Efficacy and Safety of OncoGel™ Added to Chemotherapy and Radiation Before Surgery in Subjects With Esophageal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of OncoGel™ Added to Chemotherapy and Radiation Before Surgery in Subjects With Esophageal Cancer (NCT00573131), a randomized phase 2 study. The primary outcome is Overall Tumor Response at the Primary Tumor Site Based on Measurement of Primary Tumor Volume (Excluding Involved Lymph Nodes) by Spiral CT.

2. Clinical Question

The primary endpoint assessed complete or partial response using RECIST v1.0 criteria for target lesions. The registry record gives screening and Week 12 as the assessment timeframe. The response measure is a binary outcome at the participant level: a participant either met the criteria for overall response or did not.

The comparison concerns the addition of local OncoGel to a treatment regimen that both groups received. It is not a comparison of OncoGel alone with no cancer treatment. The registry record describes OncoGel as an experimental drug-delivery system that releases paclitaxel from a gel over time, alongside systemic chemotherapy and radiation before surgery. The primary result therefore addresses the response endpoint in the setting of this combined pre-surgical regimen, rather than isolating a mechanism or effect of any single treatment component.

3. Trial Design

This was a randomized, single-masked, parallel-group phase 2 study. The registry record lists enrollment of 137 participants. One group received OncoGel, radiation therapy, and systemic chemotherapy with cisplatin plus 5-FU before surgical resection. The other group received radiation therapy and the same systemic chemotherapy before surgical resection, without OncoGel.

The registry record lists a study start of 2008-01 and primary completion in 2010-11. The status is terminated. The reason for termination is not reported in the registry record, so the status alone does not establish why the study ended or what it means for interpretation of the posted results. The trial title and descriptions indicate a pre-surgical treatment setting for esophageal cancer.

The posted primary outcome is overall tumor response at the primary tumor site, based on measurement of primary tumor volume and excluding involved lymph nodes. Spiral CT assessment was specified at screening and Week 12. The registry description defines overall response as complete response plus partial response, using RECIST v1.0 criteria. The analysis population is described as subjects who received at least one treatment with OncoGel, systemic chemotherapy, or external beam radiation therapy.

The record lists one outcome measure and no posted analysis entries. It does not report the original statistical analysis plan, the randomization method, the masking procedures, allocation concealment, or the reasons for choosing the enrollment target. Those design details should not be inferred from the labels “randomized,” “single,” or “phase 2.” The groups’ posted sample sizes for the primary result are 72 and 65, respectively, and the outcome is reported as a percentage of participants.

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: Subjects who received at least one treatment with OncoGel, systemic chemotherapy or external beam radiation therapy were included for analysis of efficacy.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Overall Tumor Response at the Primary Tumor Site Based on Measurement of Primary Tumor Volume (Excluding Involved Lymph Nodes) by Spiral CT, measured over Screening and Week 12.

The primary endpoint is the percentage of participants with an overall response at the primary tumor site, measured using spiral CT. The recorded assessment timeframe is screening and Week 12. The measurement excludes involved lymph nodes and focuses on the volume of the primary tumor.

Under the registry’s RECIST v1.0 description, complete response means disappearance of all target lesions. Partial response means a >=30% decrease in the sum of the longest diameter of target lesions. Overall response combines complete and partial response. This turns imaging measurements into a binary participant-level outcome: each participant is classified as having an overall response or not having one, and the group result is the percentage classified as responders.

A binary response endpoint can make a treatment-group comparison easier to interpret than a set of individual imaging measurements, because it expresses the fraction of participants meeting a stated response definition. It also depends on the threshold and assessment rules: changes that do not meet the partial-response criterion are not counted as overall response under the stated definition. The endpoint does not, by itself, describe how long a response lasts, changes in symptoms, surgical outcomes, survival, or response in involved lymph nodes. Those outcomes are not included in the posted primary result.

The efficacy population is described as participants who received at least one treatment with OncoGel, systemic chemotherapy, or external beam radiation therapy. The registry record does not provide additional detail about handling participants with missing imaging or incomplete follow-up, so the endpoint definition should not be taken to imply a particular missing-data procedure.

7. Results Posted in the Registry

• Group 2: 20.0% (13/65) versus 12.5% (9/72) on control; risk difference 7.5 percentage points (95% CI -4.9 to 19.9), risk ratio 1.60 (0.73–3.49); unadjusted two-proportion z-test p = 0.232.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: overall response is classified as present or absent for each participant under the stated imaging criteria. A group percentage describes the proportion classified as responders. Group 1 had 12.5% overall response among 72 participants; Group 2 had 20.0% among 65 participants. The registry also lists the response counts as 9 and 13, respectively. Percentages make the outcome accessible, while the counts and denominators show how much participant information underlies each estimate.

