemosiderosis who had previously received ICL670 or deferoxamine for 52weeks in the core study. The safety analysis set comprised participants who received at least one dose of ICL670 during the core or extension phase. The registry reports enrollment of 185, but does not describe how enrollment relates to each posted safety group beyond the analysis population description.
Intervention
ICL670, described in the official title as repeated doses of 5-30mg/kg/Day. The extension evaluates safety, tolerability, pharmacokinetics, and effects on liver iron concentration. The registry record does not give individual dosing histories or the results for pharmacokinetics or liver iron concentration.
Comparator
The posted groups are labelled ICL670 and Crossover. This is a non-randomized, single-group extension rather than a like-for-like controlled comparison. Deferoxamine appears in the description of prior core-study treatment, but the record does not provide a concurrent randomized deferoxamine control for the extension results.
Primary question
How many participants experienced adverse events or death after starting ICL670, according to the stated safety endpoint and its 0 - 60 months timeframe? The posted record gives counts of participants with serious and other adverse events by group. It does not report a comparative treatment effect, uncertainty interval, or a separate death count.
The key interpretive issue is the difference between describing events during follow-up and attributing those events to a treatment. The counts document adverse events among participants in the posted groups, but this extension was open label and non-randomized. The groups therefore should not be treated as exchangeable treatment arms. Their labels alone do not establish that participants had similar characteristics, follow-up, or prior treatment histories.
The record describes an extension to an earlier randomized, multicenter, phase II study. The extension itself is identified as non-randomized, with no masking and a single-group model. The official title also says the extension was prolonged to 4-years, while the primary endpoint timeframe is stated as 0 - 60 months. These descriptions establish the planned context and recorded timeframe, not the duration each participant was observed. Individual follow-up duration is not reported in the registry record.
3. Trial Design
This is a phase II, non-randomized, open-label extension study of ICL670 in sickle cell disease with transfusional hemosiderosis. The registry identifies a single-group model, no masking, and enrollment of 185. Its official title describes a one-year extension to a randomized, multicenter study and notes an amendment extending the extension to 4-years. The primary outcome timeframe is 0 - 60 months. These study-level descriptions do not establish how long each participant contributed safety follow-up.
The study record describes participants as having previously received ICL670 or deferoxamine for 52weeks in the core study. The primary safety analysis population included participants who received at least one dose of ICL670 during the core or extension phase. This means the reported safety population is defined by exposure to ICL670, not simply by enrollment. The record does not report a participant flow account explaining how enrollment relates to the two posted groups.
The adverse-event results are listed under ICL670 and Crossover. Because allocation is non-randomized and the study model is single-group, these labels do not create a randomized comparison. The record explicitly cautions that the arms are not a like-for-like control comparison. Differences between their event counts cannot be interpreted as a causal treatment contrast from these data.
The study’s stated aims include safety, tolerability, pharmacokinetics, and liver iron concentration. The listed secondary outcome is change in serum ferritin from start of ICL670 to end of study. The posted record does not report results for that secondary outcome or numeric findings for pharmacokinetics or liver iron concentration. It also does not report the protocol details, dose assignment rules, or statistical analysis plan. Such details are not reported in the registry record and should not be inferred from the extension title.
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 | ICL670 | Crossover |
|---|---|---|
| STARTED | 132 | 53 |
| COMPLETED | 43 | 19 |
| NOT COMPLETED | 89 | 34 |
Reasons for not completing this period:
| Reason | ICL670 | Crossover |
|---|---|---|
| Adverse Event | 12 | 2 |
| Abnormal Lab Values | 3 | 1 |
| Abnormal Test Procedure Results | 2 | 0 |
| Unsatisfactory therapeutic effect | 4 | 2 |
| Condition no longer requires drug | 5 | 4 |
| Protocol Violation | 2 | 1 |
| Withdrew consent | 33 | 11 |
| Lost to Follow-up | 11 | 6 |
| Administrative problems | 6 | 4 |
| Death | 1 | 2 |
| Stopped end of core | 7 | 0 |
| Stopped end of extension 1# | 3 | 1 |
Analysis population for the primary outcome, as defined in the registry: The primary analysis was on Safety Analysis Set which comprised all participants who received at least one dose of ICL670 during the core/extension phase of the study. All participants previously treated with ICL670/DFO for 52weeks in the core study.
