This was a well-conducted review (with the exception of a limited search) assessing the diagnostic accuracy of 18F-2-deoxy-2-fluoro-D-glucose positron emission tomography (FDG-PET) in evaluating breast cancer recurrence and metastases. The authors reported good diagnostic accuracy, and recommend that clinicians consider using the technique and that further research be undertaken. The authors' conclusions are supported by the results presented.

To assess the diagnostic performance of 18F-2-deoxy-2-fluoro-D-glucose positron emission tomography (FDG-PET) for the evaluation of breast cancer recurrence and metastases.

MEDLINE was searched from 1995 to June 2004; the search terms were reported. The electronic search was supplemented by a manual search of the bibliographies of primary studies and review articles. Studies were not excluded on the basis of language of publication.

Study designs of evaluations included in the review

No specific inclusion criteria for the study design were reported. Studies were not excluded on the basis of sample size.

Specific interventions included in the review

Studies using FDG and a dedicated PET camera were eligible for inclusion. Various camera models were used by the included studies, with resolution ranging from 6.0 to 4.6 mm.

Reference standard test against which the new test was compared

Studies that used pathology or clinical follow-up as the reference standard of diagnosis were eligible for inclusion. The majority of the included studies used a combination of pathology and clinical follow-up as the reference standard.

Participants included in the review

Studies of breast cancer recurrence were eligible for inclusion. Where a study included primary breast cancer as well as recurrence, only data relevant to recurrence and metastases were included in the analysis. The participants in the included studies were aged from 14 to 89 years.

Outcomes assessed in the review

Included studies were required to report sufficient data to allow the calculation of sensitivity and specificity. The true-positive rate (TPR; sensitivity) and false-positive rate (FPR; 1 minus the specificity) for each included study were reported.

How were decisions on the relevance of primary studies made?

Two reviewers independently assessed eligibility; any differences were resolved by consensus.

The criteria used to assess the methodological quality of the included studies evaluated the following: technical quality of FDG-PET; technical quality and application of the reference standard; independence of test interpretation and blinding of the readers; description of the clinical characteristics of the participants; and method of cohort assembly. Two reviewers extracted the data; any differences were resolved by consensus. The reviewers were not blinded to the study authors, institution, or source of publication.

Two reviewers extracted the data; any differences were resolved by consensus. The reviewers were not blinded to the study authors, institution, or source of publication.

How were the studies combined?

Pooled estimates of the TPR (sensitivity) and FPR (1 minus the specificity) were calculated. Summary receiver operating characteristic curves were generated using the method of Moses and Littenberg (see Other Publications of Related Interest). The point of maximum joint sensitivity and specificity (Q*) was also estimated as a measure of overall diagnostic accuracy. Separate analyses were conducted for patient-based data and lesion-based data.

How were differences between studies investigated?

Between-study heterogeneity was assessed using the chi-squared test. Subgroup analyses were conducted to explore potential sources of heterogeneity: the use of attenuation correction, visual interpretation of scans, and whole body scans.

A total of 18 studies were included in the review. Sixteen studies (808 participants) reported patient-based data and 7 studies (1,013 lesions) reported lesion-based data; 5 studies reported data of both types.

Study quality.

Of the 18 included studies, 7 were retrospective, 6 were prospective, and 5 did not report sufficient information to determine whether they were retrospective or prospective. Eight studies reported blinding of the FDG-PET readers to the results of the reference standard, 2 reported no blinding, and 6 did not specify. Only 4 studies reported the impact of FDG-PET findings on patient management.

Diagnostic accuracy (patient-based data).

The median sensitivity was 92.7% (range: 56 to 100) and the median specificity was 81.6% (range: 0 to 100). The pooled TPR was 90.3% (95% confidence interval, CI: 87.1, 92.9) and the pooled FPR was 12.7% (95% CI: 9.5, 16.4). Q* was 86.2% (95% CI: 82.8, 89.7). There was significant between-study heterogeneity (P<0.05). All subgroup analyses showed similar results.

Diagnostic accuracy (lesion-based data).

The median sensitivity was 91.7% (range: 57 to 97) and the median specificity was 88.9% (range: 79 to 96). The pooled TPR was 85.1% (95% CI: 81.6, 88.1) and the pooled FPR was 6.9% (95% CI: 4.9, 9.4). Q* was 89.1% (95% CI: 84.0, 94.2). There was significant between-study heterogeneity (P<0.05).

FDG-PET has a high diagnostic accuracy for the detection of breast cancer recurrence and can be considered a valuable tool for detecting recurrence and metastases.

This was a generally well-conducted review that addressed a clearly stated research question and applied appropriate inclusion criteria. The date restrictions on the literature search seemed appropriate for the technology under investigation. However, limitation of the search to one bibliographic database and reference scanning might have resulted in the omission of relevant data. Appropriate methods were used to minimise the introduction of error and bias during the review process, and the methodological quality of the included studies was assessed. The methods used to pool the data appeared broadly appropriate; although between-study heterogeneity was present in both the patient-based data and lesion-based data studies, there was no evidence of heterogeneity when outlying studies with low sensitivity values (one study that included only patients with no evidence of metastases in the patient-based data set, and one study that reported only data for bone lesions in the lesion based data set) were excluded. The authors' conclusions follow from the data presented.

Practice: The authors stated that clinicians should consider using FDG-PET in the follow-up of patients with breast cancer.

Research: The authors stated that FDG-PET has the ability to detect breast cancer recurrence at an earlier stage, and that further investigation is needed to determine whether women with early detection of recurrence, using FDG-PET, survive longer. The FPR and consequent number of unnecessary confirmatory procedures that would occur when a large population of women were monitored, using FDG-PET, after successful resection of primary breast cancer, should also be determined.

Phillips Medical Systems and Integral PET Associates.

Moses LE, Shapiro D, Littenberg B. Combining independent studies of diagnostic test accuracy into a summary ROC curve: data analytic approaches and some additional considerations. Stat Med 1993;12:1293-316.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.