ATLANTA – The routine use of imaging for the staging or surveillance of asymptomatic patients with primary cutaneous melanoma appears unwarranted, according to findings from a comprehensive literature review.
Even in high-risk patients, the use of chest x-ray (CXR), computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET), or PET-CT, yields unacceptably high rates of false positive results and has minimal impact on treatment plans, Dr. Daniel Eisen reported at the annual meeting of the American Society for Dermatologic Surgery.
The findings, which are based on an extensive systematic search of the Medline database for all relevant imaging studies conducted from 1970 to 2011, did suggest that high-frequency ultrasonography might be of diagnostic value for the early detection of subclinical lymph node metastases when used with routine follow-up in high-risk patients, but additional comparative prospective studies are needed to confirm this, said Dr. Eisen of the University of California, Davis.
Among the notable findings, CXR for staging yielded 14 true positive results in 4,320 patients (0.3%) in the reviewed studies. None of these was mentioned as amenable to surgical intervention. In addition, 326 false positives were reported in 3,447 (9.4%) patients with relevant data.
"So, [with CXR for staging] you have about a 20-fold chance of causing your patient undue psychological or iatrogenic harm," Dr. Eisen said.
CXR for surveillance yielded 198 true positive results in 10,046 patients (2%), and false positive results in 384 of 4,055 patients (9.5%).
Based on findings from studies that reported true positives as well as the number of patients who were eligible for or who underwent surgery as a result, 47 of 4,849 patients (1%) may have benefited from CXR surveillance, Dr. Eisen said.
The literature included fewer studies for CT than for CXR, likely because of the increased expense for CT, he noted.
CT was slightly more sensitive for staging, but false positives still outweighed true positives by at least threefold for staging; among 1,111 patients, there were 112 (10%) false positives vs. 41 (3.6%) true positives. CT results altered treatment in 1 of 338 patients (0.3%).
CT for routine surveillance yielded true positives in 63 of 467 patients (13.5%), and false positives in 13 of 127 patients (10.2%). There were no data on the number of patients eligible for surgical intervention based on CT surveillance, Dr. Eisen said.
Even fewer studies were available for MRI, which was almost exclusively used for cerebral imaging. MRI yielded true positives in 13 of 285 patients (4.5%), and false positives in 1 of 185 patients (0.5%). MRI for routine surveillance was addressed in only one study, which showed a true positive rate in 1 of 43 patients (2.3%).
PET for staging yielded true positives in 40 of 821 patients (4.9%), and false positives in 87 of 821 patients (10.6%). PET failed to detect disease in 159 of the patients, for a false negative rate of 19.3%, and changed management decisions in 21 of 191 patients (11%). Only 2 of these patients (1%) were eligible for surgery, however.
PET for post-treatment surveillance yielded true positives in 10 of 252 patients (4%), false positives in 16 of 218 patients (7.3%), and false negatives in 30 of 204 patients (14.7%).
Based on one study, 3 of 30 patients (10%) were found to be eligible for surgery as a result of PET imaging. PET-CT was more sensitive, but results were similarly disappointing, Dr. Eisen said.
Findings were slightly more encouraging for lymph node ultrasonography. In 11 studies, ultrasound staging yielded true positive rates in 100 of 1,035 patients (10%), and false positive results in 73 of 1,035 patients (7%); 44 patients were spared sentinel lymph node biopsies as a result of ultrasound imaging, and instead, proceeded to complete lymph node dissection, Dr. Eisen said. However, ultrasound for staging yielded a false negative result in 120 of 688 patients (17.4%).
Based on 9 studies, ultrasound for surveillance yielded true positive results in the absence of clinical findings in 79 of 1,266 patients (6.2%), false positive results in 12 of 966 patients for whom data were reported (1.2%), and false negative results in 5 of 806 patients (0.62%).
The data are imperfect, but seem to suggest that use of these imaging technologies can do more harm than good, Dr. Eisen said.
While patients may specifically request imaging to alleviate concerns that their primary cutaneous melanoma has advanced or will advance, it is important to keep in mind that, based on these results, the risk of causing unnecessary stress and iatrogenic harm outweigh the potential benefits, he concluded.