Study of Sacituzumab Govitecan-hziy Versus Treatment of Physician's Choice in Participants With HR+/HER2- Metastatic Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Sacituzumab Govitecan-hziy Versus Treatment of Physician's Choice in Participants With HR+/HER2- Metastatic Breast Cancer (NCT03901339), a randomized phase 3 study. The primary outcome is Progression-Free Survival (PFS) by Blinded Independent Central Review (BICR) Assessment.

2. Clinical Question

This is a comparison of two randomized treatment strategies in a defined metastatic breast cancer population, not a comparison with no treatment. In the control strategy, the specific single-agent treatment was chosen before randomization. The registry record identifies the study as phase 3, randomized, unmasked, and parallel-group. These design features describe how the comparison was organized; they do not by themselves settle questions about the magnitude, precision, or broader applicability of the result.

The primary result should be read as a time-to-event comparison over the recorded follow-up period. The registry lists a primary outcome timeframe of up to 42.8 months. The reported medians summarize when half of the estimated event-time distribution had been reached, while the hazard ratio summarizes the relative event rate over follow-up under the analysis approach. Neither measure describes every participant’s experience or establishes what would happen outside the trial population.

3. Trial Design

This completed phase 3 study used randomized allocation, an unmasked parallel-group design, and enrolled 543 participants. The registry lists 272 participants in the Sacituzumab Govitecan group and 271 in the Treatment of Physician’s Choice group for the primary endpoint. These endpoint group sizes match the randomized totals shown in the primary result. The record states that the primary analysis population included everyone randomized, regardless of whether study treatment was received.

The experimental strategy was sacituzumab govitecan-hziy, administered intravenously on Day 1 and Day 8 of a 21-day cycle. The comparator was treatment of physician’s choice, determined before randomization from a single-agent treatment option. Eribulin, capecitabine, gemcitabine, and vinorelbine appear in the intervention list. The registry’s available control-arm description is incomplete after the eribulin dosing text, so the full regimen details for each comparator option are not reported in the registry record.

The primary outcome was progression-free survival by blinded independent central review using RECIST 1.1. The analysis compared the time from randomization to the first documentation of progression or death, whichever occurred first. The registry reports estimation using the Kaplan-Meier method and a Log Rank comparison with a hazard ratio. The primary outcome timeframe is up to 42.8 months.

The record also lists overall survival, response outcomes, clinical benefit, investigator-assessed progression-free survival, and quality-of-life deterioration outcomes among the secondary measures. Some of these are posted with results; the registry lists 12 secondary outcomes posted. Their presence helps describe the breadth of questions examined, but does not mean that every endpoint has the same evidentiary role as the primary endpoint.

The original sample-size justification, the trial’s own design assumptions, details of any interim analyses, and a complete prespecified statistical analysis plan are not reported in the registry record. A planning illustration uses stated assumptions, but it should not be mistaken for the original trial design rationale.

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: 776 participants were screened.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The ITT Population included all participants who were randomized, regardless of whether they received study treatment or not.

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 Intent-to-Treat (ITT) Population included all participants who were randomized, regardless of whether they received study treatment or not.

6. Primary Endpoint

The registry lists the primary outcome as Progression-Free Survival (PFS) by Blinded Independent Central Review (BICR) Assessment, measured over Up to 42.8 months.

The primary endpoint was progression-free survival (PFS) assessed by blinded independent central review. It measures the time from randomization until the first documented disease progression or death, whichever occurs first. The registry specifies RECIST 1.1 for assessment. Progression was defined using changes in the sum of the longest diameters of target lesions, including an increase of greater than 20% and an absolute increase of 5 mm, or the appearance of new non-target lesions. The reference for the target-lesion change was the smallest sum recorded since baseline.

PFS is a time-to-event endpoint: it takes account of both whether an event occurred and how long participants remained without a recorded event. This is useful in a trial where the clinical question concerns the time until progression or death and where participants may have different lengths of follow-up. Kaplan-Meier estimation is a general method for describing the event-time distribution while accounting for participants whose event time is not observed before follow-up ends.

The registry reports a median PFS for each group, with a 95% confidence interval, and a hazard ratio from a Log Rank analysis. The median gives a summary point on the estimated time distribution; it does not describe the full distribution. The hazard ratio provides a relative comparison of event rates over the analysis period under the model and analysis assumptions. It is not a ratio of the two medians. The registry reports that the primary endpoint was analyzed in the intention-to-treat population, including all randomized participants whether or not they received treatment.

