A Study to Evaluate SHR-1210 in Subjects With Advanced HCC: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate SHR-1210 in Subjects With Advanced HCC (NCT02989922), a randomized phase 2 study. The primary outcome is Objective Response Rate.

2. Clinical Question

The registry frames the broader study hypothesis as whether SHR-1210 improves objective response rate and overall survival compared with standard of care. However, the posted primary outcome results compare two dosing schedules of SHR-1210; they do not report a standard-of-care group’s objective response result. The result therefore addresses the schedule comparison in the posted data, rather than establishing a comparison with standard of care.

The recorded response proportions were 11.9% for SHR-1210 Q2W and 16.7% for SHR-1210 Q3W. These are estimates from the participants included in each group’s reported endpoint denominator. The interval estimates are wide, so the observed difference should be read alongside the uncertainty around each group estimate and the direct comparison.

Although the registry’s official title describes a randomized controlled phase 2/3 study, the phase field lists phase 2. The record reports random allocation and no masking. The endpoint is a tumour-response measure, not a direct measure of survival, symptom relief, or quality of life. Overall survival appears among the secondary outcomes, but no corresponding outcome result is included in the posted results.

3. Trial Design

The registry describes a randomized, parallel-group study of SHR-1210 in people with advanced hepatocellular carcinoma who had failed or could not tolerate prior systemic treatment. The condition is also recorded as non-resectable hepatocellular carcinoma. The phase field lists phase 2, while the official title calls the study phase 2/3. The record lists Jiangsu HengRui Medicine Co., Ltd. as sponsor and reports the study status as completed.

Two SHR-1210 schedules are represented in the outcome results. The Q2W group received intravenous SHR-1210 at 3mg/kg on Day 1 every 2 weeks; the Q3W group received the same stated dose on Day 1 every 3 weeks. The registry marks the allocation as randomized, the study model as parallel, and masking as none. The posted results include 109 participants in the Q2W outcome group and 108 in the Q3W outcome group. The registry’s enrollment field is 220; the reason for any difference between that enrollment figure and the endpoint group counts is not reported in the registry record.

The registry reports a study start date of 2016-11-15 and a primary completion date of 2020-03-03. Its primary outcome is objective response rate, with a timeframe described as approximately 3 years. The outcome description defines response using confirmed complete or partial response assessed by an independent review committee according to RECIST 1.1. Duration of response, adverse events, and overall survival are listed as secondary outcomes. The posted results described for the primary endpoint do not include a result for overall survival or duration of response.

The registry record does not report the original sample-size justification or the full statistical analysis plan. It also does not explain the relationship between the broader standard-of-care hypothesis in the study summary and the two schedule groups in the posted response results. Those details cannot be inferred from the schedule comparison alone.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

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 Objective Response Rate, measured over approximate 3 years.

Objective response rate is the percentage of participants whose best overall tumour response was a confirmed complete response or partial response. In this study, the registry says responses were assessed by an independent review committee using RECIST 1.1. This endpoint reduces a sequence of tumour assessments to a responder or non-responder classification for each participant, then reports the proportion classified as responders in each group.

A response proportion can describe tumour shrinkage meeting the specified criteria, but it does not by itself describe how long a response lasts, whether participants live longer, or how symptoms change. Those are different clinical questions. Duration of response and overall survival are listed among the secondary outcomes, but their results are not included in the posted results described for this analysis.

As a binary endpoint, objective response rate is easy to interpret as a group proportion, while its uncertainty depends in part on the number of participants and the number meeting the response definition. The registry gives the outcome timeframe as approximately 3 years; it does not report more detail about assessment timing or the handling of participants without an evaluable response. The endpoint therefore needs to be read together with its confidence intervals and the limitations of the recorded analysis information.

7. Results Posted in the Registry

• SHR-1210 Q3W: 16.7% (18/108) versus 11.9% (13/109) on control; risk difference 4.7 percentage points (95% CI -4.6 to 14.0), risk ratio 1.40 (0.72–2.71); unadjusted two-proportion z-test p = 0.318.

