- Email: [email protected]
Patterns of recurrence after curative resection of pancreatic ductal adenocarcinoma
Available online at www.sciencedirect.com
EJSO 35 (2009) 600e604
www.ejso.com
Patterns of recurrence after curative resection of pancreatic ductal adenocarcinoma A. Van den broeck a, G. Sergeant a, N. Ectors b, W. Van Steenbergen c, R. Aerts a, B. Topal a,* a
Department of Abdominal Surgery, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium b Department of Pathology, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium c Department of Hepatobiliary & pancreatic diseases, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium Accepted 15 December 2008 Available online 7 January 2009
Abstract Aims: Despite curative surgery for pancreatic ductal adenocarcinoma (PDAC), most patients develop cancer recurrence and die from metastatic disease. Understanding of the patterns of failure after surgery can lead to new insights for novel therapeutic modalities. The aim of the present study is to describe the patterns of recurrence after curative resection of PDAC. Methods: A retrospective analysis was performed of 145 consecutive resections for PDAC between 1998 and 2005 (M/F 75/70; median (range) age 67 years (32e85 y)). The location of the first and consecutive recurrences, and the time interval to cancer recurrence after surgical resection was studied. The magnitude of tumour-free margin was less than a millimetre in 48 patients, whereas a positive surgical margin was observed in 27 patients. The median duration of follow-up was 18.5 (range 0.3e116.8) months. Results: Cancer recurrence was observed in 110 patients. The first location of recurrence was locoregional in 19, extra-pancreatic in 66, and combined locoregional and extra-pancreatic in 25 patients. Extra-pancreatic recurrence developed in the liver in 57, peritoneal in 35, pulmonary in 15, and retroperitoneal in 5 patients. The median (95% CI) overall (OS) and disease-free (DFS) survival was 18.7 (15.7e23.5) and 9.8 (7.5e12.4) months, respectively. The type of cancer recurrence did not significantly influence OS, while the resection margin status had a prognostic effect. Conclusion: The vast majority of patients who undergo potentially curative surgery for PDAC develop cancer recurrence located in the abdominal cavity. Surgical resection margins with tumour involvement and tumour-free margins of less then 1 mm are negative prognostic factors. Further research on better local surgical control, peri-operative locoregional treatment, and more effective adjuvant systemic therapy is necessary to improve long-term survival of patients with curable PDAC. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Cancer; Pancreas; Surgery; Recurrence
Introduction It remains a challenge to improve the survival of patients with pancreatic ductal adenocarcinoma (PDAC). The majority of patients present with locally advanced or metastatic disease and have a median survival of 6e8 months irrespective of the type of treatment modality.1 In selected patients (5e15%) with localised PDAC, surgical resection is the only treatment offering long-time survival.2 Despite the curative intent of surgical resection, cancer recurrence
* Corresponding author. Tel.: þ32 16344265; fax: þ32 16344832. E-mail address: [email protected] (B. Topal). 0748-7983/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ejso.2008.12.006
develops within two years after pancreatic surgery in over 60% of patients.3 Multiple studies have investigated the predictors of survival after surgery. However, only a few variables have been shown to correlate consistently with patient outcome. Surgical resection margin status and the presence of lymph node metastases were found to be most important determinants of survival after surgery for PDAC.2,4e6 Although the majority of patients with PDAC die from distant metastatic disease, little is known about the exact pattern of cancer recurrence. The aim of the present study is to determine the first localization of cancer recurrence and the time interval to develop recurrence following curative resection of PDAC. A better understanding of the patterns of failure,
A. Van den broeck et al. / EJSO 35 (2009) 600e604
601
Materials and methods
mesenteric (n ¼ 3) lymph node metastases on pathological examination. Patients who had positive resection margins on pathological examination (pR1) were not excluded from further analysis.
Patients
Follow-up
Between January 1998 and December 2005, one hundred and forty-five patients with PDAC underwent surgical resection with curative intent. Patients with an intraductal papillary mucinous tumour (IPMT) were excluded from the study. The male/female ratio was 75/70, with a median age of 67 years (range 32e85 y). A standard pancreaticoduodenectomy (PD; n ¼ 73) or a pylorus-preserving procedure (PPPD; n ¼ 48) was performed in 121 patients for ductal carcinoma of the pancreatic head. Patients with PDAC of the body (n ¼ 5) or tail (n ¼ 18) underwent distal pancreatectomy with splenectomy. Total pancreatectomy was carried out in 1 patient with PDAC of the head. Right para-aortic lymph node dissection was carried out routinely in pancreatic head tumours, and followed by (PP)PD in case no metastatic lymph nodes were found at frozen section pathology. In the case of metastatic para-aortic lymph nodes a (PP)PD was not performed, except in relatively young patients. Combined radiation and chemotherapy was administered in a neo-adjuvant setting in 3 and as adjuvant treatment in 38 patients. In 31 patients only chemotherapy was given as adjuvant treatment. Adjuvant therapy was given in 63 patients with lymph node metastases, while 34 patients without lymph node metastases did not receive adjuvant treatment ( p < 0.0001).