7. Results Posted in the Registry

• Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC): hazard ratio 0.65 (95% CI 0.53–0.81), reported p < 0.001 (as reported in the registry).

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.

Overall Survival (OS)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC): hazard ratio 0.79 (95% CI 0.65–0.95), reported p = 0.013 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

PFS by LIR Assessment

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC): hazard ratio 0.73 (95% CI 0.60–0.88), reported p = 0.001 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time to Deterioration (TTD) of Global Health Status/Quality of Life (QoL) Scale as Measured by European Organization for Research and Treatment of Cancer Quality of Life for Cancer Patients, Core Questionnaire Version 3.0 (EORTC QLQ-C30)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC): hazard ratio 0.75 (95% CI 0.61–0.92), reported p = 0.006 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Sacituzumab Govitecan vs Treatment of Physician's Choice (TPC) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry posts secondary findings across survival, response, disease control, investigator-assessed progression, and quality-of-life deterioration outcomes. The posted time-to-event results include overall survival, PFS by local investigator review, and time to deterioration in global health status or quality of life. Together, these outcomes broaden the description of the trial beyond the primary central-review PFS endpoint. They use different event definitions, assessment sources, and, for the quality-of-life outcome, a different analysis population from the primary endpoint.

The overall pattern in the posted time-to-event results is directionally consistent with the primary comparison: the reported hazard ratios are below 1 for Sacituzumab Govitecan versus Treatment of Physician’s Choice. This consistency can be informative, but it does not make the secondary endpoints interchangeable with the primary endpoint. For example, overall survival concerns death, while PFS concerns progression or death, and quality-of-life deterioration concerns a patient-reported scale. Each addresses a distinct question and may have different completeness, follow-up, and measurement considerations.

Secondary results generally carry less confirmatory weight when a multiplicity strategy is not reported. When many outcomes or comparisons are examined, the chance of at least one apparently notable result can increase. Statistical procedures that control this issue, and the hierarchy or testing sequence used, are not reported in the registry record. Thus, individual secondary p-values should be regarded as supportive or exploratory unless the full prespecified testing plan establishes otherwise. The record does not provide that plan.

These results can help readers see whether the primary finding sits within a broader pattern of outcomes, but they do not independently prove that every outcome has the same degree of evidence. The tables should be read for each endpoint’s definition, analysis population, timeframe, estimate, and uncertainty. The registry lists 12 secondary outcomes posted, while only a subset of the secondary results are discussed in detail on this page. Results not described are not reported in the registry record section summarized for this page.

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

• Overall Survival (OS)
• Objective Response Rate (ORR) by BICR and Local Investigator Review (LIR) Assessment
• Duration of Response (DOR) by BICR and LIR Assessment
• Clinical Benefit Rate (CBR) by BICR and LIR Assessment
• PFS by LIR Assessment
• Time to Deterioration (TTD) of Global Health Status/Quality of Life (QoL) Scale as Measured by European Organization for Research and Treatment of Cancer Quality of Life for Cancer Patients, Core Questionnaire Version 3.0 (EORTC QLQ-C30)
• TTD of Pain Score as Measured by EORTC QLQ-C30
• TTD of Fatigue Score as Measured by EORTC QLQ-C30

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a time-to-event outcome. Each participant’s time begins at randomization and ends at the first documented progression or death, whichever happens first. In general, time-to-event analyses are designed to use the timing of events rather than reduce the outcome to a simple event/no-event count. Participants who have not had a recorded event by the end of their observed follow-up can contribute information up to that point; their event time is not treated as if it were known to occur at the end of follow-up.

What the median summarizes

The registry reports median PFS in each group, estimated using the Kaplan-Meier method. In general, the Kaplan-Meier curve estimates the probability of remaining event-free over time while accommodating incomplete observation through censoring. The median is the time at which the estimated event-free proportion reaches the median point of the distribution. It is a compact summary, not a description of the whole curve. Two groups can have medians that summarize their distributions while still differing in other aspects of the curves, so the median should not be treated as a complete account of the event-time pattern.

