Giant pancreatic insulinoma. The bigger the worse? Report of two cases and literature review

Abstract

INTRODUCTION

Giant pancreatic insulinomas are rare endocrine tumors. We describe 2 cases reviewing the current literature.

PRESENTATION OF CASE

We report herein 2 female patients affected by giant insulinomas of 14 cm and 6 cm, respectively. Symptomatic hypoglycemia episodes occurred during 4 months in first case and 3 years in the second one until diagnosis. Both patients were successfully treated performing a distal pancreatectomy with splenic preservation in the first case and a Whipple's procedure in the second one.

DISCUSSION

Up to now only 7 cases have been reported previously. Insulinomas larger than 3 cm accounts for less than 5% of all. This literature review shows that despite the size hypoglycemic symptoms varies from 1 day to 3 years and only 1 out of 9 cases reported presented lymph nodes metastases. No recurrences have been described.

CONCLUSION

One of the cases here described (14 cm) is the largest presented in the literature. Despite the size, giant insulinoma is related apparently neither to metastases nor to the recurrences.

1. Introduction

Insulinomas are rare tumors and most of them small at the time of diagnosis (less than 3 cm).¹ The disease usually presents with clinical manifestation of hypoglycemia, secondary to inappropriate insulin secretion.² Only few cases of giant insulinomas have been reported up to now.^3–9 We report two more cases of giant insulinomas describing herein its characteristics and treatment. A review of the literature is performed.

2. Case reports

2.1. Case 1

A 57-year-old female was attended in the emergency department with an acute-onset low consciousness level, sudden confusion, abnormal movements and inability to recognize her relatives. She had been studied up to three times during the previous 4 months for similar episodes. Laboratory tests showed serum glucose of 24 mg/dl. Symptoms immediately disappeared after intravenous infusion of glucose. Serum hormones values were: insulin 22 μunits/mL, C-peptide 3.3 ng/mL, intact proinsulin 59.3 pmol/L. Thoracic-abdominal-pelvic CT was performed (Fig. 1). Ultrasound-guided fine-needle biopsy was suggestive for a neuroendocrine tumor like. Octreoscan study confirmed a contrast uptake by the tumor. A distal pancreatectomy with splenic preservation was performed (Fig. 2) and the patient was discharged on 8th postoperative day without complications. Pathological exam revealed a 14 cm × 10 cm × 6 cm mass identified as a neuroendocrine tumor (Fig. 3). Immunohistochemical staining revealed positivity for insulin and chromogranine. Nuclear proliferation index Ki-67 was higher than 20%. A poorly differentiated and functioning endocrine carcinoma was finally reported. The patient remains asymptomatic and without tumoral recurrence after six years of follow-up.

Fig. 1.

CT abdominal scan showed a 14 cm × 9 cm × 7 cm mass in the pancreatic tail, without pathologic lymph nodes.

Fig. 2.

Intraoperative findings and specimens.

Fig. 3.

(A) Hyperchromic nuclei with prominent nucleoli were found. (B) Ki-67 was higher than 20%. (C) Microscopic lymphatic and vascular invasion was noted.

2.2. Case 2

A 63-year-old woman was referred to emergency department with sudden lumbar pain and anxiety. The patient had a history of several hypoglycemic symptoms in the last years. On admission, laboratory tests showed a serum glucose level of 45 mg/dl, insulin 28 μunits/mL, C-peptide 3.8 ng/mL, intact proinsulin 63.2 pmol/L.

CT scan revealed a mass sized 8 mm × 6 mm × 6.5 mm in the pancreatic head, solid without cystic component and slightly heterogeneous with central calcification. PET-CT study showed increased uptakes in the head of the pancreas (SUV: 38.97). Eco-endoscopy showed a solid giant mass with high vascularization close to the common bile duct and the Wirsung and biopsy was compatible with neuroendocrine tumor. Octreoscan study was normal.

Surgical resection was planned finding a unifocal tumor in the head of pancreas (Fig. 2). A standard pancreaticoduodenectomy (Whipple's procedure) was performed. The pathology report a tumor of 8 cm with cells that were all positive by immunohistochemical staining for the neuroendocrine marker synaptophysin (Fig. 3). Amyloid was present approximately in the 60% of tumor volume. Immunohistochemical stains for insulina, gastrin, glucagon and chromogranin were all negative. The marker ki-67 < 20% and 4 mitosis for 10 high-power fields were observed. Therefore, it was classified as a well-differentiated neuroendocrine tumor. Her postoperative course was uneventful and she was dismissed on the 10th postoperative day. Her insulin, proinsulin and C-peptide levels returned normal. Currently the patient is asymptomatic with no evidence of recurrence at 11 months of follow up.

3. Discussion

Pancreatic endocrine tumors are rare with an incidence of 1–2 cases per million patients every year. The most common neuroendocrine tumor of pancreatic islet cell is insulinoma and the first report was by Nicholls in 1902.¹⁰

There is a slightly higher frequency in women and the mean age of presentation is 45 years old, while patients with multiple endocrine neoplasia type 1 (MEN 1) (5–10% of cases) occur at a younger age.¹ Our cases were slightly older.

The majority of these tumors are found to be solitary, equally distributed throughout the pancreas. They are usually small, with a mean size of 1.5 cm; only 4% are larger than 3 cm.

The mechanism by which insulinomas maintain high levels of insulin secretion in the presence of hypoglycemia is still unknown.

As in the cases herein presented, clinical presentation consists in symptoms related to hypoglycemia, which can be classified in neuroglycopenic symptoms, most common, including anxiety, dizziness, lightheadedness, confusion, visual changes, fatigue, seizures and loss of consciousness and neurogenic signs and symptoms, such as palpitations, tremors and tachycardia.