Patient follow-up was closed in January 2008, with a median follow-up time of 18.5 months (range 0.3e116.8 m). Follow-up information was obtained through review of patients’ hospital charts (n ¼ 118) or by contacting the patients’ peripheral specialist (n ¼ 24) or general practitioner (n ¼ 2). One patient was lost to follow-up. Postoperative follow-up investigations consisted of a clinical examination, biochemistry including serum carbohydrate antigen (CA) 19.9 level (normal <34 kU/l), transabdominal ultrasound and pulmonary X-ray at 3-month intervals. A computed tomography (CT) and/or magnetic resonance imaging (MRI) scan of the abdomen was performed annually. An abdominal CT-scan was performed at 6 months after surgery in 55 patients who were suspected from recurrence on transabdominal ultrasound and/or elevated serum CA 19.9 levels. An abdominal MRI-scan was performed in 3 out of these patients. A CT-scan of the thorax and/or a whole-body positron emission tomography (PET) scan with 18-fluoro-2-deoxyglucose (FDG) was performed only if other diagnostic methods were inconclusive (n ¼ 27). Any suspicion of recurrent disease was confirmed either on CT- or MRI-scan. Recurrent disease was defined as a radiological mass characterised as a malignant lesion on CT- and/or MRI-scan, or as a hot spot on FDG-PET-scan that was confirmed with CT-scan later on during follow-up.
Outcome measures
Results
Primary outcome measures were the first location and time to cancer recurrence after surgical resection. Disease-free survival (DFS) and overall survival (OS) rates were estimated using the KaplaneMeier life-table technique. Locoregional recurrence was defined as cancer recurrence at the pancreatic resection site. Cancer recurrence in the liver, peritoneum, retroperitoneum, lungs or any other site was classified as extra-pancreatic recurrence. Intra-abdominal recurrence was defined as cancer recurrence at the locoregional site, in the liver, intraabdominal lymph nodes, peritoneum or retroperitoneum. Surgical margins of the resection specimen were delineated with ink before pathological examination. A pR1-resection was defined as tumour involvement of the inked margin. The inked margins were used to determine the magnitude of the tumour-free resection margin (more or less than 1 mm). According to the UICC TNM staging 6th ed., tumours were classified histopathologically as stage Ia in 4, stage Ib in 5, stage IIa in 30, and stage IIb in 92 patients. Another 14 patients were classified as stage IV due to the presence of para-aortic (n ¼ 11) or distal
Median (range) duration of surgery was 258 (100e480) min, with a median intra-operative blood loss of 1100 (100e5000) ml. Postoperative complications occurred in 82 patients, with subsequent hospital mortality in 6 patients. Postoperative pancreatic fistula developed in 15 patients.
i.e. where and when cancer recurrence occurs, could help develop novel and more rational therapeutic approaches.
Location of cancer recurrence Cancer recurrence was observed in 110 patients. The first or primary location of cancer recurrence was locoregional in 19, extra-pancreatic in 66 and combined locoregional and extra-pancreatic in 25 patients. Extra-pancreatic recurrences developed in the liver in 57, peritoneal in 35 and pulmonary in 15 patients. Retroperitoneal recurrence was observed in 5 patients and was synchronous with locoregional recurrence (Fig. 1). At the end of follow-up, isolated locoregional or extrapancreatic recurrence was observed in 9 and 60 patients, respectively. Combined locoregional and extra-pancreatic recurrence developed in 41 patients. Extra-pancreatic
602
A. Van den broeck et al. / EJSO 35 (2009) 600e604
than 1 mm. The median DFS in patients with pR1-resection was 6.8 (5.3e9.1) as compared to 11.9 (7.7e14.0) months in patients with tumour-free (pR0) resection margins ( p ¼ 0.170). The median OS in these patients was 16.1 (8.7e17.7) vs. 21.4 (15.8e26.5) months, respectively ( p ¼ 0.005) (Fig. 2). A tumour-free resection margin of less than 1 mm negatively influenced both DFS (6.7 (5.2e8.7) vs. 12.4 (8.5e16.5) months; p ¼ 0.012) and OS (12.4 (10.4e16.8) vs. 24.0 (18.5 e 30.2); p ¼ 0.0006) (Fig. 3). Neither a pR1-resection nor a tumour-free margin of less than 1 mm resulted in a greater number of locoregional recurrences ( p > 0.527). Discussion
Figure 1. Cumulative cancer recurrence after curative surgery for pancreatic ductal adenocarcinoma according to their localisation) () indicates synchronous localisations are mentioned in each category; metachronous localisations have been excluded).
failure was most frequently located in the liver (n ¼ 59) and the peritoneal cavity (n ¼ 40). Intra-abdominal cancer recurrence was finally observed in 103 patients. Recurrence and survival rates For the entire patient population, the median DFS and OS was 9.8 months (95% confidence interval (CI) 7.6e12.4 months) and 18.7 months (CI 15.7e23.5 months), respectively. The estimated OS and DFS rates at 1, 2, and 5 years were 65.3 vs. 43.3, 39.8 vs. 23.4, and 14.2 vs. 15.9%, respectively. Liver and peritoneal metastases occurred earlier in the postoperative period than retroperitoneal and lung metastases (DFS of respectively 8.2, 9.2, 17.6 and 17.8 months). The type of cancer recurrence (locoregional, extra-pancreatic, combined locoregional/extra-pancreatic) did not significantly influence the survival ( p ¼ 0.172). On univariate analysis, adjuvant chemo(radio)therapy had no significant effect on DFS, OS, or the pattern of cancer recurrence (log-rank; p > 0.334). The median DFS ( p ¼ 0.162) and OS ( p ¼ 0.991) rates with or without chemo(radio)therapy were 7.5 (6e11.9) vs. 12.1 (8.4e16.5) and 17.7 (14.4e22.1) vs. 21.0 (12.2e28.1) months, respectively.
