A Study of Efficacy and Safety of LAG525 in Combination With Spartalizumab, or With Spartalizumab and Carboplatin, or With Carboplatin, in Patients With Advanced Triple-negative Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Efficacy and Safety of LAG525 in Combination With Spartalizumab, or With Spartalizumab and Carboplatin, or With Carboplatin, in Patients With Advanced Triple-negative Breast Cancer (NCT03499899), a randomized phase 2 study. The primary outcome is Overall Response Rate (ORR) Per Investigator's Assessment According to RECIST v1.1.

2. Clinical Question

3. Trial Design

This was a phase II, randomized, open-label, parallel-group, multicenter study of LAG525 combinations in participants with advanced triple-negative breast cancer receiving first or second line therapy. The registry lists an enrollment of 88 and the study status as completed. The study start date is 2018-07-02, and the primary completion date is 2020-02-27.

Participants were assigned to one of three groups: LAG525 + PDR001, LAG525 + PDR001 + carboplatin, or LAG525 + carboplatin. PDR001 is identified in the study title as spartalizumab. Each arm description specifies infusion administration once every 3 weeks. The registry describes the study as open-label and reports no masking.

The primary analysis population is the Full Analysis Set, defined as participants to whom study treatment was assigned by randomization. The posted primary endpoint is overall response rate per investigator’s assessment according to RECIST v1.1, evaluated up to approximately 14 months. The registry also lists secondary outcomes involving clinical benefit, response duration and timing, progression-free and overall survival, and pharmacokinetic measures for LAG525.

The registry record does not report the allocation ratio intended by the original design, detailed eligibility criteria, the rationale for the planned enrollment, or the original statistical design assumptions. The outcome counts differ by arm, but that alone does not establish the planned allocation ratio or explain why the analyzed group sizes differ. Detailed procedures for randomization, stratification, treatment discontinuation, and follow-up beyond the arm descriptions 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 132 participants were screened of which 88 participants were enrolled in the study and 87 participants received at least one dose of study treatment (1 participant was not treated due to physician decision)

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set (FAS): All participants to whom study treatment was assigned by randomization.

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.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate (ORR) Per Investigator's Assessment According to RECIST v1.1, measured over Up to approximately 14 months.

Overall response rate is the percentage of participants whose best overall response was complete response or partial response, assessed by the investigator according to RECIST v1.1. It is a binary participant-level outcome: each participant is classified according to whether a qualifying response was recorded. The registry specifies that the assessment covered up to approximately 14 months.

The registry’s definition describes complete response as disappearance of all non-nodal target lesions, with any pathological lymph nodes assigned as target lesions reduced in short axis to < 10 mm. Partial response requires at least a 30% decrease in the sum of the diameters of all target lesions, using the baseline sum as reference. These definitions make the response classification depend on specified changes in measured lesions rather than on an unstructured judgment of improvement.

Response rate is useful for describing the proportion of participants meeting those response criteria within each treatment group. It does not, by itself, describe how long a response lasts, when it first occurs, or how long participants remain alive or free from progression. The registry lists duration of response, time to response, progression-free survival, and overall survival as separate secondary outcomes. The response-rate estimates therefore address a specific endpoint and should not be read as summaries of every aspect of disease course.

7. Results Posted in the Registry

• LAG525+ PDR001+ Carboplatin: 32.4% (11/34) versus 5.0% (1/20) on control; risk difference 27.4 percentage points (95% CI 9.0 to 45.8), risk ratio 6.47 (0.90–46.45); unadjusted two-proportion z-test p = 0.020.
• LAG525 + Carboplatin: 17.6% (6/34) versus 5.0% (1/20) on control; risk difference 12.6 percentage points (95% CI -3.3 to 28.6), risk ratio 3.53 (0.46–27.24); unadjusted two-proportion z-test p = 0.182.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Clinical Benefit Rate (CBR) Per Investigator's Assessment According to RECIST v1.1

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • LAG525+ PDR001+ Carboplatin: 35.3% (12/34) versus 5.0% (1/20) on control; risk difference 30.3 percentage points (95% CI 11.6 to 49.0), risk ratio 7.06 (0.99–50.30); unadjusted two-proportion z-test p = 0.012.
  • LAG525 + Carboplatin: 20.6% (7/34) versus 5.0% (1/20) on control; risk difference 15.6 percentage points (95% CI -1.0 to 32.2), risk ratio 4.12 (0.55–31.08); unadjusted two-proportion z-test p = 0.119.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for LAG525+ PDR001+ Carboplatin than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LAG525 + Carboplatin and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With Anti-drug Antibodies (ADA) at Baseline for PDR001

