INSULINOMA-HOW TO ACHIEVE A RELIABLE LOCALIZATION OF THE TUMOR ?

Introduction: Arterial stimulation with calcium and venous sampling (ASVS) enables us to reach the goal of avoiding that any patient with insulinoma undergoes a blind surgical exploration. Since ASVS is both a functional and morphological localization procedure, its sensitivity is not influenced by factors that are causing the insensitivity of usual anatomical and morphological procedures. This results in all ASVS sensitivity data being the same, regardless of where the procedure was performed. The practical meaning of this is that when ASVS is done after other morphological procedures there are no falsely positive or negative results, and that every patient is sent to surgery with enough insulinoma localization data. Aim: Based on our own experience in preoperative localization of insulinoma we indented to show why we believe that ASVS should be performed to all patients regardless of data collected from other preoperative localization methods. Method: Retrospectively we have analyzed the accuracy of preoperative localization methods. First anatomical and morphological procedures like transabdominal ultrasoundUS, endoscopic ultrasoundEUS, computerized tomography-CT and magnetic resonance imaging-MRI were done. Then we analyzed the data collected during a functional procedure which at the same time allows regionalization ASVS. To estimate the accuracy, the results of every single method were correlated with the operative findings in all sixteen cases. Results: Prior to ASVS fourteen patients underwent US, fifteen had CT, MRI was performed in eight patients and EUS in thirteen. Using only one of these methods enabled identification of tumors in five patients, using two methods in six patients while three and four in one patient each. For three patients none of these methods was successful. ASVS revealed that all seen tumors were functional except three of the six visualized with two methods (US and EUS). In two of these three cases US and EUS localized tumors in pancreatic tail/body whilst ASVS accurately identified tumors in pancreatic head. For these patients US and EUS showed falsely positive results. In the third of these patients EUS showed tumor localized in pancreatic head, while US and ASVS accurately pointed to tail. This too was a falsely positive result of EUS. ASVS successfully provided regionalization data in three patients where other visualization methods failed. Operative and later histological findings confirmed the accuracy of ASVS in all sixteen patients including two patients that previously underwent distal pancreatectomy based on falsely positive EUS findings.


INTRODUCTION:
In adults that are not treated for diabetes mellitus, who apart from hypoglycemic episodes seem like healthy persons the usual cause for endogenic hyperinsulinemic hypoglycemias is insulinoma.Insulinomas are extremely rare with incidence of 4/1000000 (1), very small (90 % < 2cm), in 90% benign, and in 90% solitary.Practically all are tumors of the beta cells in pancreatic Langerhans islets.Insulinomas are causing hypoglycemias by uncontrolled secretion of insulin.This diagnosis comes to mind when Whipple's triad is present and when it can be provoked by prolonged fasting of course in the presence of necessary biochemical criteria for an adequate diagnosis (8).The only adequate treatment option is surgery.Preoperative tumor localization is a very important step between the diagnosis and surgery.An accurate tumor localization is essential for successful surgical outcome.The importance of the preoperative tumor localization is best represented by the commonly adopted position by relevant professionals that none of these patients should undergo a blind pancreatic exploration (10).
Insulinomas could be localized by anatomical and morphological procedures like US, EUS, CT MRI and angiography.Furthermore we could use ASVS as both functional and morphological exam.
Sensitivity data differ significantly among various morphological procedures and between diagnostic centers.For US the sensitivity is 16-64% (9,10), for CT 33-64% (3), for MRI 40-90% (3), for EUS 65-92% (3,13) and for angiography 29-50% (2).Indirectly the small size and rare occurrence can contribute to the inconsistency of the results.Different diagnostic centers use different diagnostic tools.These tools differ in power of resolution as well.Therefore, technical reasons may in part influence such difference, because insulinomas of the same size might be identified in one center and missed in center with inferior diagnostic appliance.On the other hand to achieve maximum results with owned technology, a certain level of skill, experience and devotion is much needed.Given the same technology, a diagnostic center with significantly more patients will have more chance to reach desired skill and experience.A good example of the importance of skill and experience is EUS with a sensitivity span from 65 to 92% ( 3,13).Another reason for different success rates between diagnostic centers might be the fact that every center develops and masters some of the localization procedures more than others.This comes regardless of the overall similar relevance and renown of centers in insulinoma management.In one of such famous institutions for insulinoma management US exams provide accurate localization in 65% (10) and in another with comparable experience only in16%.(9).To sum up, results of anatomical and morphological diagnostic exams differ significantly particularly if skill and experience requiring methods are used.
Never the less despite sophisticated equipment and admirable skills and experience, some tumors remain unseen-falsely negative results, while some tumors that are found will not be functional, will not be insulinomas-falsely positive results.
Because it is both a morphological and functional method ASVS substantially differs from previously mentioned procedures.The sensitivity of ASVS results is not influenced by the factors that are causing insensitivity or limitations of usual anatomical and morphological exams.Regardless of the center that reported the ASVS data, the sensitivity of results is quite uniform and reproducible.This diagnostic procedure can be performed with the same accuracy anywhere.When done after other visualization methods, ASVS enables an insight into the functionality of the found change and, if none was found, it provides sufficient regionalization of pancreatic area that contains insulinoma.