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 | ICL670 | Crossover | Total |
|---|---|---|---|---|
| Participants analysed | 132 | 53 | 185 | |
| Age, Customized: <12 years | participants | 33 | 14 | 47 |
| Age, Customized: 12 - <16 years | participants | 32 | 11 | 43 |
| Age, Customized: 16 - <65 years | participants | 67 | 28 | 95 |
| Sex: Female, Male: Female | Participants | 80 | 31 | 111 |
| Sex: Female, Male: Male | Participants | 52 | 22 | 74 |
| Region of Enrollment: France | participants | 10 | 2 | 12 |
| Region of Enrollment: United States | participants | 110 | 46 | 156 |
| Region of Enrollment: Canada | participants | 0 | 1 | 1 |
| Region of Enrollment: United Kingdom | participants | 5 | 1 | 6 |
| Region of Enrollment: Italy | participants | 7 | 3 | 10 |
6. Primary Endpoint
The registry lists the primary outcome as Number of Participants With Adverse Events After Start of ICL670, measured over 0 - 60 months.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | participants |
| Endpoint classification | Binary Endpoint |
The primary endpoint is the number of participants with adverse events after the start of ICL670. The description specifies safety assessed by the number of participants with an adverse event or death, and gives a timeframe of 0 - 60 months. This is a count of participants experiencing an event, rather than a measure of event severity, time until an event, or the total number of events each participant experienced.
Counting participants with events is one way to describe how often safety events occurred in the analysis population. It can make the burden of events easier to read than a list of individual reports, particularly when one participant could experience more than one event. However, a participant count does not capture repeated events as separate occurrences, and a count without a corresponding analysis of follow-up time does not show how quickly events arose.
The record distinguishes serious adverse events from other adverse events in the posted group summaries. These are useful descriptive safety categories, but they do not by themselves determine whether an event was caused by ICL670. Nor does a count alone describe clinical severity beyond the serious classification, duration, resolution, or relationship to treatment.
The primary description includes death, but the posted group data give no death counts. A separate death result is therefore not reported in the registry record. The registry also does not provide a confidence interval or comparative estimate for this endpoint. Accordingly, the endpoint can be interpreted descriptively from the participant counts, while its precision and any difference between groups cannot be quantified from the posted record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted results report 97 participants with serious adverse events among 132 at risk in the ICL670 group, and 34 among 53 at risk in the Crossover group. For other adverse events, the corresponding posted counts are 129 among 132 and 47 among 53. These are participant counts within the registry’s safety presentation. They are not counts of every event, and they do not tell how many events any one participant experienced.
The two groups should not be read as a controlled treatment comparison. The record identifies the study as non-randomized, open label, and single group, and expressly cautions that the arms are not a like-for-like control comparison. There is no basis in the posted record for treating the difference in group counts as an estimate of the effect of ICL670 relative to another treatment. Prior treatment, participant characteristics, exposure, and follow-up could differ, but the record does not provide the information needed to assess these possibilities.
The listed serious-event terms are sickle cell anaemia with crisis at 36.4%, pyrexia at 18.9%, abdominal pain at 10.6%, pneumonia at 9.8%, vomiting at 9.4%, and back pain at 9.4%. The registry does not specify the denominator or detailed counting rule for these term percentages in the registry record. These figures therefore describe the listed terms as recorded; they should not be interpreted as risks of a treatment-caused event or as mutually exclusive categories.
The primary outcome description covers adverse events or death, but the posted results do not give deaths separately. The endpoint timeframe is 0 - 60 months, while the official title notes an extension prolonged to 4-years. Neither statement gives the observed follow-up distribution for participants. No time-to-event estimates are reported, so the counts cannot show when events occurred or account for different exposure durations.
The registry record does not report confidence intervals, hypothesis tests, adjusted estimates, or numeric findings for the listed serum ferritin outcome. It is therefore not possible to assess statistical precision or to make a supported claim about a difference between the groups. The counts are informative as descriptive safety records, but they are not sufficient on their own for causal conclusions.
9. Statistical Methods Explained
Statistical interpretation depends on what is counted, how participants enter the analysis, and whether the comparison supports causal inference. For this study, the registry describes a participant-count endpoint and provides group counts for serious and other adverse events. No inferential analysis results are reported in the registry record. The concepts below are general explanations of how such data are usually interpreted, not claims about an unreported analysis plan.
What does a participant count represent?
A participant count records whether a person experienced an event according to the endpoint definition. It is different from an event count: a participant with repeated events may still contribute only once to a count of participants affected. A count is also different from a rate, which relates events to person-time, and from a time-to-event analysis, which considers when an event first occurs and how long participants remain under observation. The registry posts participant counts, not event rates or time-to-event estimates.