In the present study, the recurrence rate after curative surgery for PDAC was 56.7, 76.6, and 84.1% at 1, 2, and 5 years, respectively. Over 90% of patients developed cancer recurrence in the abdominal cavity as being the first site of localization, either isolated or combined, in the liver (39%), locoregional (30%), or the peritoneum (24%). Pulmonary metastases occurred in about 10% of the patients and represented the most frequent site of extra-abdominal recurrence. Recurrence rates in the current study are in accordance with the literature.3,7,8 Also the sites of cancer recurrence fit in the broad range of previously published data. The most common reported sites of recurrence are the locoregional area (7.8e69%), the liver (24e53%) and the peritoneal cavity (22e56%).3,7e10 Locoregional recurrence Isolated local recurrences are rare.3 In the current study, 19 patients developed a locoregional recurrence as the first
Surgical resection margin The magnitude of tumour-free margin was less than a millimetre in 48 patients, whereas a pR1-resection was observed in 27 patients. A positive margin was found posterior in 17, anterior in 3, lateral in 1, and at the pancreatic transsection margin in 6 patients. Cancer recurrence developed in 19/27 patients (locoregional 13, extra-pancreatic 6) with a pR1-resection, and in 38/48 patients (locoregional 19, extra-pancreatic 19) with a resection margin of less
Figure 2. Overall survival (OS) rates following pR0 vs. pR1-resection of pancreatic ductal adenocarcinoma. pR0: tumour-free surgical margins on pathological examination; pR1 positive (tumour involvement) surgical resection margin on pathological examination.
A. Van den broeck et al. / EJSO 35 (2009) 600e604
Figure 3. Overall survival (OS) rates following curative surgery for pancreatic ductal adenocarcinoma; relation with pathological surgical resection margin (pRM) of more vs. less than 1 mm.
and only site of recurrent disease, whereas at the end of follow-up only 9 patients were observed to have isolated locoregional recurrence. Early locoregional recurrence has generally been attributed to incomplete surgical resection of PDAC. Direct extension from the primary tumour to adjacent major vessels or remnant pancreas, lymph node metastases, and perineural invasion have been described as factors associated with locoregional recurrence.6,11 Long-term survival after surgical treatment of PDAC is determined by the radicality (pR0) of resection.2 Despite the curative intent and adequacy of surgery, surgical margins with tumour involvement (pR1) are still observed on pathological examination. The rate of pR1-resections reported in the literature varies from 16% to 75%, but is often underestimated by a lack of a standardized pathological protocol.12e14 Nevertheless, whatever the pR1-resection rate is, it seems not to result in increased locoregional recurrence rates.2,12,15 In the present study, a pR1-resection was observed in 19%, while the magnitude of tumour-free margins was less than 1 mm in 33% of the patients. An extensive pathological protocol wasn’t accomplished in our institution during the current study, which may explain the relatively low pR1 rate. Although a pR1-resection and tumour-free margin of less than 1 mm were not associated with increased locoregional recurrence rates, they significantly reduced overall survival. Locoregional recurrence rates were comparable in patients with or without adjuvant therapy.
603
and the peritoneal cavity, which may result in overt metastases.17 In the present study 91 patients developed extra-pancreatic recurrence as the first location of recurrent disease, whereas at the end of follow-up 101 patients were diagnosed to have extra-pancreatic recurrence. Liver metastases were most frequent and developed in the early postoperative period. More than 50% of all liver metastases developed in the first six months postoperatively, even in patients with early tumour stage. The majority of peritoneal recurrences developed between the 4th and 15th postoperative month. Lung metastases on the other hand developed less frequently and more gradually over the postoperative course. Patients with or without adjuvant therapy had comparable DFS, OS, and patterns of cancer recurrence. However, the nonrandomized nature and the small number of patients are limitations of the present study to draw final conclusions. At the present time chemotherapy is the only adjuvant therapy to improve survival.18,19 Patients with PDAC who undergo surgical resection might benefit from locoregional adjuvant therapy since the intra-abdominal cavity is the primary site of failure. Therefore, a no-touch technique and intra-operative lavage were suggested, but further research on these strategies remains necessary.17,20 Intra-operative radiation therapy upon the surgical resection site and isolated liver perfusion chemotherapy are feasible and safe, but have so far not been proven to be beneficial on local recurrence or survival.21e23 Whether intra-operative radiotherapy will improve survival of these patients remains doubtful since the liver is the primary site of cancer recurrence and patients with or without locoregional recurrence seem to have similar survival rates. Conclusion After curative surgery for PDAC most patients develop locoregional and distant cancer recurrence, and survival rates remain low. The vast majority of patients develop cancer recurrence within the abdominal cavity. Surgical resection margins with tumour involvement (pR1) and tumour-free margins of less then 1 mm are negative prognostic factors. At present, adjuvant chemotherapy is the only adjuvant treatment with a small survival benefit, but really effective adjuvant treatment is lacking. Further research on better local surgical control, peri-operative locoregional therapy, and more effective adjuvant systemic therapy is necessary to improve long-term survival of patients with curable PDAC. Conflict of interest
Extra-pancreatic recurrence The authors have no conflict of interest. Early relapse after curative surgery for PDAC may be explained by the presence of micrometastases or minimal residual disease not detectable at the time of surgery.16 Also surgical manipulation of the tumour can shed cancer cells into the portal vein, lymphatic vessels, nerve bundles
Acknowledgements Many thanks to the staff members of the departments of hepatobiliary and pancreatic diseases, digestive oncology,
604
A. Van den broeck et al. / EJSO 35 (2009) 600e604
radiotherapy, and radiology for inclusion and follow-up of patients into this study; C. Verslype, E. Van Cutsem, K. Haustermans, D. Bielen, and D. Vanbeckevoort, respectively.