8. Statistical Interpretation

9. Statistical Methods Explained

The central statistical task is to compare two proportions: the fraction of participants recorded as having a confirmed objective response in each SHR-1210 schedule group. Because the outcome is binary, each participant contributes to a response category rather than a continuously measured value. The group proportions describe the observed response frequency; confidence intervals convey uncertainty about the underlying response proportions and their difference.

What does randomization contribute?

Random allocation is intended to make the groups comparable, on average, with respect to factors that might affect outcomes. It does not guarantee that every characteristic is balanced in a particular trial, and it cannot correct for missing outcome data or differences in outcome assessment after allocation. The record says the study was randomized and parallel-group, but does not report the allocation method, concealment procedures, or baseline balance details in the registry record described for this analysis.

The registry reports no masking. For a tumour response endpoint, the stated use of an independent review committee is relevant because it places response assessment with a review body separate from the treatment groups. The record does not describe the committee’s procedures or whether its members were masked to treatment schedule, so no further claim about the review process is warranted.

How should a confidence interval be understood?

A confidence interval is a range of values compatible with the observed data under the statistical method used. It is not a guarantee that the true value lies in the interval, nor does it show the probability that a particular treatment effect is true. Wider intervals indicate less precision. The intervals for the two response proportions are broad, reflecting uncertainty around each estimate. For the risk difference, the reported interval includes both a negative and a positive value, meaning the data do not pin down the direction of the difference with precision.

The reported 95% confidence interval for Q2W is 6.5% to 19.5%; for Q3W it is 10.2% to 25.1%. Overlap between separate group intervals is not itself a formal test of the difference. The direct comparison is the more appropriate quantity for asking whether the schedules differ, and the registry reports an unadjusted p-value of 0.318 for that comparison.

What does the p-value say?

A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be under assumptions of the test. It is not the probability that the null hypothesis is true, and it is not a measure of clinical importance. The reported unadjusted p-value does not provide evidence sufficient to distinguish the observed schedule estimates from chance variation under the null model. It should be considered with the confidence interval and the observed response counts, rather than as a stand-alone conclusion.

The record labels the comparison unadjusted. It does not report whether this was the only planned comparison, whether a hierarchy of endpoints was used, or whether multiplicity adjustments were planned. Thus, the p-value should not be treated as though its role in a broader testing strategy were documented.

Why focus on proportions and their uncertainty?

For a binary outcome, the numerator is the number meeting the response definition and the denominator is the number included in the endpoint result. The registry records 13 responses among 109 participants for Q2W and 18 among 108 for Q3W. A proportion summarizes these counts in a form that can be compared across groups, while the denominator and interval indicate the amount of information behind each estimate. A point estimate alone hides that uncertainty.

The comparison is reported as a risk difference, which expresses the difference between two group proportions on an absolute scale. The registry also contains relative effect measures, but interpreting them is not necessary to understand the primary posted result. The risk-difference interval spans values on either side of no difference. Since the recorded estimate and interval are not precise enough to exclude that possibility, the point estimate should not be treated as a settled effect.

What cannot be established from the record?

The registry record does not report the endpoint analysis population, the handling of missing outcomes, the exact statistical test, or the original analysis plan. It also does not report adjustments for baseline variables or a multiplicity strategy. These details can influence the interpretation of estimates and tests. The record reports zero analysis records, so the statistical analysis details are limited to the outcome summaries and comparison statistics posted.

The registry’s study summary refers to objective response and overall survival compared with standard of care. The response figures available for this analysis instead distinguish two dosing schedules. No standard-of-care response estimate is reported in the posted results. Consequently, the schedule comparison should not be substituted for the broader comparison described in the summary. Overall survival is a secondary outcome, but its result is not reported in the registry record.

General statistical principle: a non-significant comparison does not establish that two treatments are equivalent. Demonstrating equivalence or non-inferiority ordinarily requires a prespecified margin and a design and analysis suited to that question. Such a margin and analysis are not reported in the registry record. The present results are best understood as estimates with substantial uncertainty, rather than as proof that the schedules are the same or different.