For a count to be interpretable, its denominator and analysis population matter. The safety analysis set is described as participants who received at least one dose of ICL670 during the core or extension phase. The group summaries give the numbers at risk and the numbers affected. They do not provide a detailed account of participant inclusion, exclusions, losses to follow-up, or the amount of observation contributed by each person. Those details are not reported in the registry record.
Why do design features affect comparisons?
Random allocation, when used and properly implemented, is intended to balance both known and unknown baseline factors across groups on average. This extension is identified as non-randomized, and the registry warns that its posted arms are not a like-for-like control comparison. Consequently, event counts in ICL670 and Crossover cannot be treated as if participants had been assigned to comparable treatments under a randomized design. Even if group counts differ, the difference could reflect the groups’ composition, previous treatment, duration of exposure, follow-up, or other factors rather than a treatment effect. The registry does not report enough information to separate these explanations.
Open-label status means masking was absent. In general, lack of masking can influence event recognition, reporting, or decisions about whether an event meets a classification. This is a general methodological concern, not evidence that reporting bias occurred in this study. The record does not describe procedures for event adjudication or whether reviewers were independent.
What would uncertainty estimates add?
A confidence interval is a range of values compatible with the observed data under a stated statistical model. In general, it communicates precision more clearly than a point estimate alone. For a comparative analysis, an interval around a treatment contrast can show how uncertain the estimated difference is. The registry record does not report confidence intervals or a comparative estimate for the safety counts. It is not appropriate to invent an interval or infer one from the posted information.
A hypothesis test and its p-value address a specified statistical question under assumptions. A p-value is not the probability that a treatment works, nor does it establish clinical importance. No p-value or testing procedure for the posted safety groups is reported. Without the analysis plan and a valid comparison design, a newly constructed test would not resolve the central concern that the groups are not like-for-like.
How should serious-event terms be read?
The registry lists several serious adverse event terms with percentages. In general, a percentage for an event term needs a clearly stated denominator and counting convention to be interpreted as a frequency. Categories may overlap because one participant can experience more than one kind of event. A term percentage also does not imply that the event was caused by treatment. The denominator and detailed event-counting rules for the listed terms are not reported in the registry record, so the values are best treated as a descriptive list.
Over-dispersion is a general issue when data count events: it occurs when variation exceeds what a simple count model assumes, often because observations are clustered or some people experience many events. A participant-level indicator avoids treating repeated events as independent counts, but it also discards information about recurrence. The posted primary endpoint is a participant count, and the record does not report a repeated-event model or an event-level analysis. No conclusion about over-dispersion can be drawn from these data.
What can planning illustrations establish?
A sample-size calculation is a planning exercise based on assumptions about the target effect, significance level, allocation, and analysis method. It is not an observed result and does not demonstrate that a study had adequate power for the effects actually present. The record describes planning illustrations based on fractions of an observed effect and specifies a two-sided significance level of 0.05, equal allocation, and a normal approximation. These are illustration assumptions, not evidence of the trial’s original design or statistical plan. The registry does not report a sample-size justification or identify those illustrations as the study’s prospective design calculation.
In general, planning estimates depend on the effect chosen as important and on assumptions about outcome variability and analysis. Changing those assumptions changes the required sample size. Such calculations should not be described as observed power, and they cannot remedy the limitations of a non-randomized comparison.
10. Multiplicity, Missing Data and Analysis Populations
The study record identifies a primary safety endpoint and a secondary outcome of change in serum ferritin from start of ICL670 to end of study. Its stated study aims also include tolerability, pharmacokinetics, and liver iron concentration. The record does not provide a full endpoint hierarchy, a multiplicity adjustment strategy, or a statistical analysis plan. Those details are not reported in the registry record.
Multiplicity is a general concern when many outcomes, time points, subgroups, or statistical tests are examined. The more tests considered, the greater the chance of apparently notable findings occurring by chance if the analysis does not account for them. This does not establish that multiplicity affected the study; it means the registry record does not allow readers to assess how multiple outcomes were handled. No inferential results for the secondary serum ferritin outcome are posted in the registry record.
The record gives an analysis population for the primary safety assessment: participants who received at least one dose of ICL670 during the core or extension phase. It does not report how missing outcome information was handled, whether participants with incomplete follow-up were included in each count, or whether any sensitivity analyses were conducted. It also does not describe participant withdrawals or reasons for discontinuation. These matters are not reported in the registry record.
