Diagnostic accuracy and interobserver agreement of contrast-enhanced ultrasound in the evaluation of residual lesions after treatment for malignant lymphoma and testicular cancer

Purpose: To calculate the diagnostic accuracy and interobserver agreement of contrast-enhanced ultrasound (CEUS) in the evaluation of residual lesions after treatment for malignant lymphoma and metastatic testicular cancer.

Materials and methods: Between May 2004 and October 2010 an experienced sonographer performed CEUS in 52 patients with residual lesions after treatment for malignant lymphomas and metastatic testicular cancer. Final judgement on presence or absence of active disease was based on histological findings and/or clinical follow-up. To quantify the diagnostic accuracy of the CEUS we calculated sensitivity, specificity, likelihood ratios and predictive values. A second, equally well experienced sonographer reassessed the results of the CEUS in every patient. To quantify the interobserver agreement we calculated the proportion of agreement and the Kappa statistic.

Results: Sensitivity was 72.7% (95% CI: 43.4 – 90.3%), specificity was 87.8% (95% CI: 74.5 – 94.7%), positive likelihood ratio was 5.96 (95% CI: 2.43 – 14.63) and negative likelihood ratio was 0.31 (95% CI: 0.12 – 0.82). The observers agreed in 84.6% (95% CI: 71, 4 – 92.7%) of cases, Kappa statistic was 0.76 (95% CI: 0.60 – 0.91)

Conclusion: Our preliminary study indicates that contrast-enhanced ultrasound might be helpful in the evaluation of residual lesions after chemotherapy for malignant lymphoma and metastatic testicular cancer. Based on the results we provide required sample sizes for an adequately powered phase 3 diagnostic accuracy study.