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 Extension Study of AG-348 in Adult Participants With Pyruvate Kinase Deficiency Previously Enrolled in AG-348-006 or AG348-C-007 (NCT03853798), a non-randomized phase 3 study. The primary outcome is All Cohorts: Number of Participants With at Least One Treatment Emergent Adverse Events (TEAEs), Serious TEAEs, Related TEAEs and TEAEs With Severity Greater Than or Equal to Grade 3.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03853798 |
| Phase | Phase 3 |
| Condition | Pyruvate Kinase Deficiency |
| Allocation | Non-Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 90 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Agios Pharmaceuticals, Inc. |
| Results posted | Yes |
2. Clinical Question
Population
Adults with pyruvate kinase deficiency who had previously participated in AG348-C-006 or AG348-C-007 and met the eligibility criteria for this extension study. The registry record lists enrollment of 90, but the number analyzed for each outcome is not reported in the registry record.
Intervention
Mitapivat tablets taken orally. Treatment and dose pathways depended on the participant’s prior study and treatment group. The study was open-label, so treatment assignment was not masked.
Comparator
There was no randomized control group in this extension. Participants continued or began mitapivat according to their prior study treatment and cohort pathway; these cohorts are not like-for-like control groups.
Primary question
The primary outcome asks how many participants in all cohorts experienced at least one treatment-emergent adverse event, serious treatment-emergent adverse event, related treatment-emergent adverse event, or treatment-emergent adverse event of severity grade 3 or higher during follow-up up to 197 weeks.
This is a long-term safety and tolerability question in people already enrolled in preceding mitapivat studies. The study also listed secondary outcomes concerning hemoglobin response and change in hemoglobin concentration for Cohort 1, along with mitapivat plasma concentration measures. However, the registry record does not report numerical results for those outcomes in the registry record. It therefore does not support a comparison of treatment effects against a concurrent untreated group. The safety counts describe events observed within the listed cohorts and doses; they cannot by themselves establish how those events would compare with what would have occurred without mitapivat.
3. Trial Design
This was a Phase 3, multicenter, open-label extension study with non-randomized allocation and a parallel model. It was designed to evaluate long-term safety, tolerability, and efficacy of mitapivat in adult participants with pyruvate kinase deficiency who had previously been enrolled in AG348-C-006 or AG348-C-007. The registry record lists enrollment of 90 and a completed status. The study start date is 2019-03-21, and primary completion is recorded as 2024-07-03.
The three cohorts reflect participants’ prior study and treatment. Cohort 1 included eligible participants who had received placebo in AG348-C-006 and began mitapivat at 5 mg twice daily for 4 weeks, with two potential sequential dose-level increases described in the record. The further details of those potential increases are not reported in the registry record. Cohort 2 included participants from AG348-C-006 who had received mitapivat at 5 mg, 20 mg, or 50 mg twice daily in the fixed-dose period; they continued the same dose up to Week 193, study withdrawal, or study closure. Cohort 3 followed the same continuation description for participants from AG348-C-007.
Because the study was non-randomized and had no masking, the cohort and dose groups should not be treated as exchangeable treatment arms. Their members came through different treatment histories, and the groups’ characteristics and reasons for receiving particular doses are not reported in the registry record. The study design can describe events during extension treatment, but it does not provide a randomized comparison that isolates a treatment effect. Details of eligibility criteria, dose adjustment rules beyond the abbreviated description, and the statistical design rationale 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.
Before assignment: A total of 90 participants who had completed Study AG348-C-006 (NCT03548220) or AG348-C-007 (NCT03559699), were enrolled in this extension study to receive mitapivat treatment. Participants were assigned to Cohorts 1, 2 or 3 depending on the previous treatment received in the antecedent studies.
Only the first 6 of 9 groups are shown.