For the primary endpoint, the registry gives median estimates of 5.5 months and 4.0 months, with respective 95% confidence intervals of 4.2 to 6.9 months and 3.0 to 4.4 months. These intervals describe uncertainty around each group’s median estimate. They are not a direct confidence interval for the difference between medians, and their overlap or non-overlap is not a substitute for the reported between-group analysis.

What the hazard ratio describes

The registry reports a hazard ratio of 0.653 for Sacituzumab Govitecan versus Treatment of Physician’s Choice. In general, a hazard is the instantaneous event rate among participants who have not yet experienced the event, at a given time. A hazard ratio compares those rates between groups according to the analysis. The estimate is relative, not an absolute time gain; it does not say how many months an individual participant will remain progression-free. It is also not calculated by dividing the two median values.

The proportional-hazards interpretation often associated with a single hazard ratio assumes that the relative hazards are sufficiently stable over time. The registry record does not report a diagnostic assessment of that assumption. The hazard ratio is therefore best read as the sponsor-reported summary from the stated Log Rank analysis, together with the group-specific Kaplan-Meier medians and confidence intervals. The record does not give enough detail to independently assess every modeling or diagnostic choice.

What the Log Rank comparison contributes

The Log Rank test is a general method for comparing time-to-event distributions across groups. It uses information from event times across follow-up rather than comparing only the medians. The registry reports the method and a p-value of < 0.001 for the primary comparison. In general, a p-value quantifies how unusual data at least as incompatible with a specified null hypothesis would be if that null hypothesis and the test assumptions held. It is not the probability that the observed result occurred by chance, and it does not measure the clinical importance of the result.

The registry lists a 95% confidence interval of 0.526 to 0.812 for the hazard ratio. In general, a confidence interval shows the range of effect estimates that are reasonably compatible with the observed data and the method’s assumptions. A narrower interval indicates greater statistical precision than a wider interval on the same scale, but precision does not guarantee absence of bias or establish clinical relevance. The interval is for the relative hazard estimate, not for either group’s median.

How to read the outcome definition

The primary endpoint used blinded independent central review and RECIST 1.1. The definition includes progression or death, whichever comes first, so a participant can have an endpoint event for either reason. The progression criteria refer to target-lesion size changes and new non-target lesions. In general, a prespecified and structured assessment rule helps make outcome classification more consistent across participants. The registry record specifies the assessment standard, but the full operational details and adjudication procedures are not reported in the registry record.

The primary analysis population is the intention-to-treat population: all randomized participants, regardless of treatment receipt. In general, an intention-to-treat analysis retains the randomized group assignment and is intended to preserve the comparability created by randomization. The record does not provide detailed information on missing assessments, censoring rules, handling of treatment changes, or sensitivity analyses. Those details matter when evaluating how incomplete outcome information may affect a time-to-event analysis.

These methods answer related but different questions. Kaplan-Meier medians summarize each group’s estimated event-time distribution, the hazard ratio summarizes the relative event hazard, and the Log Rank test assesses evidence for a difference between distributions under its assumptions. They should be interpreted together, without treating any single measure as a complete account of benefit, harm, or participant experience.

11. Confidence Intervals and Precision

The primary hazard ratio is 0.653, with a 95% confidence interval from 0.526 to 0.812. This interval indicates uncertainty around the relative hazard estimate: values across that range are compatible with the observed data under the analysis method and its assumptions. It is not a guarantee that the true effect lies within the interval, and it does not account for every possible source of bias or uncertainty in the study.

The interval is entirely below 1. In general, when the event is progression or death and the comparison is Sacituzumab Govitecan versus Treatment of Physician’s Choice, a hazard ratio below 1 corresponds to a lower estimated event hazard in the first group. The interval’s position should be considered alongside the point estimate and the group-specific medians, rather than as proof that the same relative difference applied at every time point. The proportional-hazards assumption and its diagnostics are not reported in the registry record.

The median PFS confidence intervals are 4.2 to 6.9 months for Sacituzumab Govitecan and 3.0 to 4.4 months for Treatment of Physician’s Choice. These intervals express uncertainty in separate group-level median estimates. They are not the confidence interval for the treatment contrast. The hazard-ratio interval is the posted interval for the between-group comparison.

