18F-FDG and 68Ga-somatostatin analogs PET/CT in patients with Merkel cell carcinoma: a comparison study

Background Merkel cell carcinoma (MCC) is an aggressive neuroendocrine skin tumor. Currently, 18F-fluoro-deoxy-glucose (18F-FDG) PET/CT is the functional imaging modality of choice. Few data are available on the use of 68Ga-somatostatin analogs. The aim of our study was to evaluate and compare the diagnostic performance of 18F-FDG and 68Ga-somatostatin analog PET/CT in MCC patients. Results Fifteen patients (12 males, 3 females; median age 73 years; range 41–81 years) with histologically proven MCC (4 with unknown primary lesion) who underwent both 18F-FDG and 68Ga-somatostatin analog PET/CT for staging, re-staging, or treatment response assessment were retrospectively evaluated. Results of both studies were qualitatively analyzed and compared on a patient- and lesion-based analysis, using histology or clinical/radiological follow-up as reference standard for final diagnosis. According to final diagnosis, 8/15 patients had at least one MCC lesion and 7/15 had no evidence of disease. On a patient-based analysis, 18F-FDG and 68Ga-somatostatin analogs correctly classified as positive 8/8 (100% sensitivity) patients and as negative 6/7 (85.7% specificity) and 5/7 (71.4% specificity) patients, respectively, with no significant difference. On a lesion-based analysis, 18F-FDG detected 67/75 lesions (89%) and 68Ga-somatostatin analogs 69/75 (92%), with no significant difference. In four patients with unknown primary MCC, both tracers failed to identify the primary MCC site. Conclusions Our preliminary data suggest that 18F-FDG and 68Ga-somatostatin analog PET/CT provide good and equivalent diagnostic performance, adding interesting insights into the complex MCC biology. However, these results do not suggest that 18F-FDG PET/CT should be replaced by 68Ga-somatostatin receptor imaging, which should be performed in addition, according to clinical indication, to the perspective of “personalized medicine.”


Background
Merkel cell carcinoma (MCC) is a rare highly aggressive skin tumor, with a quickly increasing incidence rate over the last 20 years. It is characterized by neuroendocrine features such as somatostatin receptor expression, besides a frequently high mitotic index, tumor necrosis, and vascular invasion [1][2][3]. MCC shows rapid local growth, high incidence of both regional lymph node and distant metastases, high rate of local recurrence, and a poor prognosis [1][2][3][4]. The primary tumor site cannot be found in up to 25% of patients (MCC of unknown primary origin; UPMCC) [5].
Currently, the optimal imaging algorithm of MCC is not yet defined [6,7]. 18 F-fluoro-deoxy-glucose ( 18 F-FDG) PET is a valuable functional modality to image MCC, due to the increased glucose metabolism which reflects its clinical aggressiveness [8][9][10][11][12][13][14][15]. According to the European consensus-based interdisciplinary guideline, 18 F-FDG PET/ CT or CT scan should be performed as part of the initial work up of MCC to stage the disease [16], while according to the more recent National Comprehensive Cancer Network (NCCN) guidelines, 18 F-FDG PET/CT may be preferred to CT [17]. Moreover, 18 F-FDG PET/CT is suggested as a routine imaging study during follow-up, as clinically indicated [16,17]. Due to the neuroendocrine aspects of MCC cells, somatostatin receptor scintigraphy (SRS) with 111 In-pentetreotide (Octreoscan®) has been initially used with variable results [18][19][20]. Recent evidences regarding the use of somatostatin analogs labeled with positron emitters ( 68 Ga-peptides) for PET imaging in MCC have shown positive results, suggesting a possible impact on MCC management [21][22][23][24]. A direct comparison between 18 F-FDG and 68 Ga-somatostatin receptor imaging ( 68 Ga-SRI) PET/CT in the same MCC patient has been only reported in a few case reports [25][26][27]. In this regard, potential interesting insights into tumor biology could derive from a head-to-head comparison of both tracers, besides possible implications also in the perspective of a personalized patient management.
The aim of our study was to evaluate and compare the diagnostic performance of 18 F-FDG and 68 Ga-SRI PET/ CT in patients with MCC.