References 1. Exocrine pancreas. American Joint Committee on Cancer: AJCC Cancer Staging Manual. New York: Springer; 2002. pp. 157e64. 2. Wagner M, Redaelli C, Lietz M, et al. Curative resection is the single most important factor determining outcome in patients with pancreatic adenocarcinoma. Br J Surg 2004;91:586–94. 3. Griffin JF, Smalley SR, Jewell W, et al. Patterns of failure after curative resection of pancreatic carcinoma. Cancer 1990;66:56–61. 4. Howard TJ, Krug JE, Yu J, et al. A margin-negative R0 resection accomplished with minimal postoperative complications is the surgeon’s contribution to long-term survival in pancreatic cancer. J Gastrointest Surg 2006;10:1338–45 [discussion 1345e6]. 5. Winter JM, Cameron JL, Campbell KA, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg 2006;10:1199–210 [discussion 1210e1]. 6. Ozaki H, Hiraoka T, Mizumoto R, et al. The prognostic significance of lymph node metastasis and intrapancreatic perineural invasion in pancreatic cancer after curative resection. Surg Today 1999;29: 16–22. 7. Barugola G, Falconi M, Bettini R, et al. The determinant factors of recurrence following resection for ductal pancreatic cancer. JOP 2007;8:132–40. 8. Westerdahl J, Andren-Sandberg A, Ihse I. Recurrence of exocrine pancreatic cancer e local or hepatic? Hepatogastroenterology 1993;40: 384–7. 9. Takahashi S, Ogata Y, Miyazaki H, et al. Aggressive surgery for pancreatic duct cell cancer: feasibility, validity, limitations. World J Surg 1995;19:653–9 [discussion 660]. 10. Kayahara M, Nagakawa T, Ueno K, et al. An Evaluation of radical resection for pancreatic cancer based on the mode of recurrence as determined by autopsy and diagnostic imaging. Cancer 1993;72: 2118–23.
11. Fernandez-Cruz L, Johnson C, Dervenis C. Locoregional dissemination and extended lymphadenectomy in pancreatic cancer. Dis Surg 1999;16:313–9. 12. Esposito I, Kleeff J, Bergmann F, et al. Most pancreatic cancer resections are R1 resections. Ann Surg Oncol 2008;15:1651–60. 13. Verbeke CS. Resection margins and R1 rates in pancreatic cancer e are we there yet? Histopathology 2008;52:787–96. 14. Verbeke CS, Leitch D, Menon KV, et al. Redefining the R1 resection pancreatic cancer. Br J Surg 2006;93:1232–7. 15. Raut CP, Tseng JF, Sun CC, et al. Impact of resection status on pattern of failure and survival after pancreaticoduodenectomy for pancreatic adenocarcinoma. Ann Surg 2007;246:52–60. 16. Vogel I, Kalthoff H, Henne-Bruns D, Kremer B. Detection and prognostic impact of disseminated tumor cells in pancreatic carcinoma. Pancreatology 2002;2:79–88. 17. Hirota M, Shimada S, Yamamoto K, et al. Pancreatectomy using the no-touch isolation technique followed by extensive intraoperative peritoneal lavage to prevent cancer cell dissemination: a pilot study. JOP 2005;6:143–51. 18. Oettle H, Post S, Neuhaus P, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: a randomized controlled trial. JAMA 2007;297: 267–77. 19. Regine WF, Winter KA, Abrams RA, et al. Luorouracil vs gemcitabine chemotherapy before and after fluorouracil-based chemoradiation following resection of pancreatic adenocarcinoma: a randomized controlled trial. JAMA 2008;299:1019–26. 20. Yamamoto K, Shimada S, Hirota M, et al. EIPL (extensive intraoperative peritoneal lavage) therapy significantly reduces peritoneal recurrence after pancreatectomy in patients with pancreatic cancer. Int J Oncol 2005;27:1321–8. 21. Evans DB, Termuhlen PM, Byrd DR, et al. Intraoperative radiation therapy following pancreaticoduodenectomy. Ann Surg 1993;218:54–60. 22. Sindelar WF, Kinsella TJ. Studies of intraoperative radiation therapy in carcinoma of the pancreas. Ann Oncol 1999;10:226–30. 23. Hayashibe A, Kameyama M, Shinbo M, Makimoto S. Clinical results on intra-arterial adjuvant chemotherapy for prevention of liver metastasis following curative resection of pancreatic cancer. Ann Surg Oncol 2007;14:190–4.
