Pancreatic Ductal Adenocarcinoma: A New TNM Staging System is Needed!
Major changes in the TNM classification system for pancreatic ductal adenocarcinoma were lastly performed in 2002 within the sixth edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual, with no significant changes in the seventh edition.1 However, knowledge of pancreatic cancer biology has significantly evolved in the past 14 years, and several studies have analyzed the prognostic relevance of T-stage and of lymph node (LN) status.2 Current nodal N staging of the TNM system distinguishes only between node-negative (N0) and node-positive (N1) disease, but this simple distinction does not accurately predict prognosis. For different tumors, including gastric and colorectal cancer, the number of metastatic LNs was incorporated into the TNM systems, thus improving their prognostic value.1 Recently, Strobel et al3 showed that the number of positive LNs (PLNs) has a strong prognostic impact in a large cohort of patients undergoing pancreaticoduodenectomy. These authors identified different categories of PLNs (1 PLN, 2–3, 4–7, and ≥8 PLNs), and median survival significantly decreases by increasing the number of PLNs. Similar results were found also by Basturk et al4 who applied 2 different substaging systems of LN status based on the number of PLNs already used for other GI cancer (“upper GI protocol”: N0, no metastases; N1, 1–2 PLNs; N2, ≥3 PLNs; “lower GI protocol”: N0, no metastases; N1, 1–3 PLNs; N2, ≥4 PLNs). The number of LNs harvested is of paramount importance for an accurate N-stage and a proper evaluation of PLNs, and currently 15 harvested LNs are required for this aim after pancreaticoduodenectomy.5 Of note, both Strobel et al and Basturk et al found that a higher number of median examined LNs (17–20) are necessary to adequately evaluate PLNs.3,4 Therefore, the appropriate number of LNs harvested after different pancreatic resections still remains to be clarified.
Another important issue is the site of PLNs. Different authors have found that metastases to para-aortic/interaortocaval PLN (station 16) is associated with very poor prognosis, and some proposed that these PLNs should be considered as M1 disease.6 It would be of interest for the reader to know if Strobel et al made some subgroup analysis, evaluating the prognostic role not only of the number but also of the site of PLNs. They performed a “standard lymphadenectomy” according to the recommendations of the International Study Group for Pancreatic Surgery.7 However, how often did they remove station 16 LNs? Interestingly they found that patients with only 1 PLN have a similar median survival of N0 patients, but metastatic involvement of LNs in station 16 or in peripancreatic stations (station 12/17) has likely a different prognostic significance.
Based on these observations, it becomes clear that current N-stage system is inadequate, and in the future, different N categories should be included in the AJCC/Union for International Cancer Control staging protocol considering both number and site of PLNs.
Recently, a significant criticism related to the T-stage protocol emerged as well. Saka et al8 found that 95.5% of 223 patients undergoing pancreaticoduodenectomy for pancreatic cancer had a pT3 tumor based on the current T-stage definition (pT3: spread of tumor beyond the pancreas including peripancreatic soft tissue, common bile duct, or duodenum), whereas only 4.5% had pT1/pT2 tumors. Of note, in their study, current T-stage definition did not have any correlation with survival. Therefore, they proposed a revised protocol based on tumor size, with the same cut-off adopted by European NeuroEndocrine Tumor Society for T staging of pancreatic neuroendocrine tumors (pT1: size ≤2 cm; pT2: >2 cm and ≤4 cm; pT3: >4 cm in size).