Radiation Therapy With Cisplatin or Cetuximab in Treating Patients With Oropharyngeal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Radiation Therapy With Cisplatin or Cetuximab in Treating Patients With Oropharyngeal Cancer (NCT01302834), a randomized phase 3 study. The primary outcome is Overall Survival.

2. Clinical Question

The trial asks whether the overall-survival experience differs between patients assigned to radiotherapy with cetuximab and those assigned to radiotherapy with cisplatin. It is a comparison of treatment strategies that share IMRT and differ in the concurrent systemic treatment. The registry lists the study as randomized, phase III, open-label, and parallel-group.

The stated primary endpoint counts death from any cause as an event. Survival time runs from randomization to death or last known follow-up, with participants without a recorded death censored at last follow-up. The protocol endpoint is the hazard ratio, while the five-year survival rates are reported as summary data rather than as the protocol outcome measure. The registry’s own statistical analysis results therefore deserve priority when interpreting the primary question; a simple comparison of the posted percentages is a different, unadjusted summary.

3. Trial Design

The registry describes NCT01302834 as a randomized phase III, parallel-group trial with no masking. Its title identifies radiation therapy with cisplatin or cetuximab in patients with oropharyngeal cancer; the official title describes a comparison of radiotherapy plus cetuximab versus chemoradiotherapy in HPV-associated oropharynx cancer. The sponsor is the Radiation Therapy Oncology Group. The study status is completed.

The two arms received intensity-modulated radiotherapy (IMRT). One arm received concurrent cisplatin and the other concurrent cetuximab. The posted overall-survival summaries include 406 participants in the cisplatin arm and 399 in the cetuximab arm. These endpoint-specific values should not be treated as a reconstruction of the full randomized cohort: the registry lists overall enrollment as 987, and the reasons the endpoint summaries have their reported denominators are not detailed in the record.

The study start is listed as 2011-06 and primary completion as 2018-07-12. The primary endpoint timeframe runs from randomization to last follow-up. The registry states that analysis was to occur after 180 deaths were reported, while the analysis occurred after 133 deaths; maximum follow-up at that analysis was 6.5 years. The record does not explain the reason for the difference between the planned and actual event counts.

The design information establishes the randomized comparison and treatment groups, but does not supply the original sample-size justification, detailed eligibility rules, randomization procedure, or the complete prespecified statistical plan. Those details are not reported in the registry record. The study is unmasked, so participants and investigators were not reported as blinded; the record does not describe how this feature was handled for outcomes requiring judgment.

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: Sites were required to submit participant tumor tissue for central p16 evaluation within one week of registration. If participants were determined to be p16-positive and continued on the study, then treatment arm was assigned. Of 987 participants registered, 849 were randomized.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Eligible participants

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: Eligible participants

6. Primary Endpoint

The registry lists the primary outcome as Overall Survival, measured over From randomization to last follow-up. Analysis was to occur after 180 deaths were reported. Analysis occurred after 133 deaths were reported. Maximum follow-up at time of analysis was 6.5 years..

Overall survival is the primary endpoint. An event is death from any cause, so the endpoint includes deaths regardless of cause. Survival time begins at randomization and ends at the date of death or the last known follow-up. Participants with no known death at their last follow-up are censored at that point: their observed time contributes information up to follow-up, without being treated as a death.

The registry says survival rates are estimated with the Kaplan-Meier method. This approach uses the observed event times and accounts for censoring when estimating survival over time. The posted group percentages are described as five-year rates and as summary data; the protocol endpoint is the hazard ratio reported in the registry’s statistical analysis results. The distinction matters: the rates convey survival at a specified time point, whereas a hazard ratio compares event rates over the follow-up period under the analysis model.

Overall survival is used because it is a direct endpoint based on death from any cause and does not require a judgment about whether a particular progression event is clinically meaningful. It is not, by itself, a complete account of treatment experience: it does not describe the timing or type of nonfatal cancer events, adverse events, or other outcomes. The registry lists progression-free survival and several failure outcomes as secondary measures, which add context without replacing the primary endpoint.

