Dynamic O-(2-[18F]fluoroethyl)-L-tyrosine PET imaging for the detection of checkpoint inhibitor-related pseudoprogression in melanoma brain metastases

Identifying patients with pseudoprogression, which is characterized by an initial increase of contrast-enhancing lesions that resolve or at least stabilize spontaneously on follow-up imaging without any treatment change, is critical for avoiding premature termination of potentially effective treatment. With the advent and success of checkpoint inhibitors such as ipilimumab, nivolumab, or pembrolizumab in particular, detecting pseudoprogression in patients with malignant melanoma has become a major challenge in clinical practice given a frequency as high as 7%–10% of cases.^1,2 Diagnosing progressive disease and excluding pseudoprogression in melanoma metastases using the immune-related Response Criteria (irRC)² require the initial increase of at least 25% in lesions load to be confirmed by follow-up imaging at least 4 weeks later.² However, particularly for brain metastases from malignant melanoma, follow-up imaging might not be feasible for patients with clinical deterioration at the time of initial increase of lesions load. These patients might not be able to wait 4 weeks for a follow-up investigation to decide on potentially required therapy adjustments. In this scenario, a noninvasive imaging technique for timely detection of pseudoprogression in this patient population would be very useful but has not been investigated to date.

PET using radiolabeled amino acids such as O-(2-[18F]fluoroethyl)-L-tyrosine (FET) allows imaging of amino acid transport in brain tumors and has shown promise in distinguishing pseudoprogression from true progressive tumor in glioblastoma. FET-PET has detected early and late pseudoprogression in glioblastoma patients with an accuracy of 96 % and 85%, respectively.^3,4

To assess whether FET-PET might be valuable for distinguishing pseudoprogression in patients with malignant melanoma brain metastases, we conducted a small retrospective pilot study. A total of 5 patients with brain metastases from histologically proven malignant melanoma were enrolled for FET-PET imaging while under immune checkpoint inhibitor treatment at the time of the initial increase of brain tumor burden by at least 25% as evidenced by contrast-enhanced MR imaging. Four patients were treated with ipilimumab and one patient with nivolumab at the time of PET investigation. Patients’ characteristics are given in supplementary Table 1. Pursuant to irRC, the final diagnosis was assessed built on follow-up MR imaging at least 4 weeks later. Three experienced observers analyzed each image independently (one board-certified in neuroradiology). One patient was classified as having pseudoprogression (Fig. 1A), while the remaining 4 patients presented signs of true tumor progression (Fig. 1B).

Fig. 1.

A. MRI and PET images of patient #5, who was diagnosed with pseudoprogression using immune-related response criteria (irRC). The index MRI shows a clear >25% increase in contrast-enhancing melanoma metastases (frontal and occipital). Only a low tracer uptake was observed on FET-PET. Accordingly on follow-up, MRI (obtained 8 weeks later) showed a clear decrease in tumor burden consistent with pseudoprogression. B. MRI and PET images of patient #3, who was diagnosed with true progression using irRC. The index MRI again showed a clear >25% increase in contrast-enhancing melanoma metastases (central and occipital). A very high tracer uptake was observed on FET-PET. A follow-up MRI was not feasible as the patient died 8 weeks later. C. FET-PET imaging parameters of all patients included in the present study. The kinetic pattern represents 3 patterns of time-activity curves: constantly increasing FET uptake without identifiable peak uptake (pattern I); FET uptake peaking at a midway point (> 20–40min) followed by a plateau or a small descent (pattern II); and FET uptake peaking early (≤ 20min) followed by a constant descent (pattern III). Patterns II and III are known to be associated with true progression in glioblastoma.³

Metabolic tumor activity of FET-PET was evaluated by calculating the mean and maximum tumor-to-brain ratio (TBR_mean, TBR_max, respectively) (Fig. 1C).⁵ In the 4 patients with true tumor progression, TBRmax values were considerably higher (median TBR_max, 5.4; range, 2.9–8.6) as compared with the patient diagnosed with pseudoprogression (TBR_max, 2.5). Furthermore, time-to-peak values (TTP; time in minutes from the beginning of the dynamic acquisition up to the maximum standardized uptake value of the lesion), which has been shown to be inversely correlated with pseudoprogression,⁴ was accordingly shortened in the patients with true progression (median TTP, 17min; range, 10–20min) as compared with the patient with pseudoprogression (TTP, 45min). Additionally, a time-activity curve (TAC) with a constantly increasing FET uptake was present in the patient with pseudoprogression, consistent with TAC in high-grade glioma patients with pseudoprogression.^3,4 MR and PET images of patients with true progression and pseudoprogression are presented in Fig. 1.

Our report uncovers a previously unrecognized applicability of FET-PET imaging for detecting checkpoint inhibitor-related pseudoprogression in patients with melanoma brain metastases at the early stages of increasing tumor burden. This might particularly facilitate the care of patients with rapidly deteriorating clinical courses. However, the very small sample size limits our study; therefore, a systematic analysis in a larger cohort is warranted.

Supplementary material

Supplementary material is available online at Neuro-Oncology (http://neuro-oncology.oxfordjournals.org/).

Conflict of Interest statement. none