Carcinogenicity Study of Bupropion: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Carcinogenicity Study of Bupropion (NCT01077596), a registered phase not stated study. The primary outcome is Number of Participants Diagnosed With Any of the Cancers Under Investigation Who Were Regularly Exposed to the Indicated Antidepressant.

2. Clinical Question

The rationale in the record is a lack of toxicology data on one bupropion metabolite and limited literature examining bupropion use and cancer risk. Because no particular cancer was identified as a specific concern, the study considered six common cancer sites to support site-specific comparisons. This is a rationale for the selected outcomes, not evidence that bupropion causes cancer or that any particular association was found.

3. Trial Design

This was a post-marketing, observational study, not a randomized treatment comparison. The registry lists the study as completed and describes a nested case-control design. Cancer cases were identified within a cohort of new antidepressant users, then matched with controls from that same cohort. The record does not state the allocation method; it does not describe assignment of participants to an antidepressant.

The study used two US population-based data resources with automated claims, pharmacy, and tumor registry data. The exposure period for the new-user cohort is given as January 1, 1996 through December 31, 2006. New use was defined as having no antidepressant prescription in the previous 6 months. The arm description further specifies people 18 years of age or older with medical and pharmacy benefits and at least 6 months of health plan enrollment before the first antidepressant prescription.

Exposure categories included regular bupropion use, regular SSRI use, regular TCA use, and regular use of any other antidepressant. The registry describes comparisons with other antidepressant users, but does not give the detailed matching variables, matching ratio, or exposure classification rules. Those details are not reported in the registry record.

The registry reports enrollment of 50430 and dates the study start to 2008-06 and primary completion to 2010-03. These registry entries do not establish the number of cancer cases or controls contributing to each analysis. Cancer-specific results and the analysis methods used to account for matching are 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.

Only the first 6 of 12 groups are shown.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Participants diagnosed with colorectal, lung, bladder, uterus, breast, or prostate cancer who were new antidepressant users

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.

Only the first 6 of 13 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants Diagnosed With Any of the Cancers Under Investigation Who Were Regularly Exposed to the Indicated Antidepressant, measured over January 1, 1996 - December 31, 2006.

The primary endpoint is the number of participants diagnosed with any cancer under investigation who were regularly exposed to the indicated antidepressant. The cancer sites named for this combined endpoint are colorectal, lung, bladder, uterus, breast, and prostate. Cases were nested within the cohort of new antidepressant users, with controls matched from that cohort.

A combined endpoint can address a broad question about cancer diagnoses across the selected sites, while the individual sites can be examined separately. The registry’s rationale was that no particular cancer was known to be associated with bupropion, so it selected six common cancers for the investigation. Combining sites may also make an overall comparison possible when each site-specific comparison addresses a narrower question. However, a combined count alone would not show whether an association differed by cancer site. The record lists site-specific outcomes as secondary measures.

The endpoint is stated as a count, but the registry record does not provide the endpoint results, exposure-specific denominators, or an effect estimate. It therefore does not establish the direction, magnitude, or precision of any association. The endpoint description also does not specify a statistical model or define the operational criteria for regular exposure.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a nested case-control design do?

In general, a nested case-control study starts with a defined cohort and identifies people who develop the outcome of interest as cases. Controls are selected from the same cohort, according to the study’s control-selection rules. This allows exposure among cases to be compared with exposure among controls while retaining the cohort as the source population. In this study, cancer cases were nested within a cohort of new antidepressant users, and controls were matched from that cohort.

Matching is intended to make cases and controls comparable on specified characteristics. Its interpretation depends on which characteristics were matched and how the analysis accounted for the matched sets. The matching factors and the statistical procedure used for the matching are not reported in the registry record. It is therefore not possible to assess how much matching addressed potential differences between exposure groups.

What does the comparison estimate?

In general, case-control studies compare the distribution of exposure among cases and controls. An odds ratio is a common association measure for this design, but the registry record does not specify the measure used in this study. No effect estimate is reported, so the size and direction of the comparison cannot be described. A count of exposed cases, without the corresponding control exposure information and analysis, is not enough to determine an association.

The exposure categories are regular bupropion use and regular use of other antidepressants, including SSRIs, TCAs, or any other antidepressant. The record does not explain the operational definition of regular exposure, the time window for classifying exposure, or whether exposure could change over time. These details affect what contrast a statistical estimate would represent. They are not reported in the registry record.

Why can over-dispersion matter?

General explanation: over-dispersion occurs when observed variability is greater than a statistical model expects. If a model assumes less variability than the data contain, standard errors may be too small and confidence intervals may be too narrow. Models can sometimes be adjusted to accommodate extra variability, but whether that is needed depends on the outcome and model. The registry record does not report a dispersion parameter or identify a model for these cancer comparisons, so over-dispersion cannot be evaluated from the information recorded.

How should the sample-size illustrations be read?

The record describes planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. These are illustrations, not the study’s reported enrollment by analysis group and not observed power. The record does not provide the numerical effect estimates or planning values needed to interpret specific illustrations. It also does not report the trial’s own sample-size justification, so that justification is not reported in the registry record.

In general, a sample-size calculation asks how much information a proposed design would need for a specified effect and statistical assumptions. It does not describe the precision actually achieved unless it is grounded in the analyzed data and design. A planning illustration based on an observed effect should not be interpreted as evidence that the study had a particular probability of detecting an effect.

What can be inferred from the completed status?

Completion describes the study’s status, not the presence or adequacy of a statistical analysis. The registry gives a primary completion date, but does not report whether analyses were interim or final, how missing exposure or outcome information was handled, or which analysis populations were used. These elements are needed to evaluate the analytic process and are not reported in the registry record.

