- Email: [email protected]
Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis
HPB
http://dx.doi.org/10.1016/j.hpb.2017.03.006
REVIEW ARTICLE
Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis Savio G. Barreto1,2, Sanjay Pandanaboyana3,4, Natasha Ironside3 & John A. Windsor3,4 1
Hepatobiliary and Oesophagogastric Unit, Division of Surgery and Perioperative Medicine, Flinders Medical Centre, Bedford Park, Adelaide, 2School of Medicine, Faculty of Medicine, Nursing and Health Sciences, Flinders University, South Australia, Australia, 3 Department of Surgery, Faculty of Medical and Health Sciences, University of Auckland, and 4HBP/Upper GI Unit, Department of General Surgery, Auckland City Hospital, Auckland, New Zealand
Abstract Background: Margin status is the main surgical determinant of long-term outcome in pancreatic cancer. Intraoperative frozen section (IOFS) detects microscopic positive margins at a stage when margin revision is possible. The aim of this study was to determine if IOFS driven-revision of pancreatic resection margin(s) improves overall survival (OS) in pancreatic cancer. Methods: A systematic review of major reference databases was undertaken. Patients were divided into 3 groups based on initial FS (FSR0 for negative margin and FSR1 for positive microscopic margin) and final Permanent Section report (PSR0 for negative margin and PSR1 for positive microscopic margin): Group 1 (FSR0 / PSR0), Group 2 (FSR1 / PSR0), and Group 3 (FSR1 / PSR1). Patients in Groups 2 and 3 had surgical revision of the FSR1 margin. Data was meta-analysed. Results: 4 studies included in the final analysis. No difference in OS and incidence of lymph node metastases between Groups 2 and 3 (P = 0.590 and P = 0.410). Conclusions: IOFS-based revision of R1 pancreatic resection margin does not improve OS, even when it results in an R0 margin. This suggests that any benefit of margin revision based on FS is over-ridden by markers of more advanced or aggressive disease. Received 22 November 2016; accepted 1 March 2017
Correspondence Savio G. Barreto, Department of Surgery, Flinders Medical Centre, Bedford Park, South Australia, 5042, Australia. E-mails: [email protected], [email protected]
Introduction Potentially curative surgery is only possible for a fifth of patients with pancreatic ductal adenocarcinoma (PDAC) and the majority of these patients die of systemic disease.1 Besides nodal status, the status of the resection margins on final histopathological analysis are an important determinant of outcome after pancreatoduodenectomy (PD).2 In order to achieve a negative resection margin (R0) many surgeons advocate more radical surgery including synchronous venous resection, arterial resection and/or multi-visceral resection of adjacent organs.3 Margin status has assumed greater importance with the definition of ‘borderline’ resectable PDAC.3 There remains some debate about the impact of a microscopic positive margin on overall survival.4
HPB 2017, -, 1–7
The surgeon cannot modify the biological aggressiveness of the PDAC or the fitness of the patient, but does have some control over resection margin status, and as such it is the main surgical determinant of long-term outcome. Given the importance of a negative resection margin on the outcomes of surgery for pancreatic cancer, there has been interest in using intraoperative frozen section (FS) of the transection margin to drive surgical revision of the margin in order to achieve an R0 resection. A recent study5 concluded that re-resecting a positive intraoperative FS transection margin to achieve an R0 resection improved long-term survival. In contrast, some earlier studies found no survival benefit.6,7
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
2
The aim of this systematic review and meta-analysis was to determine if intraoperative FS driven revision of the transection (neck and bile duct) margin of resection for pancreatic cancer improves overall survival.
Methodology Protocol registration The review protocol used in the present study has been registered at www.researchregistry.com with the Unique Identification number researchregistry159. Literature search A systematic review of major reference databases (MEDLINE, EMBASE, PubMed, and the Cochrane Library) was undertaken according to the PRISMA guidelines8 using a combination of text words “intraoperative”, “frozen section”, “pancreatic cancer”, “pancreatoduodenectomy”, “surgery”, “pancreaticoduodenectomy”. The search was restricted to human studies published up to August 2016 with no time or language restriction. Data from the articles were entered into a database and duplicates were removed. The abstracts were then screened for relevance. Subsequently, the reference lists of relevant trials, reviews, and international guidelines were hand-searched. Reference lists of the retrieved literature were cross-searched manually for additional publications. All available publications describing FS during surgery for pancreatic cancer were evaluated. The inclusion criteria were all types of studies (randomized, prospective observational, and retrospective observational) that investigated the role of intraoperative FS in deciding the margin of resection for pancreatic carcinoma and its impact on the pathological margin and overall survival. The articles that did not report on these two outcomes were excluded. Review articles, letters to the editor, comments and editorials were also excluded. Study groups For the analysis, the patient cohort was divided into 3 groups based on the intraoperative Frozen Section (FS) and final Permanent Section (PS) analysis. Group 1 (FSR0 / PSR0). Patients with a negative margin (R0) on frozen section (FS) that was confirmed on final histopathology (PS). Group 2 (FSR1 / PSR0). Patients with a positive margin (R1) on frozen section (FS) that was a negative margin (R0) confirmed on final histopathology (PS) after revision of margin. Group 3 (FSR1 / PSR1): Patients with a positive margin (R1) on frozen section (FS) margin and a positive margin (R1) on final histopathology (PS) after revision of margin. Outcome measure The primary outcome measure was overall survival expressed as the median difference between the three groups.
HPB 2017, -, 1–7
Pathological factors The pathological variables compared between the groups were lymph node positivity (LNP), lymphovascular (LVI) and perineural (PNI) invasion. Data extraction and quality assessment Studies were identified, and data were extracted by 2 authors (S.G.B., N.I.) independently. A third author (S.P.) further reviewed the selected studies before data extraction in accordance with Cochrane group recommendations for study selection for meta-analysis. Assessment of methodological quality of included studies The ACROBAT-NRSI grading method9 was used to assess the quality of the included non-randomised studies with studies being graded at low, moderate, serious or critical risk of bias (Supplementary file 1). Statistical analysis Statistical analysis was performed using Review Manager Version 5.2 software (Cochrane Collaboration). The odds ratio (OR) with 95% confidence interval (CI) was calculated for categorical data. Hazard ratios were used to analyse survival data. Random and fixed-effects models were used to calculate the combined outcomes of both binary and continuous data.10,11 Publication bias Heterogeneity was assessed using the Cochrane Q test, with P < 0.050 indicating significant heterogeneity. The I2 test, which describes the proportion of variation that is due to heterogeneity beyond chance, had the significance set at >33%. In the absence of heterogeneity, studies were reported using a fixed effects metaanalysis. In the presence of heterogeneity, a cause was sought. If none could be found, the results were described using a randomeffects meta-analysis.12 Potential publication bias was assessed using funnel plots. Egger’s test was used to assess funnel plot asymmetry, by calculating a linear regression of the natural logarithmic scale of the RR over the inverse of the standard error.13 Two tailed P < 0.050 was considered to indicate statistical significance.
