Although tempting to image patients with microscopic disease detected on SLNB, the a priori likelihood of nodal involvement is low, and there is a very low yield from imaging in this setting, whether this be by CT or MRI [9] or using FDG PET/CT [10]. With all modalities, there is higher number of false-positive than true-positive results in this clinical setting.

There may be a role for staging FDG PET/CT in patients with stage I/II disease and high-risk features, such as T4 disease or ulceration. In a study of 52 patients with T4 disease, the melanoma-related treatment plan was altered in 11% of patients. The main benefit of PET in this setting is identification of macroscopic locoregional nodal disease enabling the patient to proceed to lymphadenectomy and obviating the need for sentinel node biopsy.

FDG PET/CT is the imaging modality of choice when there is clinical suspicion of locoregional or systemic metastases. Indications include:

Substantially higher sensitivity and specificity of PET (92 and 94%, respectively) has been demonstrated compared to CT (58 and 45%, respectively) for detection of distant metastatic disease [14] (Fig. 5.2). The patterns of spread of melanoma are unpredictable with metastases seen in virtually any organ including the skin, subcutaneous tissue, lung, liver, brain, bone, gastrointestinal tract (small bowel > large bowel > stomach), muscle, adrenal cardiac, adrenal, gallbladder, breast, pancreas, thyroid and kidneys (Fig. 5.3).

The superior accuracy of PET/CT has been shown to confer a high management impact ranging from 22 to 62% [15–17]. In a study of 134 patients, PET detected 55% additional lesions compared to conventional modalities conferring a 62% management change [18]. In a prospective study of 103 patients with PET performed in high-risk patients prior to planned surgery, addition of PET changed management in 35% of patients, with most of these being cancellation of surgery [19]. In a study of PET/CT in patients with stage III/IV melanoma referred after review by a multidisciplinary team, major clinical impact was seen in 41% [20]. High management impact was predominantly alternation of plan from surgery to systemic therapy after identification of more extensive disease than seen on conventional workup. Although using out-dated, stand-alone PET technology and SPECT rather than SPECT/CT, an early study from our centre indicated that PET is also substantially more sensitive than Ga-67 citrate SPECT, a molecular imaging approach that was previously used for staging advanced melanoma [21]. High management impact was also demonstrated.

Stand-alone PET has lower sensitivity for pulmonary metastases, but in the PET/CT era, these are generally detected on the CT component. In a study of 50 patients, no additional benefit was demonstrated with contrast-enhanced CT compared to PET with non-enhanced non-low-dose CT [22]. Dedicated CT of the thorax can detect smaller nodules than low-dose CT, but these are invariably too small to characterise and, in our experience, frequently represent false-positive findings. Even faint FDG uptake in lesions smaller than 8 mm is highly suggestive of malignancy, but lesions of 1–4 mm may generally require surveillance or comparison with prior imaging to differentiate between benign granulomas and small metastases. In our centre, we no longer routinely perform contrast-enhanced CT but use it selectively to further interrogate equivocal findings on PET/CT. Owing to high physiologic brain activity, FDG PET/CT has very limited utility in assessing cerebral metastases, which is better assessed with MRI or by other PET radiotracers discussed later.

Ocular/uveal melanoma. These have a distinctive propensity to metastasise to the liver. A proportion of these also have low metabolic activity [23] despite aggressive behaviour, and correlation with contrast enhanced CT or MRI of the liver is therefore advised.

Data has demonstrated that assessing metabolic response on PET/CT after 2–3 cycles of conventional chemotherapy (e.g. dacarbazine) with visual dichotomisation of response as a complete or partial metabolic response versus progressive metabolic disease [24, 25] is a powerful biomarker that has the potential to replace conventional imaging endpoints such as RECIST anatomic criteria.

Management paradigms in metastatic melanoma have rapidly changed with both targeted therapies against BRAF and MEK and immunotherapy including anti-programmed death 1 (anti-PD1) monoclonal antibodies (nivolumab and pembrolizumab) and anti-CTLA4 monoclonal antibody (ipilimumab) now the treatments of choice for metastatic disease. Given the availability of several systemic therapeutic options, accurate response assessment is pivotal to guide management. Early response assessment enables an appropriate change in management thereby avoiding the morbidity and cost associated with ineffective therapy.

FDG PET/CT has been shown to be a useful marker of early biologic response to BRAF inhibition with complete or substantial reduction in metabolic activity observed as early as 15 days after commencement of therapy [26] (see Fig. 5.4). The reduction in glucose utilisation early during BRAF therapy is likely a pharmacodynamic indicator of reduced signalling through the MAP kinase pathway and is a necessary but sometimes insufficient requirement for subsequent clinical response and survival benefit. Reactivation of glucose use by previously responding tumours is a biomarker of reactivation of signalling through this pathway [27].