This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Genes in Predicting Outcome of Patients With DLBCL Treated With Rituximab and Combination Chemotherapy (R-CHOP) (NCT00450385), a na phase 2 study. The primary outcome is Determination of a List of Genes and Construction of Survival Prediction Models That Will Predict Overall Survival at 30 Months in DLBCL Patients Receiving R-CHOP Therapy..
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00450385 |
| Phase | Phase 2 |
| Condition | Lymphoma |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 57 |
| Arms in the primary outcome | n/a |
| Lead sponsor | University of Miami |
| Results posted | Yes |
2. Clinical Question
Population
Newly diagnosed patients with diffuse large B-cell lymphoma (DLBCL) treated prospectively with R-CHOP. The study required a minimum of 90 participants for gene-expression analyses; actual enrollment was 57.
Intervention
R-CHOP, comprising rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone. The treatment plan was 6 to 8 cycles.
Comparator
There was no separate comparator group. The study is recorded as a single-group trial, and allocation is listed as NA. It does not provide a like-for-like control comparison.
Primary question
Could gene-expression models or related biomarkers predict overall survival at 30 months among patients receiving R-CHOP? The registry record says no gene-expression data were collected because accrual was far below the required minimum, so the planned prediction models could not be evaluated.
The central aim was prediction, not a randomized comparison of R-CHOP with another treatment. The investigators proposed using RNA or protein gene expression, or biomarkers associated with rituximab’s anti-tumor effects, to identify information related to survival. The primary endpoint description defines overall survival time from diagnosis until death or the last follow-up examination.
Secondary questions included whether models or biomarkers could predict 24-month time to treatment failure, the overall response rate at the end of protocol therapy, and the number of participants with fixed tissue samples collected for future studies. These aims do not change the main limitation: with 57 participants enrolled against a stated minimum of 90 for gene-expression analyses, the record reports that no gene-expression data were collected. It therefore does not establish whether the proposed markers predict survival or treatment failure.
3. Trial Design
This was a phase II study sponsored by the University of Miami. The registry classifies it as a single-group study with no masking and allocation listed as NA. Enrollment was 57. The record gives the study start as 2007-04-24, primary completion as 2016-05, and status as terminated.
Participants were to receive R-CHOP for 6 to 8 cycles. The regimen description lists rituximab 375 mg/m2 on day 1, cyclophosphamide 750 mg/m2 IV on day 1, doxorubicin 50 mg/m2 on day 1, vincristine 1.4 mg/m2 IV on day 1 with a maximum of 2 mg, and prednisone 100 mg orally on days 1-5, repeated every 21 days. These are the planned regimen details in the record; they do not establish what each participant actually received.
The study’s purpose was to develop survival-prediction models from gene-expression information or biomarkers. The registry states that at least 90 participants were required for gene-expression analyses and that actual accrual of 57 was far below that minimum. As a result, no gene-expression data were collected. The primary prediction aim therefore lacks the data needed for model derivation and evaluation in this record.
The record lists six outcome measures but reports no analyses. It does not describe a control arm, a randomized allocation process, or masking. Because this is not a controlled comparison, the record cannot estimate a treatment contrast between R-CHOP and a different regimen or no treatment. The design instead set out to examine whether biological measurements could help predict outcomes among people receiving R-CHOP. The registry record does not report further detail about the planned modeling procedures or how prediction performance would be assessed.
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
| Stage | R-CHOP |
|---|---|
| STARTED | 57 |
| COMPLETED | 44 |
| NOT COMPLETED | 13 |
Reasons for not completing this period:
| Reason | R-CHOP |
|---|---|
| Death | 3 |
| Protocol Violation | 1 |
| Withdrawal by Subject | 2 |
| Lack of Efficacy | 2 |
| Adverse Event | 1 |
| Physician Decision | 1 |
| Other | 3 |
Analysis population for the primary outcome, as defined in the registry: The study required a minimum of 90 participants for gene expression analyses from which survival prediction model(s) could be derived. Due to actual participant accrual (57) being far below the minimum number of participants required, no data were collected on gene expression, and no survival prediction models were constructed.
