CRD summary

This review concluded that the strain ratio and length ratio, when used as strain measuring methods, had good diagnostic accuracy, similar to that of the elasticity score. This was a well-conducted review, but uncertainty around the blinding of test interpreters, the lack of which could have overestimated test accuracy, means the conclusions seem overly strong.

Authors' objectives

To assess the diagnostic performance of strain ratio and length ratio measurements, during ultrasound elastography, for differentiating benign and malignant breast masses.

Searching

PubMed, EMBASE, Web of Knowledge and The Cochrane Library were searched, for studies published in full in any language, from their inception to January 2012; search terms were reported. Bibliographies of identified studies were scanned.

Study selection

Studies of women of any age, with breast symptoms or an abnormal clinical breast examination, breast B-mode ultrasonography or mammography, were eligible for inclusion if they assessed the diagnostic performance of ultrasound elastography, and reported the strain ratio or length ratio, or both, for the characterisation of focal breast lesions. Histopathology (core biopsy) or cytology (fine needle aspiration) had to be used as the reference standard.

All included studies were conducted in women, used freehand compression elastography probes, and used histopathology as the reference standard, with or without cytopathology. The cut-off point for strain ratio varied from 0.5 to 4.5, and for length ratio it varied from 1.0 to 1.1. Where reported, the mean age of patients ranged from 39 to 58 years.

Two independent reviewers selected studies; discrepancies were resolved by discussion.

Assessment of study quality

Study quality was assessed, using the 14-point QUADAS, by two independent reviewers; discrepancies were resolved by discussion.

Data extraction

Data were extracted, by two independent reviewers, to construct 2x2 tables of test performance for strain ratio and length ratio; discrepancies were resolved by discussion.

Methods of synthesis

Summary estimates of sensitivity and specificity, with 95% credible intervals, were calculated using random-effects bivariate generalised linear mixed modelling. A Bayesian Markov chain, Monte Carlo simulation, with non-informative priors, was used. Positive and negative likelihood ratios, with 95% credible intervals, and the area under the curve, with 95% credible region, were calculated. Heterogeneity was assessed using Ι², and 50% was considered to indicate substantial heterogeneity. Univariate regression was used to investigate the impact of study quality. Publication bias was assessed using Deeks’ funnel plot asymmetry test.

Results of the review

Twelve studies were included in the review, with 2,270 patients (range 71 to 437). All studies used an appropriate reference standard and explained withdrawals. None reported whether the reference standard was performed without knowing the results of the index test.

Strain ratio: Nine studies reported the strain ratio. This included 1,875 patients, who had 2,087 breast masses, and 667 were malignant. Summary sensitivity was 88% (95% CrI 84 to 91), specificity was 83% (95% CrI 78 to 88), the positive likelihood ratio was 5.57 (95 % CrI 3.85 to 8.01), the negative likelihood ratio was 0.14 (95 % CrI 0.09 to 0.20), and the area under the curve was 92% (95% CrR 90 to 94). There was no evidence of publication bias and heterogeneity was low (upper confidence limit Ι²<25%).

Length ratio: Three studies reported the length ratio. This included 395 patients, who had 450 breast masses, and 131 were malignant. Summary sensitivity was 98% (95% CrI, 93 to 99), specificity was 72% (95% CrI 31 to 96), the positive likelihood ratio was 5.38 (95 % CrI 1.32 to 16.74), the negative likelihood ratio was 0.04 (95% CrI 0.004 to 0.14), and the the area under the curve was 92% (95% CrR 90 to 93). Heterogeneity assessment, with Ι², was inconclusive, with very wide confidence intervals, due to so few studies.

Authors' conclusions

The results suggested that the strain ratio and length ratio, when used as strain measuring methods, had good diagnostic accuracy, similar to that of the elasticity score.

CRD commentary

The review addressed a clear question with reproducible inclusion criteria. Several appropriate sources were searched for published studies; unpublished studies were not sought. Each stage of the review process was conducted in duplicate, reducing the risk of error and bias. Appropriate criteria were used to assess study quality, and the results were published in full, in an online appendix. Appropriate methods of synthesis were used.

This was a well-conducted review, but the included studies poorly reported the blinding of interpreters of tests, and a lack of blinding could have overestimated accuracy, so the conclusions seem overly strong.

Implications of the review for practice and research

Practice: The authors stated that in patients with a higher pretest probability of disease, the strain ratio was recommended as its sensitivity was significantly greater than the elasticity score. The length ratio had higher sensitivity, but the estimate was less reliable. The authors stated that the strain ratio or length ratio, after a positive B-mode ultrasound, might benefit patients to a similar degree as the elasticity score, and may be informative for the decision to biopsy, but they should not replace biopsy.

Research: The authors recommended that researchers reported their studies according to the STARD statement. They stated that further, rigorously designed, adequately powered studies were needed to evaluate the diagnostic performance of the combination of B-mode ultrasound and ultrasound elastography, using strain ratio or length ratio, and to investigate the accuracy of length ratio with more precision. Studies should determine the most appropriate cut-off point for reporting strain, in different measuring systems, and assess the test performance across different lesion sizes and depths from the breast surface.

Funding

Not reported.

Bibliographic details

Sadigh G, Carlos RC, Neal CH, Dwamena BA. Accuracy of quantitative ultrasound elastography for differentiation of malignant and benign breast abnormalities: a meta-analysis. Breast Cancer Research and Treatment 2012; 134(3): 923-931. [PubMed: 22418703]

Original Paper URL

http://link.springer.com/article/10.1007%2Fs10549-012-2020-x

Indexing Status

Subject indexing assigned by NLM

MeSH

Adult; Aged; Breast Diseases /ultrasonography; Diagnosis, Differential; Elasticity Imaging Techniques; Female; Humans; Middle Aged; Reproducibility of Results; Sensitivity and Specificity; Ultrasonography, Mammary; Young Adult

AccessionNumber

12012054503

Database entry date

17/04/2013

Record Status

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.