7. Results Posted in the Registry

• IMRT + Cetuximab: 77.9% (311/399) versus 84.5% (343/406) on control; risk difference -6.5 percentage points (95% CI -11.9 to -1.2), risk ratio 0.92 (0.86–0.99); unadjusted two-proportion z-test p = 0.018.

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.

Progression-free Survival

• 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:
  • IMRT + Cetuximab: 67.4% (269/399) versus 78.3% (318/406) on control; risk difference -10.9 percentage points (95% CI -17.0 to -4.8), risk ratio 0.86 (0.79–0.94); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for IMRT + Cetuximab than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time to Local-regional Failure

• 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:
  • IMRT + Cetuximab: 17.3% (69/399) versus 9.9% (40/406) on control; risk difference 7.4 percentage points (95% CI 2.7 to 12.1), risk ratio 1.76 (1.22–2.53); unadjusted two-proportion z-test p = 0.002.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for IMRT + Cetuximab than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time to Distant Metastasis

• 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:
  • IMRT + Cetuximab: 11.8% (47/399) versus 8.6% (35/406) on control; risk difference 3.2 percentage points (95% CI -1.0 to 7.3), risk ratio 1.37 (0.90–2.07); unadjusted two-proportion z-test p = 0.138.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between IMRT + Cetuximab and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists progression-free survival and several measures of cancer failure as secondary outcomes. The overall pattern is not identical across these endpoints: the progression-free and local-regional findings point in a similar direction to the primary survival summaries, while the distant-metastasis comparison is less conclusive. These outcomes describe different aspects of disease course and should not be treated as interchangeable measures of overall survival.

Secondary outcomes can help explain whether differences in the primary endpoint are accompanied by differences in intermediate or disease-specific outcomes. They may also identify areas where the data are less precise or where estimates do not clearly distinguish the groups. They do not, however, establish the mechanism behind an overall-survival pattern. The registry does not report a formal mediation analysis or other analysis linking a secondary endpoint to survival.

The reported secondary analyses include hazard ratios and confidence intervals, with log-rank methods stated for the posted time-to-event results. The comparison estimates are unadjusted recalculations. The registry does not report an adjustment for testing multiple secondary endpoints or explain whether these outcomes were prioritized in a prespecified hierarchy. When many hypotheses are assessed, the chance of at least one apparently unusual result can increase. Consequently, nominal p-values for secondary analyses carry less evidential weight than a prespecified primary analysis considered under its planned method.

The secondary results are best read as supportive context rather than as independent confirmation of the primary result. Their interpretation depends on their endpoint definitions, uncertainty, analysis methods, and the unreported multiplicity plan. The number of listed secondary outcomes is substantial, and the registry reports 44 secondary outcomes posted. That breadth makes selective emphasis on a single secondary finding especially inappropriate. The aggregate pattern is informative, but it does not erase uncertainty in the primary analysis or substitute for the missing details of the original statistical plan.

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

• Progression-free Survival
• Time to Local-regional Failure
• Time to Distant Metastasis
• Time to Secondary Primary Cancer
• Distribution of First Progression Events
• Percentage of Participants Experiencing Early Death
• Percentage of Participants With Acute Grade 3-4 Treatment-related Adverse Events: During Treatment
• Percentage of Participants With Acute Grade 3-4 Treatment-related Adverse Events: 1 Month After End of Study Treatment

9. Statistical Interpretation

10. Statistical Methods Explained

Randomized parallel-group comparison. The registry describes a phase III randomized trial with a parallel design and no masking. Randomization is intended, in general, to make treatment groups comparable at the point of assignment by allocating treatment independently of participant characteristics. It does not guarantee that every characteristic is balanced, and the record does not give the randomization method or allocation details. The open-label status means masking was not reported; the registry does not provide additional details on outcome assessment procedures.

Survival time and censoring. Overall survival is measured from randomization to death from any cause or last known follow-up. A participant without a recorded death by last follow-up is censored. In general, censoring allows information from observed follow-up to contribute without treating the participant as having an event at the censoring time. Valid survival analyses rely on assumptions about censoring and the analysis population; details beyond the endpoint definition are not reported in the registry record.