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • LAG525+ PDR001+ Carboplatin: 17.6% (6/34) versus 16.7% (3/18) on control; risk difference 1.0 percentage points (95% CI -20.5 to 22.4), risk ratio 1.06 (0.30–3.74); unadjusted two-proportion z-test p = 0.929.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LAG525+ PDR001+ Carboplatin and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With Anti-drug Antibodies (ADA) on Treatment for PDR001

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • LAG525+ PDR001+ Carboplatin: 0.0% (0/25) versus 6.7% (1/15) on control; risk difference -6.7 percentage points (95% CI -19.3 to 6.0); unadjusted two-proportion z-test p = 0.191.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LAG525+ PDR001+ Carboplatin and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes extend beyond the primary response-rate question. They include clinical benefit rate, duration and timing of response, progression-free survival, overall survival, and pharmacokinetic measurements for LAG525. The registry also lists anti-drug antibody outcomes for PDR001. Together these measures describe different dimensions of response, time course, survival, drug exposure, and immune response; they are not interchangeable confirmations of the primary endpoint.

The posted clinical benefit rates follow a pattern in which the highest observed percentage is in the LAG525 + PDR001 + carboplatin arm, with the other carboplatin-containing arm also above the LAG525 + PDR001 arm. This resembles the primary response-rate pattern at a broad descriptive level. The comparison intervals and unadjusted p-values should still be read with their own uncertainty, and a similarity in observed patterns does not establish that the endpoints share the same underlying effect or analysis properties.

For several other secondary outcomes, the registry lists endpoint definitions or participant counts but does not provide a reported analysis in the comparison summaries. The overall pattern across secondary endpoints therefore cannot be reduced to a single confirmatory conclusion. In particular, outcome listings alone do not establish results for duration of response, time to response, progression-free survival, overall survival, or the pharmacokinetic measures.

Unadjusted secondary results carry less weight when many outcomes or comparisons are considered because some apparently unusual results may occur by chance across a broad set of analyses. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for the secondary endpoints. The secondary results are consequently best regarded as descriptive or exploratory unless additional analysis-plan information establishes otherwise. Their role is to provide context and generate questions, not to replace the primary endpoint or independently settle the treatment comparison.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Clinical Benefit Rate (CBR) Per Investigator's Assessment According to RECIST v1.1
• Duration of Response (DOR) Per Investigator's Assessment According to RECIST v1.1
• Time to Response (TTR) Per Investigator's Assessment According to RECIST v1.1
• Progression Free Survival (PFS)
• Overall Survival (OS)
• Pharmacokinetics (PK) Parameter, Area Under the Plasma Concentration Versus Time Curve From Time 0 to 504 Hours (AUC0-504h) of LAG525
• PK Parameter, Cmax of LAG525
• PK Parameter, AUClast of LAG525

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a binary response outcome assessed by the investigator using RECIST v1.1. For each arm, the registry reports the number of participants in the analysis group, the response percentage, and a 95% confidence interval. It states that the intervals were computed using a two-sided exact binomial method. An exact binomial interval is a method for quantifying uncertainty in a proportion when outcomes are classified as events or non-events. The method does not turn a small group’s observed percentage into a precise estimate; its interval communicates the range of values reasonably compatible with the observed count under the method’s assumptions.

The comparison summaries use risk differences, expressing the response proportion in an arm relative to the reference arm on an absolute scale. The reported comparison estimates and their confidence intervals are unadjusted calculations from the posted group summaries. A risk difference near zero corresponds to similar observed proportions; an interval extending across zero indicates that the summary data are compatible with differences in either direction. These estimates should be interpreted alongside the arm-level rates and their uncertainty, rather than as a replacement for them.

The comparison p-values are also unadjusted. A p-value describes how compatible the observed data are with a specified null comparison under the test used; it is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The registry record does not report the statistical test used for these comparison p-values, any adjustment for multiple comparisons, or a prespecified testing sequence. The unadjusted p-values therefore do not establish that the comparisons were planned as confirmatory tests.