EJSO 35 (2009) 600e604
www.ejso.com
Patterns of recurrence after curative resection of pancreatic ductal adenocarcinoma A. Van den broeck a, G. Sergeant a, N. Ectors b, W. Van Steenbergen c, R. Aerts a, B. Topal a,* a
Department of Abdominal Surgery, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium b Department of Pathology, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium c Department of Hepatobiliary & pancreatic diseases, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium Accepted 15 December 2008 Available online 7 January 2009
Abstract Aims: Despite curative surgery for pancreatic ductal adenocarcinoma (PDAC), most patients develop cancer recurrence and die from metastatic disease. Understanding of the patterns of failure after surgery can lead to new insights for novel therapeutic modalities. The aim of the present study is to describe the patterns of recurrence after curative resection of PDAC. Methods: A retrospective analysis was performed of 145 consecutive resections for PDAC between 1998 and 2005 (M/F 75/70; median (range) age 67 years (32e85 y)). The location of the first and consecutive recurrences, and the time interval to cancer recurrence after surgical resection was studied. The magnitude of tumour-free margin was less than a millimetre in 48 patients, whereas a positive surgical margin was observed in 27 patients. The median duration of follow-up was 18.5 (range 0.3e116.8) months. Results: Cancer recurrence was observed in 110 patients. The first location of recurrence was locoregional in 19, extra-pancreatic in 66, and combined locoregional and extra-pancreatic in 25 patients. Extra-pancreatic recurrence developed in the liver in 57, peritoneal in 35, pulmonary in 15, and retroperitoneal in 5 patients. The median (95% CI) overall (OS) and disease-free (DFS) survival was 18.7 (15.7e23.5) and 9.8 (7.5e12.4) months, respectively. The type of cancer recurrence did not significantly influence OS, while the resection margin status had a prognostic effect. Conclusion: The vast majority of patients who undergo potentially curative surgery for PDAC develop cancer recurrence located in the abdominal cavity. Surgical resection margins with tumour involvement and tumour-free margins of less then 1 mm are negative prognostic factors. Further research on better local surgical control, peri-operative locoregional treatment, and more effective adjuvant systemic therapy is necessary to improve long-term survival of patients with curable PDAC. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Cancer; Pancreas; Surgery; Recurrence
Introduction It remains a challenge to improve the survival of patients with pancreatic ductal adenocarcinoma (PDAC). The majority of patients present with locally advanced or metastatic disease and have a median survival of 6e8 months irrespective of the type of treatment modality.1 In selected patients (5e15%) with localised PDAC, surgical resection is the only treatment offering long-time survival.2 Despite the curative intent of surgical resection, cancer recurrence
* Corresponding author. Tel.: þ32 16344265; fax: þ32 16344832. E-mail address: [email protected] (B. Topal). 0748-7983/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ejso.2008.12.006
develops within two years after pancreatic surgery in over 60% of patients.3 Multiple studies have investigated the predictors of survival after surgery. However, only a few variables have been shown to correlate consistently with patient outcome. Surgical resection margin status and the presence of lymph node metastases were found to be most important determinants of survival after surgery for PDAC.2,4e6 Although the majority of patients with PDAC die from distant metastatic disease, little is known about the exact pattern of cancer recurrence. The aim of the present study is to determine the first localization of cancer recurrence and the time interval to develop recurrence following curative resection of PDAC. A better understanding of the patterns of failure,
A. Van den broeck et al. / EJSO 35 (2009) 600e604
601
Materials and methods
mesenteric (n ¼ 3) lymph node metastases on pathological examination. Patients who had positive resection margins on pathological examination (pR1) were not excluded from further analysis.
Patients
Follow-up
Between January 1998 and December 2005, one hundred and forty-five patients with PDAC underwent surgical resection with curative intent. Patients with an intraductal papillary mucinous tumour (IPMT) were excluded from the study. The male/female ratio was 75/70, with a median age of 67 years (range 32e85 y). A standard pancreaticoduodenectomy (PD; n ¼ 73) or a pylorus-preserving procedure (PPPD; n ¼ 48) was performed in 121 patients for ductal carcinoma of the pancreatic head. Patients with PDAC of the body (n ¼ 5) or tail (n ¼ 18) underwent distal pancreatectomy with splenectomy. Total pancreatectomy was carried out in 1 patient with PDAC of the head. Right para-aortic lymph node dissection was carried out routinely in pancreatic head tumours, and followed by (PP)PD in case no metastatic lymph nodes were found at frozen section pathology. In the case of metastatic para-aortic lymph nodes a (PP)PD was not performed, except in relatively young patients. Combined radiation and chemotherapy was administered in a neo-adjuvant setting in 3 and as adjuvant treatment in 38 patients. In 31 patients only chemotherapy was given as adjuvant treatment. Adjuvant therapy was given in 63 patients with lymph node metastases, while 34 patients without lymph node metastases did not receive adjuvant treatment ( p < 0.0001).
