Pathologic T1 Suclassification of Ampullary Carcinoma With Perisphincteric or Duodenal Sumucosal Invasion Is It T1? DongDo You, MD; JinSeok Heo, MD; SeongHo Choi, MD; DongWook Choi, MD; Kee-Taek Jang, MD Context. In ampullary carcinoma staging, T1 is defined as a tumor limited to the ampulla of Vater or the sphincter of Oddi, and T2 is defined as invasion into the duodenal wall. However, the definition of duodenal wall invasion is vague. Ampullary carcinoma that invades eyond the sphincteric of Oddi (perisphincteric invasion) or into the duodenal sumucosa could e considered pt1 ecause sumucosal invasion is classified as pt1 in gastrointestinal tract tumors. However, there are no data regarding T suclassifications for ampullary carcinoma with perisphincteric or duodenal sumucosa invasion. Ojective. To determine the T suclassification of ampullary carcinoma that invades into perisphincteric or duodenal sumucosa. Design. Pathologically proven ampullary carcinomas with T1 or T2 were reviewed (n ¼ 105). We reclassified tumors as pt1a that were limited to within the sphincter of Oddi (n ¼ 40; 38%), as pt1 for tumors that invaded eyond the sphincter of Oddi or into the duodenal sumucosa (n ¼ 25; 24%), and as pt2 for tumors that invaded into duodenal proper muscle (n ¼ 40; 38%). Results. Lymph node metastasis and recurrence were asent in ampullary carcinoma with pt1a, whereas nodal metastasis were noted in 24% (6 of 25) and 40% (16 of 40) of the ampullary carcinomas with pt1 and pt2, respectively. Tumor recurrence/metastasis rate of ampullary carcinoma with pt1 and pt2 was 44% (11 of 25) and 40% (16 of 40), respectively. The 5-year disease-freesurvival rates from ampullary carcinoma with pt1a, pt1, and pt2 were 95% (38 of 40), 56% (14 of 25), and 58% (23 of 40), respectively (P ¼.003). The 5-year overall survival from ampullary carcinoma with pt1a, pt1, and pt2 was 98% (39 of 40), 72% (18 of 25), and 60% (24 of 40), respectively. Conclusions. The clinicopathologic outcome of ampullary carcinoma with a pt1 suclassification was worse than it was for T1a and approached the outcome for pt2. (Arch Pathol La Med. 2014;138:1072 1076; doi: 10.5858/arpa.2013-0324-OA) The ampulla of Vater (AoV) is a small confluence of the common ile duct and the major pancreatic duct. 1,2 The AoV has an anatomic layer of mucosa, the sphincter of Oddi, perisphincteric or duodenal sumucosa, and duodenal proper muscle, which correspond to the mucosa, muscularis mucosa, sumucosa, and proper muscle layers of other gastrointestinal tract organs, respectively. Because of its compact size and its variation of anatomic structure, it is sometimes difficult to identify the layering architecture of AoV. This anatomic difficulty may cause some prolems in T classification for ampullary carcinoma (AC). According to the current American Joint Committee on Cancer (AJCC) 2 staging for AC, T1 is defined as a tumor limited to the AoV or the sphincter of Oddi, and T2 is defined as invasion of the duodenal wall. We experienced some ACs with perisphincteric and/or duodenal sumucosal invasion that were difficult to clearly classify as T1 or T2. We suspect that some pathologists would consider perisphincteric or duodenal sumucosal invasion y AC as corresponding to T1 ecause the sumucosal invasion of other gastrointestinal tract tumors is classified as T1 or T1 in current AJCC staging. However, there are no data in the literature, to our knowledge, that descrie AC with perisphincteric or duodenal sumucosal invasion as corresponding to T1. The aim of this study was validation of the pathologic T suclassification of AC with perisphincteric or duodenal sumucosal invasion. If the clinicopathologic findings of AC with perisphincteric or duodenal sumucosal invasion were different from ACs confined to sphincter of Oddi muscle layer, pt1 might e a reasonale T classification for AC. Accepted for pulication Octoer 25, 2013. From the Department of Surgery, Catholic University of Korea St Vincent s Hospital, Suwon, Korea (Dr You); and the Departments of Pathology (Dr Jang) and Surgery (Drs Heo, S Choi, and D Choi), Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea. The authors have no relevant financial interest in the products or companies descried in this article. Reprints: Kee-Taek Jang, MD, Department of Pathology, Samsung MATERIALS AND METHODS Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu Seoul 135 710, Korea, (e-mail: kt12. This study was conducted in accordance with institutional review jang@samsung.com). oard requirements. 1072 Arch Pathol La Med Vol 138, August 2014 T1 Suclassification of Ampullary Carcinoma You et al
