A Study of AG-348 in Adult Participants With Pyruvate Kinase (PK) Deficiency: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of AG-348 in Adult Participants With Pyruvate Kinase (PK) Deficiency (NCT02476916), a randomized phase 2 study. The primary outcome is Percentage of Participants Experiencing at Least One Adverse Event (AEs) in the Core Period.

2. Clinical Question

The trial addressed safety and efficacy questions for AG-348 in adults with pyruvate kinase deficiency. Its primary endpoint focused specifically on safety: whether participants experienced at least one adverse event during the Core Period, up to Week 24. The registry record describes two randomized initial-dose groups, AG-348 50 mg BID and AG-348 300 mg BID. The posted primary results compare the percentage with at least one adverse event in those groups.

The primary endpoint is an event-occurrence measure, not a measure of the number, duration, severity, or causal attribution of all adverse events. The registry definition includes an untoward medical occurrence associated with drug use whether or not it was considered related to the study drug. The comparison therefore describes participants with at least one qualifying event, rather than establishing that AG-348 caused each event.

3. Trial Design

The registry record describes a Phase 2, randomized, open-label, dose-ranging study with a parallel design. Enrollment was 52 participants. The study status is completed, with a start date of 2015-06-26 and primary completion on 2017-05-08. It was sponsored by Agios Pharmaceuticals, Inc. The stated purpose was to evaluate safety and efficacy of different AG-348 dose levels in participants with pyruvate kinase deficiency, alongside pharmacokinetic and pharmacodynamic evaluation in the official title.

Participants were assigned to initial-dose groups labeled AG-348 50 mg BID and AG-348 300 mg BID. The posted primary safety results include 27 participants in the 50 mg BID group and 25 in the 300 mg BID group. The registry descriptions state that participants were assigned initial doses but could be treated across a range of doses over the Core Period, with treatment changes related to treatment-emergent adverse events and hemoglobin levels. The descriptions are truncated, so the full dose-adjustment criteria are not reported in the registry record.

The Core Period was 24 weeks, and the primary endpoint timeframe was up to Week 24. The registry also lists secondary outcomes extending to Month 102, including adverse events and changes in hemoglobin, hematocrit, reticulocyte count, and haptoglobin. The presence of these longer timeframes indicates that the record includes outcomes beyond the primary Core Period question, but the details of follow-up and the available analysis populations for each outcome are not reported in the registry record.

Open-label means that masking was not used. Participants and study personnel were not blinded according to the registry design description. Randomized allocation and a parallel structure define the trial's broad comparison, but the method used to generate the allocation sequence, allocation concealment, and any stratification are not reported in the registry record. The posted primary results compare two initial-dose assignments of the same study drug, rather than two different treatment classes.

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 52 participants were enrolled in the Core Period of the study. Participants were randomized 1:1 to receive AG-348 50 mg or AG-348 300 mg. Participants who completed the 24-week Core Period, had clinical activity, and tolerated the AG-348 dose entered the Extension Period for up to 102 months.

Core Period (up to 24 Weeks)

Reasons for not completing this period:

Extension Period (Week 25- 102 Months)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Safety Analysis Set included all participants who had received at least one dose of study drug.

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: The Full Analysis Set included all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Experiencing at Least One Adverse Event (AEs) in the Core Period, measured over Up to Week 24.

The primary endpoint is the percentage of participants experiencing at least one adverse event in the Core Period, up to Week 24. An adverse event was defined in the registry as any untoward medical occurrence associated with the use of a drug in humans, whether or not considered study-drug-related. The outcome is binary at the participant level: a participant either experienced at least one event during the stated timeframe or did not. It does not count how many separate events each participant experienced.

This type of endpoint is used to describe how commonly participants in a group had any qualifying adverse event during a specified period. It offers a broad safety summary, but it does not by itself describe event severity, seriousness, duration, recurrence, or whether an event was judged related to the study drug. Those are distinct aspects of safety that require additional outcome definitions and results.

The primary safety analysis set included all participants who had received at least one dose of study drug. The posted percentages are 96.3% for AG-348 50 mg BID and 100.0% for AG-348 300 mg BID. The registry comparison reports a risk difference of 0.037, with a confidence interval from -0.034 to 0.108, and an unadjusted p-value of 0.331. These statistics describe the observed group results and their uncertainty; they do not identify the cause of each adverse event or resolve every safety question addressed by the broader trial.