The Full Analysis Set includes participants assigned study treatment by randomization, according to the registry definition. The posted primary outcome group sizes are 20 for LAG525 + PDR001 and 34 for each other arm. That definition clarifies the stated analysis population but does not describe how missing response assessments were handled, whether all assigned participants had evaluable outcome data, or how incomplete follow-up was treated. Those details are not reported in the registry record.

The study is described as randomized, open-label, parallel, and multicenter. Random assignment can help balance participant characteristics on average, but the registry information alone does not establish balance in the realized groups or reveal all randomization procedures. Open-label means masking was absent according to the record. Because the primary outcome was investigator-assessed, knowledge of treatment assignment could be relevant to how assessments were made; the posted information does not specify blinded central review or other safeguards.

The registry lists secondary outcomes spanning response, survival, and pharmacokinetics. These outcomes answer distinct questions and can involve different analysis populations, time scales, and methods. The posted material identifies 25 secondary outcomes, but reported analyses are not listed for the secondary endpoints represented in the comparison summaries. The record’s broad outcome list does not provide enough detail to reconstruct all analysis methods or determine how incomplete observations were handled for each outcome.

The sample-size illustrations are a separate planning exercise based on observed effects, not a description of the original trial design. The listed assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. The original sample-size assumptions and justification are not reported in the registry record. A planning calculation that uses an effect observed in the same study is not evidence that the study was designed around that effect, and the figures should not be conflated with the analysis of the observed trial results.

No interim analyses, stopping rules, covariate adjustments, stratified analyses, sensitivity analyses, or analysis-plan details are reported in the registry record. Their absence from the record does not prove that none were conducted; it means the posted information does not permit those methods to be evaluated. Likewise, the registry provides outcome definitions and some calculation details, but not a complete statistical analysis plan.

11. Confidence Intervals and Precision

The primary endpoint confidence intervals show how uncertain each estimated response proportion is under the stated exact binomial method. For LAG525 + PDR001, the rate was 5.0% and the 95% interval was 0.1% to 24.9%. For LAG525 + PDR001 + carboplatin, the rate was 32.4% and the interval was 17.4% to 50.5%. For LAG525 + carboplatin, the rate was 17.6% and the interval was 6.8% to 34.5%. The interval widths indicate that the observed percentages should not be mistaken for highly precise estimates.

A confidence interval is a statistical range produced by a method that, under its assumptions, reflects sampling uncertainty. It does not mean there is a 95% probability that the fixed true response rate lies inside this particular calculated interval. The interval is also not a range containing 95% of individual participant outcomes. The registry identifies the exact binomial method for the primary endpoint intervals.

The unadjusted risk-difference interval for the triple combination versus LAG525 + PDR001 runs from 0.089 to 0.458. The interval for LAG525 + carboplatin versus LAG525 + PDR001 runs from -0.033 to 0.286. The first excludes zero under the reported calculation; the second includes zero. These intervals describe uncertainty in absolute differences and should be read with the small arm sizes and the unadjusted nature of the comparisons in mind.

Confidence intervals do not address every source of uncertainty. They do not account automatically for bias from open-label outcome assessment, missing data, multiple comparisons, or departures from the analysis assumptions. The registry record does not report enough detail to evaluate all of those issues. The intervals are therefore informative measures of statistical precision, but not a complete assessment of the reliability or generalizability of the findings.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations report participants per arm for 80% power and participants per arm for 90% power. They are not total event counts: the registry’s stated convention is that time-to-event planning entries use total events, while the other entries use participants per arm. For the response comparisons, read each entry as a hypothetical planning requirement for one arm under the stated assumptions, not as a target for the entire study.

The assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario uses the stated fraction of the observed effect as the effect to detect. The trial’s own original sample-size justification and design assumptions are not reported in the registry record.

The 80% power column gives the number of participants per arm that the planning calculation associates with an 80% chance of detecting the specified effect under those assumptions, if that effect and the model assumptions held. The 90% power column gives the corresponding planning requirement for a higher chance of detection. These are design concepts, not measures of what happened in the completed study. The figures are not observed power.