Figure 1. The serial perpendicular sections of gross examination were performed for evaluation of tumor invasion in ampullary carcinoma. Case Selection We reviewed a consecutive series of 1054 patients who underwent surgical resection with impressions of pancreatoiliary cancer and AC at Samsung Medical Center, Seoul, Korea, etween January 1995 and Decemer 2007. Of 250 cases of pathologically proven ampullary neoplasms, 124 cases (49.6%) were staged as T2 or less. Of the 124 cases, 16 cases of adenoma without carcinoma (12.9%) and 3 cases of carcinoma in situ (2.4%) were excluded, for a total of 105 T1 or T2 ACs to e included in the study. Clinical demographic and surgery methods were reviewed from electronic medical records. Pathologic T Suclassification In pathologic examination, AC was classified as ampullary only if the epicenter of the tumor was the AoV. Pancreas, distal common ile duct, and duodenal carcinomas involving the AoV were eliminated ased on gross and microscopic findings. Gross examination and tissue section was performed according to our standardized gross examination manual for ACs. First, proing along common ile duct was followed y opening of the duodenum along the antiampullary order. Second, the common ile duct was opened along the longitudinal axis, and the ampullary lesion was checked. Third, perpendicular serial sections of the AoV were performed and processed for microscopic examination (Figure 1). Conventional hematoxylin-eosin staining was conducted with 5-lm tissue sections. We reclassified pt1a as tumors limited to within sphincter of Oddi (Figure 2, A), and we reclassified pt1 as AC that invaded eyond the sphincter of Oddi (perisphincteric invasion) and/or into the duodenal sumucosa (Figure 2, B). The pt2 (Figure 2, C) classification was limited to definite duodenal proper muscle wall invasion. Pathologic findings, tumor differentiation, perineural and lymphovascular invasion, nodal metastasis, and tumor size (the sum of the preinvasive lesion and the invasive component) were reviewed. Statistical Analysis Clinicopathologic variales were presented y means, medians, or frequencies, as appropriate, using the Kruskal-Wallis, v 2, and Fisher exact tests. Disease-free status was defined as a patient with no evidence of recurrence. The primary endpoints were overall survival and disease-free survival. Survival was calculated y the Kaplan-Meier method, and the log-rank test was used to analyze differences. Only variales of P,.1 in univariate analysis were included in the multivariate analysis, which was performed using Cox proportional hazards regression. P,.05 was considered statistically significant. Figure 2. A, Ampullary carcinoma (pt1a), which is confined within the sphincter of Oddi muscle; inset, tumor invades hypertrophied sphincter of Oddi muscle. B, Ampullary carcinoma (pt1) invades eyond the sphincter of Oddi (perisphincteric invasion) and/or duodenal sumucosal region. C, Ampullary carcinoma (pt2) invades duodenal proper muscle (hematoxylin-eosin, original magnifications 310 [A through C]). Arch Pathol La Med Vol 138, August 2014 T1 Suclassification of Ampullary Carcinoma You et al 1073