7. Results Posted in the Registry

• AG-348 300 mg BID: 100.0% (25/25) versus 96.3% (26/27) on control; risk difference 3.7 percentage points (95% CI -3.4 to 10.8), risk ratio 1.04 (0.96–1.12); unadjusted two-proportion z-test p = 0.331.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: at least one adverse event during the specified period. For a binary outcome, a group percentage describes the proportion of participants classified as having the event. The posted values are 96.3% in the AG-348 50 mg BID group and 100.0% in the AG-348 300 mg BID group. The corresponding group sizes are 27 and 25, and the comparison lists 26 and 25 participants with the event. These are counts of participants meeting the endpoint, not counts of every adverse event experienced.

What does a risk difference describe?

In general, a risk difference compares the event probability in one group with the event probability in another by subtraction. It is an absolute-scale measure: it describes how far apart the two event rates are on the probability scale. The registry reports a risk difference of 0.037 for the 300 mg BID group relative to the 50 mg BID group. Its confidence interval is -0.034 to 0.108. Because that interval extends to either side of no difference, the data do not estimate the direction with high precision.

The risk difference is distinct from the displayed percentages. It summarizes the contrast between them on a statistical scale, while each percentage describes one group. The registry has not described an adjusted risk difference, so the posted estimate should be read as the stated comparison rather than as a model-adjusted result. The statistical model or calculation method used for the sponsor's comparison is not reported in the registry record.

What does a relative risk add?

In general, a relative risk compares the event probability in one group with that in another as a ratio. It expresses the group contrast relative to the reference group's event probability rather than as an absolute difference. The registry reports a relative risk of 1.038, with a confidence interval from 0.964 to 1.118. The interval crosses the ratio's no-difference reference, so it allows for a range of relative comparisons around that reference. A relative risk and a risk difference can look different because they use different scales; neither should be treated as a complete description of event severity or clinical importance.

How should the p-value be read?

The unadjusted p-value is 0.331. In general, a p-value describes how compatible the observed data are with a specified null hypothesis under the test's assumptions. It is not the probability that the null hypothesis is true, and it does not give the chance that a future participant will experience an event. It also does not describe the magnitude or precision of the estimated difference. The registry record does not report the exact test, its assumptions, or whether any adjustment was made for the many outcomes in the study. The value is explicitly unadjusted, so it should not be described as accounting for multiple comparisons.

The posted result is based on the Safety Analysis Set, defined as participants who received at least one dose of study drug. This is the stated population for the primary safety endpoint. It is not enough information to determine how all randomized participants were handled in every analysis, or whether the primary analysis included all participants as randomized regardless of treatment received. The analysis population for each other outcome is not reported in the registry record.

Why do small groups affect interpretation?

In general, uncertainty in an estimated event rate depends in part on the number of participants observed and on how common the event is. With 27 and 25 participants in the primary comparison, each participant's event status contributes materially to the group estimate. The confidence interval communicates uncertainty around the comparison rather than just uncertainty around either group percentage alone. It is especially important not to treat the displayed percentages as exact underlying event rates for all adults with pyruvate kinase deficiency.

For binary endpoints, a confidence interval gives a range of values compatible with the observed data and the method used to construct it. Different statistical procedures can produce different intervals, particularly in small samples or when an outcome is very common. The confidence-interval method is not reported in the registry record. No additional recalculated result is presented in the available record, and no sponsor-versus-recalculation discrepancy is indicated.

What do the planning illustrations mean?

In general, sample-size planning uses assumptions about an effect to estimate the number of participants that might be required under a chosen design and statistical approach. Any planning illustrations associated with the results are not observed power and are not evidence that the trial had a particular chance of detecting its observed result. The registry record gives illustration assumptions based on fractions of the observed effect, two-sided significance level 0.05, equal allocation, and a normal approximation. These are planning assumptions, not the trial's documented design rationale. The actual sample-size justification is not reported in the registry record.

The primary comparison is only one of many registered outcomes: the record lists 34 outcome measures and 33 posted secondary results. A large outcome set raises questions about multiplicity, prespecification, and selective reporting. The registry record does not report an adjustment strategy or enough detail to establish which analyses were prespecified. Accordingly, the unadjusted p-value should be read as a result for the stated comparison, not as a multiplicity-controlled conclusion about the trial as a whole.

10. Confidence Intervals and Precision

The primary group percentages are 96.3% for AG-348 50 mg BID and 100.0% for AG-348 300 mg BID. The registry reports a risk-difference estimate of 0.037 with a confidence interval from -0.034 to 0.108. It also reports a relative risk of 1.038 with a confidence interval from 0.964 to 1.118. These intervals are central to interpretation because they show that the estimated contrast is not precise enough to establish a clear direction.

For the risk difference, the interval extends below and above the no-difference point. In practical terms, values on either side remain compatible with the data under the interval method. The interval also includes contrasts that may be more substantial than the point estimate. The point estimate alone therefore should not be read as a definitive description of the underlying difference.