Results The results of the search criteria are shown in Fig. 1.5–7,14–21 Table 1 summarizes the 4 studies that included 2385 patients. All 4 studies were retrospective in design and originated from single institutions in the United States,5,7,14 with the exception of a solitary multi-institutional study.6 All the studies were published between 2014 and 2016. All of the studies were from high volume centres, defined as average >20 PDs per year. A minority of patients in the multicentre study received neoadjuvant treatment (10%).6 No mention was made
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
IdenƟficaƟon
3
Records idenƟfied through database searching (n = 291 )
AddiƟonal records idenƟfied through other sources (n = 5)
Included
Eligibility
Screening
Records aŌer duplicates removed (n = 6)
Records screened (n = 290)
Full-text arƟcles assessed for eligibility (n = 19)
Records excluded (n = 279) Full-text arƟcles excluded, with reasons (n = 15) 8 - reporƟng frozen secƟon in Intraductal papillary mucinous neoplasms 5 - failed inclusion criteria aŌer reading full text 2 - same paƟent cohort included in a later updated publicaƟon / mulƟcentre publicaƟon
Studies included in qualitaƟve synthesis (n = 4)
Studies included in quanƟtaƟve synthesis (meta-analysis) (n = 4) Figure 1 PRISMA flow diagram
of patients with borderline resectable disease in any of the 4 studies. The reasons for performing intraoperative FS (either routine practice or surgeon’s discretion) were specified in 2 of the 4 studies.7,14 The margins subjected to FS were pancreatic neck transection and portal vein margin in 1 study5 and pancreatic neck and bile duct/hepatic duct transection margin in the other 4 studies.6,7,14,15 However, only 16 of the 5 studies only published their data from FS of the pancreatic neck transection margin. In 3 studies, an R0 margin was defined as the absence of macroscopic and microscopic tumour at the resection margin.5,6,14 In the study by Pang and colleagues7 a clear margin of a minimum of HPB 2017, -, 1–7
1 mm was considered to constitute an R0 resection. The manner in which a positive FS margin was handled (i.e. extent of revision and number of revision attempts) was clearly specified in 2 studies.6,14 While at least 2 attempts at revision were undertaken in one study,14 an attempt at revising the transection margin by 1–2 cm was attempted in the other study.6 Overall survival All 4 studies were included in the analysis. Table 2 summarizes the primary outcome and pathological variables between the 3 groups. Patients in Group 1 had a significantly better survival compared with Group 2 (HR = 0.68; 95% CI: 0.53–0.87;
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
4
Table 1 Summary of Studies Table included in the analysis [Abbreviations: PDAC – Pancreatic ductal adenocarcinoma, S – single centre;
M – Multicentre; FS – intraoperative frozen section; PS – final histopathology; CI – Confidence intervals (when reported); NR – Not reached] Author (Ref)
Group 1 (FSR0 / PSR0)
Group 2 (FSR1 / PSR0)
Group 3 (FSR1 / PSR1)
Number of patients (%)
Number of patients (%)
Number of patients (%)
Median Survival in months (95% CI)
Median Survival in months (95% CI)
Median Survival in months (95% CI)
Pang et al.7
42 (36)
29 (18–NR)
7 (6)
16 (5–28)
67 (58)
23 (16–31)
Mathur et al.14
298 (67)
22
40 (9)
22
110 (24)
21
Kooby et al.6
1196 (86)
21
72 (5)
12
131 (9)
14
Nitschke et al.5
317 (75)
29
32 (8)
36
73 (17)
12
N [ 2385
1853
151
381
Table 2 Table comparing the primary outcome measure and pathological factors between the three groups (Group 1: FSR0 / PSR0;
Group 2: FSR1 / PSR0; Group 3: FSR1 / PSR1) (Abbreviations: CI – confidence intervals; OR – odds ratio, HR – hazard ratio) Primary Outcome
Groups 1v2 (HR)
95%CI
P
I2 (%)
Groups 1v3 (HR)
95%CI
P
I2 (%)
Groups 2v3 (HR)
95% CI
P
I2 (%)
Overall survival
0.68
0.53, 0.87
0.002
99%
0.72
0.64, 0.80
<0.001
99%
1.08
0.81, 1.46
0.590
99%
2
2
Pathological Factors
Groups 1v2 (OR)
95%CI
P
I (%)
Groups 1v3 (OR)
95%CI
P
I (%)
Groups 2v3 (OR)
95% CI
P
I2 (%)
Lymphovascular invasion
1.02
0.69, 1.50
0.939
0%
0.54
0.27, 1.05
0.069
28%
0.61
0.39, 0.97
0.036
77%
Perineural invasion
0.69
0.44, 1.08
0.104
0%
0.39
0.28, 0.56
<0.001
27%
0.57
0.33, 0.99
0.039
0%
Lymph node involvement
0.53
0.38, 0.72
<0.001
0%
0.41
0.23, 0.74
0.003
70%
0.83
0.54, 1.29
0.410
0%
P = 0.002) and Group 3 (HR = 0.72; 95% CI: 0.64–0.80; P < 0.001). Overall survival was not different between Groups 2 and 3 (P = 0.590) (Supplementary file 2). Pathological factors The outcomes of all three pathological factors have been summarised in Table 2. 1) Lymph node positivity (LNP) - (Supplementary file 3) Patients in Group 1 had a significantly lower LNP rate compared to Group 2 (P < 0.001) and to Group 3 (P = 0.003). 2) Lymphovascular invasion (LVI) - (Supplementary file 4) The incidence of LVI in the tumour was significantly higher in Group 3 compared to Group 2 (P = 0.036). 3) Perineural invasion (PNI) - (Supplementary file 5). The incidence of PNI in the tumour was significantly higher in Group 3 as compared to Groups 1 (P < 0.001) and 2 (P = 0.039). Publication bias The funnel plots for the overall survival comparisons, as well as the comparison of pathological factors LNP, LVI and PNI are HPB 2017, -, 1–7
shown in Supplementary File 6. Some degree of asymmetry was noted in all the analysed outcomes/variables. When the degree of heterogeneity was investigated with Egger’s test, however, the only significant differences were between Groups 1 and 2 for LVI (P = 0.042), and Group 2 and 3 for PNI (P = 0.004).
Discussion This meta-analysis demonstrates that intraoperative FS-driven revision of the transection margins does not improve overall survival for patients undergoing PD for PDAC. There was no difference in overall survival between those patients in groups 2 and 3 as well as the incidence of lymph node metastases (P = 0.590 and P = 0.410) suggesting biologically a more advanced disease. Interestingly, though, the incidence of LVI and PNI being detected in the tumour were significantly lower in Group 2 as compared to Group 3 (P = 0.036 and P = 0.039). Patients in Group 1 who had an R0 margin on both intraoperative FS and final histopathology had a significantly better overall survival when compared with those in whom a revision based on intraoperative FS resulted in a final R0 transection margin on PS (Group 2) as well as those in whom it did not (Group 3). The incidence of lymph node metastases was also significantly lower in Group 1 as compared to Groups 2 (P < 0.001) and 3 (P = 0.003).