Confidence intervals convey statistical precision, not clinical importance. A precise estimate can still require judgment about the outcome, population, treatment context, and potential harms. A confidence interval also does not show whether missing data, assessment decisions, or other design features could introduce bias. The registry does not provide a complete account of those issues. Reading the interval alongside the endpoint definition, analysis population, and study design gives a more complete view of what the posted estimate can support.

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 section is a planning illustration based on the observed effect, not a report of the trial’s original design calculation. For the time-to-event comparisons, the displayed figures are total events, not participants per arm. The table varies the effect used for planning and shows the total number of events associated with 80% or 90% power. The columns describe different planning targets: 80% power is the chance, under the stated assumptions and effect, of detecting the specified difference at the chosen significance level; 90% power represents a higher target probability under those same assumptions.

The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The event requirements use the Schoenfeld approach. The original trial’s own sample-size assumptions and justification are not reported in the registry record. These figures therefore illustrate how event requirements change under the stated planning assumptions; they do not establish what the trial originally planned.

The rows consider planning for 100%, 75%, and 50% of the observed effect. Smaller effects need more events to reach the same power target, so the event requirement rises as the effect being planned for becomes smaller. The figure is a planning aid, not a reconstruction of the study’s actual recruitment, event accumulation, or analysis history. This is not observed power.

For other endpoint types, planning figures, where shown, are expressed as participants per arm rather than total events. Read the unit attached to each table or curve before interpreting a value. A planning count is not the number of participants actually analyzed, and it cannot be compared directly with the primary endpoint’s total randomized enrollment without accounting for whether the table counts participants or events.

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

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 0.653, so ln(HR) = -0.4262.
2. d = 4 × (1.96 + 0.842)² / [ln(0.653)]² = 172.9, rounded up to 173 events in total 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
hr <- 0.653                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(3901339)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 173 (80% power), 232 (90% power)
 75% of observed effect: events = 308 (80% power), 412 (90% power)
 50% of observed effect: events = 692 (80% power), 926 (90% power)
Simulated power with 173 events: 0.785 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Group-Sequential Design Analysis

Many time-to-event trials, especially in oncology, stop or report early when interim data are overwhelming, so a single fixed-sample number understates the real design problem. The calculation below is an illustrative group-sequential design built from this trial's posted control median and hazard ratio. The registry record does not state the planned number of looks or the spending function, so the design is a planning illustration, not a reconstruction of the protocol. Assumed: one-sided α = 0.025, 90% power, three equally spaced looks on information, Lan-DeMets O'Brien-Fleming spending, 24 months of accrual, 60 months in total and a 1% monthly dropout hazard.

Under these assumptions the design needs about 252 participants in total, with the boundaries below.

Fixed design versus group-sequential alternatives

Group-sequential designs cost a small increase in maximum sample size and events compared with the fixed design, in exchange for the chance to stop early for overwhelming efficacy. O'Brien-Fleming spending keeps early boundaries very conservative and the final boundary close to the fixed-design value, whereas Pocock spending spends more alpha early.

Sensitivity to a smaller true effect

Planning for exactly the observed effect is optimistic, because observed effects in published trials tend to be somewhat larger than the true effect. The table repeats the main design (three looks, O'Brien-Fleming spending, 90% power) for smaller true hazard ratios.

For context, this trial enrolled 543 participants.

Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993. Agreement between the two packages confirms the boundary calculation.

R code and output

suppressWarnings(suppressMessages({ library(gsDesign); library(rpact) }))
med_control <- 4          # control-arm median, months (converted from the posted unit)
hr          <- 0.653          # hazard ratio used for planning
lambdaC     <- log(2) / med_control
eta         <- 0.01               # dropout hazard per month (assumption)
acc_months  <- 24                 # accrual duration (assumption)
study_months<- 60                 # total study duration (assumption)

design <- function(k, sfu, beta, h = hr) gsSurv(k = k, test.type = 1, alpha = 0.025, beta = beta, sfu = sfu,
  lambdaC = lambdaC, hr = h, hr0 = 1, eta = eta, gamma = 1, R = acc_months, T = study_months, minfup = study_months - acc_months)