What does the risk difference express?

The risk difference compares the response proportions by subtraction and is expressed on the absolute probability scale. The posted risk difference is 0.075 for Group 2 relative to Group 1. Its confidence interval runs from -0.049 to 0.199. A value of zero on this scale represents equal response proportions. Because the interval spans zero, the data do not pin down the direction of the underlying difference with precision. The interval should be read as a range of values compatible with the data under the method used, not as a list of equally likely effects or a guarantee about future studies.

How do relative measures differ?

The risk ratio compares the response probability in one group with that in the other. The posted risk ratio is 1.6, with a confidence interval from 0.733 to 3.494. A ratio of 1 represents equal probabilities. The odds ratio compares odds rather than probabilities; it is 1.75, with a confidence interval from 0.693 to 4.417. A value of 1 represents equal odds. Odds and probabilities are related but are not interchangeable, so the odds ratio should not be described as a percentage-point change or as a risk ratio. Both relative-effect intervals include their no-difference value, consistent with substantial uncertainty in this comparison.

What does the p-value test?

The registry comparison reports an unadjusted p-value of 0.232. In general, a p-value measures how incompatible the observed data are with a specified null model, assuming that model and the statistical procedure are appropriate. It is not the probability that the null model is true, and it does not measure the size or clinical importance of an effect. At the 0.05 significance level used in the planning illustration, 0.232 is not below that threshold. The record does not report a complete analysis plan, so the unadjusted result should not be mistaken for an adjusted or multiplicity-controlled analysis.

Why do confidence intervals matter?

A confidence interval communicates uncertainty around an estimate. In general, a wider interval means the data are less precise about the effect scale being reported. Here, the intervals extend across values representing no difference and also allow effects in either direction. This is important because a single point estimate—such as the risk difference or risk ratio—does not convey how much sampling uncertainty surrounds it. The interval is not a prediction interval for individual outcomes and does not ensure that a future study will produce an estimate within its limits.

What can the posted analysis establish?

These statistics describe a comparison of the recorded primary response outcome between the two groups. Randomization is intended in general to support fair comparisons by balancing factors on average, but the record does not describe the allocation procedure or provide enough information to assess implementation. The posted comparison does not report covariate adjustment. No analysis entries are posted, and additional statistical analyses are not reported in the registry record.

For binary data, estimates depend on both the event counts and the denominators. Small numbers of responses can make estimates sensitive to the outcomes of a few participants, and confidence intervals reflect sampling uncertainty. In general, confidence intervals for proportions and comparative measures can be calculated by different methods, with differing behavior in small samples. The specific interval construction used for the posted comparison is not reported in the registry record. The available estimates should therefore be interpreted using their stated values and limits, without attributing an unreported method to them.

The endpoint is defined by a response threshold, rather than by a continuous measure of tumor change. This can simplify interpretation but discards detail about the magnitude of changes among participants who do not meet the response criterion. The registry does not post secondary outcomes or analyses that would provide complementary views of tumor change, durability, or other endpoints. Statistical interpretation is consequently limited to the binary primary endpoint and posted group comparison.

10. Confidence Intervals and Precision

The primary response percentages are based on 72 participants in Group 1 and 65 in Group 2, with 9 and 13 responses, respectively. The comparative confidence intervals indicate appreciable uncertainty around the point estimates. For the risk difference, the estimate is 0.075 and the interval extends from -0.049 to 0.199. Because that interval crosses zero, the data are compatible with the absence of an absolute difference and with differences in either direction.

The risk-ratio interval, 0.733 to 3.494 around an estimate of 1.6, includes 1, the value for no relative difference. The odds-ratio interval, 0.693 to 4.417 around 1.75, likewise includes 1. On each scale, the interval covers a range of effects rather than identifying a single precise value. The results therefore do not support a precise estimate of the size or direction of a comparative response effect.

Confidence intervals are uncertainty summaries for an estimated effect under the statistical method used. In general, they are not guarantees that the true value lies within the interval, and they do not describe the range of outcomes expected for individual participants. Their interpretation also depends on the calculation method. The specific confidence-interval method is not reported in the registry record.

The unadjusted p-value is 0.232, which accords with the intervals’ inclusion of the respective no-difference values. Neither a point estimate nor a p-value alone conveys precision: the group percentages describe what was observed, while the confidence intervals show how much uncertainty surrounds the comparisons. The sample sizes and response counts are modest enough that each estimate should be read with its interval rather than as a settled quantity. The registry record does not report precision targets or the original sample-size justification.

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. Each row describes planning for a stated fraction of the observed effect. The figures are participants per arm, not total events: for 100% of the observed effect, the illustration gives 379 participants per arm for 80% power and 507 participants per arm for 90% power. For 75% of the observed effect, it gives 643 participants per arm for 80% power and 860 participants per arm for 90% power. For 50% of the observed effect, it gives 1,373 participants per arm for 80% power and 1,838 participants per arm for 90% power.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The figures do not document the trial’s original design calculations; those assumptions are not reported in the registry record.

