Introduction
Neuroendocrine tumors (NET) affect the gastrointestinal system with a frequency of
3.65 per 100,000 population [1 ] but arise ubiquitously throughout the body. Insulinomas, one of the more common
types of gastrointestinal NET, are typically found in or adjacent to the pancreas
and are characterized by inappropriate endogenous secretion of insulin. These hormonally
active NETs often cause refractory hypoglycemia that leads to significant morbidity
[2 ]
[3 ], but they typically have low potential for metastasis [3 ].
Unfortunately, for many patients with NETs, medical therapies result in variable efficacy
and can result in significant side effects. If these lesions are localized, surgical
resection is effective or even curative. However, such treatment has morbidity and
mortality rates of 20 % to 40 % and 1 % to 5 %, respectively, and its own limitations
and surgery often is not an option in elderly patients or those with significant comorbidities.
This has led to increasing interest in minimally invasive treatment options of NETs,
and of these, EUS-guided ablation with alcohol may be a promising technique [4 ]
[5 ].
Previous reports of alcohol-based ablation have shown good efficacy but with a serious
adverse events (SAEs) rate of 6 % [4 ], including post-procedure pancreatic injury with pseudocysts, acute pancreatitis,
hematomas, and ulceration of the duodenal wall [6 ]
[7 ]. SAEs associated with EUS-guided ablation generally have been felt to be secondary
to the marked inflammatory and toxic effects of dehydrated alcohol, which can extravasate
into local tissues [5 ]
[8 ]
[9 ]. This has called into question its clinical application and safety profile. A recent
randomized prospective clinical trial conducted by this group demonstrated that alcohol
is not required for EUS-guided pancreatic cyst ablation when a chemotherapy infusion
cocktail specifically designed for pancreatic neoplasia is used, and that when alcohol
is removed from the ablation process, AE rates fall significantly [9 ]. By extrapolation of these data, we speculated that solid tumor ablation would not
only be comparatively effective, but also safer, if the same admixture used in the
CHARM trial was used in place of alcohol. Here we describe use of an alcohol-free
chemoablation of multifocal insulinomas in a patient who was not a surgical candidate.
Case report
In 2004, a 49-year-old-female presented with recurrent symptomatic hypoglycemia. Triphasic
enhanced computed tomography (CT) of the abdomen and pelvis showed two lesions consistent
with insulinomas in the distal tail of the pancreas. The woman underwent a distal
pancreatectomy and splenectomy with surgical pathology revealing a multifocal well-differentiated
neuroendocrine tumor, Grade 1 ([Fig. 1 ]). There were a total of four subcentimeter tumor nodules. Immunohistochemical staining
demonstrated insulin production by the neoplastic cells. The patient was subsequently
discharged with close follow up with Endocrinology for further medical management
with diazoxide and octreotide.
Fig. 1 Resection specimen of well-differentiated neuroendocrine tumor, Grade 1, characterized
by cords and trabeculae of mitotically-inactive cells with small, uniform nuclei having
typical “salt and pepper” chromatin and inconspicuous nucleoli. There were frequent
psammoma body type calcifications (arrow) (H&E × 400).
Unfortunately, over the last 3 years, the patient had a recurrence of hypoglycemic
episodes that were not responsive to medical therapy, and for which she was forced
to self-treat by frequently sipping soft drinks, resulting in 29-kg weight gain and
depression. She was hospitalized with refractory hypoglycemia in June of 2016 with
CT imaging showing only postsurgical changes to the pancreas without any visible masses
to suggest recurrent insulinoma. Endoscopic ultrasound-guided fine-needle aspiration
(EUS-FNA) was performed in July 2016 with findings of three subcentimeter lobular
masses (9 mm × 9 mm) in the head and body of the pancreas. Color Doppler imaging was
utilized to guide FNA of the largest mass lesion with cytology diagnostic for recurrent
neuroendocrine tumor ([Fig. 2 ]). The patient was discharged with increased titration of her octreotide in addition
to a cornstarch regimen. She could not continue the diazoxide due to side effects
of excessive fluid retention. Despite these medication changes, the woman continued
to have recurrent hypoglycemic episodes.
Fig. 2 Direct smear from EUS-FNA showing cells with identical cytomorphology, with small,
uniform nuclei, “salt and pepper” chromatin, and inconspicuous nucleoli, and with
associated psammoma body type calcifications (arrow) (Papanicolaou stain, × 400).
Because of the multifocal nature of the patient’s lesions and a body mass index (BMI)
of 57.81 kg/m2 , state of immobility, and comorbidities such as chronic kidney disease, the Surgical
Oncology team felt that a surgical approach posed an excessive mortality risk. The
patient was referred to Gastroenterology for consideration of EUS-guided chemo-ablation
using the alcohol-free chemoablation admixture previously described by this group
for ablation of mucinous-type pancreatic cysts. The patient was fully informed of
all her options as well as the risks and benefits of the procedure.
The procedure was performed in an operating room due to a high-risk airway and a BMI
of 57.81 kg/m2 . Under general anesthesia, a linear EUS video endoscope (GF-UCT180; Olympus America,
Center Valley, Pennsylvania, United States) was used to perform a full pancreaticobiliary
examination. The three known NETs ([Fig. 3 ]), as well as a fourth lesion that had developed in the interim period, were treated
by injecting 1 to 2 cc of the 3-mg paclitaxel and 19 mg of gemcitabine admixture into
the center of each mass using a 22-g FNA needle (Echotip; Cook Medical, Bloomington,
Indiana, United States). Repeat EUS 3 months later identified only two masses consistent
with insulinomas and two associated pathological lymph nodes, which were all treated,
utilizing a total of 5 mL of the chemoablation cocktail. The patient maintained appropriate
hemodynamics and experienced acceptable post-procedure pain with no major complications
during either procedure.
Fig. 3 EUS with color Doppler imaging noting three lobular masses lobular masses in the
head and body of the pancreas.
At follow-up 3 months later, the patient said she felt subjectively markedly improved.
Since her last ablation, reported that she has not required any emergency care for
her symptoms. She previously reported that she would have several episodes of severe,
symptomatic hypoglycemia daily, which has now been reduced to only three to five episodes
monthly. Several of the medications to control her symptoms have since been reduced
or discontinued all together.
Conclusion
EUS-guided alcohol ablation of hormonally active NETs is effective; however, rates
of SAEs of 6 %, felt to be secondary to the inflammatory and toxic effects of alcohol,
are a limitation. Recent trials have shown that with mucinous pancreatic cyst ablation,
alcohol is not required when an admixture of gemcitabine-paclitaxel is used in its
place, and that when alcohol is removed from the ablation process, SAE rates fall
significantly [5 ]
[9 ]. In this limited case report, we have shown that alcohol-free ablation of NETs can
also be safe and effective. Although promising, it is unclear if alcohol-free EUS-guided
chemoablation of solid NETs represents an advance in treating these lesions in patients
who are not good surgical candidates. However, further investigation is certainly
warranted given the possible equivalent efficacy and improved safety when compared
to alcohol-based ablation.