Patients
All consecutive patients with a previous histological diagnosis of MCC who underwent both 18 F-FDG PET/CT and 68 Ga-SRI PET/CT at the PET/CT center of "Fondazione Policlinico Universitario A. Gemelli IRCCS" in Roma (center A) or "Arcispedale Santa Maria Nuova-IRCCS" in Reggio Emilia (center B) between September 2007 and May 2014 were considered. Inclusion criteria for this retrospective study were as follows: availability of 18 F-FDG and   68 Ga-SRI PET/CT scans performed within a 6-week interval (considering the rapid progression of MCC), no interval therapeutic procedures performed between the two scans, and availability of histopathological proof of imaging findings or, alternatively, clinical-radiological data on follow-up as reference standard for the final diagnosis. Sixteen consecutive patients were considered; one patient was excluded because he underwent surgical excision of the primary MCC lesion in the interval period between the two PET/CT scans. Therefore, 15 patients (12 males and 3 females; median age 73 years, range 41-81 years) were finally analyzed: 8 from center A and 7 from center B. This cohort of patients was also included in a previous bi-center study evaluating the role of 68 Ga-SRI PET/CT in 23 MCC patients [21]. According to standard clinical protocols in use in both institutions, this retrospective study was approved by the local ethics committees and an informed consent for both PET/CT examinations and the use of personal data was obtained from all patients. 18 F-FDG PET/CT 18 F-FDG PET/CT studies were performed according to a previously described protocol [28]. Briefly, all patients presented blood glucose levels under 8.3 mmol/L at the time of tracer injection, and they were in optimal hydration state. PET/CT images were acquired 60 ± 10 min after intravenous injection of a mean of 293 ± 72 MBq of 18 F-FDG according to body weight.

68
Ga-SRI PET/CT Patients were imaged with different radiolabeled somatostatin analogs: 68 Ga-DOTANOC was used in center A (eight patients), whereas 68 Ga-DOTATOC or 68 Ga-DO-TATATE were used in center B (five and two patients, respectively). In center A, 68 Ga was obtained from a 68 Ge/ 68 Ga generator (IGG 100; Eckert & Ziegler Isotope Products, Berlin, Germany), with a nominal activity of 1.85 GBq. DOTANOC was labeled with 68 Ga as previously described [29]. In center B, 68 Ga was obtained from a commercially available 68 Ge/ 68 Ga generator (Ciclotron, Napa, CA, USA) with a nominal activity of 1.85 GBq. DOTATOC and DOTATATE were labeled with 68 Ga according to a previously described protocol [30]. The administered activity of 68 Ga-somatostatin analogs was 2-2.5 MBq/kg. Images were acquired 60 ± 10 min after tracer injection, as previously described [21].

Images interpretation and data analysis
All 18 F-FDG and 68 Ga-SRI PET/CT images were independently evaluated by two experienced nuclear medicine physicians (ST and MS), aware of the clinical and follow-up reports. Any disagreement was resolved by consensus. A qualitative image analysis was performed: all PET/CT studies were classified as positive if at least one abnormal area of focal tracer uptake outside the physiological distribution or higher than the surrounding physiological activity was detected [31,32]. The number sites and maximum standardized uptake value (SUVmax) of all abnormal findings identified on both PET/CT scans were also recorded for each patient. The histopathological proof (as available) or, alternatively, a combination of morphological and functional studies and clinical information during follow-up were used as reference standard to verify PET/CT results. Accordingly, functional imaging findings were classified as true positive, false positive, true negative, or false negative for MCC-related lesions. The results of the two PET/CT studies were compared on a patient-based and on a lesion-based analysis.
A clinical decision was made considering all imaging modality results. The impact on patient's management was assessed evaluating whether additional information provided by 18 F-FDG PET/CT and/or 68 Ga-SRI PET/CT had influenced the clinical decision in comparison to the expected patient management based on the results of the available pre-PET morphological imaging.