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
5
A margin negative resection is considered the surgeon’s best contribution to improving long-term survival in pancreatic cancer.2 However, a critical pathological audit of surgically resected specimens indicated that 76% (84 out of 111) pancreatic head cancer resections are R1, especially along the SMA margin.22 Improved methods of pathological reporting of margins have resulted in an increase in R1 rates.23,24 Rather than having this information after the completion of the operation, FS of the transection margin allows a determination of the margin status when the surgeon has the opportunity to revise the margin to ensure a higher likelihood of an R0 resection. While this strategy is attractive, there is conflicting evidence about whether revision of a positive margin on FS can improve overall survival.5,6,14 The findings of the current study are probably best explained by an R1 pancreatic neck margin being a marker of a more biologically aggressive cancer. In this setting, any benefit gained by revision of the neck margin based on FS appears to be overridden by factors associated with the risk of systemic metastases. The absence of significant difference in survival as well as lymph node metastases between Groups 2 and 3 of this metaanalysis strongly support previous evidence that lymph node metastases are a sign of advanced/aggressive disease associated with reduced overall survival.25,26 The presence of LVI and PNI have been shown to be associated with a poor survival in pancreatic cancer.27 This was also noted in studies included in the present analysis.6,7 However, the meta-analysis of the combined data from the studies, failed to translate into a survival benefit between groups 2 and 3 in terms of PNI and LVI. The presence of significant heterogeneity between the studies (Supplementary file 6) could be responsible for this observation. This study only looked at the status of the pancreatic neck and bile duct margin, which are not the most common R1 margins with PD for PDAC.22 The direction of this analysis was guided by the availability of published data – all studies clearly reporting data only based on revision of these two margins. The most commonly involved are the posteromedial (SMA),28,29 posterior circumferential and SMV margins.30 It is not know whether FSdriven revision of these other margins would result in improved overall survival. Assessment of SMA/posteromedial margin with FS analysis would be most important given the high margin positivity at this margin. However, it must be appreciated that unlike the pancreatic neck margin where further resection is very feasible, the SMA margin is more difficult to revise if the dissection has been done optimally.28 The heterogeneity in the existing literature further precludes the deduction of firm conclusions on the role of periadventitial dissection to skeletonise the right lateral margin of the SMA in terms of improving the margin negative resection rates.29 The absence of uniform reporting of perioperative outcomes, the use of adjuvant chemotherapy and the stage of disease in the included studies limits the ability to assess the implications of
HPB 2017, -, 1–7
these factors on survival in the group of patients analysed in the meta-analysis. The findings of this meta-analysis, further strengthen the argument favouring the use of complementary strategies (such as neoadjuvant therapy) to surgery when dealing with resectable tumours that possess features compatible with biological aggressiveness (lymph node positive disease, LVI, PNI). While it is less likely to be able to predict LVI and PNI on preoperative imaging, patients with tumour size more than 2 cm31 with or without enlarged and suspicious peripancreatic lymphadenopathy32 may be considered candidates for neoadjuvant therapy in an attempt to address the likelihood of disseminated cancer cells,33 as well as help achieve better R0 resection margins.34,35 In recent years, the rising spectre of borderline resectable disease has also created interest in strategies to reduce incidence of R1 margins. The role of neoadjuvant therapy for all resectable tumours36 is one potential strategy that might achieve this. It must be appreciated though that the more widespread use of neoadjuvant therapy has created difficulties in accurately restaging pancreatic cancers, as peri-tumoural tissue could be fibrosis secondary to treatment, peri-tumoral inflammation or tumour extension.37 This means that performing intraoperative FS is likely to become more important with the widespread use of neoadjuvant therapy and the difficulty in determining whether a margin is R0 or R1 on staging modalities. An alternative to surgical revision, based on the FS result, might be ‘margin accentuation’ by irreversible electroporation (IRE).38 This technique, though not applicable to the neck and bile duct margins of transection, may have a role in accentuating the SMA and SMV/PV margins. In a cohort of 50 patients with locally advanced pancreatic cancer, intraoperative IRE was successfully performed to treat positive SMA, celiac and/or aortic margins immediately prior to resection and under the same anaesthetic. These patients had a median overall survival of 23 months (range: 8.3–36.3 months) compared with 18 months (range: 49–54.5 months) for those patients treated with IRE alone for unresectable disease.38 Whether IRE is useful in patients with an R1 margin on FS requires further investigation.
Conclusion Revision of an R1 pancreatic resection margin based on intraoperative FS report does not improve overall survival, even when the revision results in an R0 margin. This is probably because these patients are more likely to have a biologically aggressive tumour, as evidenced by an increased risk of lymph node metastases (compared to those with an R0 margin on intraoperative FS) similar to patients who ultimately have a positive/R1 margin on final histopathological section. Acknowledgement Nil.
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
6
version 1.0.0, 24 September 2014. Available from: http://www.
Study support
riskofbias.info.
Guarantor of the article: SGB & JW Financial support: Nil Potential competing interests: None
10. Demets DL. (1987 Apr-May) Methods for combining randomized clinical trials: strengths and limitations. Stat Med 6:341–350. 11. DerSimonian R, Laird N. (1986 Sep) Meta-analysis in clinical trials. Control Clin Trials 7:177–188. 12. Higgins JP, Thompson SG. (2002 Jun 15) Quantifying heterogeneity in a
Ethics committee approval
meta-analysis. Stat Med 21:1539–1558.
Not applicable.
13. Egger M, Davey Smith G, Schneider M, Minder C. (1997 Sep 13) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629 –634.
Role of funding source
14. Mathur A, Ross SB, Luberice K, Kurian T, Vice M, Toomey P et al. (2014
Not applicable.
Apr) Margin status impacts survival after pancreaticoduodenectomy but negative margins should not be pursued. Am Surg 80:353–360.
Specific author contributions SGB: Conceptualization of the study; Data collection, drafting of the
15. Lad NL, Squires MH, Maithel SK, Fisher SB, Mehta VV, Cardona K et al. (2013 Oct) Is it time to stop checking frozen section neck margins
manuscript. SP: analysis and interpretation; drafting the manuscript; critical revision for
during pancreaticoduodenectomy? Ann Surg Oncol 20:3626–3633. 16. Hernandez J, Mullinax J, Clark W, Toomey P, Villadolid D, Morton C
important intellectual content. NI: Data collection, analysis and interpretation.
et al. (2009 Jul) Survival after pancreaticoduodenectomy is not
JW: Study design; Data interpretation; critical revision for important intellec-
improved by extending resections to achieve negative margins. Ann Surg 250:76–80.
tual content. SGB, SP, NI & JW approve the final draft submitted.
17. Dillhoff M, Yates R, Wall K, Muscarella P, Melvin WS, Ellison EC et al. (2009 May) Intraoperative assessment of pancreatic neck margin at the time of pancreaticoduodenectomy increases likelihood of margin-
Conflict of interests None to declare.
negative resection in patients with pancreatic cancer. J Gastrointest Surg 13:825–830. 18. Ishii M, Kimura Y, Sugita S, Imamura M, Ito T, Nobuoka T et al. (2015
References 1. Barugola G. (2009) Resectable pancreatic cancer: who really benefits from resection? Ann Surg Oncol 16:3316–3322. 2. Howard TJ, Krug JE, Yu J, Zyromski NJ, Schmidt CM, Jacobson LE
Nov-Dec) Surgical and oncological impact of main pancreatic duct spread in invasive ductal adenocarcinoma: a clinicopathological study of 184 resected cases. Pancreatology 15:681–687.
et al. (2006 Dec) A margin-negative R0 resection accomplished with
19. Bender O, Bozkurt S, Buyukpinarbasili N, Malya FU, Battal M,
minimal postoperative complications is the surgeon’s contribution to
Karatepe O. (2015 Sep-Oct) Clinical significance of intraoperative frozen
long-term survival in pancreatic cancer. J Gastrointest Surg 10:
section analysis of pancreatic cancer surgical margin at the time of pancreaticoduodenectomy. Chirurgia (Bucur) 110:446–450.