Overall Study
| Stage | Cohort 1: 5 mg | Cohort 1: 20 mg | Cohort 1: 50 mg | Cohort 2: 5 mg | Cohort 2: 20 mg | Cohort 2: 50 mg |
|---|---|---|---|---|---|---|
| STARTED | 1 | 1 | 36 | 2 | 3 | 30 |
| COMPLETED | 0 | 0 | 25 | 2 | 3 | 22 |
| NOT COMPLETED | 1 | 1 | 11 | 0 | 0 | 8 |
Reasons for not completing this period:
| Reason | Cohort 1: 5 mg | Cohort 1: 20 mg | Cohort 1: 50 mg | Cohort 2: 5 mg | Cohort 2: 20 mg | Cohort 2: 50 mg |
|---|---|---|---|---|---|---|
| Death | 0 | 0 | 0 | 0 | 0 | 1 |
| Adverse Event | 0 | 0 | 2 | 0 | 0 | 0 |
| Withdrawal by Subject | 0 | 1 | 4 | 0 | 0 | 5 |
| Physician Decision | 1 | 0 | 0 | 0 | 0 | 0 |
| Lack of Efficacy | 0 | 0 | 2 | 0 | 0 | 0 |
| Approved Drug Available for Indication | 0 | 0 | 1 | 0 | 0 | 1 |
| Other un-specified | 0 | 0 | 2 | 0 | 0 | 1 |
Analysis population for the primary outcome, as defined in the registry: The safety analysis set included all participants who received at least 1 dose of study treatment. Due to limited number of participants with optimal dose of 5 mg or 20 mg in antecedent studies AG348-C-006 (NCT03548220) and AG348-C-007 (NCT03559699), it was pre-specified in the SAP to report data for this outcome measure (OM) by Cohort.
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: Full Analysis Set (FAS) included all participants who received at least 1 dose of study treatment. Due to limited number of participants with optimal dose of 5 mg or 20 mg in antecedent studies AG348-C-006 (NCT03548220) and AG348-C-007 (NCT03559699), it was pre-specified in the statistical analysis plan (SAP) to report data for baseline characteristics by Cohort.
| Characteristic | Unit | Cohort 1 | Cohort 2 | Cohort 3 | Total |
|---|---|---|---|---|---|
| Participants analysed | 38 | 35 | 17 | 90 | |
| Age, Continuous | Years | 38.0 (SD 15.95) | 36.9 (SD 15.77) | 36.9 (SD 15.09) | 37.4 (SD 15.56) |
| Sex: Female, Male: Female | Participants | 23 | 20 | 12 | 55 |
| Sex: Female, Male: Male | Participants | 15 | 15 | 5 | 35 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 1 | 2 | 0 | 3 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 33 | 23 | 12 | 68 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 4 | 10 | 5 | 19 |
| Race/Ethnicity, Customized: White | Participants | 31 | 23 | 13 | 67 |
| Race/Ethnicity, Customized: Asian | Participants | 3 | 5 | 2 | 10 |
| Race/Ethnicity, Customized: Native Hawaiian or Other Pacific Islander | Participants | 0 | 1 | 0 | 1 |
| Race/Ethnicity, Customized: Other | Participants | 1 | 0 | 0 | 1 |
| Race/Ethnicity, Customized: Not reported | Participants | 3 | 6 | 2 | 11 |
| Femoral Total: T-score | T-score | -0.834 (SD 1.1182) | -1.139 (SD 1.0995) | -0.813 (SD 0.6748) | -0.946 (SD 1.0416) |
| Femoral Total: Z-score | Z-score | -0.548 (SD 1.1574) | -0.916 (SD 1.0657) | -0.529 (SD 0.7137) | -0.685 (SD 1.0560) |
| Adjusted Spine: T-Score | T-score | -1.143 (SD 1.0926) | -1.808 (SD 1.1589) | -1.161 (SD 1.0469) | -1.400 (SD 1.1440) |
| Adjusted Spine: Z-score | Z-score | -0.804 (SD 1.2529) | -1.536 (SD 1.2469) | -0.848 (SD 1.0420) | -1.092 (SD 1.2506) |
6. Primary Endpoint
The registry lists the primary outcome as All Cohorts: Number of Participants With at Least One Treatment Emergent Adverse Events (TEAEs), Serious TEAEs, Related TEAEs and TEAEs With Severity Greater Than or Equal to Grade 3, measured over Up to 197 weeks.