In 1935, Whipple and Franz described a triad of clinical findings that were unique to patients with insulinoma: symptoms of hypoglycemia, a plasma glucose level of 45 mg/dl or less when they occurred, and relief of symptoms with the administration of glucose.¹¹ In addition, some authors¹ emphasized recently that concomitant measurement of C peptide levels is mandatory for establishing a diagnosis of insulinoma.

The newly 2000 classification, updated in 2010 of neuroendocrine tumors was adopted by the World Health Organization Classification (WHO) for Pancreatic Endocrine Neoplasms.¹²

If giant insulinomas are easily localized because of their size, approximately 40% of all them are not localized preoperatively, and between 3% and 10% remain occult even after intraoperative palpation and use of intraoperative ultrasound.¹⁰

Our routinely preoperative work up includes CT scan, PET scan and Eco-endoscopy; if a neuroendocrine tumor is suspected, an Octreoscan is performed, too.

Surgical excision of insulinoma is the treatment of choice. It consists on a simple enucleation of the tumor or a distal partial pancreatectomy, if localized in the tail of the pancreas. More aggressive surgical procedure is needed for larger tumors, requiring Whipple or total pancreatectomy techniques. Laparoscopic surgery can be performed depending on surgeon experience and size of the tumor.

The recurrences can occur in 6% of cases and more common in MEN 1 patients manifesting with recurrent hypoglycemia.¹ It suggests regrowth of residual insulinoma tissue left.

Insulin secreting carcinomas are potentially curable tumors when confined to the pancreas and regional lymph nodes. In addition, we agree with aggressive resection of advanced neuroendocrine tumors due to proven excellent survival rates even in presence of liver metastasis, lymph node involvement or vascular encasement,¹³ thanks to the slow-growing nature of these tumors.

In cases where insulinoma was missed during pancreatic exploration, those who are not candidate or refuse surgery, or who have unresectable metastasic disease, medical therapy with diazoxide, corticoids, glucagone or ocreotide can be the chosen treatment.¹⁴

Our literature review revealed only 7 previous reports of giant insulinomas (Table 1) with a size ranging from 9 cm to 14 cm.^3–9

Table 1.

Literature review of giant insulinoma.

Case	Author	Year	Tumor size	Duration of symptoms	Metastases	WHO classification	MEN	Prognosis
1	Arensman	1976	10 cm	1 day		NA		NA
2	Danforth	1984	9 cm	7 months	Lymph nodes	NA		7 years
3	Kataoka	1999	11 cm	9 months		NA	I	NA
4	Mittendorf	2005	9 cm	3 years		Well differentiated endocrine tumor		6 months
5	Ketari	2009	11.5 cm	Few months		Well differentiated endocrine tumor		NA
6	Sugiyama	2010	10 cm	3 months		Well differentiated endocrine carcinoma		NA
7	Oberheim	2011	13.5 cm	3 months		Well differentiated endocrine tumor		1 year
8	Present 1	2012	14 cm	4 months		Poorly differentiated endocrine carcinoma		6 years
9	Present 2	2012	6 cm	1 year		Well differentiated endocrine tumor		11 months

NA: not available.

We present 2 more cases of exceptionally giant insulinoma, one of them the largest reported in the literature up to now.

The time frame period of diagnosis of giant insulinoma reported varies from 1 day to 3 years.

Therefore, we can postulate that giant insulinoma may develop from a non-functioning pancreatic mass that, over a period of years start to have a sufficient insulin secretion to cause hypoglycemic symptoms.

Malignant behavior occurs in only 5 to 10% of cases and traditionally related to size (larger than 3 cm) and Ki67 index, with evidence of gross local invasion or metastasis to the peripancreatic lymph nodes and liver.¹²

Only one of the giant insulinomas reported had local lymph nodes invasion without recurrence at 7 years of follow up.⁴ None of these giant insulinomas presented metastasis at diagnosis, therefore, we can suppose that giant size might not be related to malignancy.

According to the WHO classification for pancreatic endocrine tumors, only 2 cases were classified as carcinoma (the first 3 cases described are not included in the WHO classification). Moreover, our poorly differentiated carcinoma is still free from recurrence after 6 years of follow up (Table 1).

In two cases (including one of the present) immunohistochemical staining for insulin was negative, although patients had resolution of hypoglycemic symptoms and normalization of laboratory abnormalities; this result suggest that a small subset of this large pancreatic mass could secrete insulin and that it might not be included in the anatomo-pathological specimen section. In fact, the case described by Mittendorf illustrates such situation, where only 50% of tumor cells were positive for insulin.⁶

Amyloid is present up to 50% of insulinomas¹; concerning giant insulinomas reported, approximately up to 70% contain amyloid with the exception of the Oberheim case that did not contain a substantial component.⁹ Investigators have postulated that giant insulinomas achieve their size because of overproduction and aberrant processing of islet amyloid polypeptide, which is secreted in concert with insulin from insulinoma cells.¹⁵

Favorable outcomes have been reported in 5 patients, including the two present cases (the remaining 4 cases lack of this data) without evidence of recurrence. In this review we found only one patient that presented MEN type I.⁵

4. Conclusions

Giant pancreatic insulinomas are extremely rare tumors. Only seven cases have been reported previously and one of the two cases herein presented is the largest reported up to now (14 cm). This literature review highlights that despite the size, they generally do not present metastases at the time of diagnosis. Treatment of choice is surgical resection with complete remission of symptoms.

Conflict of interest statement

None.

Funding

None.

Ethical approval

Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.

Author contributions

Ielpo B., Vicente E., Quijano Y., Duran H. contributed for study design. Ielpo B., Ferri V., Caruso R. contributed for data collections. Duran H., Vicente E., Diaz E., Fabra I., Puga R., Oliva C., Olivares S. contributed for data analysis.