Tale 1. Clinicopathologic Features of Ampullary Carcinomas, n ¼ 105 Features T1a, n ¼ 40 T1, n ¼ 25 T2, n ¼ 40 P a Age, y 61 60 56.5.91 Sex, M:F 19:21 19:6 21:19.06 Surgery, No. (%) Whipple 5 (12) 2 (8) 11 (28).12 PPPD 34 (85) 23 (92) 29 (72) Total pancreatectomy 1 (2) 0 (0) 0 (0) Median tumor size, cm (range) 2.0 (0.9 4.0) 1.7 (0.9 4.0) 2.0 (1.0 10.0).54 Differentiation, No. (%) Well 32 (80) 6 (24) 10 (25),.001 Moderately 6 (15) 8 (32) 13 (32) Poorly 1 (2) 10 (40) 12 (30) Other 1 (2) 1 (4) 5 (12) Perineural invasion, No. (%) 0 (0) 0 (0) 21 (52),.001 Lymphovascular invasion, No. (%) 0 (0) 4 (16) 17 (42),.001 LN metastasis at diagnosis, No. (%) 0 (0) 6 (24) 16 (40),.001 Areviations: LN, lymph node; PPPD, pylorus preserving pancreaticoduodenectomy. a All P values are for comparisons etween groups. Other types include adenosquamous carcinoma, mucinous carcinoma, and micropapillary carcinoma. RESULTS Clinicopathologic Characteristics Recurrence The median follow-up period was 65.9 months (range, 3 to approximately 158 months). Recurrence or metastasis was The median age of 105 patients was 60 years (range, 35 84 detected in 27 patients (25.7%) during the follow-up period. years), and ACs with pt2 showed younger median age There was no tumor recurrence/metastasis in AC cases with without significance. There were 58 men (55%) and 47 pt1a (0%), whereas recurrence/metastasis rates of ACs with women (45%). Men were dominant in the 3 AC groups. pt1 and pt2 were 44% (11 of 25) and 40% (16 of 40), Surgical procedures included Whipple resection in 18 respectively. Recurrence/metastasis rates of AC with pt1 patients (17%) and pylorus-preserving pancreaticoduodenectomy in 86 patients (82%). One patient (1%) underwent were significantly different from ACs with pt1a (P,.001). Locoregional recurrence in the peritoneal cavity occurred in total pancreatectomy ecause of concomitant necrotizing 14 of the 27 patients (52%), and distant metastasis occurred pancreatitis along the entire pancreas. Operative mortality in 13 patients (48%), including 3 comined metastases (oth occurred in 1 of 105 patients (1%) ecause of sepsis from liver and lung). The metastasis sites were liver (n ¼ 6; 46%) postoperative pneumonia. As per the initial pathologic and lung (n ¼ 7; 54%). The factors that potentially report, 55 cases (52%) were T1 and 50 (48%) were T2. When influenced the recurrence rate were analyzed y multivariate we reclassified AC cases according to our T suclassification analysis for disease-free-survival (Tale 2). For disease-freesurvival, the pt stage reflected recurrence. system (pt1a, pt1, and pt2), the previous 55 ACs with T1 were classified into 40 pt1a (72%) and 15 pt1 (27%). Among the previous 50 ACs with T2, 10 (20%) were pt1 Survival and 40 (80%) were pt2. Finally, there were 40 pt1a (38%), The 5-year disease-free survival rates were 95% (38 of 40), 25 pt1 (24%), and 40 pt2 (38%) cases, respectively. In this 56% (14 of 25), and 58% (23 of 40) in AC with pt1a, pt1, study, we used tumor size parameters that included the pt2, respectively (P,.001). The 5-year overall survival for preinvasive component ecause grossly visile lesions the 105 patients was 67% (n ¼ 70). The 5-year overall included the preinvasive component, and we anticipated survival rates were 98% (39 of 40), 72% (18 of 25), and 60% that tumor size (including preinvasive component) could e (24 of 40) for patients with ACs with pt1a, pt1, and pt2, a parameter to differentiate early AC that can e candidate for limited surgery, such as ampullectomy. However, the results did not show a significant difference in tumor size for Tale 2. Predictors of Disease-Free Survival predicting tumor invasion depth. In practice, it was difficult in Ampullary Carcinomas, n ¼ 105 to clearly differentiate preinvasive component from invasive Univariate Multivariate carcinoma in many cases, so we used tumor size that Analysis, a Analysis, Risk 95% included preinvasive component. Most ACs with pt1a P P Ratio CI (80%; 32 of 40) showed well- to moderately differentiated Sex.90 adenocarcinoma, whereas it was 56% (14 of 25) and 58% (23 Differentiation.04..99 of 40) in ACs with pt1 and pt2, respectively. Perineural Perineural invasion was noted only in ACs with pt2. Lymphovascular invasion.56 Lymphovascular invasion was not noted in ACs with pt1a, whereas it was invasion.02.91 identified in 16% (4 of 25) and 42% (17 of 40) in ACs with LN metastasis.007.36 pt1 and pt2, respectively. Lymph node metastasis was not pt1a versus found in ACs with pt1, whereas nodal metastasis at pt1/pt2,.001.02 9.58 1.96 46.89 diagnosis was found in 24% (6 of 25) of ACs with pt1 Areviations: CI, confidence interval; LN, lymph node. and 40% (16 of 40) of the ACs with pt2, respectively. a Log-rank test. Clinicopathologic features are summarized in Tale 1. Cox regression. 1074 Arch Pathol La Med Vol 138, August 2014 T1 Suclassification of Ampullary Carcinoma You et al