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.
| Characteristic | Unit | R-CHOP |
|---|---|---|
| Participants analysed | 57 | |
| Age, Categorical: <=18 years | Participants | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 25 |
| Age, Categorical: >=65 years | Participants | 32 |
| Sex: Female, Male: Female | Participants | 30 |
| Sex: Female, Male: Male | Participants | 27 |
| Region of Enrollment: United States | Participants | 57 |
6. Primary Endpoint
The registry lists the primary outcome as Determination of a List of Genes and Construction of Survival Prediction Models That Will Predict Overall Survival at 30 Months in DLBCL Patients Receiving R-CHOP Therapy., measured over 30 months.
| Element | Registry entry |
|---|---|
| Result type | N/A |
| Measure of spread | not applicable |
| Unit | not stated |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is overall survival at 30 months in patients with DLBCL receiving R-CHOP. The endpoint description defines overall survival time as the time from diagnosis until death or the last follow-up examination. It is intended to provide the outcome that gene-expression models would predict.
In general, overall survival is a time-to-event outcome: the event is death, and follow-up time matters because participants may be followed for different lengths of time. Participants whose last follow-up occurs before death can contribute information up to that point; their later outcome is not known from that observation. A prediction model would use biological measurements to estimate how the outcome varies across participants, but a model requires suitable outcome and predictor data to be constructed and assessed.
In this study, the stated purpose was to identify genes and construct models predicting overall survival at 30 months. The registry record reports that accrual was 57, below the required minimum of 90 for gene-expression analyses, and that no gene-expression data were collected. It therefore reports no derived survival-prediction model or primary endpoint analysis. The endpoint remains the study’s planned target, not a reported estimate of survival or demonstrated prediction performance.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The main interpretive point is that the registry record describes a prediction study whose central planned analysis could not proceed as intended. It reports enrollment of 57, while the minimum stated for gene-expression analyses was 90. The record explicitly says no gene-expression data were collected because accrual was far below that minimum. No survival prediction model or analysis of the primary outcome is reported.
Consequently, the record does not answer whether RNA expression, protein expression, or biomarkers associated with rituximab’s anti-tumor effects predict overall survival. It also does not report an estimated survival probability at 30 months, a comparison between predicted and observed outcomes, or a measure of prediction accuracy. These are not null findings: absence of an analysis is different from evidence that a marker has no predictive value.
The study is single-group, without a separate comparator. Its design can address a prediction question within people receiving R-CHOP if the relevant predictor and outcome data are available, but it does not provide a randomized estimate of the treatment’s effect relative to another treatment. The registry record does not report a controlled treatment comparison.
The record lists secondary aims concerning prediction of 24-month time to treatment failure, overall response rate at the end of protocol therapy, and tissue collection. It does not report analyses for those outcomes in the posted results. The existence of an outcome listed in a registry is not itself evidence that it was measured, analyzed, or successfully reported.
Safety counts are reported separately: the R-CHOP group lists 4 participants affected among 57 at risk for the serious adverse event measure, 3 among 57 for the other adverse event measure, and 3 deaths. These counts do not provide a randomized comparison or establish that events were caused by treatment. The listed serious adverse event terms and percentages describe entries in the record, not a complete causal assessment.
Overall, the record supports a narrow conclusion about study conduct: enrollment fell short of the stated minimum for gene-expression analyses, and those data were not collected. It does not support conclusions about the predictive performance of genes or biomarkers, nor does it provide an analyzed estimate for the primary endpoint. A planning calculation elsewhere should not be mistaken for an observed result or for power achieved by this study.
9. Statistical Methods Explained
What kind of question was the study designed to address?
The study’s central question was whether gene-expression patterns or biomarkers could predict outcomes in patients receiving R-CHOP. This is a prediction question. In general, prediction asks how accurately information about a participant can forecast an outcome for that participant or a similar patient group. It is distinct from estimating the effect of assigning one treatment rather than another.