Patient follow-up was closed in January 2008, with a median follow-up time of 18.5 months (range 0.3e116.8 m). Follow-up information was obtained through review of patients’ hospital charts (n ¼ 118) or by contacting the patients’ peripheral specialist (n ¼ 24) or general practitioner (n ¼ 2). One patient was lost to follow-up. Postoperative follow-up investigations consisted of a clinical examination, biochemistry including serum carbohydrate antigen (CA) 19.9 level (normal <34 kU/l), transabdominal ultrasound and pulmonary X-ray at 3-month intervals. A computed tomography (CT) and/or magnetic resonance imaging (MRI) scan of the abdomen was performed annually. An abdominal CT-scan was performed at 6 months after surgery in 55 patients who were suspected from recurrence on transabdominal ultrasound and/or elevated serum CA 19.9 levels. An abdominal MRI-scan was performed in 3 out of these patients. A CT-scan of the thorax and/or a whole-body positron emission tomography (PET) scan with 18-fluoro-2-deoxyglucose (FDG) was performed only if other diagnostic methods were inconclusive (n ¼ 27). Any suspicion of recurrent disease was confirmed either on CT- or MRI-scan. Recurrent disease was defined as a radiological mass characterised as a malignant lesion on CT- and/or MRI-scan, or as a hot spot on FDG-PET-scan that was confirmed with CT-scan later on during follow-up.
Outcome measures
Results
Primary outcome measures were the first location and time to cancer recurrence after surgical resection. Disease-free survival (DFS) and overall survival (OS) rates were estimated using the KaplaneMeier life-table technique. Locoregional recurrence was defined as cancer recurrence at the pancreatic resection site. Cancer recurrence in the liver, peritoneum, retroperitoneum, lungs or any other site was classified as extra-pancreatic recurrence. Intra-abdominal recurrence was defined as cancer recurrence at the locoregional site, in the liver, intraabdominal lymph nodes, peritoneum or retroperitoneum. Surgical margins of the resection specimen were delineated with ink before pathological examination. A pR1-resection was defined as tumour involvement of the inked margin. The inked margins were used to determine the magnitude of the tumour-free resection margin (more or less than 1 mm). According to the UICC TNM staging 6th ed., tumours were classified histopathologically as stage Ia in 4, stage Ib in 5, stage IIa in 30, and stage IIb in 92 patients. Another 14 patients were classified as stage IV due to the presence of para-aortic (n ¼ 11) or distal
Median (range) duration of surgery was 258 (100e480) min, with a median intra-operative blood loss of 1100 (100e5000) ml. Postoperative complications occurred in 82 patients, with subsequent hospital mortality in 6 patients. Postoperative pancreatic fistula developed in 15 patients.
i.e. where and when cancer recurrence occurs, could help develop novel and more rational therapeutic approaches.
Location of cancer recurrence Cancer recurrence was observed in 110 patients. The first or primary location of cancer recurrence was locoregional in 19, extra-pancreatic in 66 and combined locoregional and extra-pancreatic in 25 patients. Extra-pancreatic recurrences developed in the liver in 57, peritoneal in 35 and pulmonary in 15 patients. Retroperitoneal recurrence was observed in 5 patients and was synchronous with locoregional recurrence (Fig. 1). At the end of follow-up, isolated locoregional or extrapancreatic recurrence was observed in 9 and 60 patients, respectively. Combined locoregional and extra-pancreatic recurrence developed in 41 patients. Extra-pancreatic
602
A. Van den broeck et al. / EJSO 35 (2009) 600e604
than 1 mm. The median DFS in patients with pR1-resection was 6.8 (5.3e9.1) as compared to 11.9 (7.7e14.0) months in patients with tumour-free (pR0) resection margins ( p ¼ 0.170). The median OS in these patients was 16.1 (8.7e17.7) vs. 21.4 (15.8e26.5) months, respectively ( p ¼ 0.005) (Fig. 2). A tumour-free resection margin of less than 1 mm negatively influenced both DFS (6.7 (5.2e8.7) vs. 12.4 (8.5e16.5) months; p ¼ 0.012) and OS (12.4 (10.4e16.8) vs. 24.0 (18.5 e 30.2); p ¼ 0.0006) (Fig. 3). Neither a pR1-resection nor a tumour-free margin of less than 1 mm resulted in a greater number of locoregional recurrences ( p > 0.527). Discussion
Figure 1. Cumulative cancer recurrence after curative surgery for pancreatic ductal adenocarcinoma according to their localisation) () indicates synchronous localisations are mentioned in each category; metachronous localisations have been excluded).
failure was most frequently located in the liver (n ¼ 59) and the peritoneal cavity (n ¼ 40). Intra-abdominal cancer recurrence was finally observed in 103 patients. Recurrence and survival rates For the entire patient population, the median DFS and OS was 9.8 months (95% confidence interval (CI) 7.6e12.4 months) and 18.7 months (CI 15.7e23.5 months), respectively. The estimated OS and DFS rates at 1, 2, and 5 years were 65.3 vs. 43.3, 39.8 vs. 23.4, and 14.2 vs. 15.9%, respectively. Liver and peritoneal metastases occurred earlier in the postoperative period than retroperitoneal and lung metastases (DFS of respectively 8.2, 9.2, 17.6 and 17.8 months). The type of cancer recurrence (locoregional, extra-pancreatic, combined locoregional/extra-pancreatic) did not significantly influence the survival ( p ¼ 0.172). On univariate analysis, adjuvant chemo(radio)therapy had no significant effect on DFS, OS, or the pattern of cancer recurrence (log-rank; p > 0.334). The median DFS ( p ¼ 0.162) and OS ( p ¼ 0.991) rates with or without chemo(radio)therapy were 7.5 (6e11.9) vs. 12.1 (8.4e16.5) and 17.7 (14.4e22.1) vs. 21.0 (12.2e28.1) months, respectively.