In general, power is the probability that a study will detect an effect of a specified size under its assumptions if that effect is present. The 80% and 90% columns show planning targets: a design targeting 90% power is intended to have a higher chance of detecting the specified effect than one targeting 80% power, all else equal. Smaller effects are harder to distinguish from random variation and therefore require more participants under these assumptions. These planning figures are not observed power.

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 = 0.125; treatment proportion p₁ = 0.200; average p̄ = 0.163.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5217 = 1.0225.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5190 = 0.4368.
4. n = (1.0225 + 0.4368)² / (0.0750)² = 378.6, rounded up to 379 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 <- 0.125                          # control proportion
p1 <- 0.2                          # 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(573131)
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 = 379 (80% power), 507 (90% power)
 75% of observed effect: n per arm = 643 (80% power), 860 (90% power)
 50% of observed effect: n per arm = 1,373 (80% power), 1,838 (90% power)
power.prop.test: 378.6 
Simulated power at n = 379 per arm: 0.804 (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 record lists one outcome measure and no secondary outcomes. It also lists no posted analysis entries. The primary endpoint is clearly described, but the registry record does not provide a full statistical analysis plan, so details such as covariate adjustment, prespecified subgroup analyses, or the treatment of multiple planned comparisons are not reported in the registry record.

Multiplicity arises when several statistical tests or outcomes are considered: the more tests performed, the greater the chance that at least one result appears unusual by chance. Whether any adjustment would be relevant depends on the planned family of analyses. The record does not report interim analyses, additional unposted testing, or a multiplicity-adjustment strategy. The posted p-value is explicitly unadjusted and should be understood on that basis, without assuming that it is the only analysis performed outside the registry entries.

The efficacy population is described as participants who received at least one treatment with OncoGel, systemic chemotherapy, or external beam radiation therapy. That wording identifies a treated population, but the record does not state how many randomized participants were excluded from the efficacy analysis, whether the groups were analyzed according to original assignment, or how protocol deviations were handled. It does not report a detailed analysis-population definition beyond the stated treatment criterion.

Missing outcome data can affect a binary response comparison if participants without an evaluable scan are omitted or classified according to a rule. The registry record does not report the number of participants with missing Week 12 assessments, the amount of missing data, reasons for missingness, or the method used to handle it. It also does not specify whether all randomized participants were included in a primary analysis. These omissions limit assessment of possible bias from exclusions or incomplete follow-up. They do not establish that missing data occurred; they mean the handling and extent are not reported in the registry record.

These reporting gaps make it difficult to assess the analysis against a prespecified plan or determine the sensitivity of the findings to alternative population and missing-data choices. The posted response percentages and comparison remain descriptive of the reported analysis, but should not be assumed to represent an intention-to-treat analysis or a complete multiplicity-controlled evaluation.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: thrombocytopenia (6.2%), dysphagia (6.2%), leukopenia (4.2%), oesophagitis (4.2%), respiratory failure (4.2%).

The registry reports serious adverse events affecting 30 of 72 participants in Group 1 and 20 of 65 in Group 2. Other adverse events affected 67 of 72 participants in Group 1 and 55 of 65 in Group 2. These are participant counts and denominators as posted; they do not by themselves indicate the number of separate events each participant experienced, event duration, or event severity beyond the serious-event classification.

The listed serious adverse-event terms and percentages are thrombocytopenia, 6.2%; dysphagia, 6.2%; leukopenia, 4.2%; oesophagitis, 4.2%; respiratory failure, 4.2%; and neutropenia, 3.1%. The registry record does not identify these terms by treatment group in the listed summary, so the percentages should not be assigned to either group. Death counts are not reported in the registry record.

Safety counts are descriptive and do not establish that an event was caused by a particular treatment. The record does not provide details about event timing, exposure duration, severity grading, or statistical comparison between groups. The figures should be read as the posted safety summaries rather than as a complete assessment of relative safety.

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

The posted result shows overall response percentages of 12.5% in the group receiving OncoGel with chemotherapy and radiation, and 20.0% in the group receiving chemotherapy and radiation without OncoGel. The reported comparative estimates are uncertain: the risk-difference interval includes zero, and the intervals for the risk ratio and odds ratio include 1. The unadjusted p-value is 0.232. Taken together, these results do not precisely establish the direction or size of a difference in the primary response endpoint.

This comparison concerns the regimens as assigned in a randomized, parallel-group phase 2 study, in a pre-surgical setting for esophageal cancer. It does not compare OncoGel alone with no treatment, since both groups received systemic chemotherapy and radiation before surgery. The outcome is overall tumor response at the primary tumor site by spiral CT, with involved lymph nodes excluded. It is not a measure of survival, response duration, surgical success, symptoms, or quality of life; those results are not posted as outcomes in this record.