The rows that plan for 100% of the observed effect use the full observed difference as the hypothetical target. Rows planning for 75% or 50% of the observed effect use progressively smaller target effects. Smaller effects are harder to distinguish from chance variation, so the calculations require more participants per arm for the same power target. Since the observed effects can be imprecise, these illustrations should not be treated as forecasts of a future study’s results or as evidence that the original enrollment was sufficient or insufficient.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.050; treatment proportion p₁ = 0.324; average p̄ = 0.187.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5512 = 1.0802.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5161 = 0.4344.
4. n = (1.0802 + 0.4344)² / (0.2735)² = 30.7, rounded up to 31 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.05                          # control proportion
p1 <- 0.323529411764706                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3499899)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 31 (80% power), 41 (90% power)
 75% of observed effect: n per arm = 48 (80% power), 63 (90% power)
 50% of observed effect: n per arm = 87 (80% power), 116 (90% power)
power.prop.test: 30.7 
Simulated power at n = 31 per arm: 0.842 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

The study has three treatment arms and multiple listed outcomes. The primary endpoint compares two arms separately with the LAG525 + PDR001 reference group, and the registry also lists secondary response, survival, pharmacokinetic, and anti-drug antibody outcomes. Considering multiple comparisons increases the chance that at least one result may appear notable by chance, even when no underlying difference exists. This general issue is called multiplicity.

The posted risk-difference p-values are unadjusted, and the registry record does not report a multiplicity-control procedure, a prespecified order of hypothesis tests, or a division between confirmatory and exploratory comparisons. It is therefore not possible from the registry record to determine whether the reported comparisons were intended to support a family-wide claim or whether any adjustment was applied elsewhere. The unadjusted values should not be treated as if they automatically account for the full set of outcomes and comparisons.

Missing outcome information is another consideration. The Full Analysis Set is defined as all participants assigned treatment by randomization, but the registry record does not state how missing response assessments, incomplete follow-up, or other unavailable measurements were handled. It also does not report sensitivity analyses evaluating alternative assumptions about missing outcomes. The stated analysis population is useful, but it does not by itself resolve how every missing or unevaluable observation was treated.

The record does not report interim analyses, stopping rules, or the original sample-size justification. It also does not provide enough statistical-plan detail to evaluate prespecification of every outcome analysis. These matters are not necessarily absent from the study’s full documentation; they are not reported in the registry record. Without that information, interpretation should remain focused on the posted results and their stated methods rather than assuming procedures that are not documented.

15. 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: dyspnoea (8.8%), anaemia (5.9%), pleural effusion (5.9%), fanconi syndrome (5.3%), disease progression (5.3%).

The safety results distinguish the on-treatment period from the extended safety follow-up period. For serious adverse events, the record gives the number affected and the number at risk in each group. During on-treatment, serious adverse events affected 6 of 19 participants in LAG525 + PDR001, 12 of 34 in LAG525 + PDR001 + carboplatin, and 14 of 34 in LAG525 + carboplatin. During extended safety follow-up, the corresponding counts were 0 of 19, 6 of 34, and 2 of 34.

The record separately reports other adverse events, which should not be conflated with serious adverse events. During on-treatment, other adverse events affected 17 of 19, 34 of 34, and 33 of 34 participants in the three respective groups. The extended follow-up counts were 0 of 19, 11 of 34, and 3 of 34. Death counts are listed by group and period; the record does not establish from these figures that deaths were caused by study treatment.

The listed serious adverse event terms include dyspnoea, anaemia, pleural effusion, Fanconi syndrome, disease progression, and Pneumocystis jirovecii pneumonia. The registry lists percentages for these terms but does not provide further context in the posted summary about attribution or event timing. These counts describe reported safety events and are not, on their own, a comparison adjusted for differences in follow-up or participant characteristics.

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

The primary result describes the proportion of participants in each randomized group whose best overall response met investigator-assessed RECIST v1.1 criteria. It provides a direct summary of observed response within this study, and the exact binomial confidence intervals communicate statistical uncertainty around those proportions. The unadjusted risk differences compare each carboplatin-containing group with LAG525 + PDR001 on an absolute scale.

The results do not establish why the observed arm percentages differ. Randomization is a design feature, but the posted summary does not provide detailed information on realized group balance, missing assessments, or all analysis procedures. The study was open-label, and the primary endpoint relied on investigator assessment; the registry record does not describe blinded central assessment. These features matter when considering how measurements and comparisons may be affected by study conduct.

