Turning Up the Heat: Endoscopic Ablation of Pancreatic Neuroendocrine Neoplasms Free

The expanded use of cross-sectional imaging and improved radiological and histopathological identification has increased the detection rates of pancreatic neuroendocrine neoplasms (PanNENs) (1, 2). According to autopsy series, PanNENs could potentially be diagnosed in 0.1% to 10% of the population (3). A minority of PanNENs will be functional, and these patients will present with a hormonal syndrome, such as insulinoma, gastrinoma, or glucagonoma (4). Fifty percent of patients with the multiple endocrine neoplasia type 1 (MEN-1) syndrome harbor multiple PanNENs for which the optimal management strategy is unclear (5). Although small, low-grade, sporadic, nonfunctional PanNENs can be safely monitored by active surveillance, the reference standard for treatment of localized PanNENs is radical surgical resection (6). However, pancreatic surgery has been accompanied by substantial morbidity and even mortality, especially for older patients with preexisting comorbidities. Depending on the surgical technique used, the postoperative complication rates have ranged from 14% to 58%, 5% to 18%, 1% to 7%, and 3% to 6% for pancreatic fistula, delayed gastric emptying, hemorrhage, and in-hospital mortality, respectively (7).

This caveat has prompted the search for alternative, safer, less invasive, therapeutic modalities. In a recent issue of Journal of Clinical Endocrinology & Metabolism, Oleinikov et al. (8) report their experience in two centers with endoscopic ultrasound-guided radiofrequency ablation (EUS-RFA) of PanNENs. RFA was first developed for endoscopic delivery in 1999 (9). Compared with percutaneous techniques, EUS-FNA has the advantage of real-time visualization of the pancreatic lesion and its surrounding structures from close range. Recent improvements in EUS-RFA systems have enhanced its safety profile and demonstrated encouraging tumor control rates for several types of pancreatic lesions (10).

In their study, Oleinikov et al. (8) presented a retrospective series of EUS-FNA for 18 patients with 27 PanNENs. Their report constitutes a substantial contribution to the reported data for this patient subset, because previous outcome data were confined to only 35 tumors. The indications for therapy in their series were insulinoma in seven patients, tumor characteristics in six nonfunctional cases, and patient preference above active surveillance in 5 patients. The tumors were, on average, 14 mm (range, 4.5 to 29 mm), distributed throughout the pancreas, MEN-1 associated in three patients, multifocal in six patients, metastasized to lymph nodes or liver in two patients, and predominantly grade 1, except for two cases with grade three histologic features. The choice for EUS-RFA instead of surgery was by patient preference or refusal of surgery for 83% of cases. The remaining three patients were considered to have an unacceptably high surgical risk.

The responses to EUS-RFA were favorable in the series by Oleinikov et al. (8). All seven patients with insulinoma and nine lesions in total developed rapid normalization of blood glucose levels, without clinical recurrence at 3 to 21 months of follow-up. A complete radiological response was obtained in 26 of all 27 PanNENs (96%), according to periprocedural visualization of a hyperechoic area at the tumor site. One tumor could only be partially ablated owing to its proximity to the pancreatic duct. Twelve of the 18 patients had undergone CT imaging the next day, which showed necrosis. However, long-term cross-sectional imaging was only reported for 1 patient at 6 months. The safety profile of EUS-RFA for PanNENs in this cohort was encouraging, with two cases of mild pancreatitis in 18 procedures. The average hospital stay of 3 days (range, 2 to 6) suggested a more rapid clinical recovery than with surgical intervention.

Their report has opened up new avenues for minimally invasive treatment of well-differentiated, localized PanNENs, both functional and nonfunctional, sporadic and hereditary. As evidenced from this and other studies, EUS-RFA is capable of reaching tumors at all locations within the pancreas. This is of particular importance in the pancreatic head, where pancreaticoduodenectomy, with its high complication rates, is, at present, the surgical procedure of choice. Similar to an enucleation procedure, the proximity of the PanNEN to the main pancreatic duct warrants careful consideration in the patient’s evaluation for EUS-RFA. Oleinikov et al. (8) reported that tumors up to a diameter of 3 cm can be treated successfully (although follow-up time was short). They also reported that tumors adjacent to critical structures should not be treated. However, a safe distance was not defined and could probably have not been given because this still needs to be investigated systematically. The proximity of the PanNEN to the main pancreatic duct provides a similar problem; however, a safe distance between the tumor and the duct also requires study. Similarly, prophylactic stent insertion, which was performed in one of the patients in the present series, requires more evidence.