METHOD:
Data gathered during preoperative insulinoma localization in sixteen patients was analyzed retrospectively.Our group includes eleven women and five men.The age range was between 23 and 77 years.All patients underwent surgery from 2002 to 2013.Two patients in this group were unsuccessfully operated in other institutions.For every patient, after performing the usual anatomical and morphological exams, we did ASVS.The accuracy of each localization method was compared to operative and histological findings.
The first two ASVS exams (1996 and 2002) were done in accordance with the original protocol (4).Later from 2006 we have modified the procedure in a way that all venous sampling was undertaken from the right hepatic vein (5), arterial stimulation with calcium was applied both in proximal and distal part of the lineal artery( 6) and the dose of calcium was fixed to 1.35mEq-3ml of 10% calcium gluconate (7).
The procedure: Catheter used for venous sampling was inserted through femoral vein and fixed to position in right hepatic vein.The catheter used for angiography of celiac plexus was inserted through femoral artery.Then selective catheterization was performed for each of the following arteries: a.gastroduodenalis, a. mesenterica superior, proximal and distal part of a.lienalis and a. hepatica.After each selective catheterization angiography and stimulation with 3ml of 10% of calcium gluconate were performed.Blood samples from the right hepatic vein were taken 30 sec prior to immediately prior to and 30, 60, 90 and 120 seconds after the stimulation with calcium.A double or higher rise in insulin concentration compared to the starting value was considered diagnostically significant.So when a diagnostic rise in insulin concentration after calcium stimulation was noted in a.gastroduodenalis and a.mesenterica superior we considered that insulinoma was regionalized in pancreatic head or procesus uncinatus.When the above was found after the stimulation of both ends of a. lienalis insulinoma was regionalized in pancreatic tail.On the other hand when it was seen after the stimulation of only proximal part of a. lienalis the tumor was in the body of pancreas.A double or higher rise in insulin concentration after the stimulation of a.hepatica suggests metastatic disease in liver.If angiographic findings (angiography during procedure) are positive and correlate with functional and regionalization data, then ASVS adds localization value to the test.It becomes more than a functional and regionalization test.
Prior to ASVS in our group 13 patients underwent US, 15 CT , 8 MR and 13 EUS.

RESULTS:
All 16 patients had ASVS done in a previously described way.In every case, after the stimulation with calcium, a diagnostic rise of insulin was noted and that enabled an accurate localization or regionalization of tumors.This was confirmed with operative and histological findings.Prior to ASVS, 7 suspected tumors were seen with US but for two ASVS showed no functionality-falsely positive findings.CT and MRI identified 5 and 2 changes respectively and all of them were confirmed functional by ASVS.When EUS was used 10 suspected tumors were noted but only 7 were confirmed functional by ASVS and other 3 were not functional-falsely positive results ( By using only one anatomical and morphological method suspected tumors were identified in 5 patients (N o 2, 5, 6, 7, and 9).In those cases, ASVS confirmed functionality for each one.The findings obtained by two diagnostic methods identified suspected change in six cases (N o 1, 3, 8, 12 and 16).For three of them (N o 1, 3 and 12 ) ASVS showed functionality but in other three cases it did not.US and EUS indicated that tumor was in pancreatic body, whilst ASVS accurately showed that insulinomas were present in pancreatic head.For both of them, US and EUS results were falsely positive.In the third patient of this group a suspected change was identified by EUS on pancreatic head whilst US result suggested body/tail segment.ASVS showed that US finding was accurate and that EUS gave a falsely positive result.In our group, two patients had accurate localization achieved by three or four morphologic exams and their functionality was also confirmed with ASVS.
Patients N o 14 and 15 had unsuccessful distal pancreactomy in other institutions 1.5 and 5 years earlier respectively.The diagnoses were based on falsely positive EUS results.
Angiography was performed in all patients.For three patients it was accurate and corresponded to the functional findings.In three cases it was falsely positivedid not corresponded to the functionality findings.Eventually, for ten patients angiography was negative although tumor was found later on-falsely negative results.
All patients were monitored after the surgery for at least one year and had no hypoglycemic episodes.

