Sorafenib Tosylate With or Without Recombinant Interferon Alfa-2b in Treating Patients With Metastatic Kidney Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sorafenib Tosylate With or Without Recombinant Interferon Alfa-2b in Treating Patients With Metastatic Kidney Cancer (NCT00126594), a randomized phase 2 study. The primary outcome is Objective Response Rate (ORR) Evaluated Using Response Evaluation Criteria in Solid Tumors (RECIST).

2. Clinical Question

This was a phase II trial sponsored by the National Cancer Institute (NCI). The registry lists the study as completed. Its stated purpose was to evaluate sorafenib with or without low-dose interferon in metastatic renal cell carcinoma. The primary outcome description specifies tumor restaging at 8 weeks following baseline, with responding or stable participants restaged at 8 week intervals, up to 12 months. These outcome-assessment details describe the registered measure; additional protocol details are not reported in the registry record.

3. Trial Design

The registry describes a randomized, open-label, parallel-group phase II trial. The listed enrollment is 80, and the response outcome has 40 participants in each group. One group received sorafenib alone; the other received sorafenib with low-dose recombinant interferon alfa-2b. Masking is listed as none, so the registry record does not describe a blinded comparison. The sponsor is the National Cancer Institute (NCI).

The trial began in 2005-06, had primary completion in 2013-08, and is listed as completed. The intervention description specifies oral sorafenib 400 mg twice daily on days 1-28 in both groups. The combination group also received interferon alfa-2b 0.5 million units subcutaneously twice daily on days 1-28. Laboratory biomarker analysis is also listed as an intervention, but further details about that analysis are not reported in the registry record.

The primary outcome was objective response rate (ORR), defined as the percentage of participants with complete response or partial response under RECIST. The registry specifies tumor restaging at 8 weeks after baseline, followed by restaging at 8 week intervals for responding or stable participants, up to 12 months. It also describes radiologic testing for progressive disease every 8 weeks. The primary analysis population is identified as intent-to-treat.

Four secondary outcomes are listed: selected grade 3-4 adverse events under NCI CTCAE Version 4.0, progression-free survival, median overall survival, and duration of response for participants with stable disease. The registry record does not report the trial's original sample-size justification, detailed statistical analysis plan, or planned interim analyses. The outcomes and design features should therefore be distinguished from any additional protocol information that is not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analysis performed for the intent-to-treat population.

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 (ORR) Evaluated Using Response Evaluation Criteria in Solid Tumors (RECIST), measured over Tumor restaging performed at 8 weeks following baseline, responding or stable participants restaged at 8 week intervals, up to 12 months..

The primary endpoint, objective response rate, measures the percentage of participants whose tumors meet the registry's definition of complete or partial response. Under the stated RECIST definitions, complete response means disappearance of all target lesions. Partial response requires a reduction of >30% in the sum of the longest diameters of target lesions, relative to baseline. Objective response combines these two categories; stable disease and progressive disease do not count as objective response under this endpoint.

The registry defines progressive disease using an increase of >20% in the sum of lesion diameters relative to the smallest sum recorded since treatment began, or the appearance of one or more new lesions. Stable disease means that neither the shrinkage threshold for partial response nor the increase threshold for progressive disease is met. These categories clarify what is counted in the response measure; they do not describe every outcome of treatment.

Response rate is a binary-outcome summary: each participant either meets the complete- or partial-response definition or does not. It is used to describe the frequency of tumor responses in each group and to compare the groups using proportions. The posted record gives response percentages and 95% confidence intervals. It does not report additional detail about adjudication or the handling of participants without an evaluable response, beyond identifying the primary analysis population as intent-to-treat.

Assessment timing matters because a response is determined through scheduled tumor restaging. The registered timeframe calls for restaging at 8 weeks following baseline and, for responding or stable participants, at 8 week intervals up to 12 months. The response-rate result should therefore be interpreted as an endpoint assessed under that stated schedule, rather than as a measure of every possible clinical outcome.