Kaplan-Meier estimates. The registry says the survival rates are estimated using the Kaplan-Meier method. In general, this method estimates the probability of remaining event-free over time while incorporating observed event times and censored observations. The posted arm percentages are summary rates; the registry describes the five-year rate as summary data rather than the protocol endpoint. A rate at one time point and a comparison of event hazards over follow-up answer related but different questions.

Hazard ratio. The protocol endpoint is a hazard ratio (HR), reported in the registry’s statistical analysis results. In general, an HR compares the event rate at a given time between groups among participants still at risk, according to the fitted time-to-event analysis. It is not a ratio of the two five-year survival percentages, and it does not directly state a difference in the probability of survival at a particular time. The primary result is labeled non-inferiority and reports HR 1.45 and a 95% confidence interval upper limit of 1.94. The lower confidence limit, p-value, non-inferiority margin, and specific analysis method are not reported for that primary result. Without the margin and full result, the registry entry does not permit an independent evaluation of the non-inferiority criterion.

Risk difference and p-value. The unadjusted recalculation expresses the difference in the posted event summaries as a risk difference for cetuximab relative to cisplatin. In general, a risk difference is an absolute contrast between two group proportions, rather than a relative comparison. Its 95% confidence interval describes the range of values compatible with the data and calculation method under repeated sampling assumptions. The unadjusted p-value measures how inconsistent the data are with a null comparison under that method; it does not measure the probability that the null hypothesis is true, or the size or clinical importance of an effect. The recalculated risk contrast is not the same estimand as the primary time-to-event hazard ratio.

Secondary time-to-event results. The registry reports log-rank analyses for progression-free survival, time to local-regional failure, and time to distant metastasis, along with hazard ratios and confidence intervals. In general, the log-rank test compares event-time patterns between groups, while a hazard ratio describes a relative event rate under a time-to-event model. The secondary results have their own estimates and uncertainty; they do not become confirmatory merely because they are reported alongside the primary endpoint. The registry does not report a multiplicity adjustment for these outcomes.

Adverse-event summaries. The safety listing gives group-specific counts at risk and counts affected for serious and other adverse-event categories, as well as death counts and selected serious-event terms with percentages. These are descriptive summaries. The registry entry does not provide enough detail to infer from these values alone whether an event was caused by treatment, how recurrent events were handled, or how the selected terms were chosen.

Planning calculations. The posted planning illustrations use the observed effect or stated fractions of it, with two-sided significance level 0.05, equal allocation, and a normal approximation. The non-time-to-event illustrations are participants per arm; time-to-event illustrations are total event counts. In general, power is the probability of detecting a specified effect under the assumptions used for planning. These calculations are not a reconstruction of the original trial design, whose assumptions are not reported in the registry record.

11. Confidence Intervals and Precision

The arm-specific overall-survival estimates are 84.6% (95% confidence interval 80.6% to 88.6%) with cisplatin and 77.9% (95% confidence interval 73.4% to 82.5%) with cetuximab. Each interval conveys uncertainty in its own estimated survival rate. The intervals should not be read as a direct test of the difference between groups; a comparison requires an interval or test for the contrast itself, based on its specified analysis.

The primary registry analysis reports HR 1.45 and an upper 95% confidence limit of 1.94, but does not report the lower limit or a p-value. That incomplete interval is an important limitation for judging precision of the sponsor’s primary estimate from the registry entry. The primary result is labeled non-inferiority, but the margin is not reported, so the interval cannot be evaluated against the non-inferiority criterion from this record alone.

The unadjusted risk-difference recalculation has a 95% confidence interval from -0.119 to -0.012. This interval gives the uncertainty for that specific unadjusted summary contrast, not the hazard ratio. Its endpoints are on a difference scale, while the sponsor’s result is on a time-to-event scale. The two intervals therefore should not be compared as if they measured the same quantity.