In the present study, the recurrence rate after curative surgery for PDAC was 56.7, 76.6, and 84.1% at 1, 2, and 5 years, respectively. Over 90% of patients developed cancer recurrence in the abdominal cavity as being the first site of localization, either isolated or combined, in the liver (39%), locoregional (30%), or the peritoneum (24%). Pulmonary metastases occurred in about 10% of the patients and represented the most frequent site of extra-abdominal recurrence. Recurrence rates in the current study are in accordance with the literature.3,7,8 Also the sites of cancer recurrence fit in the broad range of previously published data. The most common reported sites of recurrence are the locoregional area (7.8e69%), the liver (24e53%) and the peritoneal cavity (22e56%).3,7e10 Locoregional recurrence Isolated local recurrences are rare.3 In the current study, 19 patients developed a locoregional recurrence as the first
Surgical resection margin The magnitude of tumour-free margin was less than a millimetre in 48 patients, whereas a pR1-resection was observed in 27 patients. A positive margin was found posterior in 17, anterior in 3, lateral in 1, and at the pancreatic transsection margin in 6 patients. Cancer recurrence developed in 19/27 patients (locoregional 13, extra-pancreatic 6) with a pR1-resection, and in 38/48 patients (locoregional 19, extra-pancreatic 19) with a resection margin of less
Figure 2. Overall survival (OS) rates following pR0 vs. pR1-resection of pancreatic ductal adenocarcinoma. pR0: tumour-free surgical margins on pathological examination; pR1 positive (tumour involvement) surgical resection margin on pathological examination.
A. Van den broeck et al. / EJSO 35 (2009) 600e604
Figure 3. Overall survival (OS) rates following curative surgery for pancreatic ductal adenocarcinoma; relation with pathological surgical resection margin (pRM) of more vs. less than 1 mm.
and only site of recurrent disease, whereas at the end of follow-up only 9 patients were observed to have isolated locoregional recurrence. Early locoregional recurrence has generally been attributed to incomplete surgical resection of PDAC. Direct extension from the primary tumour to adjacent major vessels or remnant pancreas, lymph node metastases, and perineural invasion have been described as factors associated with locoregional recurrence.6,11 Long-term survival after surgical treatment of PDAC is determined by the radicality (pR0) of resection.2 Despite the curative intent and adequacy of surgery, surgical margins with tumour involvement (pR1) are still observed on pathological examination. The rate of pR1-resections reported in the literature varies from 16% to 75%, but is often underestimated by a lack of a standardized pathological protocol.12e14 Nevertheless, whatever the pR1-resection rate is, it seems not to result in increased locoregional recurrence rates.2,12,15 In the present study, a pR1-resection was observed in 19%, while the magnitude of tumour-free margins was less than 1 mm in 33% of the patients. An extensive pathological protocol wasn’t accomplished in our institution during the current study, which may explain the relatively low pR1 rate. Although a pR1-resection and tumour-free margin of less than 1 mm were not associated with increased locoregional recurrence rates, they significantly reduced overall survival. Locoregional recurrence rates were comparable in patients with or without adjuvant therapy.
603
and the peritoneal cavity, which may result in overt metastases.17 In the present study 91 patients developed extra-pancreatic recurrence as the first location of recurrent disease, whereas at the end of follow-up 101 patients were diagnosed to have extra-pancreatic recurrence. Liver metastases were most frequent and developed in the early postoperative period. More than 50% of all liver metastases developed in the first six months postoperatively, even in patients with early tumour stage. The majority of peritoneal recurrences developed between the 4th and 15th postoperative month. Lung metastases on the other hand developed less frequently and more gradually over the postoperative course. Patients with or without adjuvant therapy had comparable DFS, OS, and patterns of cancer recurrence. However, the nonrandomized nature and the small number of patients are limitations of the present study to draw final conclusions. At the present time chemotherapy is the only adjuvant therapy to improve survival.18,19 Patients with PDAC who undergo surgical resection might benefit from locoregional adjuvant therapy since the intra-abdominal cavity is the primary site of failure. Therefore, a no-touch technique and intra-operative lavage were suggested, but further research on these strategies remains necessary.17,20 Intra-operative radiation therapy upon the surgical resection site and isolated liver perfusion chemotherapy are feasible and safe, but have so far not been proven to be beneficial on local recurrence or survival.21e23 Whether intra-operative radiotherapy will improve survival of these patients remains doubtful since the liver is the primary site of cancer recurrence and patients with or without locoregional recurrence seem to have similar survival rates. Conclusion After curative surgery for PDAC most patients develop locoregional and distant cancer recurrence, and survival rates remain low. The vast majority of patients develop cancer recurrence within the abdominal cavity. Surgical resection margins with tumour involvement (pR1) and tumour-free margins of less then 1 mm are negative prognostic factors. At present, adjuvant chemotherapy is the only adjuvant treatment with a small survival benefit, but really effective adjuvant treatment is lacking. Further research on better local surgical control, peri-operative locoregional therapy, and more effective adjuvant systemic therapy is necessary to improve long-term survival of patients with curable PDAC. Conflict of interest
Extra-pancreatic recurrence The authors have no conflict of interest. Early relapse after curative surgery for PDAC may be explained by the presence of micrometastases or minimal residual disease not detectable at the time of surgery.16 Also surgical manipulation of the tumour can shed cancer cells into the portal vein, lymphatic vessels, nerve bundles
Acknowledgements Many thanks to the staff members of the departments of hepatobiliary and pancreatic diseases, digestive oncology,
604
A. Van den broeck et al. / EJSO 35 (2009) 600e604
radiotherapy, and radiology for inclusion and follow-up of patients into this study; C. Verslype, E. Van Cutsem, K. Haustermans, D. Bielen, and D. Vanbeckevoort, respectively.
