First and most important, the surgical approach of Wagnetz et al. might have tended to bias the results. The impact of intraoperative ultrasound depends on the results of preoperative imaging, which are essential in the workup. MRI and CT are more accurate, and intraoperative ultrasound is potentially less useful to stage the disease. However, the impact of intraoperative ultrasound depends on the surgical approach as well as on the rigor of the intraoperative ultrasound protocol. The surgical plan was completely decided preoperatively in at least 26% of the patients for whom intraoperative ultrasound was performed only in the liver segments spared by an anticipated major or extended resection. The stated reason for this approach was the need to minimize examination time during the operation, which we do not consider valid. Moreover, even though a parenchymal sparing approach was used by the authors, as stated in the Methods section, the majority of patients underwent traditional major or extended resection (201 major or extended vs 88 wedge resections). This result may explain the poor results of intraoperative ultrasound. For instance, in a case of anticipated right hepatectomy for some tumors in segments VI and VII, the findings of a new tumor in segment V does not change the surgical strategy unless a systematic ultrasound-guided parenchymal sparing approach is applied. Indeed, this point is enhanced by the finding of approximately 600 resected liver segments found to be free of disease by the pathologist. Perhaps a large part of those segments would have been spared using a different surgical approach. Thus, the surgical policy may predict the impact of intraoperative ultrasound, as has been widely highlighted.
Second, the impact of intraoperative ultrasound versus preoperative imaging techniques should be also evaluated considering intrahepatic recurrence at 6 months after surgery, which can be used as a surrogate for residual disease (false-negative findings at preoperative imaging as well as intraoperative ultrasound during surgery). Unfortunately, the authors did not perform this analysis.
Third, our experience, also confirmed by others, is that the impact of intraoperative ultrasound is positively correlated with the number of tumors—especially for colorectal liver metastases. In patients with multiple tumors (> four tumors), the probability of finding a new tumor at intraoperative ultrasound is definitively increased. Of note, the authors did not mention tumor size, tumor numbers, and tumor location. These features are of paramount importance to understand the population of the study and draw valid conclusions.
“IntraoperativeUltrasound of the Liver in Primary and Secondary Hepatic Malignancies: Comparison With Preoperative 1.5-T MRI and 64-MDCT,” AJR March 2011.