The results also do not establish that one combination should be used in clinical practice, that it is superior for every relevant outcome, or that the observed percentages will apply to other populations or settings. No medical recommendation follows from these statistics. The record describes a phase II study with a defined population, treatment arms, and response endpoint; it does not provide a no-treatment comparison or a standard-care comparator.

Response rate is not equivalent to duration of response, progression-free survival, or overall survival. Although the registry lists those as secondary outcomes, the outcome listings alone do not convey their results. Nor do response data alone describe adverse effects, treatment burden, or an overall balance of benefits and harms. Safety counts are a separate part of the record and require their own context.

The most defensible reading is therefore bounded: the highest observed primary response percentage was in the LAG525 + PDR001 + carboplatin group, and the estimates and unadjusted comparisons carry substantial uncertainty. The registry record does not provide enough information to turn this pattern into a definitive claim about treatment effect, nor does it document all features needed for a full confirmatory assessment.

17. Why This Trial Matters Statistically

This study illustrates why a response percentage and its uncertainty need to be considered together. The arm-level estimates are easy to compare descriptively, but the confidence intervals show that each estimate remains uncertain. The observed ordering of percentages does not by itself establish the size or reliability of a difference.

It also illustrates the distinction between an arm-level outcome and a comparative estimate. Risk differences express how far apart the observed proportions are on an absolute scale, while the associated confidence intervals show uncertainty about those differences. A p-value answers a different question and should not be used as a measure of effect size or clinical importance.

The three-arm structure creates multiple comparisons, and the registry lists a broad set of secondary outcomes. Without a reported multiplicity strategy, unadjusted comparisons require restraint. This is especially important when a study’s results include several possible ways to characterize response, survival, and treatment exposure.

Finally, the sample-size illustrations demonstrate how planning depends on the effect selected for detection and the intended power. They are distinct from the trial’s original design assumptions, which are not reported in the registry record. Together, these points make the study useful for understanding the roles of endpoint definition, precision, comparison strategy, and transparent reporting in clinical-trial interpretation.

18. 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 posted primary endpoint is based on investigator assessment in an open-label study. The registry record does not report blinded central review, so the information available does not establish whether response assessments were independently verified. The endpoint is also specific to RECIST v1.1 response classification and does not capture response duration or survival on its own.

The primary analysis groups are modest in size, and the response confidence intervals are broad. The unadjusted comparisons are based on the posted summaries, while the record does not report a multiplicity procedure, detailed statistical analysis plan, or prespecified testing hierarchy. This limits the strength of inferential claims based on individual p-values.

The Full Analysis Set is defined, but missing-data handling and the number of participants with evaluable assessments are not described in sufficient detail to assess the impact of incomplete response data. The registry also does not report the original sample-size justification, interim-analysis procedures, or detailed randomization methods. Those omissions do not show that such procedures were absent from the full study documentation; they mean they cannot be assessed from the registry record.

The study evaluates three active combinations without a no-treatment or standard-care comparator in the posted arms. The arm comparisons therefore address differences among these combinations, not whether any combination differs from an external control. The findings are specific to the listed advanced triple-negative breast cancer population and first or second line setting, and the record does not establish broader applicability.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study of Efficacy and Safety of LAG525 in Combination With Spartalizumab, or With Spartalizumab and Carboplatin, or With Carboplatin, in Patients With Advanced Triple-negative Breast Cancer — NCT03499899.
• Registry record last updated 2023-01-30; status last verified 2023-01.
• Results first posted 2022-05-13.
• 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.

22. Record Summary

NCT03499899 is a completed, randomized, open-label, parallel-group, multicenter phase II study of LAG525 combinations for advanced triple-negative breast cancer in first or second line therapy. The registry lists enrollment of 88 and three arms: LAG525 + PDR001, LAG525 + PDR001 + carboplatin, and LAG525 + carboplatin.

The primary endpoint was investigator-assessed overall response rate according to RECIST v1.1, with two-sided exact binomial 95% confidence intervals. The highest posted primary response percentage was in the triple-combination arm. The unadjusted comparisons show uncertainty, and the record does not report multiplicity adjustment or the original sample-size justification.

Secondary outcomes cover clinical benefit, response timing and duration, survival, pharmacokinetics, and anti-drug antibodies. The record also reports adverse-event counts for on-treatment and extended safety follow-up periods. Interpretation is limited by the unadjusted comparisons, uncertainty in the small-group estimates, and details not reported in the registry record about missing data and the full analysis plan.