The Overall Response Rate, or ORR for short, is the proportion of patients in oncological trials who are evaluable or considered according to the predefined analysis principle and have an objectively documented complete or partial tumor response. It thus describes the number of individuals in whom the tumor burden decreases sufficiently according to defined criteria. ORR is a measure of antitumor activity, but not direct evidence of longer survival.
Definition, calculation and RECIST reference
The EMA describes ORR as the proportion of individuals in whom a complete remission or a partial remission was observed. The numerator therefore includes confirmed or evaluated complete and partial responses according to the rules defined in the protocol. The denominator as well as the handling of non-evaluable, missing or early terminated cases must be determined in advance. When evaluating ORR, the intention-to-treat principle must generally be observed.
An objective response is based on a measurable decrease in tumor burden using defined target lesions or other justified indicators. The ORR condenses individual findings into a proportion and therefore does not answer how long a response lasts, when it begins or what effects occur outside of tumor measurement. Confidence intervals and a transparent presentation of all individuals considered are important for the interpretation.
Evaluation according to RECIST and protocol
The EMA guideline requires ORR to be documented according to international standards such as RECIST or WHO criteria. Which criteria are used, when imaging takes place, how target and non-target lesions are evaluated and how new lesions are included must be defined in the protocol. Adaptations to standard criteria can be appropriate in special situations, but must then be justified and described in advance.
The comparability of ORR values requires a consistent application of the same criteria. Differences in imaging intervals, confirmation of a response, reading methods or selection of lesions can influence the rate. An external independent review of the tumor response is recommended depending on the trial objectives. It can help in particular to ensure a consistent assessment across trial sites and treatment arms.
Classification as an efficacy measure
ORR is particularly suitable for describing the observed tumor activity of a therapy, for example in early development phases or in single-arm trials. The rate must be considered in the clinical context: tumor type, prior treatments, duration of response, safety, symptomatology and available alternatives influence its relevance. A high rate alone describes neither durability nor overall patient-relevant benefit.
For a comprehensible evaluation, data on the duration of response, time-to-event endpoints and survival data are usually reported in addition to ORR, if available. If ORR results are highlighted in a per-protocol population, data of all individuals enrolled in the trial should also be presented. This transparency prevents the impression of activity from being determined only by selected, easily evaluable cases.
Differentiation from overall survival and progression-free survival
Overall survival, or OS for short, measures the time from a defined starting point to death from any cause. It is therefore a time-to-event endpoint and not a proportion of individuals with tumor regression. An individual can show an objective response without this resulting in a longer overall survival for the entire trial group. Conversely, a therapy can influence survival without generating a high ORR.
Progression-free survival, PFS, captures the time until documented tumor progression or death. It thus takes into account both the course without progression and the temporal dimension. ORR exclusively captures the achievement of a predefined regression; stable disease and the duration until progression are not included in the rate. ORR, OS and PFS therefore answer different questions and must not be read as interchangeable evidence for the same benefit.
Relevance for clinical trials
For ORR trials, oncological assessment criteria, imaging time points, data queries and procedures for unclear findings must interact precisely. Incomplete imaging, deviating criteria or undocumented confirmations can significantly influence the rate. In single-arm designs, comparability with external data must also be justified particularly carefully.
Full-service CROs such as Mediconomics support the operationalization of RECIST-related requirements in the protocol, for eCRFs and imaging data flows as well as in the planning of the ORR analysis. Data management, biostatistics, medical writing, monitoring and project management coordinate evaluation windows, query processes, independent assessments and the transparent presentation of ORR, duration of response, PFS and OS.
Frequently Asked Questions (FAQ)
Which responses make up the ORR?
It comprises complete and partial tumor responses according to the defined evaluation criteria.
Is stable disease part of the ORR?
No. Stable disease can be clinically relevant, but is not part of the sum of complete and partial response.
Does a high ORR prove a survival benefit?
No. ORR describes tumor regression. For statements on survival or progression, the corresponding time-to-event endpoints must be evaluated separately.
Regulatory references
- EMA “Guideline on the evaluation of anticancer medicinal products in man” – defines requirements for ORR and the documentation according to RECIST.
- ICH E9 “Statistical Principles for Clinical Trials” – describes principles for analysis populations and target variables.
- ICH E8(R1) “General Considerations for Clinical Studies” – classifies clinically relevant endpoints and surrogate relationships.
- ICH E6(R3) “Good Clinical Practice” – requires reliable data collection and documented processes.