1338–1345. discussion 45–6. 3. Barreto SG, Windsor JA. (2016 Mar) Justifying vein resection with pancreatoduodenectomy. Lancet Oncol 17:e118–e124. 4. Giovinazzo F, Turri G, Katz MH, Heaton N, Ahmed I. (2016 Feb) Metaanalysis of benefits of portal-superior mesenteric vein resection in
20. Nelson DW, Blanchard TH, Causey MW, Homann JF, Brown TA. (2013 May) Examining the accuracy and clinical usefulness of intraoperative frozen section analysis in the management of pancreatic lesions. Am J Surg 205:613–617. discussion 7.
pancreatic resection for ductal adenocarcinoma. Br J Surg 103:179–191.
21. Fatima J, Schnelldorfer T, Barton J, Wood C, Wiste H, Smyrk T. (2010)
5. Nitschke P, Volk A, Welsch T, Hackl J, Reissfelder C, Rahbari M et al.
Pancreatoduodenectomy for ductal adenocarcinoma: implications of
(2016 Jun 8) Impact of intraoperative Re-resection to achieve R0 status on survival in patients with pancreatic cancer: a single-center experience with 483 patients. Ann Surg (Epub ahead of print). 6. Kooby DA, Lad NL, Squires, MH, 3rd, Maithel SK, Sarmiento JM,
positive margin on survival. Arch Surg 145:167–172. 22. Esposito I, Kleeff J, Bergmann F, Reiser C, Herpel E, Friess H et al. (2008 Jun) Most pancreatic cancer resections are R1 resections. Ann Surg Oncol 15:1651–1660.
Staley CA et al. (2014 Sep) Value of intraoperative neck margin analysis
23. Gomez-Mateo Mdel C, Sabater-Orti L, Ferrandez-Izquierdo A. (2014
during Whipple for pancreatic adenocarcinoma: a multicenter analysis
Sep 15) Pathology handling of pancreatoduodenectomy specimens:
of 1399 patients. Ann Surg 260:494–501. discussion -3. 7. Pang TC, Wilson O, Argueta MA, Hugh TJ, Chou A, Samra JS et al. (2014 Apr) Frozen section of the pancreatic neck margin in pancreatoduodenectomy for pancreatic adenocarcinoma is of limited utility. Pa-
approaches and controversies. World J Gastrointest Oncol 6:351–359. 24. Verbeke CS. (2008 Jun) Resection margins and R1 rates in pancreatic cancer–are we there yet? Histopathology 52:787–796. 25. Lim JE, Chien MW, Earle CC. (2003 Jan) Prognostic factors following curative resection for pancreatic adenocarcinoma: a population-based,
thology 46:188–192. 8. Moher D, Liberati A, Tetzlaff J, Altman DG. (2009 Jul 21) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA
linked database analysis of 396 patients. Ann Surg 237:74–85. 26. van Rijssen LB, Narwade P, van Huijgevoort NC, Tseng DS, van Santvoort HC, Molenaar IQ et al. (2016 Jul) Prognostic value of lymph
statement. PLoS Med 6:e1000097. 9. Sterne J, Higgins J, Reeves B. (29/02/2015) On behalf of the develop-
node metastases detected during surgical exploration for pancreatic or
ment group for ACROBAT-NRSI. In: A Cochrane risk of bias assessment
periampullary cancer: a systematic review and meta-analysis. HPB 18:
tool: for non-randomized studies of interventions (ACROBATNRSI),
559–566.
HPB 2017, -, 1–7
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
7
27. Chen JW, Bhandari M, Astill DS, Wilson TG, Kow L, Brooke-Smith M
34. McClaine RJ, Lowy AM, Sussman JJ, Schmulewitz N, Grisell DL,
et al. (2010 Mar) Predicting patient survival after pancreaticoduode-
Ahmad SA. (2010 Feb) Neoadjuvant therapy may lead to successful
nectomy for malignancy: histopathological criteria based on perineural
surgical resection and improved survival in patients with borderline
infiltration and lymphovascular invasion. HPB 12:101–108. 28. Sanjay P, Takaori K, Govil S, Shrikhande S, Windsor J. (2012) ’Artery-
resectable pancreatic cancer. HPB 12:73–79. 35. Pingpank JF, Hoffman JP, Ross EA, Cooper HS, Meropol NJ,
first’ approaches to pancreatoduodenectomy. Br J Surg 99:1027–1035.
Freedman G et al. (2001 Mar-Apr) Effect of preoperative chemo-
29. Butler JR, Ahmad SA, Katz MH, Cioffi JL, Zyromski NJ. (2016 Apr)
radiotherapy on surgical margin status of resected adenocarcinoma of
A systematic review of the role of periadventitial dissection of the superior mesenteric artery in affecting margin status after pancreatoduodenectomy for pancreatic adenocarcinoma. HPB 18:305–311. 30. Menon KV, Gomez D, Smith AM, Anthoney A, Verbeke CS. (2009 Feb)
the head of the pancreas. J Gastrointest Surg 5:121–130. 36. Crippa S, Reni M, Balzano G, Doglioni C, Falconi M. (2016) Justifying vein resection with pancreatoduodenectomy. Lancet Oncol 17:e177. 37. Ferrone CR, Marchegiani G, Hong TS, Ryan DP, Deshpande V,
Impact of margin status on survival following pancreatoduodenec-
McDonnell EI et al. (2015 Jan) Radiological and surgical implications of
tomy for cancer: the Leeds Pathology Protocol (LEEPP). HPB 11:
neoadjuvant treatment with FOLFIRINOX for locally advanced and
18 – 24.
borderline resectable pancreatic cancer. Ann Surg 261:12–17.
31. Marchegiani G, Andrianello S, Malleo G, De Gregorio L, Scarpa A, Mino-
38. Martin R, Kwon D, Chalikonda S, Sellers M, Kotz E, Scoggins C et al.
Kenudson M et al. (2016 Jun 17) Does size matter in pancreatic
(2015) Treatment of 200 locally advanced (stage III) pancreatic adeno-
Cancer?: reappraisal of tumour dimension as a predictor of outcome
carcinoma patients with irreversible electroporation: safety and efficacy.
beyond the TNM. Ann Surg (Epub ahead of print).
Ann Surg 262:486–494.
32. Shrikhande S, Barreto S, Goel M, Arya S. (2012 Oct) Multimodality imaging of pancreatic ductal adenocarcinoma: a review of the liter-
Appendix A. Supplementary data
ature. HPB 14(10):658 –668.
Supplementary data related to this article can be found at http://dx.doi.org/
33. Sohal DP, Walsh RM, Ramanathan RK, Khorana AA. (2014 Mar)
10.1016/j.hpb.2017.03.006.
Pancreatic adenocarcinoma: treating a systemic disease with systemic therapy. J Natl Cancer Inst 106. dju011.
