Prudent applications of the many expensive and sometimes time-consuming modalities which may assess pancreatic lesions, continues to challenge clinicians. Local expertise and availability and costs are all parameters which may impact on decisions for staging before resections of these lesions.
Worldwide, some medical centers may inform patients that they had benign not malignant disease (such as autoimmune or focal pancreatitis) only after undergoing a futile major operation. In Israel, which has 18 EUS centers for a population of 7.5 million persons, such futile surgery is widely considered to be unacceptable. Conversely, negative FNA could result in delaying surgery for potentially resectable tumors. As demonstrated in the first case, for T2N0MO pancreatic lesions, in contrast to gastroenterologist and oncologist opinions (73.3-89.5%), only 51.4 percent of surgeons wrote that FNA is indicated prior to surgery. This is one of the most contentious issues dividing physicians. While a positive FNA confirms need for surgery, negative FNA does not necessarily refute the need for surgery. The frequency which one accepts of finding post-hoc that surgery was performed only to find a focal pancreatitis of any cause is at the focus of the debate. In Israel, every effort to prevent unnecessary surgery generally includes attempting FNA and having cytological confirmation if possible. In places where FNA is unavailable, it might seem much more acceptable to attempt surgery based on imaging alone. The false positive rate of surgery may be compared to the rate of operating on what turn out to be "white or non-inflamed appendix". In young male patients, the negative appendectomy rate should be low, perhaps under10%; this means that less than 10% of misdiagnoses proven at surgery are considered an indication that too many true appendicitis cases are likely being missed. The morbidity and mortality risks of pancreatic surgery are considerably higher than for appendectomy. Also, the urgency for operation of the pancreas is less than for appendectomy, for which a delay in hours could prove disastrous. Thus, the reluctance to accept unnecessary operations may arguably be lower for pancreatic operations, and the thoroughness of establishing need for surgery for pancreatic lesions should be maximal.
Most physicians in this survey (68.7%) wrote that it is recommended to perform EUS-guided FNA to evaluate an EUS demonstrated pancreatic lesion prior to surgical intervention. It is interesting to consider why surgeons support this measure to a lesser extent in comparison to their peers (only 51.4% of surgeons). The very physicians most at risk of disappointing patients were found to be most likely to decide on operations which might be futile. The threshold of data needed to decide to operate by surgeons was found to be lower than among others.
In the second case, it appears that most physicians believe that cytology should be attempted prior to commencing chemotherapy in a patient with an inoperable pancreatic tumor according to CT. Supporting this decision are two key considerations: 1- that cytological diagnosis could determine the specific tumor type and thus change the proper choice of chemotherapy in 8-10% of cases which are not primary pancreatic adenocarcinomas and 2- to avoid treating with chemotherapy a benign inflammatory lesion of the pancreas. Once again, the consideration of how heavy a preponderance of evidence indicates a need for a specific therapy (in this case chemotherapy) involves the tendency of physicians to make risky decisions which could potentially be avoided by cytology. The risk rate of patients receiving chemotherapy wrongly treated for a benign condition/misdiagnosis has not been reported specifically for pancreatic lesions. The risks of the specific chemotherapy being offered vs. the difficulties involved in getting cytological or histological confirmation are part of an equation involving clinical cultural and legal environments.
The third case reveals that most physicians, regardless to their specialty, believe that a symptomatic pancreatic cyst, even if smaller than 30 mm, should be aspirated rather than referred directly for surgery. Considerations supporting EUS-FNA include the possibilities raised by results of testing the cystic fluid for CEA, amylase and cytologically determining the presence of glycogen-rich cells- all which would raise the accuracy of evaluating the risk of malignancy. Pseudocysts and serous cystadenomas are cases for which a watch and wait strategy would dominate, as opposed to premalignant or malignant cysts which indicate need for require surgical intervention.
In the fourth case described, the majority of physicians answering the survey indicated that the management of an asymptomatic and stable pancreatic cyst, smaller than 30 mm, is to watch and wait, meaning, follow-up only. Some disagreement was found within the sub-groups of gastroenterologists. Of those performing EUS by themselves, most believe that the cyst should be drained, whereas, most of those who do not perform EUS by themselves, recommended more conservative management. This controversy appears to exist between the ASGE and ACG guidelines. The former, emphasizing the risk that every cystic lesion of the pancreas, regardless of its size, may be malignant (or premalignant) recommends diagnostic evaluation and mentions the clinical useful information added by assessing the cystic fluid for tumor markers, amylase and lipase. The latter, however, allows cross sectional imaging follow-up as long as the lesion is smaller than 5 mm and remains asymptomatic and not growing. Increasingly, a cyst which has all of the EUS features of a serous cystadenoma is considered a low risk lesion, and deferring FNA and making the diagnosis based on EUS imaging alone is considered an option [9, 10].