The relative-risk interval likewise crosses its no-difference reference. It allows for relative contrasts in either direction, rather than supporting a sharply determined ratio. The two intervals use different scales, and their numerical limits are not interchangeable. The risk difference describes an absolute contrast; the relative risk describes a ratio. Both show uncertainty in the comparison.

A confidence interval is not a guarantee that the true value lies within the displayed limits, nor does it assign equal probability to every value in the interval. In general, it is a range generated by a statistical procedure with a stated coverage property under its assumptions. The procedure used for the posted intervals is not reported in the registry record. The small group sizes and the very common event make cautious interpretation especially important, but the record does not provide enough detail to determine how the interval method addressed these features.

The unadjusted p-value of 0.331 is consistent with substantial uncertainty, but it should not substitute for the intervals or be treated as proof of equivalence. The registry does not report a prespecified equivalence or noninferiority margin. It also does not report a target precision for the primary safety comparison. Thus, the record supports describing the range of uncertainty, but not concluding that the initial-dose groups have the same event rate.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 34 outcome measures and 33 posted secondary results, in addition to the primary endpoint. These include longer-term adverse-event reporting and laboratory measures such as hemoglobin, hematocrit, reticulocyte count, and haptoglobin. When a study evaluates many outcomes or comparisons, the chance of at least one apparently notable result can increase. In general, multiplicity procedures can address that issue, but the registry record does not report a multiplicity-adjustment plan or whether the outcomes were organized into prespecified testing families.

The primary comparison has an unadjusted p-value of 0.331. It should therefore be understood as unadjusted, not as evidence that accounts for testing across the study's outcome set. The registry record does not report interim analyses, a statistical analysis plan, or the timing of any analysis decisions. The criteria used to select which secondary results were posted are also not reported in the registry record.

The primary safety analysis set is defined as all participants who had received at least one dose of study drug. That definition identifies the stated population for this endpoint, but the record does not report how missing endpoint information was handled, whether any participants had missing adverse-event data, or how withdrawals were treated. The missing-data approach for secondary outcomes is likewise not reported in the registry record.

Randomization is reported, but the record does not state whether the primary safety analysis followed randomized assignment for every participant or how deviations from assigned initial dose were handled. It also does not describe allocation concealment, the randomization sequence, or an analysis plan for dose adjustments. These omissions limit assessment of analysis choices and potential sources of bias; they do not establish that a particular bias occurred.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: haemolytic anaemia (7.4%), haemolysis (4.0%), hypertriglyceridaemia (4.0%), cellulitis (4.0%), gastroenteritis (4.0%).

The serious adverse-event summary covers the Core Period plus Extension Period, a longer window than the primary endpoint's Core Period through Week 24. The registry lists 12 participants with a serious adverse event among 27 at risk in the AG-348 50 mg BID group and 7 among 25 at risk in the AG-348 300 mg BID group. It lists 27 participants with other adverse events among 27 at risk in the 50 mg BID group and 25 among 25 at risk in the 300 mg BID group. No deaths are listed in either group.

These are participant counts, not counts of all serious-event episodes. The group labels refer to the initial-dose groups, while the descriptions indicate that dosing could range across levels during the study. The registry does not state in this summary whether participants with multiple serious events are counted once or how event attribution was determined. The serious-event figures should therefore be read as the posted participant-level summary, not as a complete characterization of each event.

The listed serious terms and percentages include Haemolytic anaemia at 7.4%, and Haemolysis, Hypertriglyceridaemia, Cellulitis, Gastroenteritis, and Infected bite at 4.0% each. The record does not provide enough detail in this summary to determine the event counts behind each percentage, how overlapping terms were handled, or whether the terms were considered related to study drug. These longer-period summaries are distinct from the primary Core Period endpoint and should not be substituted for it.

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

The posted primary results describe a high frequency of at least one adverse event in both randomized initial-dose groups during the Core Period: 96.3% with AG-348 50 mg BID and 100.0% with AG-348 300 mg BID. The estimated risk difference is 0.037, with a confidence interval from -0.034 to 0.108, and the unadjusted p-value is 0.331. The interval crosses the no-difference point, so the observed comparison does not precisely determine whether the group event rates differed or in which direction.

This result does not establish that the dose groups are equivalent. A failure to obtain a small unadjusted p-value is not a demonstration of equality, particularly when the interval remains broad enough to include contrasts in either direction. The registry record does not report an equivalence margin or a noninferiority margin, so the comparison cannot be interpreted as a formal equivalence or noninferiority analysis.

Nor does the primary endpoint show that AG-348 caused every reported adverse event. The event definition includes occurrences whether or not considered study-drug-related. The endpoint also does not separate mild from severe events, serious from non-serious events, or events that occurred once from those that recurred. Those details require separate summaries and definitions.