| Element | Registry entry |
|---|---|
| Result type | Count Of Participants |
| Measure of spread | not applicable |
| Unit | Participants |
| Endpoint classification | Binary Endpoint |
The primary endpoint is the number of participants with at least one treatment-emergent adverse event (TEAE), serious TEAE, related TEAE, or TEAE of severity grade 3 or higher. It is assessed across all cohorts up to 197 weeks. A TEAE is defined as an adverse event with an initial onset during the on-treatment period or worsening from baseline. The definition includes serious and non-serious events; an adverse event does not need to be caused by the study drug to meet the general adverse-event definition.
Severity was evaluated using National Cancer Institute Common Terminology Criteria for Adverse Events, Version 4.03. Grade 1 is mild, grade 2 moderate, grade 3 severe or medically significant but not immediately life-threatening, grade 4 life-threatening or disabling, and grade 5 death related to an adverse event. Seriousness and severity describe different aspects of an event: an event may be serious under regulatory criteria without necessarily being grade 3 or higher, and severity grading does not by itself determine relatedness.
Counting participants with at least one event is a practical safety summary because it expresses how many people experienced an event within the outcome definition, rather than how many separate episodes occurred. It does not capture the full number, duration, recurrence, or timing of events. The registry record’s primary outcome description includes several event categories, but the group-level counts available in the posted results do not separately provide every category named in the endpoint. The full endpoint results and analysis details are not reported in the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted safety results list participant counts by cohort and dose for serious events and for a second event category labeled “other,” alongside numbers at risk and deaths. The serious-event counts vary across the listed groups, from no recorded serious events in some groups to multiple events in others. For example, the listed serious-event counts include 10 of 36 in Cohort 1 at 50 mg, 6 of 30 in Cohort 2 at 50 mg, and 4 of 16 in Cohort 3 at 50 mg. These are descriptive counts within non-randomized groups, not estimates from a controlled comparison.
The listed counts for the “other” category are also descriptive. The registry record does not define that category in the adverse-event group listing, so it should not be equated automatically with non-serious TEAEs or with all TEAEs. The primary outcome includes related events and events of severity grade 3 or higher, but the available group listing does not provide separate counts for those categories. Thus, the posted counts do not resolve every component of the primary endpoint.
One death is listed in Cohort 2 at 50 mg, and zero deaths are listed in the other dose groups. The registry record does not state in this listing whether the recorded death was related to treatment or whether it met the endpoint’s grade 5 definition. It is not appropriate to infer causation from the death count alone.
Differences in counts across cohorts should not be interpreted as evidence that one dose is safer or less safe than another. Participants entered through different antecedent studies and treatment histories, and the groups were not randomized against each other. The time at risk and other features that could affect event detection are not described for each group in the posted results. Even when a group count is small or zero, that observation does not establish that an event cannot occur.
The registry record reports 28 outcome measures and zero analyses in its analysis listing. That does not mean no statistical work was performed; it means no analysis records are listed there. Numerical efficacy results, estimates of uncertainty, formal between-group tests, and a complete set of primary endpoint results are not reported in the registry record. The posted safety counts therefore support a descriptive reading, not a precise comparative conclusion.
9. Statistical Methods Explained
What does a participant count represent?
In general, a count endpoint records how many participants meet a stated event definition. For a safety endpoint expressed as participants with at least one event, each person is counted according to whether the event criterion was met during the specified observation period. That differs from counting every event episode: a participant with recurrent events can still contribute one person to a “at least one” count. This summary is useful for describing how many participants were affected, but it does not show the number of episodes, their duration, or their clinical course.