# Main design: 3 looks, Lan-DeMets O'Brien-Fleming spending, one-sided alpha 0.025, 90% power
x <- design(3, sfLDOF, 0.10)
cat(sprintf("N_TOTAL|%s\n", format(ceiling(sum(x$gamma * x$R)), big.mark = ",")))
for (i in 1:3) cat(sprintf("GS|%d|%.1f|%.0f|%.3f|%.5f|%.5f\n", i, x$T[i], x$n.I[i], x$upper$bound[i], pnorm(-x$upper$bound[i]), cumsum(x$upper$spend)[i]))
cat("\n"); print(gsBoundSummary(x))

# Comparison of designs
cmp <- list(list("Fixed design, one analysis", 1, sfLDOF), list("Two looks, O'Brien-Fleming", 2, sfLDOF),
            list("Three looks, O'Brien-Fleming", 3, sfLDOF), list("Three looks, Pocock", 3, sfLDPocock))
cat("\n")
for (b in c(0.20, 0.10)) for (d in cmp) {
  y <- design(d[[2]], d[[3]], b)
  cat(sprintf("DES|%s|%d|%.0f|%s|%.1f\n", d[[1]], round(100 * (1 - b)), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ","), tail(y$T, 1)))
}

# Sensitivity: what if the true effect is smaller than the observed one? (3 looks, O'Brien-Fleming, 90% power)
cat("\n")
for (fr in c(1, 0.75, 0.5)) {
  y <- design(3, sfLDOF, 0.10, exp(log(hr) * fr))
  cat(sprintf("SENS|%.2f|%.3f|%.0f|%s\n", fr, exp(log(hr) * fr), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ",")))
}

# Independent cross-check of the boundaries with rpact
r <- getDesignGroupSequential(kMax = 3, alpha = 0.025, beta = 0.10, sided = 1, typeOfDesign = "asOF", informationRates = x$timing)
cat(sprintf("\nCross-check, upper boundaries: gsDesign %s | rpact %s\n", paste(round(x$upper$bound, 3), collapse = ", "), paste(round(r$criticalValues, 3), collapse = ", ")))

Method: LachinFoulkes 
    Analysis               Value Efficacy
   IA 1: 33%                   Z   3.7103
      N: 146         p (1-sided)   0.0001
  Events: 78        ~HR at bound   0.4313
   Month: 14    P(Cross) if HR=1   0.0001
             P(Cross) if HR=0.65   0.0338
   IA 2: 67%                   Z   2.5114
      N: 236         p (1-sided)   0.0060
 Events: 156        ~HR at bound   0.6687
   Month: 22    P(Cross) if HR=1   0.0060
             P(Cross) if HR=0.65   0.5603
       Final                   Z   1.9930
      N: 252         p (1-sided)   0.0231
 Events: 234        ~HR at bound   0.7705
   Month: 60    P(Cross) if HR=1   0.0250
             P(Cross) if HR=0.65   0.9000



Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993

15. Analyses Reported by the Sponsor

The registry posts a sponsor-reported Log Rank hazard ratio of 0.653 for the primary comparison, with a 95% confidence interval from 0.526 to 0.812 and p < 0.001. The reported analysis lists the same estimate, interval, method, and p-value. This is the analysis to use when describing the registry’s primary comparison. The posted medians are separate group summaries and cannot, by themselves, reproduce the hazard ratio or its test.

The record does not present a separate recalculated primary hazard ratio that would replace the sponsor-reported result. A simple unadjusted re-calculation from summary statistics, where one is displayed elsewhere on the page, is not equivalent to the trial’s participant-level time-to-event analysis. Such a calculation cannot reconstruct the underlying event times, censoring, risk sets, or model details. The registry does not provide those participant-level data.

No discrepancy is flagged between the posted primary comparison and the reported analysis in this record. Accordingly, the sponsor’s own reported analysis remains the source for the hazard ratio and p-value. If a discrepancy were present, the sponsor’s own analysis would take precedence over a simple summary-statistic re-calculation; that is not the situation indicated. The distinction is important because a recalculation based on summary values can serve as a check or illustration, but it should not be presented as an independent replication of the trial analysis.

16. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and a range of secondary outcomes, including survival, response, investigator-assessed progression, and quality-of-life measures. Testing several outcomes raises a general multiplicity question: if many hypotheses are tested, some may appear statistically notable by chance even when no corresponding effect exists. A prespecified testing hierarchy, adjustment procedure, or clear distinction between confirmatory and exploratory analyses can help readers interpret this issue.

