In this retrospective study, we performed a liver lesion-based comparison of histopathologic reports and SSR-PET/CT ([68Ga]Ga-DOTATOC and [68Ga]Ga-DOTA-TATE) findings in patients with diagnosed NET. Our results demonstrate high diagnostic accuracy of SSR-PET/CT in the identification of liver metastases of NET patients with an excellent PPV of 91.0% and an even higher PPV of 95.5% when compared to re-biopsy of initially negative histopathological results.

Several previous studies have demonstrated the high diagnostic accuracy of SSR-PET/CT in the evaluation of NETs. In a large study of 728 NET patients, [68Ga]Ga-DOTA-TATE PET/CT showed high sensitivity, specificity, accuracy, positive predictive value, and negative predictive value (97.0%, 95.1%, 96.6%, 98.5%, and 90.4%) in the detection of overall NET lesions [11]. Specifically identifying hepatic involvement in NET patients, similar high diagnostic accuracy was reported for SSR-PET/CT compared to the reference standard liver MRI with a sensitivity of 97.0% and a specificity of 97.7%, a PPV of 99.0% and NPV of 93.0% [12]. Moreover, SSR-PET/CT was found to have an important value in the diagnosis of pulmonary neuroendocrine tumors (carcinoids) and its distant metastases [13, 14]. However, one patient in our cohort with a carcinoid of the lung and a Ki-67 of 15% (G2) showed SSR-negative liver metastases. Haug et al. also reported on low tracer of liver metastases in a patient with atypical carcinoid tumor of intermediate grade [15].

In 12 (8.6%) SSR-positive lesions, no tumor tissue was detected by initial histopathology in our cohort, whereas tumor cells of NETs were identified after re-biopsy in most cases (6/8 cases). Thus, the underlying reasons were most likely inadequate sampling in the sense of non-target biopsies. This finding emphasizes that the absence of detectable metastatic lesions on liver biopsy does not necessarily rule out the presence of liver metastases of NET patients. Tumor heterogeneity of NETs is well known, and low Ki-67 may be challenging to be detected on histological examinations. Moreover, not all liver metastases are easily accessible for biopsies, for example if located close to large vessels. In consideration of these possible false-negative biopsy results, histologic verification of PET-positive liver lesions appears redundant in cases of patients with histologically proven NETs, and functional imaging with SSR-PET/CT might reduce unnecessary further biopsies. However, one must be aware that biopsies are nowadays not only used for pure histological confirmation of metastases but are also required for additional molecular diagnostics in precision medicine [16].

In 6 (4.3%) cases, the index lesion was not labeled as NET metastasis in SSR-PET/CT, however, subsequent dynamic, contrast-enhanced MRI of the liver unequivocally confirmed these lesions as metastatic NETs. This finding emphasizes the ongoing essential role of liver MRI in specific cases. Notably, all 6 of these lesions were categorized as G2 in the corresponding histopathological reports, with Ki-67 values ranging between 2% and 15%. The observed heterogeneity of SSR-expression patterns in NETs is likely attributed to the inherent tumor/metastatic variability, resulting in diverse SSR-expression profiles even within the same patient across different lesions [17, 18]. Furthermore, it is well known that NETs of low differentiation lack SSR-expression and SSR-PET/CT has low sensitivity but can be localized by [18F]FDG PET/CT [2, 19, 20]. Liver metastases from NETs are often diagnosed by CT- or ultrasound-guided biopsy, in which tumor tissue is obtained randomly from a specific lesion, depending on accessibility, and the Ki-67 is determined without targeting regions that may have a higher proliferation rate [21]. In our analysis, in 2 patients with NET G2 (Ki-67 2% and Ki-67 10%, respectively) additional [18F]FDG PET/CT was performed and showed high [18F]FDG-uptake indicating intertumoral heterogeneity with tumor cells of lower differentiation. For these cases, a proposed grading scheme which describes the joint results of both, the [18F]FDG- and SSR-PET scans in dual tracer imaging (NETPET), showed promising results as a prognostic biomarker and correlated significantly with overall survival [1, 22]. Therefore, [18F]FDG PET/CT may have a clinical role in intermediate grade neuroendocrine neoplasms based on clinical presentation (e.g., patients with CT progression or with SSR-negative lesions). Despite above mentioned limitations in a small number of patients, SSR-PET/CT represents a whole-body, one-stop-shop diagnostic tool for NETs.

There are various limitations to this study. Firstly, this was a retrospectively conducted single-center study. The use of SSR-analogues ([68Ga]Ga-DOTATOC and [68Ga]Ga-DOTA-TATE) varied across patients and examinations. Nevertheless, previous studies have consistently reported no differences in diagnostic accuracy when evaluating sensitivities and uptake values among different SSR-ligands, including considerations of tumor origin and grading [23, 24]. Additionally, the effective doses of both radiopharmaceuticals are comparable, rendering them equivalent even from a radiation dosimetry perspective [25]. Furthermore, not all patients with SSR-negative NET lesions received additional [18F]FDG PET/CT for further diagnostic evaluation. The lack of [18F]FDG PET/CT data for some SSR-negative patients could potentially influence the overall diagnostic accuracy and completeness of our findings. Although we have not directly compared the diagnostic performance of SSR imaging and [18F]FDG PET/CT in this study, acknowledging this limitation is essential for a comprehensive understanding of the scope and implications of our results. Future studies that incorporate both SSR imaging and [18F]FDG PET/CT data for all patient groups could offer a more comprehensive assessment of the diagnostic capabilities of these imaging modalities in the context of SSR-negative NET lesions.