The comparison is by randomized initial-dose assignment, while the registry descriptions state that participants could be treated across a range of doses during the Core Period. The primary result therefore does not isolate the effect of maintaining either initial dose throughout the period. Details of the dose histories and full adjustment criteria are not reported in the registry record.

Finally, this primary endpoint answers a limited safety question over a specified timeframe in the study's safety population. It does not establish longer-term safety, efficacy, or applicability to people outside the population studied. The listed secondary outcomes and extended-period adverse-event summaries address other questions, but their methods and missing-data handling are not reported in the registry record. The most defensible reading is descriptive: adverse events were common in both groups, and the between-group estimate is uncertain.

14. Why This Trial Matters Statistically

This trial illustrates why a binary safety endpoint and its comparison need to be read together. The percentages indicate that at least one adverse event was recorded for nearly all participants in both initial-dose groups, but the risk-difference and relative-risk intervals show that the contrast between groups remains uncertain. Reporting only the percentages would conceal the uncertainty in the comparison; reporting only a p-value would conceal both the size of the observed estimates and the range of plausible contrasts.

The study also highlights the importance of distinguishing randomized initial assignment from actual treatment exposure. The registry describes dose-ranging treatment and possible movement across dose levels during the Core Period. A comparison labeled by initial dose is not automatically a comparison of sustained exposure to two fixed doses. Knowing the dose history and adjustment rules would be important for interpreting dose-specific safety patterns, but those details are not fully reported in the registry record.

Its open-label design and small group sizes are further context for understanding the evidence. Open-label status means masking was not used, while modest group sizes contribute to uncertainty around estimates. Neither feature alone establishes that the results are biased, but both matter when assessing how confidently the observed difference can be interpreted. The unadjusted p-value and confidence intervals should be considered alongside the outcome definition and analysis population.

Finally, the registry lists many outcome measures and extended follow-up outcomes. That makes transparency about the statistical analysis plan, missing data, multiplicity, and the relationship among primary and secondary outcomes particularly useful. Those details are not reported in the registry record, so the posted primary comparison is best treated as one result within a broader study rather than a complete account of every safety or efficacy question.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The primary comparison includes 27 participants in the AG-348 50 mg BID group and 25 in the AG-348 300 mg BID group. The confidence intervals are broad enough to allow for differences in either direction, limiting precision. The registry does not report a target precision or a sample-size justification for this comparison. Planning illustrations based on the observed effect are not evidence of observed power and should not be used to imply that the trial was adequately powered for this endpoint.

The treatment descriptions indicate that participants could receive a range of doses after assignment to an initial dose. The precise criteria and actual exposure histories are not fully reported in the registry record. As a result, the primary comparison cannot be interpreted as a clean comparison of sustained exposure to the two labeled initial doses. The registry also does not report the randomization method, allocation concealment, or the details of any dose-adjusted analysis.

The study was open-label. The record does not report whether outcome assessment involved independent review or other procedures to reduce the influence of unmasked treatment knowledge. This is relevant context for recorded adverse events, but it does not establish that event reporting was biased.

The primary endpoint combines all participants with at least one adverse event, without distinguishing seriousness, severity, recurrence, duration, or study-drug relationship. The posted serious-event summary uses a Core Period plus Extension Period timeframe, which differs from the primary endpoint timeframe. The registry does not report enough detail to reconcile individual event-level patterns across these summaries.

Important analysis details are not reported in the registry record, including the method used for the confidence intervals and p-value, missing-data handling, multiplicity procedures, and a statistical analysis plan. These limitations constrain independent assessment of the analysis, but should not be mistaken for evidence that a particular method or result is incorrect.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of AG-348 in Adult Participants With Pyruvate Kinase (PK) Deficiency — NCT02476916.
• Registry record last updated 2026-05-01; status last verified 2026-04.
• Results first posted 2020-06-11.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry record describes NCT02476916, a completed Phase 2, randomized, open-label, parallel dose-ranging study of AG-348 in adults with pyruvate kinase deficiency. Enrollment was 52. The primary endpoint was the percentage experiencing at least one adverse event up to Week 24, analyzed in participants who had received at least one dose of study drug.

The posted primary results are 96.3% for AG-348 50 mg BID and 100.0% for AG-348 300 mg BID. The reported risk difference is 0.037, with a confidence interval from -0.034 to 0.108; the unadjusted p-value is 0.331. The interval reflects substantial uncertainty about the direction and size of the group contrast. The record also lists secondary outcomes and adverse-event summaries extending through the Extension Period, but important analysis and missing-data details are not reported in the registry record.