The multiplicity strategy, any adjustment to secondary p-values, and a formal order for testing outcomes are not reported in the registry record. This means that the posted secondary p-values should not automatically be read as if each were an isolated confirmatory test. The primary endpoint remains the main registered efficacy question in the available summary, while secondary findings offer additional information with the caveat that their joint testing context is not described.

The registry defines the primary analysis population as the intention-to-treat population, including all randomized participants whether or not they received study treatment. That definition is useful, but the detailed missing-data and censoring procedures are not reported in the registry record. For a time-to-event endpoint, details about censoring and incomplete follow-up can affect interpretation. For response and quality-of-life measures, missing assessments and the handling of participants who discontinue or do not complete questionnaires can also matter. The record does not give enough detail to assess those procedures.

The record does not report a complete statistical analysis plan, sensitivity analyses for missing outcomes, or diagnostic checks for the time-to-event assumptions. It also does not report whether interim analyses were conducted or how any such analyses would have affected testing. These omissions do not change the posted numerical results, but they limit a full appraisal of how the analysis was planned and how robust it may be to alternative reasonable assumptions. Readers should distinguish between what the registry reports and what remains undocumented.

17. 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: diarrhoea (4.9%), febrile neutropenia (4.1%), neutropenia (3.0%), abdominal pain (2.2%), nausea (2.0%).

The registry lists serious adverse events for participants with safety information available: 268 at risk in the Sacituzumab Govitecan group, of whom 74 were affected, and 249 at risk in the Treatment of Physician’s Choice group, of whom 48 were affected. It also lists other adverse events, with 268 at risk and 264 affected in the Sacituzumab Govitecan group, and 249 at risk and 234 affected in the Treatment of Physician’s Choice group. These are counts, not exposure-adjusted rates, and the numbers at risk differ between the safety groups.

The listed serious adverse event terms include diarrhoea at 4.9%, febrile neutropenia at 4.1%, neutropenia at 3.0%, abdominal pain at 2.2%, nausea at 2.0%, and pneumonia at 2.0%. The record summary does not give enough detail to assign these term frequencies to a particular arm, so they should be read as the listed serious-event terms rather than as a direct between-arm comparison.

These figures describe reported adverse-event information; they do not, on their own, establish that an event was caused by treatment. Differences in exposure, observation, event definitions, and reporting can affect comparisons. The registry also lists deaths in the adverse-event data, but these counts should not be treated as equivalent to treatment-related deaths without the relevant attribution and analysis details, which are not reported in the registry record. Safety counts complement the efficacy results and should be interpreted as descriptive information in the context of the trial’s follow-up and reporting.

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

The posted primary result shows that the randomized groups had different estimated PFS distributions under the study’s central-review endpoint and Log Rank analysis. The median PFS estimates were 5.5 months for Sacituzumab Govitecan and 4.0 months for Treatment of Physician’s Choice. The hazard ratio was 0.653, with a 95% confidence interval from 0.526 to 0.812. These are summaries of the trial data, not predictions for a particular person.

The result does not mean that every participant experienced the same change in time to progression or death. A median is a group-level summary, and a hazard ratio is a relative measure of event hazard over follow-up. Neither gives an individual’s likely outcome. The p-value does not measure the size of an effect or its importance to patients, and statistical evidence alone does not settle the balance between potential benefit and harm.

The trial compares sacituzumab govitecan with a strategy of physician’s choice among listed single-agent options; it does not compare the intervention with no treatment. The result also does not establish that the intervention is superior to every individual medicine used in the comparator strategy. The comparator was selected before randomization, but the posted summary does not report outcomes separated by the chosen comparator medicine.

These findings apply most directly to the population described in the study record: participants with HR-positive, HER2-negative metastatic breast cancer who had failed at least two prior chemotherapy regimens. They should not be assumed to apply unchanged to populations with different disease characteristics, prior treatment, or care settings. The registry reports additional outcomes, including survival and quality-of-life measures, but each has its own definition and uncertainty. The primary PFS result cannot stand in for all those questions.

Finally, the result should not be read as a complete account of the trial’s design or analysis. The original sample-size rationale, multiplicity plan, detailed missing-data procedures, and interim-analysis information are not reported in the registry record. Those gaps constrain how fully the statistical process can be assessed from the posted summary alone.