In general, a confidence interval represents the range of parameter values reasonably compatible with the data and analysis assumptions; it does not guarantee that the true value lies within the particular interval. A narrower interval indicates greater precision for the estimate under the same method, while a wider interval indicates more uncertainty. Precision also depends on the endpoint, follow-up, number of observed events, censoring, and analysis method. The registry reports that the primary analysis occurred after 133 deaths rather than the planned 180 deaths and that maximum follow-up was 6.5 years. These registry entries provide context for the reported analysis, but the record does not quantify how much each affected precision.

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 planning illustrations translate an observed comparison into hypothetical sample-size requirements under stated assumptions. For the binary percentage outcomes, the figures are participants per arm. For time-to-event outcomes, the figures are the total number of events, not the number of participants. The planning assumptions are two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect.

The 80% power column describes the hypothetical number of participants per arm, or total events for a time-to-event outcome, associated with an 80% chance of detecting the specified effect if the planning assumptions hold. The 90% power column uses a higher target chance of detection and therefore generally requires a larger planned information size. These are design calculations, not observed trial outcomes.

Rows that plan for 75% or 50% of the observed effect assume a smaller effect than the full observed effect. Smaller effects are harder to distinguish from random variation, so they require more participants or events under the same planning method. For example, the overall-survival illustrations list 560 participants per arm for 80% power at 100% of the observed effect, 962 at 75%, and 2,087 at 50%; the corresponding 90% power figures are 749, 1,287, and 2,793. These figures are planning illustrations calculated from the observed effect, not the original trial’s design targets. The original design assumptions and sample-size justification are not reported in the registry record. This is not observed power: it does not measure how much power the completed trial actually had.

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 = 0.845; treatment proportion p₁ = 0.779; average p̄ = 0.812.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5524 = 1.0827.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5505 = 0.4633.
4. n = (1.0827 + 0.4633)² / (-0.0654)² = 559.1, rounded up to 560 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 <- 0.844827586206897                          # control proportion
p1 <- 0.779448621553885                          # 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(1302834)
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 = 560 (80% power), 749 (90% power)
 75% of observed effect: n per arm = 962 (80% power), 1,287 (90% power)
 50% of observed effect: n per arm = 2,087 (80% power), 2,793 (90% power)
power.prop.test: 559.1 
Simulated power at n = 560 per arm: 0.809 (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. Analyses Reported by the Sponsor

The registry’s own primary result reports a hazard ratio of 1.45 for overall survival, labels the analysis non-inferiority, and gives an upper 95% confidence limit of 1.94. The lower confidence limit and p-value are not reported, and the non-inferiority margin is not reported in the registry record. Those missing elements prevent a full assessment of the sponsor’s primary inferential claim from the registry entry alone.

The re-calculated comparison is a simple unadjusted risk-difference contrast from posted group summary statistics. It gives a different type of estimate from the sponsor’s hazard ratio: it does not incorporate event timing in the same way as a time-to-event analysis. Its p-value and confidence interval therefore do not replace or revise the sponsor’s primary analysis. The registry lists no discrepancy between the posted summaries and the reported analysis; where a re-calculation and the sponsor’s analysis address different estimands, the sponsor’s own analysis takes precedence for the protocol endpoint.

The arm-specific percentages help describe the observed survival summaries, while the hazard ratio is the protocol-level comparative measure. Neither should be interpreted without its method and uncertainty. In particular, the record does not provide enough of the primary HR interval or non-inferiority specification to independently reproduce the sponsor’s analysis or evaluate its decision rule.

15. Multiplicity, Missing Data and Analysis Populations

The registry reports one primary endpoint and a broad set of secondary outcome measures. It lists 48 outcome measures and 22 analyses, and reports 44 secondary outcomes posted. Evaluating multiple outcomes creates a multiplicity issue: when many statistical comparisons are made, some may appear notable by chance even if there is no underlying difference. The registry record does not report whether the trial used a multiplicity adjustment, a testing hierarchy, or another method to control the overall false-positive risk.

