Daniel G. Coit, MD, of Memorial Sloan Kettering Cancer Center, New York, and Chair of the National Comprehensive Cancer Network (NCCN) Melanoma Guidelines Panel, discussed the findings reported by Lewin et al at the ASCO Annual Meeting.

“This is a small retrospective study examining a prospective database of patients with resected stage III melanoma who had PET [positron-emission tomography] scans as a part of a consistent surveillance program, looking at the accuracy of PET. The ‘gold standard’ of true-positive was the histologic status of the positive PET finding. There is no consistent working definition of a false-negative—ie, relapse after a negative PET—and, therefore, this biases the results toward a higher [negative predictive value] (true-negative divided by true-negative plus false-negative),” he suggested.

The study does confirm the relatively high false-positive rate of PET, with almost one-third (10 of 35) of the positive scans ultimately found to be unrelated to recurrent melanoma. It does not provide information as to what proportion of all recurrences were detected by patients or physicians without tests. Importantly, the study also lacks a contemporary comparator to PET scans to guide the use of this information, he maintained.

Dr. Coit also questioned whether immediate surgical resection of low-volume asymptomatic systemic recurrence detected by PET is likely to impact long-term outcomes. However, he acknowledged, “while we are not looking for the earliest opportunity to jump in … in the era of effective systemic therapy, we do need to reexamine whether early detection of systemic recurrence is associated with improved survival beyond increased lead time.” ■

Disclosure: Dr. Coit reported no potential conflicts of interest.