Note from the National Guideline Clearinghouse (NGC): The National Institute for Health and Clinical Excellence (NICE) commissioned an independent academic centre to perform an assessment of the manufacturer's submission on the technology considered in this appraisal and prepare an Evidence Review Group (ERG) report. The ERG report for this technology appraisal was prepared by the School of Health and Related Research (ScHARR), University of Sheffield (see the "Availability of Companion Documents" field).
Clinical Effectiveness
Description and Critique of Manufacturer's Approach to Validity Assessment
The submission does not reference a quality assessment tool. Quality is assessed in relation to the three criteria required by the single technology assessment (STA) specification: randomisation, blinding, and adequacy of follow-up.
It is not clear from published sources that a secure randomisation method was used in either BCIRG 001 or PACS 01. However, section 2.4.1 of the manufacturer's submission states that both studies used a secure randomisation method in which the randomisation sequence was kept away from the clinical area and administered by staff not directly involved in patient care. It is not clear how the manufacturer obtained this additional information in relation to PACS 01, given that they state that they did not have access to unpublished data from this trial, and that they state that information on the method of randomisation is not available for PACS 01.
The patients and clinical staff do not appear to have been blinded in either BCIRG 001 or PACS 01; this is claimed to be normal for cancer trials. It is not clear from the publications relating to either study whether the outcome assessors were blinded to treatment allocation, but the submission's response suggests that, in BCIRG 001, they were not. The US Food and Drug Administration (FDA) state that, although blinding is not essential when the outcome being measured is overall survival, it is preferred when the outcome is disease-free survival, and is necessary to minimise bias in the assessment of drug toxicity.
Description and Critique of Manufacturer's Outcome Selection
The primary outcome measure used in BCIRG 001 was disease-free survival, defined as time from randomisation to date of a clinical relapse, a second cancer (except skin cancer other than melanoma, ductal or lobular carcinoma in situ of the breast, or in situ carcinoma of the cervix), or death, whichever occurred first. The secondary outcome measures were overall survival (defined as time from randomisation until death from any cause), health-related quality of life, and toxic effects.
These outcome measures are appropriate. In adjuvant therapy, the prolongation of disease-free survival appears to represent intrinsic benefit rather than acting only as a surrogate for overall survival. However, the FDA advises that the magnitude of that benefit should be carefully weighed against the toxicity of the treatment. As noted earlier, an overall survival gain is generally felt to be required to compensate for the toxicity of the therapy.
The tools used in BCIRG 001 to measure health-related quality of life, are appropriate for this purpose.
Description and Critique of the Statistical Approach Used
The submission appears to contain unbiased estimates of relative treatment effects expressed in terms of hazard ratios, adjusted when necessary to take account of possible imbalances in prognostic factors. Meta-analysis was not undertaken as only one trial was identified which used docetaxel in its licensed application.
Refer to Section 4.1 of the ERG Report (see the "Availability of Companion Documents" field) for more information.
Economic Evaluation
Overview of Manufacturer's Economic Evaluation
The economic evaluation model has three components:
- Adjuvant chemotherapy decision tree model
A decision tree is used to calculate expected cost and quality-adjusted life year (QALY) outcomes associated with the adjuvant chemotherapy treatments under consideration.
- Model of long term disease progression
A state transition model (Markov model) is used to generate estimates of disease free survival, quality adjusted life years, and monitoring costs. These outcomes are incurred up to disease relapse or death over the lifetime of the model defined as 40 years in the base case.
- Consequences of disease relapse
Recurrence of locoregional breast cancer or distant metastatic disease is assumed to be associated with constant cost, survival and quality of life outcomes.
Refer to Section 5 of the ERG Report (see the "Availability of Companion Documents" field) for a complete discussion of the modelling of survival effects, quality of life and costs within these three components as well as the sensitivity analysis.