Statistical analysis
A descriptive analysis was performed. For patient-based analysis, the diagnostic performances of 18 F-FDG and 68 Ga-SRI PET/CT in detecting MCC lesions were calculated in terms of sensitivity and specificity, and compared by the chi-square test with Yates' correction. Results were reported with 95% confidence intervals (CI). For lesion-based analysis, the Wilcoxon signed rank test was used to compare differences in the number of true positive lesions detected by 18 F-FDG and 68 Ga-somatostatin analogs. A p value < 0.05 was considered significant. Data were analyzed by MedCalc Statistical Software version 15.11.4.

Results
Detailed demographic and clinical features of the study population are listed in Table 1. Histological diagnosis of MCC was obtained by surgical removal of the primary lesion in 11/15 patients and by excisional biopsy of lymph node lesions in the remaining 4/15 patients with no identifiable primary tumor site (UPMCC). Clinical indications for PET/CT examination were staging at initial diagnosis (4/15), post-surgical staging (3/15), re-staging for suspected recurrence at imaging (3/15) or for clinical progression (2/15), and post-chemotherapy evaluation (3/15). In this comparative study, 18 F-FDG PET/CT was the first imaging of choice for evaluating MCC, according to current international guidelines [17]. PET/CT results were confirmed by histology in one case (lymph node surgical excision) and by a combination of imaging data (diagnostic CT, MRI, ultrasound imaging, chest X-ray, further PET/CT scan) and clinical information during follow-up in the remaining 14 patients. The median follow-up time was 34 months (range 7-122 months). At the end of follow-up, 8/15 patients were dead (MCC-related death in 7 cases) and the remaining 7 patients were alive.

Patient-based analysis
Eight out of 15 patients were classified as having loco-regional or distant MCC lesions according to reference measures (histology and/or clinical-radiological follow-up), whereas in the remaining 7/15 cases no evidence of disease was observed (as reported in Table 2). Both studies identified at least one abnormal area of tracer uptake in all eight patients with MCC-related lesions, resulting in 100% sensitivity (95% CI 67.5-100%). Regarding the seven patients with no evidence of disease, six were correctly classified as true negative by 18 F-FDG PET/CT (85.7% specificity; 95% CI 48.7-97.4%) and five by 68 Ga-SRI (71.4% specificity; 95% CI 35.9-91.8%), with no significant difference in specificity (p = 0.59). Both 18 F-FDG and 68 Ga-somatostatin analogs ( 68 Ga-DOTATATE) misclassified the same patient (#9) as false positive, showing abnormal uptake at the site of primary MCC removal, then attributed to cutaneous post-surgical inflammation. Discordant results between 18 F-FDG (true negative) and 68 Ga-SRI (false positive) PET/ CT were observed in patient #10 (focal 68 Ga-DOTATOC liver activity, not confirmed by the subsequent follow-up examinations). Regarding the four patients diagnosed as UPMCC, both PET/CT scans correctly classified one patient as negative (#13) and the other three cases as positives (#4, #8, and #14), but neither 18 F-FDG nor 68 Ga-SRI PET/ CT were able to identify potential primary MCC sites.

Patient management
Additional information provided by 18 F-FDG and/or 68 Ga-SRI PET/CT influenced clinical decisions and led to a change in patient's management in 4/15 patients (27%), as detailed in Table 2. In one patient (patient #1), the clinical strategy was influenced by specific findings identified by 18 F-FDG alone (liver metastases which lead to an upstaging to stage IV), in another one (patient #8) by 68 Ga-somatostatin analogs (due to the evidence of SSTR expression, he could be addressed to a palliative/ second-line treatment with "cold" somatostatin analogs), while in the remaining two cases (patients #3 and #6) both tracers had the same clinical impact (not confirming radiologically suspected metastases).

