Phase I Imaging Study Evaluating Gem/Cis or Gem/Carbo for Participants With Non-Small Cell Lung Cancer (MK-0000-083 AM3): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase I Imaging Study Evaluating Gem/Cis or Gem/Carbo for Participants With Non-Small Cell Lung Cancer (MK-0000-083 AM3) (NCT00599755), a na phase 1 study. The primary outcome is Metabolic Response Conversion Rate Between 3 and 6 Weeks After Starting Chemotherapy at a Threshold of a 20% Decrease in SUVmean.

2. Clinical Question

The study focused on imaging change over time during chemotherapy, rather than a comparison of chemotherapy with no treatment or with a separate control regimen. Its endpoint asks about a change in metabolic-response classification between two scans. That is a specific question about FDG uptake and the chosen threshold; it is not, by itself, a direct measure of longer-term clinical outcomes.

3. Trial Design

The registry record describes a multicenter Phase Ib trial in participants with stage IIIB or IV non-small cell lung cancer. The study was completed, with an enrollment of 68. It is classified as a single-group study with allocation listed as not applicable and masking listed as none. The record does not describe randomized assignment to the two chemotherapy options, so the regimens should not be treated as randomized comparison arms.

The chemotherapy options were gemcitabine with cisplatin or gemcitabine with carboplatin. Imaging included FDG-PET, with CT or MRI also named in the intervention record. The official title describes imaging before and after chemotherapy. The primary endpoint concerns the change in metabolic-response classification from week 3 to week 6 after starting chemotherapy, among participants who were non-metabolic responders relative to baseline at week 3.

The registry record lists secondary outcomes concerning repeatability of baseline FDG SUVmean and changes in FDG-PET uptake over baseline-to-week-3, week-3-to-week-6, and baseline-to-week-6 intervals. These measures provide context for the study’s imaging focus, but their listed presence does not establish that every participant had an evaluable scan at every time point.

Enrollment is not necessarily the denominator for the primary endpoint: its stated population is participants with evaluable scans and a specified week-3 response classification. The registry record does not report the number in that population, the number with missing or unevaluable scans, or how such cases were handled. It also does not report a sample-size justification or the method used to decide between chemotherapy options. Those omissions limit reconstruction of how the study was planned and how its endpoint was analyzed.

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: Participants with evaluable scans classified as non-metabolic responders relative to baseline at 3 Weeks after starting chemotherapy

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 Metabolic Response Conversion Rate Between 3 and 6 Weeks After Starting Chemotherapy at a Threshold of a 20% Decrease in SUVmean, measured over Weeks 3 and 6 following chemotherapy.

The primary endpoint is the number of participants who changed from being classified as non-metabolic responders relative to baseline at week 3 to being classified as metabolic responders relative to week 3 at week 6. The classification uses a pre-specified threshold of a 20% decrease in mean standardized uptake value, or SUVmean, for FDG measured by PET. The endpoint population is restricted to participants with evaluable scans who met the week-3 nonresponse definition.

The record describes SUVmean as calculated by summing radioactivity from volumes of interest within each tumor and normalizing for injected dose and lean body mass. The endpoint therefore converts a continuous imaging measure into a response category using the stated threshold. It focuses on whether a participant crosses from one classification to another over the specified interval, not on the average amount of uptake change across all participants.

In general, a threshold-based endpoint can make a result easier to interpret as a classification, but participants close to the threshold may be placed in different categories after relatively small measurement changes. A conversion endpoint also depends on the initial classification: only participants classified as nonresponders at week 3 are included in the stated endpoint population. The registry record does not report the observed number of conversions, a rate, or an uncertainty interval. It therefore does not establish how often conversion occurred or how precisely it was estimated.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the endpoint count?

The endpoint is a count of participants who meet a sequence of conditions: they must have evaluable scans, be classified as non-metabolic responders relative to baseline at week 3, and then be reclassified as metabolic responders relative to week 3 at week 6. The classification rule is a 20% decrease in SUVmean. This is a conditional endpoint: it concerns a defined subset rather than everyone enrolled. The registry record does not report the size of that subset or the observed count.

In general, a count is most interpretable when its denominator is also clear. A count without the number of eligible participants does not show how common the event was in the population being analyzed. A rate would express the count relative to that denominator, while an uncertainty interval would indicate how much sampling uncertainty surrounds the estimate. The registry record reports neither the conversion count nor the denominator, rate, or interval.

Why does the threshold matter?

The endpoint classifies response according to a stated threshold for change in SUVmean. In general, thresholds turn a continuous measurement into categories. This can simplify interpretation, but it discards detail about the size of changes within each category. Participants whose measured changes lie near the boundary may be sensitive to measurement variability or scan conditions. The record names a repeatability outcome, but does not report its result, so the degree of measurement variability cannot be assessed from the registry record.

More generally, a threshold needs to be specified before outcome assessment to avoid choosing a classification rule after seeing results. The registry description calls the threshold pre-specified. That statement identifies the intended rule, but the registry record does not report further details of the analysis plan, such as how borderline or technically unevaluable scans were handled.