Without a stated plan, the unadjusted p-values for secondary results should be treated as exploratory rather than as a collection of confirmatory tests. The record does not identify which secondary outcomes, if any, were designated as key confirmatory endpoints, nor does it state how the order of testing was handled. The primary endpoint result is separately identified, but its non-inferiority margin and full confidence interval are not reported.

The registry defines overall-survival censoring at last known follow-up. It does not report detailed missing-data procedures for other endpoints, how loss to follow-up was handled beyond the primary endpoint description, or sensitivity analyses for missing outcomes. For time-to-event outcomes, censoring is a standard component of analysis, but the assumptions and checks used in this trial are not reported in the registry record.

The primary endpoint population is described as eligible participants. The registry does not provide enough detail in these results to establish whether analyses followed an intention-to-treat population, a per-protocol population, or another specified analysis set, nor does it explain exclusions from each endpoint denominator. Those details matter for randomized comparisons, especially for a non-inferiority analysis, where the analysis population and protocol adherence can affect interpretation.

The primary analysis was planned after 180 deaths and occurred after 133 deaths; maximum follow-up was 6.5 years. The registry does not state whether interim analyses were conducted, whether stopping rules applied, or why the actual analysis occurred at the reported event count. These points are not reported in the registry record and should not be inferred from the stated event totals.

16. 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: dehydration (10.8%), nausea (9.5%), vomiting (8.5%), dysphagia (5.8%), mucositis oral (5.6%).

The safety listing gives group-specific numbers at risk for serious adverse events and counts of participants affected. For IMRT with cisplatin, the serious-event listing reports 398 at risk and 177 affected; for IMRT with cetuximab, it reports 394 at risk and 115 affected. The record also lists other adverse-event counts and deaths by group. These are descriptive safety summaries, and the death counts should not be assumed to be treatment-related.

The selected serious adverse-event terms include dehydration (10.8%), nausea (9.5%), vomiting (8.5%), dysphagia (5.8%), mucositis oral (5.6%), and acute kidney injury (5.3%). The record does not explain in this listing whether these percentages represent participants with at least one event, how repeated events were counted, or how the selected terms were chosen. They should therefore be read as registry-reported frequency summaries, not as causal estimates or as a complete comparison of safety.

Differences in the listed counts do not by themselves establish that a treatment caused more or fewer events. Interpretation would require the event definitions, observation period, exposure, severity and attribution rules, and appropriate comparative analyses. Those details are not reported in the registry record in the safety summary.

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

The registry records a randomized comparison of IMRT with concurrent cisplatin against IMRT with concurrent cetuximab, with overall survival as the primary endpoint. The posted survival-rate estimates are numerically different between groups, and the registry reports a primary hazard-ratio analysis. These findings describe the observed trial data under the analyses recorded; they are not a complete account of the treatment effects or their clinical relevance.

The primary result is labeled non-inferiority, but the registry entry does not report the non-inferiority margin or the lower confidence limit for the HR. Without those details, the non-inferiority conclusion cannot be independently assessed from the registry record. The unadjusted recalculation is not a substitute, because it compares summary proportions rather than the protocol’s time-to-event endpoint.

A statistical difference does not automatically establish that a difference is important to patients. Clinical importance depends on the magnitude and timing of outcomes, adverse effects, patient priorities, and other context not fully captured by the registry results. Conversely, an uncertain or non-significant comparison does not prove that the groups are identical. Estimates and confidence intervals communicate uncertainty; they do not resolve every question about benefit and harm.

The findings also do not establish why outcomes differed, whether all patient subgroups had the same experience, or whether the results apply to populations not represented by the trial. The registry does not provide detailed eligibility criteria or subgroup results in the registry record for this analysis. Nor do aggregate event summaries determine whether an individual outcome was caused by treatment.

The proper reading is therefore bounded: the record documents a randomized comparison, reported survival estimates, a sponsor-reported primary analysis, secondary outcomes, and descriptive adverse-event information. It does not supply every statistical detail needed to reproduce the primary analysis or judge all assumptions. Statistical interpretation should remain tied to the endpoints, analysis methods, uncertainty, and limitations actually reported.