References 1. Exocrine pancreas. American Joint Committee on Cancer: AJCC Cancer Staging Manual. New York: Springer; 2002. pp. 157e64. 2. Wagner M, Redaelli C, Lietz M, et al. Curative resection is the single most important factor determining outcome in patients with pancreatic adenocarcinoma. Br J Surg 2004;91:586–94. 3. Griffin JF, Smalley SR, Jewell W, et al. Patterns of failure after curative resection of pancreatic carcinoma. Cancer 1990;66:56–61. 4. Howard TJ, Krug JE, Yu J, et al. A margin-negative R0 resection accomplished with minimal postoperative complications is the surgeon’s contribution to long-term survival in pancreatic cancer. J Gastrointest Surg 2006;10:1338–45 [discussion 1345e6]. 5. Winter JM, Cameron JL, Campbell KA, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg 2006;10:1199–210 [discussion 1210e1]. 6. Ozaki H, Hiraoka T, Mizumoto R, et al. The prognostic significance of lymph node metastasis and intrapancreatic perineural invasion in pancreatic cancer after curative resection. Surg Today 1999;29: 16–22. 7. Barugola G, Falconi M, Bettini R, et al. The determinant factors of recurrence following resection for ductal pancreatic cancer. JOP 2007;8:132–40. 8. Westerdahl J, Andren-Sandberg A, Ihse I. Recurrence of exocrine pancreatic cancer e local or hepatic? Hepatogastroenterology 1993;40: 384–7. 9. Takahashi S, Ogata Y, Miyazaki H, et al. Aggressive surgery for pancreatic duct cell cancer: feasibility, validity, limitations. World J Surg 1995;19:653–9 [discussion 660]. 10. Kayahara M, Nagakawa T, Ueno K, et al. An Evaluation of radical resection for pancreatic cancer based on the mode of recurrence as determined by autopsy and diagnostic imaging. Cancer 1993;72: 2118–23.
11. Fernandez-Cruz L, Johnson C, Dervenis C. Locoregional dissemination and extended lymphadenectomy in pancreatic cancer. Dis Surg 1999;16:313–9. 12. Esposito I, Kleeff J, Bergmann F, et al. Most pancreatic cancer resections are R1 resections. Ann Surg Oncol 2008;15:1651–60. 13. Verbeke CS. Resection margins and R1 rates in pancreatic cancer e are we there yet? Histopathology 2008;52:787–96. 14. Verbeke CS, Leitch D, Menon KV, et al. Redefining the R1 resection pancreatic cancer. Br J Surg 2006;93:1232–7. 15. Raut CP, Tseng JF, Sun CC, et al. Impact of resection status on pattern of failure and survival after pancreaticoduodenectomy for pancreatic adenocarcinoma. Ann Surg 2007;246:52–60. 16. Vogel I, Kalthoff H, Henne-Bruns D, Kremer B. Detection and prognostic impact of disseminated tumor cells in pancreatic carcinoma. Pancreatology 2002;2:79–88. 17. Hirota M, Shimada S, Yamamoto K, et al. Pancreatectomy using the no-touch isolation technique followed by extensive intraoperative peritoneal lavage to prevent cancer cell dissemination: a pilot study. JOP 2005;6:143–51. 18. Oettle H, Post S, Neuhaus P, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: a randomized controlled trial. JAMA 2007;297: 267–77. 19. Regine WF, Winter KA, Abrams RA, et al. Luorouracil vs gemcitabine chemotherapy before and after fluorouracil-based chemoradiation following resection of pancreatic adenocarcinoma: a randomized controlled trial. JAMA 2008;299:1019–26. 20. Yamamoto K, Shimada S, Hirota M, et al. EIPL (extensive intraoperative peritoneal lavage) therapy significantly reduces peritoneal recurrence after pancreatectomy in patients with pancreatic cancer. Int J Oncol 2005;27:1321–8. 21. Evans DB, Termuhlen PM, Byrd DR, et al. Intraoperative radiation therapy following pancreaticoduodenectomy. Ann Surg 1993;218:54–60. 22. Sindelar WF, Kinsella TJ. Studies of intraoperative radiation therapy in carcinoma of the pancreas. Ann Oncol 1999;10:226–30. 23. Hayashibe A, Kameyama M, Shinbo M, Makimoto S. Clinical results on intra-arterial adjuvant chemotherapy for prevention of liver metastasis following curative resection of pancreatic cancer. Ann Surg Oncol 2007;14:190–4.