The registry record identifies the primary outcome as a count of participants and gives the safety analysis set as participants who received at least 1 dose of study treatment. The description also says that data for certain small-dose groups in the antecedent studies were to be reported in a particular way under the SAP, but that sentence is truncated. The full population handling and analysis details are not reported in the registry record. The available safety listing gives an at-risk count for each cohort-dose group, but it does not fully explain the relationship between those group counts and the planned analysis population.
How should an event proportion be understood?
General statistical interpretation: when participant counts are related to a denominator, the resulting proportion describes the observed frequency of participants meeting the event definition in that group during its follow-up. It is not automatically a risk comparison between groups. A meaningful comparison requires comparable groups, clearly defined observation windows, consistent event ascertainment, and an analysis that accounts for the design. In a non-randomized study, differences may reflect prior treatment, participant characteristics, follow-up, or other factors as well as any treatment-related differences. The registry record does not report an adjusted comparative analysis.
Confidence intervals are generally used to express uncertainty around an estimate. They indicate a range of values compatible with the data under the method’s assumptions; they are not a guarantee that the true value lies in that range. Small groups typically yield less precise estimates, and a zero count does not imply zero underlying risk. No confidence intervals for the posted safety counts are reported in the registry record. The posted results also do not report hypothesis tests for comparisons among cohorts or doses.
Why do the cohort pathways matter?
General statistical interpretation: random allocation helps make treatment groups comparable on average, whereas this extension’s allocation was non-randomized. Here, cohorts represent different histories in AG348-C-006 or AG348-C-007, including participants who previously received placebo and participants who continued mitapivat at an existing dose. These histories mean that a direct contrast between cohort counts is not equivalent to a randomized dose comparison. Parallel grouping alone does not make groups comparable.
The study was open-label, so treatment was not masked. In general, lack of masking can affect reporting, clinical assessment, and decisions about recording or evaluating events, although the extent of any such effect cannot be established from the registry record. The record does not describe event adjudication procedures, exposure-adjusted analyses, or methods for handling differences in follow-up time.
What do the planning illustrations mean?
General statistical interpretation: a sample-size planning illustration uses assumptions to show the approximate group size needed to detect a chosen effect under a specified design. It is not the same as the study’s actual observed result, and it is not observed power. The registry record’s planning assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is expressed as a fraction of the observed effect. For non-time-to-event outcomes, the illustrations are described as participants per arm. Because this study is non-randomized and its listed cohorts are not like-for-like control groups, such illustrations should not be read as evidence that the trial was designed or powered for a valid randomized comparison. The registry record does not report the trial’s own sample-size justification.
What analysis details are missing?
The record lists zero analyses, while the study includes 28 outcome measures. Statistical tests, model specifications, covariate adjustment, multiplicity procedures, missing-data handling, and analysis-specific participant counts are not reported in the registry record. General statistical interpretation: when many outcomes are examined, multiplicity procedures may be used to control the chance of false-positive findings across a family of tests. Whether such procedures were planned or used in this study is not reported in the registry record. Without the specified analysis methods and complete outcome results, the safety counts should be treated as descriptive evidence rather than as inferential estimates.
10. Multiplicity, Missing Data and Analysis Populations
The registry record lists 28 outcome measures, including the primary safety outcome and secondary measures concerning hemoglobin response, hemoglobin concentration, and mitapivat pharmacokinetics. It lists zero analyses. The statistical plan, prespecified hierarchy among outcomes, and any multiplicity adjustment are not reported in the registry record. General statistical context: when a study assesses many outcomes or makes many comparisons, the chance of at least one false-positive result can increase; a prespecified testing strategy can address that issue. No conclusion about the strategy used in this study can be drawn from the available record.
The record identifies the safety analysis set as participants who received at least 1 dose of study treatment. However, the full analysis population for each endpoint, participant flow into each cohort, and outcome-specific denominators are not reported in the registry record. The primary outcome population description is truncated after noting a prespecified SAP approach for participants at certain doses in the antecedent studies. The complete rule is not reported in the registry record.