Figure 3. Overall survival curves for ampullary carcinoma with pt1a (n ¼ 40), pt1 (n ¼ 25), and pt2 (n ¼ 40). Overall survival curve for ampullary carcinoma with pt1 was different from pt1a and closer to pt2. respectively (P,.001; Figure 3). Univariate analysis for overall survival showed a positive tendency for histologic differentiation, lymphovascular invasion, lymph node metastasis, and pt suclassification (P,.05). When multivariate analysis was performed using those factors, pt suclassification turned out to e an independent, significant prognostic factor (Tale 3). COMMENT The most confusing point in T classification of ACs is the vague definition of T2: Tumor invades duodenal wall. What is the definition of duodenal wall? It seems that the duodenal wall includes the duodenal mucosa, the sumucosa, and the proper muscle layer, except for the duct units of the AoV and the surrounding sphincter of Oddi muscle layer. However, there is no precise description or definition for the duodenal wall in the protocol, which might lead to personal variation in T classification for AC staging. In this study, we used a more precise definition of pt2: Tumor invades duodenal proper muscle and separated ACs with perisphincteric and duodenal sumucosal invasion ecause we think our definition of pt2 is comparale to the T2 classification of other gastrointestinal tract tumors, which invade the muscularis propria in the stomach, small intestine, and colorectum. If a pathologist considers the duodenal mucosa and sumucosa as part of the duodenal wall, a small invasive carcinoma arising from papilla of Vater could e classified as T2, whereas a much larger invasive carcinoma confined to the sphincter of Oddi muscle would have to e regarded as T1. 3 Although there are no data, to our knowledge, we suspect there might e meaningful personal variation in definitions and interpretations of duodenal wall invasion in ACs. In this study, we found some ACs with perisphincteric or duodenal sumucosal invasion that were classified as T1 and other cases where they were classified T2. We suspect that overestimation of the T classification was caused y misinterpretation of duodenal wall invasion. We think vague and prolematic definitions of T1 and T2 in ACs should e more precisely classified into our proposed pt1a as a tumor confined within the sphincter of Oddi muscle, pt1 as tumor invades perisphincteric and/or duodenal sumucosa, and pt2 as tumor invades muscularis propria of duodenum. We suspect that our more precise description and definition could lead to more-reproducile T classification in AC staging. Pathologists are familiar with the layering architecture of mucosa, sumucosa, proper muscle, suserosa, and serosa in the gastrointestinal tract organs. Although the AoV is small, it has all those layering architectures; however, differentiation of the anatomic layering is more difficult for the AoV than it is for other gastrointestinal tract organs. In some cases, it was difficult to define tumor invasion eyond the sphincter of Oddi, which corresponds to sumucosal invasion (T1) in stomach cancer, ecause the sphincter of Oddi muscle was irregularly distriuted and hypertrophied. Thus, we defined pt1 as an ovious tumor invasion eyond the sphincter of Oddi muscle or duodenal sumucosa space. When there is tumor invasion into perisphincteric (eyond sphincter of Oddi muscle layer) or duodenal sumucosal region, we think it is reasonale to interpret it as pt1 ecause the sumucosal invasion corresponds to T1 of the stomach and small intestine in current AJCC staging. 