HPB 2017, -, 1–7
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
http://dx.doi.org/10.1016/j.hpb.2017.03.006
REVIEW ARTICLE
Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis Savio G. Barreto1,2, Sanjay Pandanaboyana3,4, Natasha Ironside3 & John A. Windsor3,4 1
Hepatobiliary and Oesophagogastric Unit, Division of Surgery and Perioperative Medicine, Flinders Medical Centre, Bedford Park, Adelaide, 2School of Medicine, Faculty of Medicine, Nursing and Health Sciences, Flinders University, South Australia, Australia, 3 Department of Surgery, Faculty of Medical and Health Sciences, University of Auckland, and 4HBP/Upper GI Unit, Department of General Surgery, Auckland City Hospital, Auckland, New Zealand
Abstract Background: Margin status is the main surgical determinant of long-term outcome in pancreatic cancer. Intraoperative frozen section (IOFS) detects microscopic positive margins at a stage when margin revision is possible. The aim of this study was to determine if IOFS driven-revision of pancreatic resection margin(s) improves overall survival (OS) in pancreatic cancer. Methods: A systematic review of major reference databases was undertaken. Patients were divided into 3 groups based on initial FS (FSR0 for negative margin and FSR1 for positive microscopic margin) and final Permanent Section report (PSR0 for negative margin and PSR1 for positive microscopic margin): Group 1 (FSR0 / PSR0), Group 2 (FSR1 / PSR0), and Group 3 (FSR1 / PSR1). Patients in Groups 2 and 3 had surgical revision of the FSR1 margin. Data was meta-analysed. Results: 4 studies included in the final analysis. No difference in OS and incidence of lymph node metastases between Groups 2 and 3 (P = 0.590 and P = 0.410). Conclusions: IOFS-based revision of R1 pancreatic resection margin does not improve OS, even when it results in an R0 margin. This suggests that any benefit of margin revision based on FS is over-ridden by markers of more advanced or aggressive disease. Received 22 November 2016; accepted 1 March 2017
Correspondence Savio G. Barreto, Department of Surgery, Flinders Medical Centre, Bedford Park, South Australia, 5042, Australia. E-mails: [email protected], [email protected]
Introduction Potentially curative surgery is only possible for a fifth of patients with pancreatic ductal adenocarcinoma (PDAC) and the majority of these patients die of systemic disease.1 Besides nodal status, the status of the resection margins on final histopathological analysis are an important determinant of outcome after pancreatoduodenectomy (PD).2 In order to achieve a negative resection margin (R0) many surgeons advocate more radical surgery including synchronous venous resection, arterial resection and/or multi-visceral resection of adjacent organs.3 Margin status has assumed greater importance with the definition of ‘borderline’ resectable PDAC.3 There remains some debate about the impact of a microscopic positive margin on overall survival.4
HPB 2017, -, 1–7
The surgeon cannot modify the biological aggressiveness of the PDAC or the fitness of the patient, but does have some control over resection margin status, and as such it is the main surgical determinant of long-term outcome. Given the importance of a negative resection margin on the outcomes of surgery for pancreatic cancer, there has been interest in using intraoperative frozen section (FS) of the transection margin to drive surgical revision of the margin in order to achieve an R0 resection. A recent study5 concluded that re-resecting a positive intraoperative FS transection margin to achieve an R0 resection improved long-term survival. In contrast, some earlier studies found no survival benefit.6,7
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
2
The aim of this systematic review and meta-analysis was to determine if intraoperative FS driven revision of the transection (neck and bile duct) margin of resection for pancreatic cancer improves overall survival.
Methodology Protocol registration The review protocol used in the present study has been registered at www.researchregistry.com with the Unique Identification number researchregistry159. Literature search A systematic review of major reference databases (MEDLINE, EMBASE, PubMed, and the Cochrane Library) was undertaken according to the PRISMA guidelines8 using a combination of text words “intraoperative”, “frozen section”, “pancreatic cancer”, “pancreatoduodenectomy”, “surgery”, “pancreaticoduodenectomy”. The search was restricted to human studies published up to August 2016 with no time or language restriction. Data from the articles were entered into a database and duplicates were removed. The abstracts were then screened for relevance. Subsequently, the reference lists of relevant trials, reviews, and international guidelines were hand-searched. Reference lists of the retrieved literature were cross-searched manually for additional publications. All available publications describing FS during surgery for pancreatic cancer were evaluated. The inclusion criteria were all types of studies (randomized, prospective observational, and retrospective observational) that investigated the role of intraoperative FS in deciding the margin of resection for pancreatic carcinoma and its impact on the pathological margin and overall survival. The articles that did not report on these two outcomes were excluded. Review articles, letters to the editor, comments and editorials were also excluded. Study groups For the analysis, the patient cohort was divided into 3 groups based on the intraoperative Frozen Section (FS) and final Permanent Section (PS) analysis. Group 1 (FSR0 / PSR0). Patients with a negative margin (R0) on frozen section (FS) that was confirmed on final histopathology (PS). Group 2 (FSR1 / PSR0). Patients with a positive margin (R1) on frozen section (FS) that was a negative margin (R0) confirmed on final histopathology (PS) after revision of margin. Group 3 (FSR1 / PSR1): Patients with a positive margin (R1) on frozen section (FS) margin and a positive margin (R1) on final histopathology (PS) after revision of margin. Outcome measure The primary outcome measure was overall survival expressed as the median difference between the three groups.
HPB 2017, -, 1–7
Pathological factors The pathological variables compared between the groups were lymph node positivity (LNP), lymphovascular (LVI) and perineural (PNI) invasion. Data extraction and quality assessment Studies were identified, and data were extracted by 2 authors (S.G.B., N.I.) independently. A third author (S.P.) further reviewed the selected studies before data extraction in accordance with Cochrane group recommendations for study selection for meta-analysis. Assessment of methodological quality of included studies The ACROBAT-NRSI grading method9 was used to assess the quality of the included non-randomised studies with studies being graded at low, moderate, serious or critical risk of bias (Supplementary file 1). Statistical analysis Statistical analysis was performed using Review Manager Version 5.2 software (Cochrane Collaboration). The odds ratio (OR) with 95% confidence interval (CI) was calculated for categorical data. Hazard ratios were used to analyse survival data. Random and fixed-effects models were used to calculate the combined outcomes of both binary and continuous data.10,11 Publication bias Heterogeneity was assessed using the Cochrane Q test, with P < 0.050 indicating significant heterogeneity. The I2 test, which describes the proportion of variation that is due to heterogeneity beyond chance, had the significance set at >33%. In the absence of heterogeneity, studies were reported using a fixed effects metaanalysis. In the presence of heterogeneity, a cause was sought. If none could be found, the results were described using a randomeffects meta-analysis.12 Potential publication bias was assessed using funnel plots. Egger’s test was used to assess funnel plot asymmetry, by calculating a linear regression of the natural logarithmic scale of the RR over the inverse of the standard error.13 Two tailed P < 0.050 was considered to indicate statistical significance.