What can a single-group design establish?

This study is classified as single-group, with allocation not applicable. In general, measurements made at multiple times in one group can describe changes over time, but they do not provide the same protection against confounding as a randomized comparison. Changes may reflect the treatment period, measurement variation, participant characteristics, or other influences. Without a randomized comparator, an observed before-and-after pattern cannot by itself isolate the causal effect of a regimen.

The registry record names two chemotherapy combinations but does not state that participants were randomly allocated between them or give comparative results. The combinations therefore should not be interpreted as experimental and control groups. The named CT or MRI and FDG-PET procedures likewise do not constitute a control arm for the primary endpoint.

How should uncertainty be assessed?

In general, a confidence interval gives a range of values compatible with the data and the chosen statistical model. Its width reflects how informative the data are under that model. For a count or rate, interpretation also depends on the denominator and on the exact analysis population. The registry record does not report an interval for the primary endpoint, so its precision is not available. A missing interval does not establish that the estimate is precise or imprecise; it leaves precision unreported.

A p-value, when reported, addresses compatibility of data with a statistical null model under specified assumptions. It is not a measure of effect size, clinical importance, or the probability that a hypothesis is true. No primary-endpoint p-value or statistical test is reported in the registry record. The record also does not report an adjusted analysis or identify a covariate model for this endpoint.

What does over-dispersion mean in general?

In general, over-dispersion occurs when observed variation in count data exceeds what a statistical model assumes. If ignored, standard errors and uncertainty intervals can be misleading. The primary endpoint is described as a participant count, but the registry record does not specify a count model, report a dispersion estimate, or indicate that over-dispersion was evaluated. It would therefore be inappropriate to claim that such a method was or was not used.

What do sample-size illustrations mean?

In general, a sample-size calculation asks how much information a planned study might need to detect a specified effect under stated assumptions. The planning illustration associated with these analyses is described as using the observed effect, two-sided significance level 0.05, equal allocation between two hypothetical arms, and a normal approximation; the target effect is a stated fraction of the observed effect. It is a planning illustration, not the trial’s original design or observed power. The registry record does not report the trial’s own sample-size justification. Because this trial is recorded as single-group, the equal-allocation assumption in such an illustration should not be mistaken for the actual trial design. Any sample-size results from that approach describe hypothetical planning scenarios, not a retrospective measure of how likely the completed trial was to detect an effect.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Gemcitabine and Cisplatin or Gemcitabine and Carbo")
n      <- c(25)
events <- c(10)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Gemcitabine and Cisplatin or Gemcitabine and Carbo  25  10  40.0  21.1  61.3  23.4  59.3  23.4  59.4

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary endpoint and secondary imaging outcomes, including baseline repeatability and changes in FDG-PET uptake over several scan intervals. It reports six outcome measures and three analyses, but does not describe how statistical testing across outcomes or analyses was handled. In particular, a multiplicity adjustment, testing hierarchy, or strategy for controlling false-positive findings is not reported in the registry record.

In general, testing several outcomes can increase the chance of finding at least one apparently unusual result even when no underlying effect exists. Adjustment methods or a pre-specified hierarchy can address that issue, but the registry record does not identify such methods. The number of outcomes listed does not, by itself, show that all were tested in the same way or that any particular adjustment was required or applied.

The endpoint population requires evaluable scans and a week-3 nonresponse classification. The registry record does not report the number of participants meeting those criteria, missing-scan counts, reasons for missingness, or rules for excluding or imputing missing measurements. It also does not report sensitivity analyses for missing data. In general, missing-data handling matters when availability of a measurement is related to participant characteristics or outcome status; the direction and size of any resulting bias cannot be inferred without relevant information.

The record does not describe a statistical analysis plan, analysis-set definitions beyond the primary endpoint population, or handling of protocol deviations. It is therefore not possible to determine from the registry record whether analyses used all enrolled participants, only participants with complete imaging, or another defined set for outcomes other than the primary endpoint.

12. 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: pneumonia (4.4%), atrial fibrillation (2.9%), lower respiratory tract infection (2.9%), neoplasm malignant (2.9%), anaemia (1.5%).

The registry record lists one chemotherapy group with 68 participants at risk. It reports serious adverse events for 24 participants and other adverse events for 61 participants. These are separate registry summaries; they should not be added together, because the record does not state whether the same participant could be counted in both categories. The record does not report deaths in the field for deaths, so death information is not reported in the registry record.

The listed serious adverse-event terms and percentages are pneumonia, 4.4%; atrial fibrillation, 2.9%; lower respiratory tract infection, 2.9%; neoplasm malignant, 2.9%; anaemia, 1.5%; and neutropenia, 1.5%. These are the registry’s listed terms and percentages. The record does not provide further detail about event severity beyond the serious-event classification, timing, recurrence, or attribution to study treatment.

In general, an adverse event occurring during a trial is not automatically evidence that treatment caused it. Safety counts also depend on how events were collected, defined, and reported. The registry record does not provide a control group for a like-for-like safety comparison, and these figures alone do not establish comparative risk between the chemotherapy options.