2 We think that vague definition of duodenal wall in AJCC leads to confusion in proper T classification of ACs. The need to stage early ACs may increase ecause of surgical ampullectomy and endoscopic papillectomy. 4,5 We think the current definition of T2 duodenal wall invasion may lead to misinterpretation of ACs, especially in cases of early AC. For example, the muscle layer of the sphincter of Oddi is usually hypertrophied in AC, and it could e misinterpreted as duodenal proper muscle, which might lead to overestimation of the T classification of the AC. The results from this study provided many clinicopathologic findings for ACs with pt1 (perisphincteric or duodenal sumucosa invasion) that were different from those of pt1a. Ampullary carcinoma with pt1 showed more frequent lymph node metastasis and a greater tumor recurrence/metastasis rate, which suggest that ACs with pt1 should e separated from ACs with pt1a. Data for disease-free survival and OS also showed that ACs with pt1 are different from ACs with pt1a. We initially considered the suclassification of pt2a ecause clinicopathologic findings of ACs with perisphincteric or duodenal sumucosal invasion were more closely linked to ACs with pt2 than to pt1a. However, we decided to use pt1 for uniformity in T classification of tumor staging of the gastrointestinal tract organs. In general, T2 means invasion Tale 3. Predictors of Overall Survival in Ampullary Carcinomas, n ¼ 105 Univariate Analysis, a P Multivariate Analysis, P Risk Ratio 95% CI Sex.96 Differentiation.04.78 Perineural invasion.21 Lymphovascular invasion.002.57 LN metastasis.03.75 pt1a versus pt1/pt2,.001.007 30.16 3.29 276.47 Areviations: CI, confidence interval; LN, lymph node. a Log-rank test. Cox regression. Arch Pathol La Med Vol 138, August 2014 T1 Suclassification of Ampullary Carcinoma You et al 1075
into muscularis propria of the corresponding organs. Furthermore, there is no pt2a suclassification in AJCC staging. Although ACs with perisphincteric or duodenal sumucosal invasion are classified as pt1, we think the malignant potential of perisphincteric and/or duodenal sumucosal invasion is greater than it is in sumucosal invasion of other gastrointestinal tract organ tumors. Why does perisphincteric or duodenal sumucosal invasion in AC have a different malignant potential from sumucosal invasion of other gastrointestinal tract organ tumors? When compared with the sumucosal space of other gastrointestinal tract organs, the perisphincteric and duodenal sumucosal space is relatively small, compact, and closer to next layering of duodenal proper muscle and pancreas. Thus, we think the malignant potential of invasive cancer eyond the perisphincteric muscle layer or into duodenal sumucosal region may e greater than that of other gastrointestinal tract tumors. We think the T1a classification should e strictly applied in AC as a tumor limited within the sphincter of Oddi muscle and a hypertrophied sphincter of Oddi muscle should not e interpreted as a duodenal proper muscle component. We think the current definition of limited to AoV in T1 is also vague and may cause misinterpretation in the T classification of ACs. The AoV is not a single unit of organ ut a comination unit of duodenal papillae with the distal end of pancreatic and the common ile duct. 1 The regional definition of AoV is not clear, and it is not certain whether the description of limited to AoV includes the perisphincteric and duodenal sumucosal regions. Thus, there is the possiility that the diagnostic description of limited to AoV may include relatively heterogeneous cases. When we see the survival curves of ACs in the current AJCC staging ook, we think there is a weak point: The survival curve of stage 1B is etter than that of stage 1A. 2 That pattern could e interpreted as T2 eing etter than T1 in AC survival analysis ecause only the T classification is different etween stage 1A (T1N0M0) and 1B (T2N0M0). We suspect that unreasonale pattern suggests the prolem in the T classification of early stage AC. We suspect that previous T1 cases used in survival analysis of AC in AJCC staging may include some cases of AC with perisphincteric or duodenal sumucosal