7. Results Posted in the Registry

• Sorafenib Tosylate, Recombinant Interferon Alfa-2b: 25.0% (10/40) versus 30.0% (12/40) on control; risk difference -5.0 percentage points (95% CI -24.5 to 14.5), risk ratio 0.83 (0.41–1.70); unadjusted two-proportion z-test p = 0.617.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary response measure. For each participant, the outcome is whether the tumor response met the registry's definition of complete response or partial response under RECIST. A response proportion summarizes the fraction of participants in a group who met that definition. The posted arm-level estimates are 30.0% for sorafenib alone and 25.0% for sorafenib plus interferon, with 95% confidence intervals of 16.56% to 46.53% and 12.69% to 41.2%, respectively.

What does a confidence interval describe?

In general, a confidence interval expresses the uncertainty in an estimate using a range of values compatible with the observed data under the interval's statistical method. A narrower interval indicates greater precision; a wider interval indicates less precision. It is not a range containing a known fixed proportion of individual outcomes, and it does not say that every value inside the interval is equally plausible. For the two response proportions, the posted intervals are broad, reflecting uncertainty in the estimated rates.

How is the group contrast expressed?

The risk difference subtracts one group's response proportion from the other's. In the posted comparison, the direction is combination minus sorafenib alone, giving an estimate of -0.05. A value of zero represents no difference in response proportions. The posted 95% confidence interval runs from -0.245 to 0.145 and includes zero. That interval therefore includes differences in both directions, rather than ruling out either direction.

The risk ratio compares the response proportion in the combination group with the response proportion in the sorafenib-alone group. A value of 1 represents equal proportions. The posted risk ratio is 0.833, with a 95% confidence interval from 0.407 to 1.705. The odds ratio compares the odds of response between groups, rather than comparing proportions directly. Its estimate is 0.778, with a 95% confidence interval from 0.291 to 2.082. For both relative measures, the confidence interval includes 1.

What does the p-value add?

The unadjusted p-value for the comparison is 0.617. In general, a p-value describes how inconsistent the observed data, or data more extreme, would be with a specified null hypothesis under the statistical test's assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The registry record does not report a multiplicity-adjusted p-value for this comparison. The word unadjusted is important: it identifies the posted comparison as not adjusted for multiple testing.

The record describes the primary population as intent-to-treat. In general, an intent-to-treat analysis evaluates participants according to their randomized group, regardless of subsequent treatment received, when the analysis has been specified and implemented that way. The registry's population statement does not provide further detail about the response status assigned to participants with missing assessments or other departures from follow-up. Those handling details are not reported in the registry record.

Why can proportions be uncertain?

A proportion is estimated from a finite group of participants. If a different set of participants had been enrolled under the same conditions, the observed percentage could vary. Confidence intervals account for sampling uncertainty under a chosen method. With the listed 40 participants per group, the response estimates have intervals that span a substantial range of percentages. The posted summaries do not specify the confidence-interval construction method, so that method is not reported in the registry record.

The between-group estimates are also simple summaries of the response data. The displayed recalculated comparisons are unadjusted re-calculations from posted summary statistics; they do not incorporate an adjustment for other characteristics. The registry record does not report a covariate-adjusted estimate for the primary comparison. Because the discrepancy flag is false, no discrepancy between a sponsor analysis and the posted result is identified in this record.

Statistical significance and estimation answer related but different questions. A p-value summarizes evidence relative to a null model, while an estimate and its confidence interval show the magnitude and uncertainty of the contrast. In this trial, the interval around the risk difference spans zero and extends in both directions. Interpreting only the p-value would omit that range of compatible values; interpreting only the point estimate would omit its uncertainty.

These general explanations describe common statistical meanings, not additional claims about the trial's analysis plan. The registry record does not report the original sample-size justification, its planned statistical test in detail, the confidence-interval method, or whether interim analyses were planned. Those aspects cannot be inferred from the posted outcome numbers.