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

The registry record establishes that this was a Phase Ib imaging study of gemcitabine with cisplatin or carboplatin in non-small cell lung cancer, and it gives a specific FDG-PET response-conversion definition. It does not report the observed primary endpoint count or rate. Accordingly, the posted results in the record do not show how many eligible participants converted from nonresponse at week 3 to metabolic response at week 6, nor how precisely that quantity was estimated.

The design is single-group rather than a randomized comparison. It can frame a question about imaging classifications over time among participants receiving chemotherapy, but it does not provide a like-for-like causal comparison between the two named regimens. The registry record also does not establish whether changes in FDG uptake corresponded to clinical benefit, or whether the imaging endpoint predicts another outcome. Those claims would require evidence not reported in this record.

In general, an imaging endpoint describes the measurement it was designed to capture. Here, the measure is FDG uptake summarized as SUVmean and classified using the specified threshold. A participant’s classification can change without that result alone establishing why it changed or what it means for other health outcomes. The listed secondary outcomes may address repeatability and changes across scan intervals, but their names do not provide their results.

The safety information describes reported adverse-event counts and selected serious-event terms. It is useful as a record of what was reported, but it does not support comparing the regimens or assigning causation. The registry record does not report deaths in the corresponding field, and it does not describe additional details needed to assess event timing or attribution.

The appropriate statistical conclusion is therefore limited: the record specifies a longitudinal imaging question and its classification rule, but does not report the primary result needed to answer that question. It also lacks the analysis details required to assess missing-data handling, multiplicity, and the precision of the endpoint estimate.

14. Why This Trial Matters Statistically

This trial is statistically notable for defining a response-conversion question rather than simply reporting a change in average FDG uptake. The endpoint asks whether participants who were nonresponders relative to baseline at week 3 were classified as responders relative to week 3 at week 6. That definition makes the starting classification and the evaluability of scans central to understanding the result.

The threshold-based approach also highlights a general measurement issue. Turning continuous SUVmean values into response categories can help create a clear rule, but it can conceal the magnitude of individual changes and make classification sensitive to measurement variation near the threshold. The registry record lists baseline repeatability as an outcome, making measurement consistency relevant to the study’s question, but it does not report the repeatability findings.

The single-group design matters because an observed change over time is not the same as a randomized treatment contrast. The registry describes two chemotherapy options, but does not establish randomized allocation or report comparative estimates. This limits what can be inferred about differences between those options.

Finally, the registry record illustrates the difference between a detailed endpoint definition and a reported endpoint result. It identifies the target population and classification rule, yet does not report the conversion count, denominator, or uncertainty interval. Without those results, readers cannot assess the observed frequency or its precision. The record is informative about the statistical question and its design context, but incomplete for judging the primary endpoint’s empirical answer.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The principal limitation for interpreting the primary endpoint is that the registry record does not report its observed count, denominator, rate, confidence interval, or test result. The stated enrollment of 68 cannot substitute for the endpoint denominator because the endpoint applies only to participants with evaluable scans who met the week-3 nonresponse definition. The number satisfying those conditions is not reported.

The study is classified as single-group, with allocation not applicable. Although two chemotherapy combinations are named, the registry record does not report a randomized comparison or provide regimen-specific endpoint results. The design therefore does not support a direct comparison of those options based on the information in the record.

The endpoint relies on FDG-PET SUVmean and a threshold-based classification. The record does not report scan-evaluability criteria, handling of measurements near the threshold, or the results of the listed repeatability outcome. These details would help readers assess how stable the classifications are under measurement variation.

Missing-data rules, the analysis plan, covariate adjustment, and multiplicity procedures are not reported in the registry record. Nor does the record give details linking the imaging endpoint to clinical outcomes. These omissions constrain assessment of analysis choices, potential bias, and the broader meaning of an imaging conversion. The listed safety summaries also lack details about timing and treatment attribution, and deaths are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Phase I Imaging Study Evaluating Gem/Cis or Gem/Carbo for Participants With Non-Small Cell Lung Cancer (MK-0000-083 AM3) — NCT00599755.
• Registry record last updated 2018-01-16; status last verified 2017-12.
• Results first posted 2012-06-27.
• 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.

19. Record Summary

NCT00599755 is a completed, multicenter Phase Ib, single-group imaging study in stage IIIB and IV non-small cell lung cancer. The registry record reports enrollment of 68 and describes chemotherapy with gemcitabine and cisplatin or gemcitabine and carboplatin, alongside FDG-PET and CT or MRI imaging.

The primary endpoint is the number of scan-evaluable participants classified as non-metabolic responders relative to baseline at week 3 who were classified as metabolic responders relative to week 3 at week 6, using a 20% decrease in SUVmean. The registry record does not report the endpoint result or its precision. It also does not report the statistical analysis plan, missing-data handling, or multiplicity procedures. Safety summaries list serious and other adverse-event counts, while death information is not reported in the registry record.