Missing-data methods are not reported in the registry record. In general, missing outcome information can affect estimates if participants with unrecorded outcomes differ from those with observed outcomes. For safety outcomes, length of follow-up and time exposed may also affect the opportunity to observe an event. The posted counts do not report person-time or a method that adjusts for differing observation duration. Withdrawal reasons, missingness by cohort, and sensitivity analyses are not reported in the registry record.
These gaps limit assessment of how complete and comparable the outcome data are. They also prevent verification of whether the reported counts follow a prespecified analysis plan across all listed measures.
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 |
|---|---|---|---|
| Cohort 1: 5 mg | 1 | 1 | 100.0% |
| Cohort 1: 20 mg | 0 | 1 | 0.0% |
| Cohort 1: 50 mg | 10 | 36 | 27.8% |
| Cohort 2: 5 mg | 0 | 2 | 0.0% |
| Cohort 2: 20 mg | 2 | 3 | 66.7% |
| Cohort 2: 50 mg | 6 | 30 | 20.0% |
| Cohort 3: 20 mg | 0 | 1 | 0.0% |
| Cohort 3: 50 mg | 4 | 16 | 25.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Cohort 1: 5 mg | 0 of 1 |
| Cohort 1: 20 mg | 0 of 1 |
| Cohort 1: 50 mg | 0 of 36 |
| Cohort 2: 5 mg | 0 of 2 |
| Cohort 2: 20 mg | 0 of 3 |
| Cohort 2: 50 mg | 1 of 30 |
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 |
|---|---|
| Cohort 1: 5 mg | 0 of 1 |
| Cohort 1: 20 mg | 1 of 1 |
| Cohort 1: 50 mg | 35 of 36 |
| Cohort 2: 5 mg | 2 of 2 |
| Cohort 2: 20 mg | 2 of 3 |
| Cohort 2: 50 mg | 27 of 30 |
Most frequent serious adverse event terms
| Event term | Organ system | Cohort 1: 5 mg | Cohort 1: 20 mg | Cohort 1: 50 mg | Cohort 2: 5 mg | Cohort 2: 20 mg | Cohort 2: 50 mg |
|---|---|---|---|---|---|---|---|
| Sepsis | Infections and infestations | 0 of 1 | 0 of 1 | 3 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
| Pneumonia | Infections and infestations | 0 of 1 | 0 of 1 | 2 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
| Tonsillitis | Infections and infestations | 0 of 1 | 0 of 1 | 2 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
| Gastroenteritis | Infections and infestations | 0 of 1 | 0 of 1 | 1 of 36 | 0 of 2 | 0 of 3 | 1 of 30 |
| Syncope | Nervous system disorders | 0 of 1 | 0 of 1 | 1 of 36 | 0 of 2 | 1 of 3 | 0 of 30 |
| Haemolysis | Blood and lymphatic system disorders | 0 of 1 | 0 of 1 | 1 of 36 | 0 of 2 | 0 of 3 | 1 of 30 |
| Cytomegalovirus infection | Infections and infestations | 1 of 1 | 0 of 1 | 0 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
| Diverticulitis | Infections and infestations | 0 of 1 | 0 of 1 | 1 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
| Erysipelas | Infections and infestations | 0 of 1 | 0 of 1 | 1 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
| Pharyngitis | Infections and infestations | 0 of 1 | 0 of 1 | 1 of 36 | 0 of 2 | 0 of 3 | 0 of 30 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Cohort 1: 5 mg | Cohort 1: 20 mg | Cohort 1: 50 mg | Cohort 2: 5 mg | Cohort 2: 20 mg | Cohort 2: 50 mg |
|---|---|---|---|---|---|---|---|
| Headache | Nervous system disorders | 0 of 1 | 1 of 1 | 20 of 36 | 2 of 2 | 0 of 3 | 6 of 30 |
| COVID-19 | Infections and infestations | 0 of 1 | 0 of 1 | 12 of 36 | 1 of 2 | 2 of 3 | 13 of 30 |
| Fatigue | General disorders | 0 of 1 | 1 of 1 | 10 of 36 | 1 of 2 | 1 of 3 | 6 of 30 |
| Pyrexia | General disorders | 0 of 1 | 0 of 1 | 10 of 36 | 2 of 2 | 0 of 3 | 7 of 30 |
