Reply To A Letter To The Editor Regarding The International Study Group On Pancreatic Surgery Definition And Classification Of Chyle Leak After Pancreatic Operation.
Published 2017 · Medicine
To the Editors: We would like to thank Debs and colleagues for their interest in the consensus definition and classification of chyle leak (CL) after pancreatic operation by the International Study Group on Pancreatic Surgery (ISGPS). They raise some interesting questions about the consensus, which we address consecutively. The authors question whether the diagnosis of CL can already be determined at the third postoperative day, especially in patients for whom an enteral diet was not started. Indeed, the detection of CL frequently is related to the start of feeding. However, early oral feeding after pancreatectomy is now well established within the enhanced recovery after surgery strategy and used routinely in many centers. Thus, as it is thought the lymphatic disruption leading to CL is present from the time of the operation, the clinical manifestation varies depending on the diet. Dietary advancement to lipid-containing food varies among centers but to our knowledge does not occur prior to day 3. Therefore, a definition that waits for 5 days to make the diagnosis will not be relevant in some centers. This is similar to the determination of a pancreatic fistula. In a recent large study, CL was first diagnosed at a median of 5 days after the operation in the context of a fast track concept with early oral nutrition. Waiting until 5 days would, therefore, delay early detection and treatment of a CL in #50% of patients. Indeed, clinicians should be aware that in patients for whom an enteral diet was not started early, CL may be diagnosed at a later stage. Second, the authors mention the potential relation among CL, deferred initiation of adjuvant chemotherapy, and consequently worse survival. Although other studies have found that survival rates in pancreatic cancer seem not to be affected by postoperative complications, there is only scarce and conflicting data on the impact specifically of CL after surgery for pancreatic cancer. In fact, this is one of the reasons the ISGPS decided also to harmonize the definition and classification of CL to facilitate further studies on the clinical relevance and treatment of CL after pancreatic operations. Third, the authors enquire if the duration of drainage should be added as a factor in the grading system. Discharge with an operative drain in place indeed leads to a diagnosis of grade B CL as it indicates a deviation from a normal clinical pathway. This “deviation” principle has been used in other definitions. At the time of the design of the CL definition, the postoperative pancreatic fistula definition similarly scored discharge with a drain as a grade B fistula. In the recent 2016 update, extended drainage is only scored as a grade B pancreatic fistula when the drain remains in place for >3 weeks. Such a duration of drainage for CL is not very common, and we therefore decided to maintain the more strict definition for CL. Additionally, the criteria for drain removal for an isolated CL varies greatly among centers and may not necessarily correlate with clinical significance. In some cases, a drain may be kept in place longer than necessary for a variety of reasons, but this does not correlate with clinical condition. Duration of drainage was not included as a factor for grading, because there is at present no evidence-based advice on the timing of removal of the drains for CL. Fourth, the authors ask how hospital stay could not be affected as listed for a grade A CL. As with other ISGPS definitions, grade A refers to a disorder that is biochemically present but has no clinical impact (i.e., it is not clinically relevant). For instance, a CL that is completely treated within the patient s 1to 2-week hospital stay after pancreatectomy. Hospital stay may vary widely throughout the world (East versus West). Finally, the authors question why the clinical condition of the patient was not incorporated in the grading system. Infection as grading was working quite well in the study by van der Gaag et al, but as stated, the current grading system for CL was designed in line with other ISGPS grading systems to reflect changes in management or deviations from a normal clinical pathway. Patients who are clinically ill due to a CL will generally require an intervention; if not, they are classified as a grade A CL. Moreover, the major impact to a prolonged CL, or chylous ascites, is malnutrition, disruption of fat-soluble vitamin absorption and immune cell depletion---all conditions that manifest during relatively longer periods of time than is the intent of ISGPS grading systems.