10. Confidence Intervals and Precision

The response estimate for SHR-1210 Q2W is 11.9%, with a 95% confidence interval from 6.5% to 19.5%. For SHR-1210 Q3W, the estimate is 16.7%, with a 95% confidence interval from 10.2% to 25.1%. These ranges are broad in relation to their point estimates, so the exact response proportion in the population remains uncertain.

The interval for the risk difference also spans values below and above no difference. This is important: the observed Q3W proportion is higher, but the direct comparison does not estimate the difference precisely enough to rule out either direction. The unadjusted p-value of 0.318 is consistent with that uncertainty. It should not be interpreted as proof that there is no difference, nor as a probability that the observed difference is due to chance.

Confidence intervals are useful because they show a range of effect sizes compatible with the data under the analysis method. They do not account for every possible source of uncertainty. For example, the registry does not report the handling of missing or unevaluable responses, the analysis population, or whether the posted estimates were adjusted. The intervals cannot resolve those unreported design and analysis details.

The response counts were 13 in Q2W and 18 in Q3W, against endpoint group counts of 109 and 108. The relatively limited number of recorded responses means that the estimate may be sensitive to a small change in response classification. The independent review committee and RECIST 1.1 definition describe how response was assessed, but the registry record does not provide further assessment details. The intervals should therefore be read as a measure of statistical precision, not as a complete account of all uncertainty in the study.

11. 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 planning figures are illustrations based on the observed effect, not the trial’s original design calculations. For this binary-outcome comparison, the listed values are participants per arm, not total events. The assumptions recorded for the illustration are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario uses the stated fraction of the observed effect as the effect to detect.

The row for 100% of the observed effect corresponds to planning for the full observed effect; the 75% and 50% rows illustrate planning for smaller effects. The table gives 855 participants per arm for 80% power and 1,144 participants per arm for 90% power when planning for 100% of the observed effect. For 75% of the observed effect, the corresponding figures are 1,468 and 1,965 participants per arm. For 50% of the observed effect, they are 3,183 and 4,261 participants per arm.

In plain language, the 80% power column gives the number of participants per arm in the planning calculation intended to give an 80% chance of detecting the specified effect if that effect is real and the assumptions hold. The 90% column uses a higher target chance of detection and therefore requires more participants per arm for each listed scenario. Smaller effects are harder to distinguish from random variation, so planning to detect a smaller fraction of the observed effect requires more participants.

The accompanying curve should be read as a planning illustration relating effect size to required sample size, not as a summary of observed trial performance. These figures are not observed power. The registry record does not report the original sample-size assumptions or justification, so the planning illustration cannot be used to claim what the study was designed to detect.

12. 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 = 0.119; treatment proportion p₁ = 0.167; average p̄ = 0.143.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4950 = 0.9702.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4939 = 0.4157.
4. n = (0.9702 + 0.4157)² / (0.0474)² = 854.9, rounded up to 855 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 <- 0.119266055045872                          # control proportion
p1 <- 0.166666666666667                          # 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(2989922)
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 = 855 (80% power), 1,144 (90% power)
 75% of observed effect: n per arm = 1,468 (80% power), 1,965 (90% power)
 50% of observed effect: n per arm = 3,183 (80% power), 4,261 (90% power)
power.prop.test: 854.9 
Simulated power at n = 855 per arm: 0.803 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists objective response rate as the primary outcome and lists duration of response, adverse events, and overall survival among secondary outcomes. The record also states that there are five outcome measures, while the results information summarized for this analysis includes a primary endpoint comparison and three posted secondary outcomes. The registry record does not report a complete testing hierarchy or explain how the outcome measures relate to one another statistically.