10. Confidence Intervals and Precision

The arm-level response estimates are 30.0% for sorafenib alone and 25.0% for sorafenib plus interferon. Their 95% confidence intervals are 16.56% to 46.53% and 12.69% to 41.2%, respectively. These intervals indicate that each percentage is estimated with substantial uncertainty. The observed proportions are useful summaries, but they should not be read as exact values for a broader population.

The risk difference is -0.05, with a 95% confidence interval from -0.245 to 0.145. Since the interval contains zero and extends on both sides of it, the posted data are compatible with a lower response proportion in the combination group, a higher response proportion in that group, or little difference. The interval is wide enough that the estimate does not precisely determine the direction or size of the group contrast.

The risk ratio confidence interval, 0.407 to 1.705, includes 1, as does the odds ratio confidence interval, 0.291 to 2.082. For each relative measure, 1 represents no relative difference. These intervals reinforce that the estimated relative effect is uncertain. They do not prove that the groups are equivalent; showing equivalence would require a defined equivalence margin and an appropriate analysis, neither of which is reported in the registry record.

In general, a 95% confidence interval is a method for expressing uncertainty in an estimate. It is not a guarantee that the true value lies inside any particular calculated interval. The precision observed should be read from the full interval, not from the number of decimal places in the point estimate. More decimal places do not remove the uncertainty represented by a broad interval.

The registry record does not report a target precision for the primary endpoint or the original sample-size justification. Therefore, the interval can be described as broad, but it cannot be judged against an unreported design target. The trial's response estimates and their intervals describe the posted evidence; they do not alone establish a clinically meaningful difference.

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 sample-size figures are planning illustrations based on the observed effect, not the trial's original design assumptions. They use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. The registry record does not report the trial's own original sample-size assumptions or justification.

For planning at 100% of the observed effect, the illustration gives 1,251 participants per arm for 80% power and 1,674 participants per arm for 90% power. For planning at 75% of the observed effect, it gives 2,256 participants per arm for 80% power and 3,019 participants per arm for 90% power. For planning at 50% of the observed effect, it gives 5,144 participants per arm for 80% power and 6,886 participants per arm for 90% power.

The 80% power column describes the planned number of participants per arm under the stated assumptions for a design intended to have an 80% chance of detecting the specified effect if that effect were true. The 90% column describes the corresponding planning target for a 90% chance. These are design probabilities under assumptions, not probabilities that the trial result is correct. In general, smaller effects are harder to distinguish from random variation, so detecting a smaller effect requires more participants under otherwise similar assumptions. The listed values reflect that pattern.

The curve should be read as the same planning relationship across effect sizes: as the assumed effect becomes smaller, the required number of participants per arm increases. The figures are participants per arm, not total participants or total events. This is not observed power; it is a prospective planning illustration based on the observed effect and stated assumptions.

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 = 0.300; treatment proportion p₁ = 0.250; average p̄ = 0.275.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6315 = 1.2377.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6305 = 0.5306.
4. n = (1.2377 + 0.5306)² / (-0.0500)² = 1,250.7, rounded up to 1,251 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 <- 0.3                          # control proportion
p1 <- 0.25                          # 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(126594)
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 = 1,251 (80% power), 1,674 (90% power)
 75% of observed effect: n per arm = 2,256 (80% power), 3,019 (90% power)
 50% of observed effect: n per arm = 5,144 (80% power), 6,886 (90% power)
power.prop.test: 1250.7 
Simulated power at n = 1,251 per arm: 0.801 (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 the objective response rate as the primary outcome and four secondary outcomes: selected grade 3-4 adverse events using NCI CTCAE Version 4.0, progression-free survival, median overall survival, and duration of response for participants with stable disease. It lists six outcome measures overall. Multiple outcomes can raise a multiplicity concern because testing more than one hypothesis can increase the chance of at least one apparently unusual result by chance.

The posted primary comparison has an unadjusted p-value of 0.617. The registry record does not report a multiplicity adjustment, a testing hierarchy, or a prespecified strategy for controlling error across outcomes. It also does not report a detailed statistical analysis plan. It is therefore not possible to determine from this record whether adjustment was planned or used for other analyses. The primary response comparison should be interpreted as the posted unadjusted comparison, rather than as evidence that all outcome testing was controlled under a single family-wise procedure.