Discussion
To the best of our knowledge, this is the first original study directly comparing the diagnostic performance of 18 F-FDG and 68 Ga-SRI PET/CT in a series of patients with MCC. We found that 18 F-FDG and 68 Ga-somatostatin analogs provide good and equivalent diagnostic performance, with no significant differences either on a patient-based analysis or on a lesion-based analysis. From literature data, only few MCC patients have been imaged using both PET tracers, with better or equivalent performance of 68 Ga-SRI compared to 18 F-FDG PET/CT [25][26][27]; however, they are all case reports and, therefore, not reliably comparable with our findings. Similarly, our findings are not comparable with those reported by Lu et al. [33] in the largest series of nine MCC patients who were investigated comparing 18 F-FDG PET/CT to 111 In-pentetreotide SRS. In this series, 18 F-FDG PET/CT performed better than SRS, but as suggested by the authors, this finding could be more related to the better spatial resolution and image quality of PET than to a real advantage of a metabolic tracer (glucose analog) over a receptor tracer (radiolabeled somatostatin analog). This hypothesis seems to be confirmed by our results. Indeed, in our series of patients investigated by PET/CT, 18 F-FDG and 68 Ga-somatostatin analogs provided comparable results in terms of sensitivity (100%) and specificity (86 and 71%, respectively). Also considering the overall lesion detection rate, no significant differences between the two tracers were found (89% for 18 F-FDG vs 92% for 68 Ga-somatostatin analogs), although 68 Ga-SRI PET/CT failed to detect two liver lesions that were, on the contrary, detected by 18 F-FDG leading to a change of both patient's staging and therapy planning (patient #1). A possible explanation of this failure could be linked to the heterogeneous somatostatin receptor (SSTR) expression among MCC lesions, as observed by Gardair et al. [34]. These authors analyzed 105 tissue samples from 98 MCC patients by immunohistochemistry and demonstrated  different degrees of expression and distribution of SSTR subtypes. Therefore, in our case (patient #1), we cannot exclude a heterogeneous SSTR expression in liver metastases (SSTR-PET negative) compared to the primary MCC lesion (SSTR-PET positive), although immunohistochemistry to confirm our hypothesis was not available.
Notably, some organs may be not well explored with 68 Ga-SRI since tracer is physiologically taken up by their cells. Malignant lesions have different patterns and intensities of 68 Ga-somatostatin analog uptake compared to non-malignant tissues, also in sites characterized by high physiological background (e.g., the uncinated process of the pancreas and adrenals), as described by Kroiss et al. [35] in one of the largest cohort of NET patients imaged with 68 Ga-DOTATOC. Accordingly, in our series, also adrenal metastases (patient #14) were easily identified at 68 Ga-SRI PET/CT since the pattern of uptake was really inhomogeneous, its intensity was higher than liver, and the glands were enlarged (Fig. 2). In any case, the physiological tracer distribution should be always taken into account, with awareness that some metastases, especially small ones, could remain undiagnosed.
When comparing SUVmax values of the two tracers, we observed that the whole 18 F-FDG uptake was higher than 68 Ga-somatostatin analogs, suggesting that in our MCC population the tumor aggressiveness-as reflected by glucose avidity-seems to be a predominant characteristic. In other neuroendocrine tumors, particularly those of the gastro-entero-pancreatic tract, a positive 68 Ga-SRI (representing the expression of SSTRs) with a faint or absent 18 F-FDG uptake is typically observed in well-differentiated tumors with good prognosis, and a positive 18 F-FDG with a faint or absent 68 Ga-SRI uptake (representing the loss of SSTRs) is typically observed in more aggressive undifferentiated tumors [36,37]. From our findings, MCC seems to deviate from this general "rule," and shows a similar uptake pattern using both PET tracers. In support of these observations, Gardair et al. [34] argued that the expression of at least one SSTR in the majority of MCC lesions was an unexpected finding, as MCCs are poorly differentiated, highly proliferative tumors. Moreover, SSTR expression was not associated with clinical characteristics, Ki67 proliferative index, or clinical