The incidence of PanNENs increases with age, leading to a greater percentage of patients presenting with comorbidities (11). Given the high complication rates of surgical resection, an increasing group of patients would benefit from EUS-RFA to treat PanNEN. The safety profile of EUS-RFA reported thus far appears favorable compared with that of surgery (12). However, a publication bias could not be excluded as a confounding factor, and sufficiently powered prospective series are needed before periprocedural safety can be firmly established. Also, RFA will prevent a complete histological evaluation of the tumor, possibly leading to misclassification of the diagnosis or grade using fine needle aspiration or biopsy in a few cases (13). Another concern is the long-term safety of ablation, because the current reported data have been restricted to follow-up durations on clinical grounds, on average, in the range of 1 year. More information is needed on the recurrence rates and consequences of RFA for the feasibility of future surgical resections.

Insulinomas will show benign biological behavior in 90% of cases and are clinically easy to recognize owing to the occurrence of insulin-induced hypoglycemia (14). Consequently, they constitute a tumor subset that is particularly suitable for minimally invasive interventions. Patients can be closely monitored for glucose levels and hypoglycemic symptoms as markers for response directly after the intervention and during long-term follow-up. Small gastrin- or ACTH-producing PanNENS are also potential tumor subtypes suitable for EUS-RFA. In contrast, other functional PanNENs often present in advanced stages of disease.

The optimal size cutoff for therapeutic intervention for nonfunctional PanNENs remains a matter of debate. The malignant potential increases incrementally at sizes >2 cm, providing a rationale for guidelines to advise resection of PanNENs >2 cm (15). Resection of small lesions with the accompanying risk of complications should be weighed against active surveillance, because the tumor growth rates for both sporadic and MEN-1–associated PanNENs will generally be indolent at 0.01 to 0.04 cm annually (16–18). Active surveillance of nonfunctional PanNENs <2 cm was shown to be a safe strategy in a systematic review of five retrospective series that included 540 patients (19). In their series of 11 nonfunctional PanNENs, Oleinikov et al. (8) performed EUS-RFA in five cases of grade 1 lesions with a diameter ranging from 7 to 14 mm because of patient preference. It is debatable whether the availability of a minimally invasive technique should stimulate its use for small nonfunctional PanNENs that might never necessitate any therapeutic intervention. Owing to the short follow-up of patients and lack of repeated imaging over time, the long-term consequences of EUS-RFA for PanNEN remain uncertain.

In conclusion, EUS-RFA represents a novel, minimally invasive therapy for small functional and nonfunctional PanNENs. It constitutes a possible alternative to surgical resection for highly selected patients with increased perioperative risk with nonfunctional and functional (especially insulinomas) PanNENs <3 cm. The experimental character of this treatment requires intensive discussion with the patient, including consideration of the availability of established alternative strategies such as follow-up without intervention and surgery, including minimally invasive strategies (20). At present, the lack of prospective studies with sufficient follow-up should prevent the judicious use of this technique for ablation of PanNENs. The outcome of the upcoming international, multicenter RAPNEN (endoscopic ultrasound-guided radiofrequency ablation for the treatment pancreatic neuroendocrine neoplasms) trial (ClinicalTrial.gov identifier, NCT03834701) on EUS-RFA in patients with PanNEN might provide further insights regarding the safety and efficacy of this promising interventional technique.

Acknowledgments

Clinical Trial Information: ClinicalTrials.gov no. NCT03834701 (registered 1 February 2019).

Additional Information

Disclosure Summary: The authors have nothing to disclose.

Abbreviations:

Abbreviations:
 
• EUS-RFA

  endoscopic ultrasound-guided radiofrequency ablation

 
• MEN-1

  multiple endocrine neoplasia type 1

 
• PanNEN

  pancreatic neuroendocrine neoplasm

References and Notes