| Arthralgia | Musculoskeletal and connective tissue disorders | 0 of 1 | 0 of 1 | 9 of 36 | 0 of 2 | 2 of 3 | 5 of 30 |
| Upper respiratory tract infection | Infections and infestations | 0 of 1 | 0 of 1 | 9 of 36 | 0 of 2 | 0 of 3 | 4 of 30 |
| Alanine aminotransferase increased | Investigations | 0 of 1 | 0 of 1 | 8 of 36 | 1 of 2 | 0 of 3 | 4 of 30 |
| Nasopharyngitis | Infections and infestations | 0 of 1 | 0 of 1 | 6 of 36 | 1 of 2 | 0 of 3 | 5 of 30 |
| Diarrhea | Gastrointestinal disorders | 0 of 1 | 0 of 1 | 6 of 36 | 0 of 2 | 1 of 3 | 4 of 30 |
| Middle insomnia | Psychiatric disorders | 0 of 1 | 0 of 1 | 8 of 36 | 0 of 2 | 0 of 3 | 2 of 30 |
Serious events with the highest arm-level rate: cytomegalovirus infection (100.0%), syncope (33.3%), cholecystitis chronic (33.3%), sepsis (8.3%), covid-19 pneumonia (6.2%).
The serious-event listing presents, for each cohort and dose, the number at risk and the number with a serious event. Examples include 10 of 36 in Cohort 1 at 50 mg, 6 of 30 in Cohort 2 at 50 mg, and 4 of 16 in Cohort 3 at 50 mg. Other listed groups have smaller counts, including zero serious events in the listed Cohort 1 groups at 20 mg and Cohort 2 at 5 mg. Cohort 3 at 5 mg has zero participants at risk in the listing. These counts should be read within their individual groups rather than combined into a comparison across doses.
The death listing records one death in Cohort 2 at 50 mg and zero in the other listed groups. The event’s relationship to treatment is not reported in the registry record. The record also gives selected serious-event terms with percentages, including Cytomegalovirus infection at 100.0%, Syncope at 33.3%, and Cholecystitis chronic at 33.3%. The denominator used for those percentages is not reported in the registry record, so the percentages should not be interpreted as risks among all treated participants.
Seriousness, severity, and relatedness are distinct classifications. The posted group counts do not provide a complete breakdown for every primary endpoint category, and the “other” category is not defined in the listing. The figures are descriptive and do not establish causality or a comparative safety effect.
12. What the Result Does, and Does Not, Mean
The record’s central statistical contribution is a long-term, cohort-by-cohort description of safety events among participants continuing or beginning mitapivat after earlier study participation. It specifies the primary safety question, gives event counts and at-risk counts for listed cohort-dose groups, and identifies a follow-up timeframe of up to 197 weeks. It also records one death in Cohort 2 at 50 mg and gives selected serious-event terms with percentages.
These observations do not amount to a randomized estimate of mitapivat’s effect on safety. Participants were not randomly allocated to the listed groups, and the cohorts reflect different prior studies and treatment histories. The counts cannot establish that observed differences between groups were caused by dose, nor can they separate treatment effects from differences in participants or observation. The event listing also does not report all components of the primary outcome separately.
The results do not establish whether mitapivat improves hemoglobin outcomes or produces a particular pharmacokinetic profile. Although such outcomes appear among the secondary measures, numerical results for them are not reported in the registry record. Nor does the record report confidence intervals, formal tests, adjusted estimates, or a complete statistical analysis for the safety outcome.