According to this survey most physicians would avoid the risks of FNA. This approach, compared to endosonographers, may be due to concerns about the risks, unawareness of the benefits, and/or considerations of availability and costs of EUS-FNA. The risks of FNA, have recently been estimated in a large meta-analysis. Mortality was found to be about 1/5000, while significant morbidities including post-FNA pain and post-FNA pancreatitis were more common, especially among the prospective studies .
As evident by the fifth and final case, approximately 70 percent of physicians wrote that cytological evidence should be sought for a suspected GIST. Modern immunohistochemistry tests for c-KIT and PDGFRA can diagnose GIST. Impacts of FNA in such a situation are that low-grade small GISTs may require no therapy, whereas high-grade GISTs respond very well to imatinib treatment. It is crucial to note that there are many other intramural/submucosal lesions which could be diagnosed by EUS- FNA in this situation, some of which would have therapeutic impact (e.g. carcinoid).
A final caveat is that FNA is far from a perfect diagnostic modality. The ability to obtain accurate samples from EUS-guided FNA varies from 50 to about 90%, depending on the organ being sampled, the expertise, and possibly the presence of a cytologist during the procedure . Even for cases of thyroid nodules, for which FNA is done from outside of the body, under relatively direct vision, and being able to regulate all body part movements which could adversely affect the procedure and lead to false negative due to sample error there is a significant (roughly 20%) rate of non-informative cytology results .
In the case of a pancreatic lesion visualized by CT, performing EUS-guided FNA prior to surgical or oncological treatment may provide (in 85% of cases) supporting cytological evidence. The role of cytology in a case in which, according to the CT, the pancreatic tumor is inoperable, (e.g. due to vascular involvement), is to determine the proper chemotherapy/radiotherapy, especially for atypical tumors.
The symptoms associated with pancreatic cysts are hard to establish. A pancreatic cyst in a symptomatic patient is an indication for cytological evaluation, and/or for drainage, with different methods being available depending on cyst etiology, size, and other variables [14, 15]. The search for more dependable cyst fluid biomarkers than CEA continues: most recently MUC7 was found to have significant value in determining likelihood of a need for surgery and malignant potential of cysts .
Conservative measures in an asymptomatic patient presenting with a small and stable cyst, according to radiography, is widely considered to be sufficient.
A gastric submucosal lesion may be a GIST, a carcinoid or a benign harmless lesion, and so it should be aspirated in order to determine its stage and the treatment. EUS-FNA has become an essential tool in the diagnosis and staging of pancreatic tumors. The major advantage of EUS-FNA lies in the ability of EUS to detect an unresectable disease, to prevent unnecessary surgical exploration and to diagnose small lesions undetectable by other imaging modalities. EUS-FNA can be both therapeutic and diagnostic. The diagnostic values of the FNA are many - these have subjective value. As Brugge, and others since, have summarized, "The chief advantage of EUS-guided FNA is the ability to target, small, intra-pancreatic masses" . Most recently, EUS-guided interventions using FNA needles are being attempted. Novel EUS based techniques are emerging as reasonably safe minimally invasive alternative to the surgical or radiological approaches . As FNA comes into increasing utilization, the measurable impacts in these various real-life situations can be better established.
Most recently, the European Society for Gastrointestinal Endoscopy published its guidelines for EUS-FNA . These guidelines take into consideration the many hundreds of developments which have been discovered and published in the five years since the International Society of Pancreatologists (ISP) reported their consensus guidelines . The ISP guidelines, as mentioned above, went largely unnoticed by the majority of endosonographers, at least insofar as reported in a major survey by Buscaglia et al in 2009 . The present study differs significantly from the ESGE guideline, in that the ESGE guidelines are written only by endoscopists/gastroenterologists, whilst the present work presents the differences in perspectives of surgeons, oncologists and gastroenterologists. Bringing the perspectives of the referring physicians and end-users of the "products" of EUS-FNA may help elucidate the professional needs of the multidisciplinary team who treat these patients.