The registry identifies a single-group design and no separate comparator. That structure is consistent with an intended within-group prediction analysis, but it cannot by itself establish a treatment effect relative to a control group. The record does not report a randomized comparison. The proposed predictors were RNA or protein gene expression and biomarkers associated with rituximab’s anti-tumor effects. The record does not describe the exact model-building or validation procedure.
What does the primary endpoint represent?
Overall survival time is defined from diagnosis until death or the last follow-up examination, with the prediction target specified at 30 months. In general, time-to-event methods account for both whether an event is observed and how long each participant was followed. When follow-up ends before an event is observed, the observation is typically treated as censored: it contributes information through the last known follow-up but does not show what happened afterward.
The registry record does not report an analysis of this endpoint. It says no gene-expression data were collected because the 57 enrolled participants were far below the stated minimum of 90 required for gene-expression analyses. Without the planned predictor data, the record cannot report a model based on those data or its predictive performance.
How should a prediction model be assessed?
In general, constructing a prediction model and assessing its performance are separate tasks. Model construction estimates a relationship between predictors and outcomes in development data. Assessment asks how well predictions correspond to outcomes, ideally using data not used to build the model or another suitable validation approach. A model can appear to fit its development data while performing less well on new participants. The registry record does not describe a validation plan, a performance measure, or results from model development. Those details are not reported in the registry record.
In general, a confidence interval expresses uncertainty around an estimate under the assumptions of the analysis. A narrow interval indicates greater precision than a wide interval, but neither width nor the estimate can be interpreted without knowing the model, data, and outcome definition. No primary endpoint estimate or confidence interval is reported in this record, so there is no such result to interpret.
What do the planning illustrations mean?
Any sample-size calculations associated with the posted material are planning illustrations, not observed power and not part of the trial’s reported analysis. The stated assumptions for those illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. For time-to-event illustrations, the quantity is the total number of events. For other illustrations, it is participants per arm.
These assumptions describe hypothetical planning calculations. They do not change the actual design, which is single-group, and should not be read as evidence that the trial had two arms or achieved a particular probability of detecting an effect. The registry record does not report the study’s own sample-size justification or a calculation establishing its planned enrollment. The record does report a minimum of 90 participants for gene-expression analyses and actual accrual of 57.
Why can over-dispersion matter?
In general, over-dispersion occurs when observed variability is greater than a statistical model expects. If ignored, uncertainty may be understated and inferences can appear more precise than the data justify. Whether over-dispersion was relevant to any analysis in this study is not reported in the registry record; no analysis is listed.
What can be learned from the available analysis record?
The registry lists six outcome measures and zero analyses. An outcome definition describes what investigators intended to measure; an analysis result would show how collected observations were used to estimate or test a quantity. The two should not be conflated. Here, the record’s explanation for the missing gene-expression analysis is that accrual fell far below the stated minimum and no gene-expression data were collected. The primary endpoint is therefore a planned endpoint, not a reported model result.
10. Multiplicity, Missing Data and Analysis Populations
Multiplicity refers, in general, to the challenge of interpreting results when several hypotheses or outcomes are examined. Testing many outcomes or candidate predictors can increase the chance of at least one apparently notable finding by chance. Statistical plans may address this through prespecified priorities or adjustment methods. The registry record does not report a multiplicity strategy, any adjustment method, or an analysis plan specifying how multiple genes, biomarkers, or outcomes would be handled.
The record lists a primary overall-survival prediction aim and secondary aims involving 24-month time to treatment failure, overall response rate at the end of protocol therapy, and tissue collection. It reports six outcome measures but no analyses. The presence of several registered outcomes makes it important not to infer which were measured or analyzed: results for these secondary outcomes are not reported in the registry record.
Missing data are also central to interpretation. The record states that no gene-expression data were collected because enrollment of 57 was far below the required minimum of 90 for gene-expression analyses. This is not simply a small amount of missing predictor information within an otherwise reported model analysis; the planned gene-expression modeling could not be performed as described. The record does not report how missing outcome information, loss to follow-up, or other incomplete observations were handled.