In general, descriptive safety counts can help characterize events observed during treatment, but they do not determine what would have happened to the same participants under a different treatment or without treatment. A zero count in a group is not proof that the event is impossible, and a recorded event is not proof that the study drug caused it. The strongest interpretation supported by the record is therefore limited: events were recorded in the listed groups during the extension, while the design and reporting do not support definitive causal or between-dose conclusions.
13. Why This Trial Matters Statistically
Extension studies address a different statistical question from short-term randomized trials: they describe experience during continued follow-up among people who have already participated in earlier studies. This can be useful for documenting longer-term safety observations and retaining information about participants’ prior treatment pathways. In this study, the registry record specifies follow-up up to 197 weeks and presents counts across cohorts that entered from AG348-C-006 or AG348-C-007.
The design also makes clear why long-term description should not be mistaken for a controlled comparison. Non-randomized allocation and open-label treatment leave potential differences among cohorts and doses that a simple count cannot disentangle. The study’s structure therefore matters as much as the event totals when interpreting the posted results. Counts are informative about what was recorded in each group, but not by themselves about comparative or causal effects.
The distinction between a participant count and an event count is also important. The primary outcome concerns participants with at least one event in specified categories. That summary does not show recurrence, event duration, or how events changed over time. The listed data also distinguish serious-event counts from another category whose definition is not reported in the registry record, underscoring the need to know exactly what each endpoint captures.
Finally, the record’s extensive outcome list alongside zero listed analyses illustrates the difference between registering outcomes and reporting interpretable analysis results. Without complete outcome values, analysis methods, uncertainty estimates, and missing-data details, readers cannot assess precision or multiplicity handling. The study remains statistically informative as a description of its extension design and posted safety observations, but the limits of its non-randomized structure and incomplete analysis reporting should remain central to interpretation.
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 principal design limitation for comparative interpretation is non-randomized allocation. Cohorts were defined by prior study and treatment history, so group differences may reflect those histories or other unreported characteristics. The study was also open-label, and the registry record does not describe whether adverse events were independently adjudicated or how event ascertainment was standardized.
The posted safety results do not provide a complete outcome-by-outcome analysis of the primary endpoint. The endpoint includes all TEAEs, serious TEAEs, related TEAEs, and TEAEs of severity grade 3 or higher, while the available group listing gives serious and “other” counts without defining the latter category or separately reporting every endpoint component. The relationship between the listed counts and the full safety analysis set is not fully described.
Follow-up is described as up to 197 weeks, but group-specific exposure duration and person-time are not reported in the registry record. Counts alone may not reflect differing opportunities to observe events. The number analyzed for each outcome, participant flow, withdrawals, missing-data handling, and sensitivity analyses are also not reported in the registry record. The primary outcome population description is truncated, and the complete SAP details are not reported.
Finally, the registry record lists 28 outcome measures but zero analyses. Numerical efficacy results and inferential safety analyses are not reported in the registry record. These reporting limitations, together with the absence of a randomized control group, prevent a precise causal assessment of treatment effects or robust comparisons among cohort-dose groups.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Extension Study of AG-348 in Adult Participants With Pyruvate Kinase Deficiency Previously Enrolled in AG-348-006 or AG348-C-007 — NCT03853798.
- Registry record last updated 2025-11-18; status last verified 2025-11.
- Results first posted 2025-11-18.
- 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
This completed Phase 3 extension study evaluated long-term safety, tolerability, and efficacy of mitapivat in adults with pyruvate kinase deficiency previously enrolled in AG348-C-006 or AG348-C-007. It was open-label, non-randomized, and parallel, with enrollment listed as 90. The study started on 2019-03-21, and primary completion is recorded as 2024-07-03.
The primary outcome counted participants with specified treatment-emergent adverse events up to 197 weeks. The posted safety listing gives serious-event and other-event counts by cohort and dose, as well as deaths and selected serious-event terms. Cohorts reflect different prior study and treatment histories, so they are not like-for-like control groups. The registry record lists 28 outcome measures and zero analyses; full inferential results, analysis details, and missing-data methods are not reported in the registry record.