Without those details, the posted counts should be read as the reported descriptive safety results, not as a complete account of all possible missing-data assumptions or analyses. No intention-to-treat approach, imputation method, or missing-at-random assumption should be attributed to the study based on this record.
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 |
|---|---|---|---|
| ICL670 | 97 | 132 | 73.5% |
| Crossover | 34 | 53 | 64.2% |
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 |
|---|---|
| ICL670 | 129 of 132 |
| Crossover | 47 of 53 |
Most frequent serious adverse event terms
| Event term | Organ system | ICL670 | Crossover |
|---|---|---|---|
| Sickle cell anaemia with crisis | Congenital, familial and genetic disorders | 48 of 132 | 19 of 53 |
| Pyrexia | General disorders | 25 of 132 | 9 of 53 |
| Abdominal pain | Gastrointestinal disorders | 14 of 132 | 4 of 53 |
| Pneumonia | Infections and infestations | 13 of 132 | 3 of 53 |
| Vomiting | Gastrointestinal disorders | 7 of 132 | 5 of 53 |
| Chest pain | General disorders | 8 of 132 | 4 of 53 |
| Back pain | Musculoskeletal and connective tissue disorders | 6 of 132 | 5 of 53 |
| Cholelithiasis | Hepatobiliary disorders | 7 of 132 | 2 of 53 |
| Pain in extremity | Musculoskeletal and connective tissue disorders | 6 of 132 | 3 of 53 |
| Headache | Nervous system disorders | 7 of 132 | 2 of 53 |
Most frequent non-serious adverse event terms
| Event term | Organ system | ICL670 | Crossover |
|---|---|---|---|
| Headache | Nervous system disorders | 61 of 132 | 30 of 53 |
| Nausea | Gastrointestinal disorders | 52 of 132 | 19 of 53 |
| Pyrexia | General disorders | 45 of 132 | 22 of 53 |
| Back pain | Musculoskeletal and connective tissue disorders | 48 of 132 | 15 of 53 |
| Vomiting | Gastrointestinal disorders | 45 of 132 | 17 of 53 |
| Upper respiratory tract infection | Infections and infestations | 43 of 132 | 19 of 53 |
| Diarrhoea | Gastrointestinal disorders | 39 of 132 | 19 of 53 |
| Oropharyngeal pain | Respiratory, thoracic and mediastinal disorders | 40 of 132 | 15 of 53 |
| Pain in extremity | Musculoskeletal and connective tissue disorders | 41 of 132 | 13 of 53 |
| Nasopharyngitis | Infections and infestations | 37 of 132 | 16 of 53 |
Serious events with the highest arm-level rate: sickle cell anaemia with crisis (36.4%), pyrexia (18.9%), abdominal pain (10.6%), pneumonia (9.8%), vomiting (9.4%).
The registry lists serious adverse event counts of 97 among 132 at risk for ICL670 and 34 among 53 at risk for Crossover. It lists other adverse event counts of 129 among 132 and 47 among 53, respectively. These values describe the number of participants recorded with each category, rather than the total number of event episodes. The primary endpoint description includes adverse events or death, but a separate death count is not reported in the registry record.
The listed serious-event terms are sickle cell anaemia with crisis (36.4%), pyrexia (18.9%), abdominal pain (10.6%), pneumonia (9.8%), vomiting (9.4%), and back pain (9.4%). They are descriptive terms, not evidence that ICL670 caused those events. The record does not specify the denominator or whether participants could contribute to more than one term.
Differences between group counts do not constitute a treatment comparison because the extension was non-randomized and the record warns that the groups are not like-for-like controls. The counts also do not show event timing, duration, or individual follow-up. They should therefore be read as safety observations within the reported groups, with important limits on comparison and attribution.
12. What the Result Does, and Does Not, Mean
The results support a descriptive reading of adverse events among participants in the registry’s safety analysis. The record reports participants with serious and other adverse events in the ICL670 and Crossover groups and names several serious-event terms. It does not report a separate death count, comparative effect estimate, confidence interval, or inferential test for these group results. The posted counts can establish what was recorded in those categories, but not the precision of a treatment contrast.
The design is central to interpretation. This is an open-label, non-randomized extension, and the registry explicitly cautions that the arms are not a like-for-like control comparison. As a result, the group counts do not establish whether one group had a higher or lower treatment-caused risk. They also cannot isolate the effect of ICL670 from differences in prior therapy, participant characteristics, follow-up, or exposure, for which the record does not supply enough detail.