outcome, suggesting that the expression of SSTR is not itself a favorable prognostic index in MCC, differently from other neuroendocrine tumors. To explain this peculiar functional PET behavior of MCC, we may refer to one of the postulated theories on the controversial origin of MCC cell: a totipotent stem cell that acquires neuroendocrine characteristics (as SSTRs expression) during malignant transformation [38,39]. The pattern of uptake was inhomogeneous with both 18 F-FDG and 68 Ga-somatostatin analogs and, in the whole, a "reverse" and complementary distribution of uptake was evident with the two tracers, with clearly parts of the tumor that take up one or the other tracer (e.g., the "hottest" part of the right adrenal at 18 F-FDG appears substantially "cold" at 68 Ga-SRI) In our series, UPMCC occurred in 4/15 patients (26%), localizing at the iliac-inguinal region in all cases. Our results are in accordance with the estimated incidence of UPMCC (as high as 25% of all MCC) and with the clinical site of presentation, being inguinal lymph node area one of the most common involved site at diagnosis [5,40]. As previously mentioned, also in our series, both tracers could confirm lymph node involvement or detect unknown metastatic localizations, but were unable to identify the primary tumor [12]. Actually, in UPMCC, it has not yet been delineated if the tumor arises de novo from neural cells located within the involved lymph nodes or if the primary lesion undergoes spontaneous regression [5,40].
We are aware that the retrospective nature of our study and the variable treatment protocols among different clinical institutions may lead to non-univocal results in terms of impact on patient management. However, we observed that additional information provided by 18 F-FDG and/or 68 Ga-SRI PET/CT influenced clinical decisions in 4/15 patients (27%). In particular, in one case (patient #1) 18 F-FDG lead to a change of tumor staging (from III to IV) detecting two additional foci of increased uptake in the liver, not evident neither at 68 Ga-DOTANOC nor at previous diagnostic CT and subsequently confirmed as metastatic lesions. With respect to this finding, although no statistically significant difference between the diagnostic performance of the two tracers was found, we retain that the overall performance of 18 F-FDG and 68 Ga-SRI PET/CT cannot be considered completely equal from a clinical perspective. Therefore, as shown by our data and reported in the NCCN guidelines [17], we retain that, in clinical practice, 18 F-FDG PET/CT should not be replaced by 68 Ga-SRI, which should be performed in addition, if clinically indicated. Anyway, it is important to consider that the decision on which tracer to employ at first line should be strongly influenced by the information that is considered more relevant for clinical management, in the perspective of "personalized medicine." In this regard, 18 F-FDG PET/CT may also add prognostic information, as recently suggested [11,41], while 68 Ga-SRI PET/CT may be useful for both therapy planning and response assessment in a "theranostic strategy" when peptide receptor radionuclide therapy (PRRT) is considered as a further potential treatment option [24,25,27,35,42]. Moreover, with the advent of immunotherapy with immune checkpoint inhibitors, PET/CT imaging with the two different tracers may turn out to play a role in a tailored-treatment approach providing useful and complementary information also in the setting of response assessment and follow-up for advanced MCC [43,44].
Although our study represents to date the largest comparative series of MCC patients imaged by 18 F-FDG and 68 Ga-SRI PET/CT, it is affected by some limitations, mainly represented by its retrospective nature and the limited number of patients. However, the low incidence of MCC needs to be considered. Moreover, phenotypic tumor properties (i.e., proliferative index) and immunohistochemical characteristics were available only in few patients, preventing the correlation of tumor features and PET/CT results. Another potential limitation of this study concerns the use of three different 68 Ga-SRI with different receptor affinities and pharmacokinetics. However, although some studies comparing different 68 Ga-labeled peptides found the superiority of one radiopharmaceutical over the other, at present, the observed differences are not considered clinically relevant [45].