18. Why This Trial Matters Statistically

This trial is statistically notable because it compares two concurrent treatment approaches within a randomized, parallel-group phase III design and uses overall survival as its primary endpoint. The outcome has a clear event definition—death from any cause—and the registry specifies how survival time and censoring are defined. That clarity helps distinguish the primary endpoint from the disease-control and failure outcomes reported secondarily.

The record also illustrates why the estimand matters. Kaplan-Meier survival percentages describe survival at a time point, whereas the protocol’s hazard ratio compares event rates over follow-up. A simple unadjusted risk difference based on posted summary values can be useful as a descriptive contrast, but it does not answer the same question as the sponsor’s time-to-event analysis. Keeping these quantities separate prevents a recalculation from being mistaken for a reproduction of the primary result.

Precision and reporting completeness are central to interpretation. The sponsor’s primary result includes an upper confidence limit but not the lower limit or p-value, and the non-inferiority margin is not reported. The primary analysis also occurred after 133 deaths rather than the planned 180 deaths. These features do not by themselves invalidate the result, but they limit how fully a reader can evaluate its inferential basis from the registry record.

The many secondary outcomes demonstrate another recurring challenge in clinical research: a broad outcome set can add context while increasing the risk of chance findings. Without a reported multiplicity plan, secondary p-values should not be treated as equivalent to the primary test. The trial thus offers a useful example of reading endpoint definitions, analysis scale, precision, and reporting gaps together rather than relying on any single number.

19. 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.

Several trial-specific limitations affect how completely the registry results can be evaluated. The primary statistical result reports HR 1.45 and an upper 95% confidence limit of 1.94, but the lower confidence limit and p-value are not reported. Although the analysis is labeled non-inferiority, the non-inferiority margin and decision rule are not reported in the registry record. This prevents independent assessment of the primary non-inferiority criterion from the listed result.

The primary analysis was planned after 180 deaths and occurred after 133 deaths, with maximum follow-up of 6.5 years. The reason for this difference, and whether interim analyses or stopping rules were involved, are not reported. The actual event count and follow-up duration are relevant context for precision, but the registry does not quantify their effects.

The trial is listed as unmasked. The registry does not detail how open-label conduct may have affected endpoint assessment, treatment delivery, follow-up, or reporting. Overall survival is based on death from any cause, but the secondary outcomes may involve other event definitions; the record does not provide additional assessment procedures in these results.

The complete analysis populations, exclusions, missing-data procedures beyond primary survival censoring, and original sample-size justification are not reported. The posted sample-size illustrations are based on observed effects and stated planning assumptions, not the original trial design. The broad secondary outcome set is not accompanied by a reported multiplicity strategy, limiting confirmatory interpretation of nominal secondary p-values.

Finally, the registry summary does not provide detailed eligibility criteria or subgroup analyses in the registry record described. The scope of applicability to populations outside the enrolled trial group cannot be determined from these results alone. These limitations are about what the registry record reports; they do not establish that unreported methods were or were not used.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Radiation Therapy With Cisplatin or Cetuximab in Treating Patients With Oropharyngeal Cancer — NCT01302834.
• Registry record last updated 2026-07-20; status last verified 2025-09.
• Results first posted 2020-01-09.
• 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.

23. Record Summary

NCT01302834 is a completed randomized phase III, open-label, parallel-group trial sponsored by the Radiation Therapy Oncology Group. It compared IMRT with concurrent cisplatin against IMRT with concurrent cetuximab in a trial described as addressing HPV-associated oropharynx cancer. Overall survival was the primary endpoint, defined from randomization to death from any cause or last known follow-up.

The registry reports Kaplan-Meier survival-rate summaries and a sponsor-reported primary hazard-ratio analysis labeled non-inferiority. The HR lower confidence limit, p-value, and non-inferiority margin are not reported. The primary analysis occurred after 133 deaths rather than the planned 180, and maximum follow-up was 6.5 years. Secondary outcomes and descriptive adverse-event summaries are also listed. The original statistical plan, detailed missing-data procedures, and multiplicity handling are not reported in the registry record.