invasion, which might cause a worse survival curve for stage 1A than for stage 1B. Thus, we think that there should e some improvement or more precise description in T suclassification of AC. We think pathologic T1a should e strictly defined as tumors limited within the sphincter of Oddi muscle. However. we experienced some AC cases that were difficult to clearly define as pt1a ecause the sphincter of Oddi muscle is usually hypertrophied and often merges into the duodenal proper muscle in some areas, and microscopic foci of invasive carcinoma can e missed in the histopathologic examination. In this study, we used our standardized gross protocol of perpendicular serial sections for ampullary neoplasms. We think a standardized gross protocol for AC is crucial for the exact evaluation of tumor invasion in AC. Adsay et al 3,6 also suggested that perpendicular gross sectioning with 2 proes along the common ile and pancreatic ducts may e est way to examine invasion y AC. The depth of tumor invasion (T classification) has een reported as an independent prognostic factor in many previous studies. 7 12 In conclusion, we showed that clinicopathologic features of AC with perisphincteric and/or duodenal sumucosal invasion is different from AC with pt1a. Ampullary carcinoma with perisphincteric or duodenal sumucosal invasion should e separated as pt1, which corresponds to sumucosal invasion of other gastrointestinal tract organ tumors. Evaluation of the tumor invasion through a standardized gross protocol of perpendicular, serial sectioning is needed for ojective and reproducile measurement of tumor invasion in ampullary neoplasms. References 1. Alores-Saavedra J, Henson DE, Klimstra DS. Tumors of the Gallladder, Extrahepatic Bile Ducts, and Ampulla of Vater. Washington, DC: Armed Forces of Institute of Pathology; 2000. Atlas of Tumor Pathology; 3rd series, fascicle 27. 2. Edge SB, Byrd BD, Compton CC, Fritz AG, Greene FL, Trotti A, eds. AJCC Cancer Staging Manual. 7th ed. New York, NY: Springer; 2010. 3. Adsay NV, Bagci P, Tajiri T, et al. Pathologic staging of pancreatic, ampullary, iliary, and gallladder cancers: pitfalls and practical limitations of the current AJCC/UICC TNM staging system and opportunities for improvement. Semin Diagn Pathol. 2012;29(3):127 141. 4. Yoon YS, Kim SW, Park SJ, et al. Clinicopathologic analysis of early ampullary cancers with a focus on the feasiility of ampullectomy. Ann Surg. 2005;242(1):92 100. 5. Nguyen N, Shah JN, Binmoeller KF. Outcomes of endoscopic papillectomy in elderly patients with ampullary adenoma or early carcinoma. Endoscopy. 2010;42(11):975 977. 6. Adsay V, Ohike N, Tajiri T, et al. Ampullary region carcinomas: definition and site specific classification with delineation of four clinicopathologically and prognostically distinct susets in an analysis of 249 cases. Am J Surg Pathol. 2012;36(11):1592 1608. 7. Klempnauer J, Ridder GJ, Pichlmayr R. Prognostic factors after resection of ampullary carcinoma: multivariate survival analysis in comparison with ductal cancer of the pancreatic head. Br J Surg. 1995;82(12):1686 1691. 8. Mori K, Ikei S, Yamane T, et al. Pathological factors influencing survival of carcinoma of the ampulla of Vater. Eur J Surg Oncol. 1990;16(3):183 188. 9. Neoptolemos JP, Talot IC, Carr-Locke DL, et al. Treatment and outcome in 52 consecutive cases of ampullary carcinoma. Br J Surg. 1987;74(10):957 961. 10. Roder JD, Schneider PM, Stein HJ, Siewert JR. Numer of lymph node metastases is significantly associated with survival in patients with radically resected carcinoma of the ampulla of Vater. Br J Surg. 1995;82(12):1693 1696. 11. Tsao JI, Rossi RL, Lowell JA. Pylorus-preserving pancreatoduodenectomy: is it an adequate cancer operation. Arch Surg. 1994;129(4):405 412. 12. Iacono C, Verlato G, Zamoni G, et al. Adenocarcinoma of the ampulla of Vater: T-stage, chromosome 17p allelic loss, and extended pancreaticoduodenectomy are relevant prognostic factors. J Gastrointest Surg. 2007;11(5):578 588. 1076 Arch Pathol La Med Vol 138, August 2014 T1 Suclassification of Ampullary Carcinoma You et al