Multiplicity arises when several outcomes, comparisons, or analyses are examined. With repeated statistical tests, the chance of finding at least one apparently unusual result can increase unless the analysis plan addresses that issue. The posted comparison is explicitly unadjusted. The registry record does not report whether any multiplicity adjustment was planned, whether other comparisons were conducted, or whether the primary result was intended to stand alone within a prespecified testing sequence. The unadjusted p-value should therefore be interpreted as the result of the reported comparison, not as evidence that all possible testing concerns have been addressed.

The endpoint analysis population is not reported in the registry record. Nor are the rules for missing, unevaluable, or unconfirmed response assessments. These details matter because a percentage depends on both which participants are included and how missing outcomes are classified. The posted group counts do not establish what happened to any participant outside the reported endpoint denominators.

The record does not describe an intention-to-treat, per-protocol, or other named analysis population, and it does not report sensitivity analyses for missing data. It also does not give enough information to determine whether any adjustment for baseline characteristics was used. These absences do not alter the arithmetic of the posted summary, but they limit assessment of how robustly the result represents the randomized comparison.

Because the study summary describes a standard-of-care comparison while the posted response estimates compare two SHR-1210 schedules, the registry information also leaves the relationship between the broader hypothesis and these particular results unclear. No standard-of-care endpoint result or detailed analysis plan is reported. These are reporting limitations, not grounds for inferring what analyses may have been performed.

14. 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: disease progression (11.9%), hepatic function abnormal (5.6%), hepatic cancer (3.7%), upper gastrointestinal haemorrhage (3.7%), hepatic failure (3.7%).

The registry records serious adverse events affecting 50 of 109 participants in the SHR-1210 Q2W group and 46 of 108 in the SHR-1210 Q3W group. The corresponding counts recorded for participants with any adverse event are 108 of 109 and 107 of 108. These are group counts and denominators as posted; they should not be read as evidence that every event had the same severity, duration, or relationship to treatment.

The record also lists deaths of 27 in Q2W and 29 in Q3W. The information described does not establish that these deaths were caused by treatment, nor does it supply a time-to-event analysis or contextual details for interpreting them. The record’s serious adverse event terms include disease progression at 11.9%, hepatic function abnormal at 5.6%, hepatic cancer at 3.7%, upper gastrointestinal haemorrhage at 3.7%, hepatic failure at 3.7%, and multiple organ dysfunction syndrome at 2.8%.

These figures describe events recorded in the registry, not a complete comparative safety analysis. The registry record does not report event definitions beyond the serious-event label, exposure-adjusted rates, attribution, timing, or a statistical comparison between groups. Counts should therefore be interpreted descriptively and in the context of the reported denominators, not as a causal assessment.

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

The posted primary outcome results show confirmed objective response proportions of 11.9% for SHR-1210 Q2W and 16.7% for SHR-1210 Q3W. The Q3W point estimate is higher, but the intervals around both estimates are broad, and the direct risk-difference interval spans no difference. The unadjusted p-value is 0.318. Together, these results describe a numerical contrast while leaving substantial uncertainty about the size and direction of any underlying schedule difference.

They do not establish that one schedule is superior, that the schedules are equivalent, or that either schedule improves outcomes compared with standard of care. The study summary states a broader standard-of-care hypothesis, but the posted primary response results compare Q2W with Q3W. A result for a standard-of-care group is not reported in the registry record.

Objective response rate concerns confirmed tumour response under the RECIST 1.1 assessment described in the record. It does not, on its own, answer questions about survival, duration of response, quality of life, or other outcomes. Overall survival and duration of response are listed as secondary outcomes, but their results are not included in the results. The registry’s adverse-event counts provide descriptive safety information, but they do not establish that events were caused by SHR-1210 or show a full comparative safety analysis.

General statistical principle: a result that does not provide clear evidence of a difference is not evidence that the treatments are the same. To support an equivalence conclusion, a trial would need a suitable design and a prespecified equivalence margin; the registry record does not report such a margin or analysis. Equally, a higher observed proportion in one group does not prove a real advantage when the comparison remains imprecise.

The most defensible interpretation is limited to what was posted: the observed response proportion was higher in Q3W than Q2W, but uncertainty remains considerable. Unreported analysis details, including the analysis population and missing-data approach, further limit how fully the result can be assessed. No clinical recommendation follows from these registry statistics alone.