The analysis populations are not reported in the registry record. It does not state which participants were included in any endpoint analysis, and it lists zero analyses. The record also does not report an interim-analysis plan, a prespecified analysis population, or a method for handling missing observations. Without these details and without reported analyses, there is no basis to assess multiplicity control, missing-data assumptions, or the robustness of any model estimates.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| R-CHOP | 4 | 57 | 7.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| R-CHOP | 3 of 57 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| R-CHOP | 3 of 57 |
Most frequent serious adverse event terms
| Event term | Organ system | R-CHOP |
|---|---|---|
| Abdominal pain | Gastrointestinal disorders | 2 of 57 |
| Dyspnea | Respiratory, thoracic and mediastinal disorders | 2 of 57 |
| Diarrhea | Gastrointestinal disorders | 1 of 57 |
| Febrile neutropenia | Blood and lymphatic system disorders | 1 of 57 |
| Fever | General disorders | 1 of 57 |
| Infection - Other | Infections and infestations | 1 of 57 |
| Muscle weakness, generalized or specific area (not due to neuropathy) | Musculoskeletal and connective tissue disorders | 1 of 57 |
| Urinary frequency | Renal and urinary disorders | 1 of 57 |
Most frequent non-serious adverse event terms
| Event term | Organ system | R-CHOP |
|---|---|---|
| Constipation | Gastrointestinal disorders | 2 of 57 |
| Fatigue | General disorders | 2 of 57 |
| Headache | Nervous system disorders | 2 of 57 |
| Leukocytes (total WBC) | Blood and lymphatic system disorders | 2 of 57 |
| Platelet count decreased | Blood and lymphatic system disorders | 2 of 57 |
| Alopecia | Skin and subcutaneous tissue disorders | 1 of 57 |
| Anorexia | Gastrointestinal disorders | 1 of 57 |
| Cholesterol, serum-high (hypercholestremia) | Metabolism and nutrition disorders | 1 of 57 |
| Constitutional Symptoms - Other | General disorders | 1 of 57 |
| Dizziness | Nervous system disorders | 1 of 57 |
Serious events with the highest arm-level rate: abdominal pain (3.5%), dyspnea (3.5%), diarrhea (1.8%), febrile neutropenia (1.8%), fever (1.8%).
The registry record lists safety information for the R-CHOP group. For the serious adverse event measure, 57 participants were at risk and 4 were affected. For the other adverse event measure, 57 were at risk and 3 were affected. The record also lists 3 deaths. These are participant counts as recorded; they are not rates calculated.
The listed serious adverse event terms include abdominal pain at 3.5%, dyspnea at 3.5%, diarrhea at 1.8%, febrile neutropenia at 1.8%, fever at 1.8%, and infection - other at 1.8%. They are registry entries and should be read as reported terms and percentages, not as a complete account of every event or a causal judgment about treatment.
In general, safety counts describe events observed and reported under the registry’s categories. They do not establish whether treatment caused an event, and a count alone does not indicate how events compare with another treatment. This study has no separate comparator group. The record does not provide a randomized safety comparison or further details about event severity, timing, or attribution. The death count should therefore be read as a reported count, not as an estimate of treatment-related mortality.
12. What the Result Does, and Does Not, Mean
The record documents a phase II, single-group study intended to develop gene-expression or biomarker models for predicting survival among patients with DLBCL receiving R-CHOP. It reports enrollment of 57 and states that at least 90 participants were required for gene-expression analyses. Because actual accrual was far below that minimum, no gene-expression data were collected.
That information supports a conclusion about the feasibility of carrying out the planned gene-expression work at the achieved accrual. It does not establish whether any gene or biomarker predicts overall survival, whether a prediction model would be accurate, or whether the same approach would perform in another group of patients. No primary endpoint analysis or prediction performance result is reported.
The absence of a result should not be interpreted as evidence of no association or no predictive value. A negative finding would require an analysis of relevant data and an estimate of the association or predictive performance. The record instead explains that the planned gene-expression data were not collected. The posted results therefore cannot distinguish between a potentially useful predictor and one with little predictive value.
The study is not a randomized comparison of R-CHOP with another treatment. Its design does not answer a comparative treatment-effect question. The listed safety counts and adverse event terms are descriptive registry entries; without a comparator and further attribution information, they do not establish causal effects.