The record lists 7 outcome measures and 56 analyses. These counts indicate that the registration encompasses multiple outcomes and analyses, but do not reveal which comparisons were prespecified as primary within each cancer site, whether all planned analyses were conducted, or how the analysis results were interpreted. The statistical plan is not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The study includes a combined primary cancer endpoint and secondary outcomes for colorectal, lung, bladder, uterine, breast, and prostate cancer. The registry also lists 56 analyses. When many outcomes or comparisons are examined, the chance of finding an apparently unusual result by chance can increase. General approaches to multiplicity include prespecifying a primary comparison, adjusting significance thresholds, or interpreting results with the full set of analyses in view. The registry record does not report whether any multiplicity approach was planned or used.

The record does not specify which analysis was the principal comparison for each site, whether the site-specific comparisons were confirmatory or exploratory, or how the combined endpoint related to the individual-site analyses. It also does not report a statistical analysis plan or results for the listed comparisons. Therefore, the role of multiplicity in interpreting any findings cannot be evaluated.

Missing-data handling is not reported in the registry record. The record does not describe how unavailable claims, pharmacy, or tumor registry information was handled, whether any participants or observations were excluded for incomplete data, or whether sensitivity analyses addressed missingness.

The analysis populations and denominators are also not reported. The enrollment entry does not identify how many cases or controls contributed to each cancer-specific comparison or how many had usable exposure information. The record does not report the number of matched sets retained for analysis. These omissions limit assessment of completeness, selection, and the precision of estimates.

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

The registry record establishes that a completed observational study was designed to examine the association between regular bupropion use and diagnoses of selected cancers, compared with other antidepressant use. It identifies a new-user cohort, a nested case-control design, two US population-based data resources, and the cancer sites under investigation. It also records the rationale for examining those sites.

The record does not report the cancer results needed to determine whether exposure was associated with a higher or lower frequency of diagnosis, or whether any comparison was compatible with no association. No effect estimates, confidence intervals, p-values, or cancer-specific exposure counts are reported. The available record therefore cannot establish the magnitude or precision of an association.

Even if an association estimate were available, an observational comparison would not by itself prove that antidepressant exposure caused a cancer outcome. Treatment choice and cancer diagnosis may be related to other characteristics, and the registry does not give enough information about matching, covariate adjustment, or exposure classification to assess those issues. General interpretation of such evidence must distinguish an observed association from a causal effect.

The zero values in the listed adverse-event group fields do not represent comparative cancer findings. They should not be read as evidence that no cancer diagnoses occurred or that the study found no association. The study’s completed status likewise does not substitute for reported results. The conclusions available from the registry record are about the study question and design, not about whether bupropion is associated with the cancers examined.

12. Why This Trial Matters Statistically

This study is statistically notable for applying a nested case-control design to a post-marketing question about medication exposure and cancer diagnoses. The design uses cases and controls drawn from a cohort of new antidepressant users, rather than describing a randomized assignment. This makes the definition of the source cohort, case identification, control selection, exposure classification, and matching central to interpreting any estimate.

The study also addresses a broad outcome question alongside site-specific comparisons. A combined endpoint can provide an overall view across selected cancer sites, while separate analyses can reveal whether results vary across those sites. These questions are not interchangeable: a combined result cannot by itself describe every site, and a collection of site-specific results raises multiplicity and interpretation considerations.

The registry’s focus on six cancers was tied to a stated rationale: there was no evidence that bupropion was associated with a particular cancer, and the selected sites were common cancers in the United States. This makes the distinction between the study rationale and the eventual findings important. The registry record does not report findings that would resolve the association question.

Finally, the record illustrates why completed status, enrollment, and outcome evidence should be kept separate. Enrollment of 50430 does not provide the case-control comparison, and completion does not supply effect estimates or their precision. For biostatistical interpretation, the design description is informative, but the absence of reported cancer results and key analysis details limits what can be concluded.

13. 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 registry record does not provide cancer association estimates, confidence intervals, p-values, or the exposure counts needed to assess the primary or site-specific comparisons. As a result, the direction, magnitude, and precision of the findings cannot be evaluated. The enrollment figure does not indicate the number of cases, controls, or matched sets used for any outcome.

Important operational details are also absent. The record does not describe how regular antidepressant exposure was defined, the exposure time window, how changes in use were classified, or the specific comparator used for each analysis. It does not identify the matching factors, the statistical model, or any covariate adjustment. These details are especially relevant to interpretation of an observational nested case-control study.

Outcome ascertainment is described through use of tumor registry data, but the record does not report validation procedures or the handling of incomplete data. Missing-data methods, analysis populations, exclusions, and sensitivity analyses are not reported in the registry record. The record also does not state whether multiple comparisons were addressed or whether all planned analyses were completed.

The study was not described as a randomized comparison, and allocation is not stated. This limits causal interpretation regardless of the statistical method used. These limitations concern what can be assessed from the registry record; they do not establish that the underlying study lacked methods or analyses that were not recorded there.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Carcinogenicity Study of Bupropion — NCT01077596.
• Registry record last updated 2017-05-30; status last verified 2017-05.
• Results first posted 2011-02-10.
• 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.

17. Record Summary

The registry record describes a completed post-marketing study of regular bupropion use and cancer diagnoses among new antidepressant users. It gives an enrollment of 50430, a study start of 2008-06, and primary completion in 2010-03. The design was nested case-control, using cases and matched controls from a cohort of new antidepressant users in two US population-based data resources.

The primary outcome covered colorectal, lung, bladder, uterine, breast, or prostate cancer. Listed exposure groups included regular bupropion, SSRI, TCA, and other antidepressant use. Cancer association estimates and key analysis details are not reported in the registry record. The listed adverse-event fields contain zero entries, but those entries are not the cancer comparison results.