Results The results of the search criteria are shown in Fig. 1.5–7,14–21 Table 1 summarizes the 4 studies that included 2385 patients. All 4 studies were retrospective in design and originated from single institutions in the United States,5,7,14 with the exception of a solitary multi-institutional study.6 All the studies were published between 2014 and 2016. All of the studies were from high volume centres, defined as average >20 PDs per year. A minority of patients in the multicentre study received neoadjuvant treatment (10%).6 No mention was made
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
IdenƟficaƟon
3
Records idenƟfied through database searching (n = 291 )
AddiƟonal records idenƟfied through other sources (n = 5)
Included
Eligibility
Screening
Records aŌer duplicates removed (n = 6)
Records screened (n = 290)
Full-text arƟcles assessed for eligibility (n = 19)
Records excluded (n = 279) Full-text arƟcles excluded, with reasons (n = 15) 8 - reporƟng frozen secƟon in Intraductal papillary mucinous neoplasms 5 - failed inclusion criteria aŌer reading full text 2 - same paƟent cohort included in a later updated publicaƟon / mulƟcentre publicaƟon
Studies included in qualitaƟve synthesis (n = 4)
Studies included in quanƟtaƟve synthesis (meta-analysis) (n = 4) Figure 1 PRISMA flow diagram
of patients with borderline resectable disease in any of the 4 studies. The reasons for performing intraoperative FS (either routine practice or surgeon’s discretion) were specified in 2 of the 4 studies.7,14 The margins subjected to FS were pancreatic neck transection and portal vein margin in 1 study5 and pancreatic neck and bile duct/hepatic duct transection margin in the other 4 studies.6,7,14,15 However, only 16 of the 5 studies only published their data from FS of the pancreatic neck transection margin. In 3 studies, an R0 margin was defined as the absence of macroscopic and microscopic tumour at the resection margin.5,6,14 In the study by Pang and colleagues7 a clear margin of a minimum of HPB 2017, -, 1–7
1 mm was considered to constitute an R0 resection. The manner in which a positive FS margin was handled (i.e. extent of revision and number of revision attempts) was clearly specified in 2 studies.6,14 While at least 2 attempts at revision were undertaken in one study,14 an attempt at revising the transection margin by 1–2 cm was attempted in the other study.6 Overall survival All 4 studies were included in the analysis. Table 2 summarizes the primary outcome and pathological variables between the 3 groups. Patients in Group 1 had a significantly better survival compared with Group 2 (HR = 0.68; 95% CI: 0.53–0.87;
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
4
Table 1 Summary of Studies Table included in the analysis [Abbreviations: PDAC – Pancreatic ductal adenocarcinoma, S – single centre;
M – Multicentre; FS – intraoperative frozen section; PS – final histopathology; CI – Confidence intervals (when reported); NR – Not reached] Author (Ref)
Group 1 (FSR0 / PSR0)
Group 2 (FSR1 / PSR0)
Group 3 (FSR1 / PSR1)
Number of patients (%)
Number of patients (%)
Number of patients (%)
Median Survival in months (95% CI)
Median Survival in months (95% CI)
Median Survival in months (95% CI)
Pang et al.7
42 (36)
29 (18–NR)
7 (6)
16 (5–28)
67 (58)
23 (16–31)
Mathur et al.14
298 (67)
22
40 (9)
22
110 (24)
21
Kooby et al.6
1196 (86)
21
72 (5)
12
131 (9)
14
Nitschke et al.5
317 (75)
29
32 (8)
36
73 (17)
12
N [ 2385
1853
151
381
Table 2 Table comparing the primary outcome measure and pathological factors between the three groups (Group 1: FSR0 / PSR0;
Group 2: FSR1 / PSR0; Group 3: FSR1 / PSR1) (Abbreviations: CI – confidence intervals; OR – odds ratio, HR – hazard ratio) Primary Outcome
Groups 1v2 (HR)
95%CI
P
I2 (%)
Groups 1v3 (HR)
95%CI
P
I2 (%)
Groups 2v3 (HR)
95% CI
P
I2 (%)
Overall survival
0.68
0.53, 0.87
0.002
99%
0.72
0.64, 0.80
<0.001
99%
1.08
0.81, 1.46
0.590
99%
2
2
Pathological Factors
Groups 1v2 (OR)
95%CI
P
I (%)
Groups 1v3 (OR)
95%CI
P
I (%)
Groups 2v3 (OR)
95% CI
P
I2 (%)
Lymphovascular invasion
1.02
0.69, 1.50
0.939
0%
0.54
0.27, 1.05
0.069
28%
0.61
0.39, 0.97
0.036
77%
Perineural invasion
0.69
0.44, 1.08
0.104
0%
0.39
0.28, 0.56
<0.001
27%
0.57
0.33, 0.99
0.039
0%
Lymph node involvement
0.53
0.38, 0.72
<0.001
0%
0.41
0.23, 0.74
0.003
70%
0.83
0.54, 1.29
0.410
0%
P = 0.002) and Group 3 (HR = 0.72; 95% CI: 0.64–0.80; P < 0.001). Overall survival was not different between Groups 2 and 3 (P = 0.590) (Supplementary file 2). Pathological factors The outcomes of all three pathological factors have been summarised in Table 2. 1) Lymph node positivity (LNP) - (Supplementary file 3) Patients in Group 1 had a significantly lower LNP rate compared to Group 2 (P < 0.001) and to Group 3 (P = 0.003). 2) Lymphovascular invasion (LVI) - (Supplementary file 4) The incidence of LVI in the tumour was significantly higher in Group 3 compared to Group 2 (P = 0.036). 3) Perineural invasion (PNI) - (Supplementary file 5). The incidence of PNI in the tumour was significantly higher in Group 3 as compared to Groups 1 (P < 0.001) and 2 (P = 0.039). Publication bias The funnel plots for the overall survival comparisons, as well as the comparison of pathological factors LNP, LVI and PNI are HPB 2017, -, 1–7
shown in Supplementary File 6. Some degree of asymmetry was noted in all the analysed outcomes/variables. When the degree of heterogeneity was investigated with Egger’s test, however, the only significant differences were between Groups 1 and 2 for LVI (P = 0.042), and Group 2 and 3 for PNI (P = 0.004).
Discussion This meta-analysis demonstrates that intraoperative FS-driven revision of the transection margins does not improve overall survival for patients undergoing PD for PDAC. There was no difference in overall survival between those patients in groups 2 and 3 as well as the incidence of lymph node metastases (P = 0.590 and P = 0.410) suggesting biologically a more advanced disease. Interestingly, though, the incidence of LVI and PNI being detected in the tumour were significantly lower in Group 2 as compared to Group 3 (P = 0.036 and P = 0.039). Patients in Group 1 who had an R0 margin on both intraoperative FS and final histopathology had a significantly better overall survival when compared with those in whom a revision based on intraoperative FS resulted in a final R0 transection margin on PS (Group 2) as well as those in whom it did not (Group 3). The incidence of lymph node metastases was also significantly lower in Group 1 as compared to Groups 2 (P < 0.001) and 3 (P = 0.003).