The primary outcome population is identified as intent-to-treat. The record does not provide further details about how missing response assessments were handled, whether all randomized participants had evaluable response data, or what assumptions were used for any missing outcomes. Missing-data methods can affect estimates when outcome information is unavailable, so these specifics would be important to assess but are not reported in the registry record.

The registry lists four secondary outcomes, but their detailed estimates and analyses are not included in the registry record described for the posted response comparison. Results for those outcomes are therefore not reported in the registry record for this account. The record also does not state whether additional outcomes were assessed but not posted, or whether any analyses were exploratory. Such details should not be inferred from the number of outcomes listed.

In general, reporting the analysis population, missing-data handling, and multiplicity plan helps readers understand how outcome estimates were constructed and how to interpret statistical evidence across several outcomes. Here, the primary population is named, but the other details are not reported in the registry record. This limits assessment of how completely the analysis procedures can be evaluated from the registry information.

14. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: death (72.5%), hyperuricemia (15.0%), pancreatitis (2.5%), appendicitis (2.5%), serum amylase increased (2.5%).

The registry's adverse-event summary lists serious adverse events affecting 29 of 40 participants in the sorafenib-alone group and 30 of 40 in the sorafenib-plus-interferon group. It also lists other adverse events affecting 40 of 40 participants in each group. These counts describe the posted safety summaries; they do not by themselves establish whether an event was caused by treatment or compare the severity or duration of events between groups.

The leading serious-event terms listed are Death at 72.5%, Hyperuricemia at 15.0%, Pancreatitis at 2.5%, Appendicitis at 2.5%, Serum amylase increased at 2.5%, and Lipase Increased at 2.5%. These are event terms from the registry summary, not a treatment-effect estimate. The group-level death counts are listed as null, so separate counts of deaths by group are not reported in the registry record.

Safety summaries require attention to the category being counted, the population at risk, and the event definition. The record lists the number affected for serious and other adverse events, but further detail about timing, recurrence, attribution, and event-specific group counts is not reported in the registry record. The selected grade 3-4 adverse events are also listed as a secondary outcome, but detailed results for that outcome are not reported in the registry record.

These figures describe reported adverse events during the study context. They are not, on their own, a complete account of comparative safety or evidence that any event was caused by either treatment.

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

The response results describe the proportion of participants meeting the registry's RECIST definition of objective response. The posted rate was 30.0% with sorafenib alone and 25.0% with sorafenib plus interferon. The risk-difference estimate was -0.05, with a 95% confidence interval from -0.245 to 0.145, and the unadjusted p-value was 0.617. These summaries show that the observed rates were not identical, while also showing substantial uncertainty around the estimated difference.

The interval includes zero, so the data do not establish a difference in response proportions. But lack of statistical evidence for a difference is not proof that the treatments have equal effects. The confidence interval includes differences in either direction, and the registry does not report a prespecified equivalence margin or a non-inferiority objective. The posted result should not be recast as demonstrating equivalence.

The endpoint is response rate, not overall survival, progression-free survival, symptom change, or quality of life. Those outcomes are not interchangeable. Although progression-free survival and median overall survival are listed as secondary outcomes, their detailed results are not reported in the registry record for this analysis. The response result cannot be used to infer findings on those outcomes.

The safety summary lists serious and other adverse events, but counts alone do not establish causation. The registry record lists group-level adverse-event counts and leading serious-event terms; it does not report group-specific death counts. The response comparison and safety summary should each be read within their stated outcome definitions and available details.

The design was randomized and parallel-group, with no masking listed. Randomization is relevant to the intended comparison, while the absence of masking and unreported analysis details are part of the context for evaluating the evidence. The registry identifies the primary analysis population as intent-to-treat but does not report missing-data handling or the original statistical plan. These limits make it important to avoid conclusions beyond the posted estimates.

