| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
1 Department of Plastic Surgery, Indiana University School of Medicine, Indianapolis, Indiana
2 Wagner and Associates Plastic and Reconstructive Surgery Consultants, Suite 570, 8040 Clearvista Parkway, Indianapolis, Indiana 46256
Correspondence: Address correspondence and reprint requests to: Jeffrey D. Wagner, MD, Wagner and Associates Plastic and Reconstructive Surgery Consultants, Suite 570, 8040 Clearvista Parkway, Indiana-polis, IN 46256; E-mail: jdwagner{at}insightbb.com.
Melanoma patients with advanced disease are often considered for major therapeutic surgical interventions. Accurate identification of all sites of metastatic disease is critical, because only patients rendered completely free of disease are likely to experience disease-free intervals or a potential survival benefit from surgery.
Metabolic imaging with fluorine 18–labeled deoxyglucose (FDG)-positron emission tomography (PET) has recently become more widely applied in a variety of human malignancies. Several reports show FDG-PET to be a more sensitive indicator of metastatic melanoma than conventional diagnostic imaging modalities,1–4 particularly in patients with recurrent disease. However, various design shortcomings in early reports may have overestimated the utility of PET.5 Recent studies focusing on specific clinical scenarios are providing information to develop guidelines for the effective use of PET in patients with melanoma.
In this issue of Annals of Surgical Oncology, Brady et al.6 provide a thoughtful report of their investigations of the clinical utility of preoperative PET imaging in patients planned for surgery for high-risk or recurrent/metastatic melanoma. Their work provides further evidence that PET has a role in management of the complex advanced melanoma patient. Although this study showed FDG-PET to be more sensitive than conventional imaging for detection of metastatic disease, the combination of PET and conventional imaging had a greater effect on clinical decision making than either modality alone. The authors showed that the addition of PET to computed tomographic (CT) imaging affected clinical decision making in one third of patients. Collectively, this and other studies7–10 show that properly selected patients with metastatic/recurrent melanoma may benefit from adjunctive PET imaging.
Should FDG-PET imaging be obtained in all melanoma patients? The collective literature clearly does not support this practice. Prospective studies of preoperative FDG-PET imaging in stage I and II melanoma patients do not show a significant clinical effect.11,12 FDG-PET imaging is very unlikely to upstage early-stage melanoma patients already staged by conventional staging protocols that include sentinel node biopsy. Our group found that PET was only 21% sensitive for identification of occult regional lymph node metastases compared with sentinel node biopsy. Preoperative FDG-PET was not useful in predicting the histology of the residual lymph node basin after a positive sentinel node. Most importantly, FDG-PET rarely predicts the presence of occult distant metastases in this population. Increased FDG uptake at a site of possible distant metastatic disease was 10 times more likely to be false positive than true positive.
Despite increasing use of FDG-PET in melanoma, it is important to point out that FDG-PET is not generally considered to be the single most sensitive diagnostic test for metastatic disease in most organ systems. For example, brain metastases are best sought by using magnetic resonance imaging. Lung and liver metastases are better imaged with spiral/ infusion CT scans. Small lymph nodes and soft tissue foci are probably best identified by ultrasonography. Technetium bone scans and magnetic resonance imaging are the most sensitive tests for bone metastases. The lower limit of sensitivity for FDG-PET detection with modern PET equipment is approximately 6 to 8 mm, which is below that of the best conventional imaging modalities. Therefore, a negative whole-body FDG-PET scan alone is not the most definitive indicator of disease-free status. This is particularly true after systemic therapy, where transiently decreased tumor metabolism may yield a false-negative result.
Nonspecific FDG uptake is also frequently seen with PET imaging: it occurs in many inflammatory lesions, infections, reactive lymphadenopathy, recent surgical wounds, exercised skeletal muscle, and benign tumors in a wide variety of tissues. The high background utilization of glucose can limit the utility of FDG-PET in the gastrointestinal and genitourinary tracts, in the central nervous system, and in the mediastinum around the heart. Nonetheless, some of these false-positive (nonmelanoma) findings may be clinically important. The importance of histological verification of PET abnormalities merits emphasis, and conventional imaging is usually necessary for this purpose. When suspicious or equivocal PET findings cannot be confirmed to represent metastatic disease, patients should be considered potentially curable, and therapy should not be delayed or withheld.
It is worth noting that studies of this nature are very difficult to design and perform. Sensitivity and specificity of PET for detection of distant metastases are elusive values that must be derived from comparison to clinical follow-up and histology. Follow-up and biopsies rely on conventional imaging, known to be relatively unsatisfactory in detection of small foci of disease. The apparent clinical utility of PET can be heavily influenced by other clinical information, the experience of the PET team, the threshold used by clinicians for assigning a PET scan finding as positive or negative, and the duration of follow-up. Equivocal PET findings may go unconfirmed or dismissed unless repeat imaging and adequate follow-up are designed into the study. If clinical follow-up is inadequate, the true-negative rate will be overestimated and artificially inflate diagnostic performance. Extensive diagnostic evaluations before referral can also introduce bias that could inflate the utility of PET. Another inherent difficulty with studies of this type is the somewhat subjective nature of PET interpretations. This study was conducted by in a tertiary referral setting by investigators highly experienced in melanoma care and PET imaging. A high level of investigator expertise is more likely to result in correct decisions, possibly because of many factors—not just the additional information provided by PET. These results may be difficult to reproduce in less-experienced settings.
Perhaps the most important contribution of this work is the clear demonstration that PET and conventional imaging are complimentary—not competitive—modalities. A compelling case can be made for the use of both PET and CT in selected patients with advanced melanoma before surgery for metastatic disease. The weaknesses of metabolic imaging are improved by the strengths of anatomically based imaging, and vice versa. Surgeons reluctant to base clinical decisions on equivocal CT or equivocal PET findings alone are likely to find the complimentary information provided by a combination of modalities useful and reassuring. The collective literature illustrates that the best application of aggressive PET/CT staging before therapy is for patients with stage III and IV disease.
We now have the groundwork for further prospective studies of new technologies, such as PET-CT. Clinical decisions may be altered by imaging information, but these decisions may not necessarily be correct or produce better clinical outcomes in the long run. The next generation of studies should move beyond measures of sensitivity and specificity, which are not always accurate measures of effect, to clinical outcome-focused measures of cost-benefit and cost-effectiveness. Data of this nature are necessary to determine whether downstream clinical decisions based on diagnostic imaging studies result in improved care and better outcomes.
Received for publication December 2, 2005. Accepted for publication December 20, 2005.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