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
5
A margin negative resection is considered the surgeon’s best contribution to improving long-term survival in pancreatic cancer.2 However, a critical pathological audit of surgically resected specimens indicated that 76% (84 out of 111) pancreatic head cancer resections are R1, especially along the SMA margin.22 Improved methods of pathological reporting of margins have resulted in an increase in R1 rates.23,24 Rather than having this information after the completion of the operation, FS of the transection margin allows a determination of the margin status when the surgeon has the opportunity to revise the margin to ensure a higher likelihood of an R0 resection. While this strategy is attractive, there is conflicting evidence about whether revision of a positive margin on FS can improve overall survival.5,6,14 The findings of the current study are probably best explained by an R1 pancreatic neck margin being a marker of a more biologically aggressive cancer. In this setting, any benefit gained by revision of the neck margin based on FS appears to be overridden by factors associated with the risk of systemic metastases. The absence of significant difference in survival as well as lymph node metastases between Groups 2 and 3 of this metaanalysis strongly support previous evidence that lymph node metastases are a sign of advanced/aggressive disease associated with reduced overall survival.25,26 The presence of LVI and PNI have been shown to be associated with a poor survival in pancreatic cancer.27 This was also noted in studies included in the present analysis.6,7 However, the meta-analysis of the combined data from the studies, failed to translate into a survival benefit between groups 2 and 3 in terms of PNI and LVI. The presence of significant heterogeneity between the studies (Supplementary file 6) could be responsible for this observation. This study only looked at the status of the pancreatic neck and bile duct margin, which are not the most common R1 margins with PD for PDAC.22 The direction of this analysis was guided by the availability of published data – all studies clearly reporting data only based on revision of these two margins. The most commonly involved are the posteromedial (SMA),28,29 posterior circumferential and SMV margins.30 It is not know whether FSdriven revision of these other margins would result in improved overall survival. Assessment of SMA/posteromedial margin with FS analysis would be most important given the high margin positivity at this margin. However, it must be appreciated that unlike the pancreatic neck margin where further resection is very feasible, the SMA margin is more difficult to revise if the dissection has been done optimally.28 The heterogeneity in the existing literature further precludes the deduction of firm conclusions on the role of periadventitial dissection to skeletonise the right lateral margin of the SMA in terms of improving the margin negative resection rates.29 The absence of uniform reporting of perioperative outcomes, the use of adjuvant chemotherapy and the stage of disease in the included studies limits the ability to assess the implications of
HPB 2017, -, 1–7
these factors on survival in the group of patients analysed in the meta-analysis. The findings of this meta-analysis, further strengthen the argument favouring the use of complementary strategies (such as neoadjuvant therapy) to surgery when dealing with resectable tumours that possess features compatible with biological aggressiveness (lymph node positive disease, LVI, PNI). While it is less likely to be able to predict LVI and PNI on preoperative imaging, patients with tumour size more than 2 cm31 with or without enlarged and suspicious peripancreatic lymphadenopathy32 may be considered candidates for neoadjuvant therapy in an attempt to address the likelihood of disseminated cancer cells,33 as well as help achieve better R0 resection margins.34,35 In recent years, the rising spectre of borderline resectable disease has also created interest in strategies to reduce incidence of R1 margins. The role of neoadjuvant therapy for all resectable tumours36 is one potential strategy that might achieve this. It must be appreciated though that the more widespread use of neoadjuvant therapy has created difficulties in accurately restaging pancreatic cancers, as peri-tumoural tissue could be fibrosis secondary to treatment, peri-tumoral inflammation or tumour extension.37 This means that performing intraoperative FS is likely to become more important with the widespread use of neoadjuvant therapy and the difficulty in determining whether a margin is R0 or R1 on staging modalities. An alternative to surgical revision, based on the FS result, might be ‘margin accentuation’ by irreversible electroporation (IRE).38 This technique, though not applicable to the neck and bile duct margins of transection, may have a role in accentuating the SMA and SMV/PV margins. In a cohort of 50 patients with locally advanced pancreatic cancer, intraoperative IRE was successfully performed to treat positive SMA, celiac and/or aortic margins immediately prior to resection and under the same anaesthetic. These patients had a median overall survival of 23 months (range: 8.3–36.3 months) compared with 18 months (range: 49–54.5 months) for those patients treated with IRE alone for unresectable disease.38 Whether IRE is useful in patients with an R1 margin on FS requires further investigation.
Conclusion Revision of an R1 pancreatic resection margin based on intraoperative FS report does not improve overall survival, even when the revision results in an R0 margin. This is probably because these patients are more likely to have a biologically aggressive tumour, as evidenced by an increased risk of lymph node metastases (compared to those with an R0 margin on intraoperative FS) similar to patients who ultimately have a positive/R1 margin on final histopathological section. Acknowledgement Nil.
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
6
version 1.0.0, 24 September 2014. Available from: http://www.
Study support
riskofbias.info.
Guarantor of the article: SGB & JW Financial support: Nil Potential competing interests: None
10. Demets DL. (1987 Apr-May) Methods for combining randomized clinical trials: strengths and limitations. Stat Med 6:341–350. 11. DerSimonian R, Laird N. (1986 Sep) Meta-analysis in clinical trials. Control Clin Trials 7:177–188. 12. Higgins JP, Thompson SG. (2002 Jun 15) Quantifying heterogeneity in a
Ethics committee approval
meta-analysis. Stat Med 21:1539–1558.
Not applicable.
13. Egger M, Davey Smith G, Schneider M, Minder C. (1997 Sep 13) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629 –634.
Role of funding source
14. Mathur A, Ross SB, Luberice K, Kurian T, Vice M, Toomey P et al. (2014
Not applicable.
Apr) Margin status impacts survival after pancreaticoduodenectomy but negative margins should not be pursued. Am Surg 80:353–360.
Specific author contributions SGB: Conceptualization of the study; Data collection, drafting of the
15. Lad NL, Squires MH, Maithel SK, Fisher SB, Mehta VV, Cardona K et al. (2013 Oct) Is it time to stop checking frozen section neck margins
manuscript. SP: analysis and interpretation; drafting the manuscript; critical revision for
during pancreaticoduodenectomy? Ann Surg Oncol 20:3626–3633. 16. Hernandez J, Mullinax J, Clark W, Toomey P, Villadolid D, Morton C
important intellectual content. NI: Data collection, analysis and interpretation.
et al. (2009 Jul) Survival after pancreaticoduodenectomy is not
JW: Study design; Data interpretation; critical revision for important intellec-
improved by extending resections to achieve negative margins. Ann Surg 250:76–80.
tual content. SGB, SP, NI & JW approve the final draft submitted.
17. Dillhoff M, Yates R, Wall K, Muscarella P, Melvin WS, Ellison EC et al. (2009 May) Intraoperative assessment of pancreatic neck margin at the time of pancreaticoduodenectomy increases likelihood of margin-
Conflict of interests None to declare.
negative resection in patients with pancreatic cancer. J Gastrointest Surg 13:825–830. 18. Ishii M, Kimura Y, Sugita S, Imamura M, Ito T, Nobuoka T et al. (2015
References 1. Barugola G. (2009) Resectable pancreatic cancer: who really benefits from resection? Ann Surg Oncol 16:3316–3322. 2. Howard TJ, Krug JE, Yu J, Zyromski NJ, Schmidt CM, Jacobson LE
Nov-Dec) Surgical and oncological impact of main pancreatic duct spread in invasive ductal adenocarcinoma: a clinicopathological study of 184 resected cases. Pancreatology 15:681–687.
et al. (2006 Dec) A margin-negative R0 resection accomplished with
19. Bender O, Bozkurt S, Buyukpinarbasili N, Malya FU, Battal M,
minimal postoperative complications is the surgeon’s contribution to
Karatepe O. (2015 Sep-Oct) Clinical significance of intraoperative frozen
long-term survival in pancreatic cancer. J Gastrointest Surg 10:
section analysis of pancreatic cancer surgical margin at the time of pancreaticoduodenectomy. Chirurgia (Bucur) 110:446–450.