19. Why This Trial Matters Statistically

This trial is statistically informative because it uses randomized allocation to compare two treatment strategies in a defined population and reports a time-to-event primary outcome. Randomization is intended, in general, to make groups comparable at baseline on average, while the intention-to-treat population retains participants according to randomized assignment. The registry states that all randomized participants were included in the primary population, regardless of treatment receipt.

The primary endpoint also combines two clinically relevant event types: documented progression and death. A time-to-event analysis can use both the occurrence and timing of events, rather than reducing follow-up to a simple tally. Here, Kaplan-Meier estimation summarizes each group’s event-time distribution, and the reported Log Rank hazard ratio provides a relative comparison across follow-up. Reading these measures together helps avoid over-relying on a single summary.

The study is unmasked, but the primary PFS assessment was made by blinded independent central review. This combination matters when considering outcome assessment: central review provides a defined independent assessment approach, while the unmasked design remains relevant to other aspects of treatment delivery and participant experience. The registry specifies the review method but does not report every operational detail needed to evaluate possible sources of assessment variation.

The secondary outcomes extend the clinical questions beyond PFS, including overall survival and quality-of-life deterioration. Their presence is useful for a broader account, but the multiplicity strategy is not reported. This makes the distinction between the primary comparison and secondary findings important for statistical interpretation. In addition, the sample-size planning illustration demonstrates how event requirements depend on the effect targeted and the power goal, without claiming to reproduce the original design calculation.

Overall, the record offers a clear primary time-to-event comparison while leaving some analysis-planning details undocumented. That combination makes it a useful example of both what a registry result can communicate and what requires more complete protocol or analysis-plan information.

20. 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 result is tied to the population and comparator strategy described in the record. Participants had HR-positive, HER2-negative metastatic breast cancer and had failed at least two prior chemotherapy regimens. Treatment of physician’s choice could involve different single-agent options, and the posted summary does not separate the primary result by which comparator medicine was selected. The overall comparison therefore concerns the randomized strategy, not a direct estimate against each listed medicine individually.

The trial was unmasked. Although the primary PFS assessment used blinded independent central review, the registry record does not describe all procedures for assessment, adjudication, or handling of disagreements. The record also does not provide detailed information about censoring, missing assessments, sensitivity analyses, or diagnostic evaluation of the time-to-event analysis assumptions. These omissions limit an independent appraisal of robustness.

The multiplicity approach for secondary outcomes is not reported in the registry record. Consequently, the posted secondary p-values cannot be interpreted as fully independent confirmatory evidence without information on the planned testing strategy. A complete statistical analysis plan, including the original sample-size justification and any interim-analysis procedures, is also not reported.

Safety information is presented as counts in groups with different numbers at risk. The available summary does not provide all details needed to compare exposure-adjusted event rates or establish causality. The listed serious adverse event terms are not mapped to an arm in the summary. These limitations do not invalidate the listed counts, but they constrain comparisons and attribution.

The timeframe of up to 42.8 months does not describe the follow-up duration for each participant. The posted medians and hazard ratio also compress more complex time-to-event data into summary measures. Full event-time curves, risk sets, and participant-level data would be needed for a more detailed assessment than the registry summary allows.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: Study of Sacituzumab Govitecan-hziy Versus Treatment of Physician's Choice in Participants With HR+/HER2- Metastatic Breast Cancer — NCT03901339.
• Registry record last updated 2024-10-21; status last verified 2024-09.
• Results first posted 2024-10-21.
• 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.

24. Record Summary

This completed randomized phase 3 study compared sacituzumab govitecan with treatment of physician’s choice in participants with HR-positive, HER2-negative metastatic breast cancer who had failed at least two prior chemotherapy regimens. It enrolled 543 participants in an unmasked, parallel-group design.

The primary endpoint was PFS by blinded independent central review in the intention-to-treat population. Median PFS was 5.5 months in the Sacituzumab Govitecan group and 4.0 months in the Treatment of Physician’s Choice group. The reported Log Rank hazard ratio was 0.653 (95% CI 0.526 to 0.812; p < 0.001).

The registry also posts secondary outcome and adverse-event information. The original sample-size justification, multiplicity plan, detailed missing-data procedures, and interim-analysis information are not reported in the registry record.