The endpoint counts people with an event, not repeated event episodes, event rates over person-time, or time to first event. The timeframe of 0 - 60 months does not show the actual observation duration for each participant. The official title notes an extension prolonged to 4-years, but that does not establish that every participant was followed for the same period.
The study also lists serum ferritin change as a secondary outcome and describes aims involving pharmacokinetics and liver iron concentration. Numeric results for these outcomes are not reported in the registry record. The safety summaries therefore should not be generalized to conclusions about these other measures or about every aspect of the study’s stated objectives.
In short, the posted safety data provide a record of observed participant counts. They do not, on their own, establish causation, comparative effectiveness, or the absence of risk. Conclusions beyond the descriptive results would require information and analyses not reported in the registry record.
13. Why This Trial Matters Statistically
This study illustrates a common distinction in clinical research: an extension can provide additional safety observations while offering limited evidence for a controlled treatment comparison. The study is described as non-randomized, open label, and single group. Although the registry presents two labelled groups, it cautions that they are not like-for-like controls. Recognizing that distinction prevents descriptive group counts from being mistaken for a causal estimate.
The primary endpoint is also a useful example of how outcome definition shapes interpretation. A participant count answers how many people were recorded as having an event category. It does not indicate the total number of episodes, how long participants were observed, or when events occurred. Those questions require different summaries and, often, different statistical methods. The available record gives participant counts, so interpretation should remain aligned with that measure.
The extension’s stated timeframe and its actual follow-up are not interchangeable. The primary outcome is assigned a timeframe of 0 - 60 months, and the official title notes that the extension was prolonged to 4-years. Neither description reports individual follow-up duration. This makes it important not to treat a study-level timeframe as a uniform observation period or a rate denominator.
Finally, the record shows why statistical transparency matters. The primary safety population is described, but details on missing data, multiplicity, analysis methods, and uncertainty estimates are not reported. Without them, readers can describe the posted results but cannot fully assess precision, robustness, or how all planned outcomes were handled. The case is a reminder that careful interpretation depends not only on the event counts, but also on design, analysis population, follow-up, and reporting.
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 most important limitation for interpretation is the non-randomized, open-label design. The registry explicitly says the posted groups are not a like-for-like control comparison. Differences in their adverse-event counts therefore cannot be interpreted as a causal contrast. The record does not give the information needed to assess how the groups differed at entry or in prior treatment and follow-up.
The safety analysis population is defined by receipt of at least one ICL670 dose during the core or extension phase. The registry does not report a detailed participant flow, group-specific follow-up duration, withdrawal information, or reasons for discontinuation. Although the endpoint timeframe is 0 - 60 months, actual observation time by participant is not reported in the registry record.
The outcome presentation gives participant counts with serious and other adverse events, not recurrent event totals or person-time rates. Death is included in the primary endpoint description, but a separate death count is not reported. The listed serious-event term percentages lack a reported denominator and counting convention in the registry record presented, limiting interpretation of their frequency.
No confidence intervals, comparative estimates, hypothesis tests, missing-data methods, multiplicity strategy, or statistical analysis plan are reported. Results for the secondary serum ferritin outcome and numeric findings for pharmacokinetics or liver iron concentration are also not reported in the registry record. Consequently, the scope of interpretation is restricted to the descriptive safety counts and terms that are posted.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Evaluation of the Safety, Tolerability, Pharmacokinetics (PK) and Effects on Liver Iron Concentration of ICL670 Relative to Deferoxamine(DFO). — NCT01090323.
- Registry record last updated 2011-06-06; status last verified 2011-05.
- Results first posted 2011-01-20.
- 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 record identifies NCT01090323 as a completed phase II, non-randomized, open-label extension study sponsored by Novartis Pharmaceuticals. It describes ICL670 evaluation in people with sickle cell disease and transfusional hemosiderosis. Enrollment is recorded as 185. The study began in 2004-07 and primary completion is recorded as 2009-07.
The primary outcome is the number of participants with adverse events after starting ICL670, with a timeframe of 0 - 60 months. The safety population comprised participants who received at least one ICL670 dose during the core or extension phase. Posted results give serious and other adverse-event counts in groups labelled ICL670 and Crossover; the registry warns these are not like-for-like control groups.
The record lists change in serum ferritin as a secondary outcome, but does not report its result. It also does not report separate death counts, comparative estimates, uncertainty intervals, or a statistical analysis plan. The available safety findings should therefore be interpreted descriptively.