1338–1345. discussion 45–6. 3. Barreto SG, Windsor JA. (2016 Mar) Justifying vein resection with pancreatoduodenectomy. Lancet Oncol 17:e118–e124. 4. Giovinazzo F, Turri G, Katz MH, Heaton N, Ahmed I. (2016 Feb) Metaanalysis of benefits of portal-superior mesenteric vein resection in
20. Nelson DW, Blanchard TH, Causey MW, Homann JF, Brown TA. (2013 May) Examining the accuracy and clinical usefulness of intraoperative frozen section analysis in the management of pancreatic lesions. Am J Surg 205:613–617. discussion 7.
pancreatic resection for ductal adenocarcinoma. Br J Surg 103:179–191.
21. Fatima J, Schnelldorfer T, Barton J, Wood C, Wiste H, Smyrk T. (2010)
5. Nitschke P, Volk A, Welsch T, Hackl J, Reissfelder C, Rahbari M et al.
Pancreatoduodenectomy for ductal adenocarcinoma: implications of
(2016 Jun 8) Impact of intraoperative Re-resection to achieve R0 status on survival in patients with pancreatic cancer: a single-center experience with 483 patients. Ann Surg (Epub ahead of print). 6. Kooby DA, Lad NL, Squires, MH, 3rd, Maithel SK, Sarmiento JM,
positive margin on survival. Arch Surg 145:167–172. 22. Esposito I, Kleeff J, Bergmann F, Reiser C, Herpel E, Friess H et al. (2008 Jun) Most pancreatic cancer resections are R1 resections. Ann Surg Oncol 15:1651–1660.
Staley CA et al. (2014 Sep) Value of intraoperative neck margin analysis
23. Gomez-Mateo Mdel C, Sabater-Orti L, Ferrandez-Izquierdo A. (2014
during Whipple for pancreatic adenocarcinoma: a multicenter analysis
Sep 15) Pathology handling of pancreatoduodenectomy specimens:
of 1399 patients. Ann Surg 260:494–501. discussion -3. 7. Pang TC, Wilson O, Argueta MA, Hugh TJ, Chou A, Samra JS et al. (2014 Apr) Frozen section of the pancreatic neck margin in pancreatoduodenectomy for pancreatic adenocarcinoma is of limited utility. Pa-
approaches and controversies. World J Gastrointest Oncol 6:351–359. 24. Verbeke CS. (2008 Jun) Resection margins and R1 rates in pancreatic cancer–are we there yet? Histopathology 52:787–796. 25. Lim JE, Chien MW, Earle CC. (2003 Jan) Prognostic factors following curative resection for pancreatic adenocarcinoma: a population-based,
thology 46:188–192. 8. Moher D, Liberati A, Tetzlaff J, Altman DG. (2009 Jul 21) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA
linked database analysis of 396 patients. Ann Surg 237:74–85. 26. van Rijssen LB, Narwade P, van Huijgevoort NC, Tseng DS, van Santvoort HC, Molenaar IQ et al. (2016 Jul) Prognostic value of lymph
statement. PLoS Med 6:e1000097. 9. Sterne J, Higgins J, Reeves B. (29/02/2015) On behalf of the develop-
node metastases detected during surgical exploration for pancreatic or
ment group for ACROBAT-NRSI. In: A Cochrane risk of bias assessment
periampullary cancer: a systematic review and meta-analysis. HPB 18:
tool: for non-randomized studies of interventions (ACROBATNRSI),
559–566.
HPB 2017, -, 1–7
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006
HPB
7
27. Chen JW, Bhandari M, Astill DS, Wilson TG, Kow L, Brooke-Smith M
34. McClaine RJ, Lowy AM, Sussman JJ, Schmulewitz N, Grisell DL,
et al. (2010 Mar) Predicting patient survival after pancreaticoduode-
Ahmad SA. (2010 Feb) Neoadjuvant therapy may lead to successful
nectomy for malignancy: histopathological criteria based on perineural
surgical resection and improved survival in patients with borderline
infiltration and lymphovascular invasion. HPB 12:101–108. 28. Sanjay P, Takaori K, Govil S, Shrikhande S, Windsor J. (2012) ’Artery-
resectable pancreatic cancer. HPB 12:73–79. 35. Pingpank JF, Hoffman JP, Ross EA, Cooper HS, Meropol NJ,
first’ approaches to pancreatoduodenectomy. Br J Surg 99:1027–1035.
Freedman G et al. (2001 Mar-Apr) Effect of preoperative chemo-
29. Butler JR, Ahmad SA, Katz MH, Cioffi JL, Zyromski NJ. (2016 Apr)
radiotherapy on surgical margin status of resected adenocarcinoma of
A systematic review of the role of periadventitial dissection of the superior mesenteric artery in affecting margin status after pancreatoduodenectomy for pancreatic adenocarcinoma. HPB 18:305–311. 30. Menon KV, Gomez D, Smith AM, Anthoney A, Verbeke CS. (2009 Feb)
the head of the pancreas. J Gastrointest Surg 5:121–130. 36. Crippa S, Reni M, Balzano G, Doglioni C, Falconi M. (2016) Justifying vein resection with pancreatoduodenectomy. Lancet Oncol 17:e177. 37. Ferrone CR, Marchegiani G, Hong TS, Ryan DP, Deshpande V,
Impact of margin status on survival following pancreatoduodenec-
McDonnell EI et al. (2015 Jan) Radiological and surgical implications of
tomy for cancer: the Leeds Pathology Protocol (LEEPP). HPB 11:
neoadjuvant treatment with FOLFIRINOX for locally advanced and
18 – 24.
borderline resectable pancreatic cancer. Ann Surg 261:12–17.
31. Marchegiani G, Andrianello S, Malleo G, De Gregorio L, Scarpa A, Mino-
38. Martin R, Kwon D, Chalikonda S, Sellers M, Kotz E, Scoggins C et al.
Kenudson M et al. (2016 Jun 17) Does size matter in pancreatic
(2015) Treatment of 200 locally advanced (stage III) pancreatic adeno-
Cancer?: reappraisal of tumour dimension as a predictor of outcome
carcinoma patients with irreversible electroporation: safety and efficacy.
beyond the TNM. Ann Surg (Epub ahead of print).
Ann Surg 262:486–494.
32. Shrikhande S, Barreto S, Goel M, Arya S. (2012 Oct) Multimodality imaging of pancreatic ductal adenocarcinoma: a review of the liter-
Appendix A. Supplementary data
ature. HPB 14(10):658 –668.
Supplementary data related to this article can be found at http://dx.doi.org/
33. Sohal DP, Walsh RM, Ramanathan RK, Khorana AA. (2014 Mar)
10.1016/j.hpb.2017.03.006.
Pancreatic adenocarcinoma: treating a systemic disease with systemic therapy. J Natl Cancer Inst 106. dju011.
HPB 2017, -, 1–7
© 2017 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Barreto SG, et al., Does revision of resection margins based on frozen section improve overall survival following pancreatoduodenectomy for pancreatic ductal adenocarcinoma? A meta-analysis, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2017.03.006