The point estimates are observed summaries from the registry’s stated efficacy population. The record does not give enough information to confirm how missing assessments, participants who received no treatment, protocol deviations, or other analysis-population issues were handled. It also does not report the full analysis plan or the method used to calculate the confidence intervals. These limits matter when deciding how much confidence to place in the apparent difference between the group percentages.

The study status is terminated, but the reason for termination is not reported in the registry record. The status should not be treated as evidence for or against an effect. The posted statistics are best read as an imprecise comparison of the reported primary endpoint in this study, not as a definitive conclusion about the broader clinical role of OncoGel or about outcomes beyond the endpoint measured.

16. Why This Trial Matters Statistically

This study illustrates why a treatment comparison should be read through both its point estimates and its uncertainty. The response percentages differ, but the comparative confidence intervals cover the no-difference value and values on either side of it. Reporting the risk difference, risk ratio, and odds ratio makes clear that the apparent pattern depends on the scale used to express the effect, while the wide intervals emphasize that the underlying effect is not precisely determined by these data.

It also highlights the importance of understanding what the endpoint includes. Overall response is a threshold-based classification from spiral CT at the primary tumor site, excluding involved lymph nodes. A binary response measure is interpretable as a fraction of participants meeting a defined criterion, but it does not capture all tumor changes or other outcomes. Statistical conclusions should remain tied to the outcome actually measured.

The trial compares an added local treatment within a combined regimen. Both groups received chemotherapy and radiation before surgery, so the intended contrast is the addition of OncoGel, not treatment versus no treatment. That distinction is essential when interpreting the question addressed by the comparison. The registry’s treatment-received description of the efficacy population and the absence of a detailed analysis plan also show why population definitions and missing-data reporting matter for evaluating randomized evidence.

Finally, the planning illustration demonstrates that sample-size needs depend on the effect a future study aims to detect: planning for smaller fractions of the observed effect requires more participants under the stated assumptions. Such illustrations are not a substitute for the original design rationale, which is not reported in the registry record. Together, these features make the study a useful example of the separation between observed estimates, statistical precision, and the strength of conclusions a registry result can support.

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 posted primary comparison has limited precision, with confidence intervals that include the no-difference value on the risk-difference, risk-ratio, and odds-ratio scales. The number of participants contributing to the group results is 72 and 65, with 9 and 13 responses. The point estimates therefore require cautious interpretation and should not be treated as precise measures of an underlying effect.

The record describes the efficacy population as participants who received at least one treatment, rather than explicitly stating that all randomized participants were analyzed in their assigned groups. It does not report the number of participants with missing imaging, how missing assessments were handled, or whether exclusions differed between groups. The registry record also does not report an analysis plan, the specific confidence-interval method, interim analyses, or any adjustment strategy for multiple comparisons.

The primary endpoint is based on response at the primary tumor site and excludes involved lymph nodes. It is assessed by spiral CT at screening and Week 12 and does not convey response duration or other clinical outcomes. No secondary outcomes are posted. The registry does not describe the procedures used for randomization or masking beyond the design labels, and the reason for termination is not reported in the registry record. These are limits on interpretation of the available record; they do not establish that a particular bias, missingness problem, or procedural issue occurred.

Finally, the planning sample-size figures are based on the observed effect under stated illustrative assumptions, not on the study’s original design justification. The original assumptions and rationale are not reported in the registry record. The posted efficacy and safety summaries should therefore be interpreted as limited registry results rather than as a complete account of all trial analyses or outcomes.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Efficacy and Safety of OncoGel™ Added to Chemotherapy and Radiation Before Surgery in Subjects With Esophageal Cancer — NCT00573131.
• Registry record last updated 2022-06-22; status last verified 2022-06.
• Results first posted 2014-05-16.
• 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

NCT00573131 is a randomized, single-masked, parallel-group phase 2 study of OncoGel added to systemic cisplatin and 5-FU chemotherapy and radiation before surgery for esophageal cancer. The comparator group received chemotherapy and radiation without OncoGel. The registry lists enrollment of 137, a study start of 2008-01, primary completion in 2010-11, and status as terminated; the reason for termination is not reported.

The primary outcome is overall response at the primary tumor site by spiral CT, assessed at screening and Week 12. Response was reported as 12.5% among 72 participants in Group 1 and 20.0% among 65 in Group 2. The risk difference is 0.075, with a confidence interval from -0.049 to 0.199, and the unadjusted p-value is 0.232. No secondary outcomes or analysis entries are posted. The registry record does not report the original sample-size rationale or detailed missing-data and analysis procedures.