16. Why This Trial Matters Statistically

This trial is statistically interesting because it reports two schedules of the same investigational treatment rather than a primary response comparison with the standard-of-care group described in the study summary. That distinction changes the question the posted figures can answer. They compare response proportions for Q2W and Q3W; they do not quantify the stated standard-of-care contrast.

The endpoint is a binary response measure assessed by an independent review committee using RECIST 1.1. Such measures can be clinically interpretable, but a proportion compresses a complex course of disease assessment into a responder classification. The estimate is therefore most informative when considered with its definition, denominator, and interval rather than as a stand-alone percentage.

The results illustrate why statistical uncertainty matters. The Q3W response proportion is numerically higher, yet the direct comparison’s confidence interval extends across no difference and its unadjusted p-value is 0.318. A point estimate alone could overstate how much the data establish. Conversely, a non-significant result does not prove sameness. The appropriate conclusion is that the posted comparison is uncertain.

The planning illustration also highlights the relation between effect size and information needs: smaller effects require more participants per arm to detect under its assumptions. These calculations are separate from the original trial design, whose assumptions are not reported in the registry record. Keeping planning illustrations distinct from observed results prevents a common error: treating an illustrative power calculation as evidence about what the completed study could or could not detect.

Finally, the limited reporting of analysis populations, missing data, and multiplicity makes it harder to evaluate the robustness of the results. Clear reporting of these features is important for biostatisticians and clinical researchers who need to distinguish the observed estimates from the assumptions used to interpret them.

17. 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 registry results leave several trial-specific questions unresolved. The primary outcome description gives an approximate 3-year timeframe, but more detailed timing and response-assessment schedules are not reported in the registry record. The record also does not explain how participants with missing or unevaluable assessments were handled, or identify the population used for the endpoint analysis. Those details are needed to assess whether the posted denominators reflect all randomized participants or a defined subset.

The registry reports randomization and no masking, but it does not give the method of sequence generation or allocation concealment in the information. It states that an independent review committee assessed responses using RECIST 1.1, but committee procedures and masking are not described. The study summary refers to a comparison with standard of care, while the posted response results distinguish two SHR-1210 schedules; the relationship between these descriptions is not explained.

The analysis information is also limited. The registry reports no analysis records and does not report the original sample-size justification, analysis plan, multiplicity strategy, or adjustment for baseline characteristics. The available comparison is unadjusted. Although the planning illustration lists assumptions, those are not the trial’s original design assumptions and cannot fill that reporting gap.

Safety information is likewise descriptive. The record lists group counts for serious adverse events, any adverse events, and deaths, but does not provide attribution, timing, exposure-adjusted analyses, or a direct statistical comparison in the information. These limitations restrict causal interpretation and comparisons between schedules. They do not justify assuming that unreported analyses were or were not conducted.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Study to Evaluate SHR-1210 in Subjects With Advanced HCC — NCT02989922.
• Registry record last updated 2024-03-15; status last verified 2017-12.
• Results first posted 2024-03-15.
• 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.

21. Record Summary

The registry describes a randomized, parallel-group phase 2 study of SHR-1210 in advanced, non-resectable hepatocellular carcinoma after failure or intolerance of prior systemic treatment. The official title describes it as phase 2/3, whereas the phase field lists phase 2. The study is marked completed and reports an enrollment of 220.

The posted primary outcome is objective response rate assessed by an independent review committee according to RECIST 1.1, with a timeframe of approximately 3 years. The recorded response proportions were 11.9% for SHR-1210 Q2W and 16.7% for SHR-1210 Q3W. The unadjusted comparison reports p = 0.318, and its confidence interval spans no difference.

Duration of response, adverse events, and overall survival are listed as secondary outcomes. The record includes descriptive adverse-event counts, but does not report overall survival or duration-of-response results in the information. Important analysis details, including the endpoint population and handling of missing data, are not reported in the registry record.