In statistical terms, the result is an uncertain estimate of an arm-to-arm response contrast. It does not establish that adding interferon increases response, decreases response, or leaves response unchanged. It does not provide medical advice or determine which treatment should be used.

16. Why This Trial Matters Statistically

This trial poses a clear statistical comparison: a binary response outcome in two randomized, parallel groups, with sorafenib in both groups and interferon added in one. That structure makes it possible to describe response proportions and estimate an absolute and relative contrast. The posted rates, 30.0% and 25.0%, are straightforward to read, while their confidence intervals and the comparison interval show why point estimates alone are not enough.

The risk-difference interval crosses zero and is broad, and the relative-effect intervals include 1. This illustrates the distinction between observing a numerical difference and estimating that difference precisely. The result does not establish a treatment difference, but it also does not demonstrate equality. That distinction is important when interpreting a non-significant comparison: an imprecise estimate may leave a range of effects compatible with the data.

The trial also illustrates how endpoint definitions shape interpretation. Objective response is based on specified RECIST categories and a restaging schedule. It is a particular measure of tumor response, not a substitute for the other outcomes listed in the registry. The presence of secondary outcomes makes reporting and multiplicity planning relevant, yet those analysis details are not reported in the registry record.

Finally, the planning illustration helps distinguish trial results from design calculations. It shows how assumed effect size and desired power influence the number of participants needed per arm under stated assumptions. Its figures are not observed trial outcomes or evidence about the trial's achieved power. Together, the outcome estimate, its uncertainty, the registered endpoint definition, and the limits of reported analysis detail provide the basis for a careful statistical reading.

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 primary response estimates are based on 40 participants in each group, and the confidence intervals are broad. The comparison interval includes zero and spans values in both directions, so the magnitude and direction of any group difference remain uncertain. The registry record does not report a target precision or original sample-size justification against which to assess that uncertainty.

The registry lists no masking. The record does not provide additional information about whether response assessments were independently reviewed or about the procedures used to limit assessment bias. These details are not reported in the registry record and should not be assumed.

The primary population is described as intent-to-treat, but the handling of missing or unevaluable response assessments is not reported in the registry record. The confidence-interval construction method and full statistical analysis plan are also not reported. The posted comparison is unadjusted, and the record does not specify a multiplicity strategy for the primary and secondary outcomes.

Four secondary outcomes are listed, but their detailed results are not reported in the registry record for this response analysis. The trial's registered outcome timeframe extends to 12 months for responding or stable participants, but additional details about follow-up completeness and outcome ascertainment are not reported. The registry also lists adverse-event summaries without group-specific death counts; those counts are not reported in the registry record.

These limitations concern what can be evaluated from the registry information. They do not establish that a particular method was or was not used when details are absent. The interpretation should remain focused on the posted response estimates, their uncertainty, and the explicitly reported design features.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Sorafenib Tosylate With or Without Recombinant Interferon Alfa-2b in Treating Patients With Metastatic Kidney Cancer — NCT00126594.
• Registry record last updated 2016-09-16; status last verified 2016-07.
• Results first posted 2016-09-16.
• 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

NCT00126594 is a completed randomized phase II trial sponsored by the National Cancer Institute (NCI), comparing sorafenib alone with sorafenib plus low-dose recombinant interferon alfa-2b in metastatic kidney cancer. The registry lists enrollment of 80 and 40 participants per group for the primary response outcome. Masking is listed as none.

The primary outcome was RECIST objective response rate in the intent-to-treat population. The posted response rates were 30.0% with sorafenib alone and 25.0% with the combination. The risk difference was -0.05, with a 95% confidence interval from -0.245 to 0.145; the unadjusted p-value was 0.617. The interval includes zero and allows for uncertainty in either direction.

Secondary outcomes listed include selected grade 3-4 adverse events, progression-free survival, median overall survival, and duration of response for participants with stable disease. The registry record does not report the original sample-size justification, detailed missing-data procedures, multiplicity plan, or detailed results for those secondary outcomes.