Likewise, registered secondary outcomes are study aims, not evidence of completed analyses. Results for time to treatment failure, response, and tissue collection are not reported in the registry record. The most defensible reading is that the central biological prediction objective remained unassessed in the posted results, with the record identifying low accrual and the absence of gene-expression data as the reason.
13. Why This Trial Matters Statistically
This study illustrates an important distinction between registering a scientific aim and producing evidence that answers it. The objective was to use gene expression or biomarkers to predict survival in patients treated with R-CHOP. The registry record also specifies a minimum of 90 participants for gene-expression analyses, while actual enrollment was 57. It states that no gene-expression data were collected because accrual was far below that minimum. The proposed prediction question therefore remains unanswered in the posted results.
For biostatisticians and clinical researchers, the record highlights how design feasibility is linked to the data needed for the planned analysis. A model requires predictor measurements and outcome information in the same analytic setting. When the planned predictor data are not collected, the obstacle is not resolved by fitting a different model to the registered aim; the data necessary for that analysis are absent.
The case also underscores that a prediction study and a comparative treatment study answer different questions. This single-group study was designed around prognosis or prediction within participants receiving R-CHOP, not a randomized contrast between treatment strategies. Its results cannot be used as a controlled estimate of treatment effect.
Finally, the record demonstrates why null results, unavailable results, and incomplete analyses must be distinguished. No gene-expression analysis is reported, so there is no estimate to interpret as positive or negative. Clear reporting of accrual, missing predictor data, and the absence of analyses helps readers understand the boundary between the study’s original objective and what the registry record can support.
14. 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 is that the planned gene-expression analyses were not conducted as described. The registry states that the minimum required for those analyses was 90 participants, while enrollment was 57, and no gene-expression data were collected. This prevents evaluation of the primary prediction objective using the planned biological measurements.
The registry record reports zero analyses, so it provides no primary endpoint estimate, prediction-performance measure, or analysis of the listed secondary outcomes. It also does not report a validation procedure, the model specification, handling of missing observations, or the analysis populations. These details are necessary for judging the reliability and scope of any prediction result, but no such result is posted.
The study is single-group and has no separate comparator. This limits interpretation to the intended within-group prediction question and prevents a like-for-like comparison of R-CHOP with another treatment. The lack of masking is recorded, but the registry does not provide additional information about how this feature affected data collection or assessment.
The record gives a primary completion date of 2016-05 and a terminated status, but does not explain the termination beyond the accrual shortfall described for gene-expression analyses. The reason for termination, the timing and completeness of follow-up, and the handling of any participants who did not complete protocol therapy are not reported in the registry record.
Safety information is also limited to the listed counts and terms. The record does not provide a controlled safety comparison or details sufficient to assess event attribution. These limitations mean the posted information is useful for understanding study intent, accrual, and the stated barrier to the primary analysis, but does not establish predictive performance or comparative treatment effects.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Genes in Predicting Outcome of Patients With DLBCL Treated With Rituximab and Combination Chemotherapy (R-CHOP) — NCT00450385.
- Registry record last updated 2017-06-23; status last verified 2017-05.
- Results first posted 2017-06-23.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT00450385 is a phase II, single-group study sponsored by the University of Miami, titled “Genes in Predicting Outcome of Patients With DLBCL Treated With Rituximab and Combination Chemotherapy (R-CHOP).” The study aimed to develop gene-expression or biomarker models predicting overall survival at 30 months among patients with DLBCL receiving R-CHOP. The planned regimen was 6 to 8 cycles.
The record reports enrollment of 57, against a stated minimum of 90 for gene-expression analyses. It says no gene-expression data were collected because accrual was far below that minimum. The study status is terminated, and the record lists zero analyses. No primary prediction model result or analysis of the registered secondary outcomes is reported.
The design has no separate comparator and is not a randomized treatment comparison. Safety entries include 4 participants affected among 57 at risk for the serious adverse event measure, 3 among 57 for the other adverse event measure, and 3 deaths. These descriptive counts do not establish causation or comparative safety.