DISCUSION:
The diagnosis of endogenous hyperinsulinemic hypoglycemias, commonly caused by insulinomas, is based on the established criteria, so that its detection is not that difficult (8).
On the other hand preoperative localization of insulinomas inside pancreatic tissue is difficult.Exact localization is necessary since surgery is the only valuable option for a definitive treatment.Having accurate tumor localization enhances chances for successful tumor resection, shortens the time of operation, and therefore reduces the number of unsuccessful operations and a need for reoperations to the minimum.Reoperations correlate with higher morbidity.Therefore it is necessary to do everything to avoid that a single patient undergoes a blind pancreatic exploration (10).
For preoperative localization we can use several anatomical and morphological exams ( US, EUS, CT, MRI, angiography) and ASVS that is both a functional and regionalization test.
Based on previously given limitations of anatomical and morphological exams it is not possible to localize all insulinoma cases using only these procedures, causing some of insulinoma cases to remain unseen-falsely negative results.It is not rare that all morphological methods fail to localize tumor (occult insulinomas).Sending patients to would in fact blind pancreatic exploration.Another problems are falsely positive results.Some changes that are seen are not functional, therefore are not insulinomas.A patient sent to surgery based on falsely positive results will not be successfully operated.Difficulties arising from falsely positive or falsely negative results can be overcome if after anatomical and morphological exams ASVS is performed.Unlike these methods ASVS is a functional exam that can point to a specific region of pancreas as a possible tumor site-regionalization.If there is a concurrence of angiographic and functional part of the test in a suspected region (characteristic angiographic finding in the area with diagnostic rise in insulin levels) ASVS becomes a functionally-anatomical localization exam.Since ASVS is functional test limitations of other anatomical and morphological methods do not influence its results.Consequently, the reports about ASVS sensitivity are reproducible and uniform regardless of the institution where it was done.In all institutions that are performing ASVS almost all insulinomas will be regionalized or localized.
ASVS enables localization or regionalization of insulinomas that were unseen with other methods and if a suspected change was identified, it provides additional information about the functionality of the explored lesionit determents if it is actually an insulinoma.So, ASVS, on one hand prevents patients from undergoing a blind surgical exploration of pancreas and, on the other hand, it prevents surgery based on falsely positive results.
By performing ASVS after other anatomical and morphological methods we have used all its benefits.We had falsely negative findings in three patients, but after ASVS, the accurate regionalization was made and that prevented a blind pancreatic exploration and led to a successful surgical outcome.
Falsely positive results may lead to unsuccessful surgery.An example of this is the outcome of our two patients that were diagnosed with insulinomas in pancreatic body based on falsely positive EUS results.Other anatomical and morphological exams were negative.
Distal pancreactomies were unsuccessful in both cases.One and a half and five years after that these patients were admitted to our institution and after performing ASVS, we have determined the presence of insulinomas in pancreatic head,which was followed by a successful surgical enucleation.If ASVS had been done prior to the surgery, both patients could have been accurately diagnosed (EUS findings would have been seen as falsely positive) and properly operated during their first hospitalization.
Being operated based on falsely positive findings could have been an outcome of three more patients if ASVS had not been done to them.Fortunately, ASVS was used to check the functionality of suspected pancreatic changes identified by anatomical and morphological exams and such scenario was avoided.Two of them had suspected tumors based on US and EUS findings in pancreatic body but ASVS accurately regionalized insulinomas in pancreatic head.Both patients had successful surgical enucleation of insulinomas from pancreatic head.Chances for these patients to be misguidedly operated based on falsely positive results were much higher since not just one, but two methods gave wrong localization of tumor.AS for he third patient, if ASVS had not been done she would have been in danger of undergoing a complex Whipple pancreatic operation.In this case the EUS result indicated that the tumor was in pancreatic head and US pointed to body and tail.EUS is considered to be the most sensitive tool for localization of insulinomas in pancreatic head, superior to US, so it is not such a dilemma which result would have been chosen as true.If surgery was in order it could turn into Whipple section, since tumor in pancreatic head can easily be missed by palpation even when the surgeon is experienced.
Never the less ASVS showed the EUS result to be falsely positive and the US finding to be true.Distal pancreatctomy was performed and the unnecessary Whipple operation avoided.
Five patients had pancreatic changes identified only by one method.Unlike the unsuccessfully operated patients these five patients had ASVS done before surgery and, although it could go the other way, all tumors were functional.There were no falsely positive results and patients were successfully operated.
The common ground for all medical workers that are treating insulinoma patients is a desire to achieve an exact preoperative localization of tumor.Since such a goal can not be reached universally with the same equipment and methods, every institution in this line of work need to develop their own methodology.The only measure of success is a concurrence of operative and histological findings with preoperative localization procedures.All approaches are equally good if their results are the same, and ideal if all insulinomas are accurately localized.
Our approach to perform ASVS in every patient after other anatomical and morphological procedures, regardless of the results, is based on two facts.Firstly, by doing ASVS we were ableto localize insulinomas in patients with falsely negative results and, secondly, in this way we could check functionality of identified changes and avoid to be misled by falsely positive results.We will continue with such practice since we consider it to be the only way to send patients to surgery without a fear of failure.ASVS will be performed until it is proven that it can be replaced with non-invasive localization methods such as scintigraphy of GLP-1 receptors.

Table N o 2 :
Basic demographic characteristics, preoperative diagnostic results, operative and histological findings .P-positive, N-negative, FP-false positive.
table N o 1) Basic demographic characteristics, preoperative diagnostic results and operative and histological findings for patients are given in table N o 2. For 3 patients (N o 4, 10 and 14) without preoperative localization with anatomical and morphological exams ASVS provided accurate regionalization of insulinomas.