- Email: [email protected]
Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
PAN523_proof ■ 26 February 2016 ■ 1/8
Pancreatology xxx (2016) 1e8
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Review article
Q5
Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis
Q4
Neil Bhardwaj a, *, Ashley R. Dennison a, Guy J. Maddern b, Giuseppe Garcea a a b
Q1
University Hospitals Leicester, Department of HPB Surgery, Leicester, LE5 4PW, UK University of Adelaide, Discipline of Surgery, The Queen Elizabeth Hospital, Woodville, Adelaide, Australia
a r t i c l e i n f o
a b s t r a c t
Article history: Available online xxx
Introduction: IPMN is a relatively new clinical entity and surgeons are continuing to develop their understanding of this complex pathology. Little is known of the natural disease process post-resection of an IPMN, particularly the impact of gland histology and margin status on the chance of recurrence and survival in benign and invasive IPMN. Methods: An online search was conducted to evaluate and include those studies which reported on gland histology, margin status and disease recurrence in resected benign and malignant IPMN. A Meta analysis was then performed using a random effects model. Results: The chance of recurrence in non-invasive margin positive IPMN is similar to margin negative IPMN. The chance of recurrence is higher in invasive gland IPMN compared to non-invasive gland. The vast majority of recurrences occurred in patients with positive margins demonstrating invasion. Conclusion: All patients with intra- or post-operative evidence of invasive carcinoma at the resection margin should undergo further resection to achieve a negative margin. Patients with evidence of IPMN at the transaction margin (even with changes of high grade dysplasia/CIS) may not achieve any benefit from further resection. Patients with recurrence in benign/non-invasive IPMN should undergo re-resection, whereas patients with recurrence in invasive IPMN should not. Copyright © 2016, IAP and EPC. Published by Elsevier India, a division of Reed Elsevier India Pvt. Ltd. All rights reserved.
Keywords: Survival Resection IPMN Outcomes Resection margin
Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Definition of histology and margin status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of margin status on recurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of gland histology on recurrence and survival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
* Corresponding author. E-mail address: [email protected] (N. Bhardwaj). http://dx.doi.org/10.1016/j.pan.2016.02.008 1424-3903/Copyright © 2016, IAP and EPC. Published by Elsevier India, a division of Reed Elsevier India Pvt. Ltd. All rights reserved.
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 2/8
2
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
remnant transaction margin after a partial pancreatectomy for IPMN.
Introduction Intraductal papillary neoplasm (IPMN) of the pancreas was first described as a discreet entity in 1982 [1]. In 1996, the World Health Organisation revised pancreatic tumour classification and recognized IPMN as a separate pathological tumour to mucinous cystic tumours; these guidelines were recently updated in 2012 [2]. Morphologically, IPMNs are divided into three types; main-duct, branch-duct or combined, mainly based on imaging and/or histology. It is now well established that IPMNs exhibit a spectrum of neoplastic transformation similar to the adenoma-carcinoma sequence observed in colon cancer [3]. Histological grading of pancreatic cancer is graded from “hyperplasia” or adenoma (currently classified as “low-grade dysplasia”) to invasive carcinoma. The definition of “malignancy” has been variable, as there remains some debate as to whether to include “carcinoma in situ” (CIS) in the malignant category or whether that definition should be reserved for invasive carcinomas only. This makes it difficult to compare studies and may negatively impact a surgeon's ability to accurately predict patient prognosis. There has been increased understanding of IPMN pathology and behaviour with accepted guidelines and acknowledged risk factors. Non-invasive tumours have a favourable outcome compared to invasive tumours [4,5] and it would be reasonable to assume that IPMN (or even invasive carcinoma) at the pancreatic cut margin (on frozen section) would merit further resection to obtain clear margins. However, despite the recent advances in the understanding of IPMN, little is known regarding the potential risk of recurrence in patients with low grade, medium grade or high grade dysplasia at the pancreatic cut surface. There have also been reports of invasive carcinoma developing after resection of non-invasive IPMN [6e8]. The aim of this meta-analysis is to establish the future recurrence risk associated with resection of benign and malignant glands and in the various grades of dysplasia (low, medium and high) in the
Abstracts idenƟfied through database searching (n = 82 )
Methods Search strategy This was performed as per the guidelines issued in the PRISMA and MOOSE statement [9,10]. A systematic literature search of the Medline, Embase, Cochrane and Pubmed database was conducted over the last 20 years using the terms intraductal papillary mucinous neoplasm/IPMN/intraductal papillary mucinous tumour. The latest search was performed in 2014 and there were no language restrictions. Abstracts of all relevant clinical studies were reviewed and if relevant, the article was reviewed in detail and included in the analysis. Further articles were also identified from relevant reviews and cross-referenced with those already included to avoid duplication of studies (Fig. 1 e PRISMA flow diagram). Inclusion criteria All studies that reported on resection of primary IPMN, margin status, recurrence rate and detailed benign and malignant primary histological diagnoses were included. In addition, only studies reported in English with a minimum follow up of 2 years were included. There was insufficient detail in the studies examined to consistently record if the lesions resected were main-duct, branchduct or mixed type IPMN. Exclusion criteria Case reports and review series, in addition to all studies not reporting margin status, separate benign and malignant histology, follow up and recurrence rates were excluded. If more than one
AddiƟonal records idenƟfied through other sources (n = 10 )
Abstracts aŌer duplicates removed (n = 86)
Full text arƟcles screened (n = 50 )
Abstracts excluded as failure to meet criteria (n = 36)
Full-text arƟcles excluded as per exclusion criteria (n = 34 )
Studies included in qualitaƟve synthesis (n = 16 ) Fig. 1. PRISMA flow diagram.
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 3/8
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
study was reported from the same institute then the most recent publication was included. Definition of histology and margin status Resected IPMNs were either classified as malignant or benign and similarly transection margin status was classified into malignant (invasive carcinoma), non-invasive benign or negative. Benign tumours were further classified according to the latest guidelines into adenoma, also known as mild dysplasia, moderate dysplasia (previously known as borderline) or high grade dysplasia, often referred to as carcinoma-in-situ (CIS) [2]. However, according to the latest WHO guidelines of 2010, CIS should be abandoned as a term in favour of high grade dysplasia [11]. Some studies referred to Pancreatic intra-epithelial neoplasms (PanIN) when describing resection margins. PanINs are pathologically distinct entities to IPMNs and are recognised pre-cursers to pancreatic adenocarcinoma [12,13]. Although they are a distinct histological diagnosis, they do share certain common characteristics with IPMNs; being also a precursor for pancreatic adenocarcinoma. They are usually small (<5 mm) intraductal non-invasive lesions formed by metaplasia and proliferation of ductal epithelium. They are similarly graded into mild, moderate or severe dysplasia (PanIN-1, PanIN-2 and PanIN-3, respectively) [3]. For the purposes of this metaanalysis these lesions, when described, in the remnant margin were categorised mild, moderate, severe similar to their IPMN counterpart. Recurrence Recurrence was defined as pathological evidence of recurrence on repeat resection or radiological evidence of recurrence on surveillance imaging. All recurrences: at the resected margin, distant to the margin but in the gland and/or distant to the gland were grouped under one category. No recurrences occurred distant to the gland without evidence of intra-pancreatic recurrence.
3
further refinement of the data was required due to the small numbers and therefore histological status of the gland and the cut margin were grouped together and defined as either non-invasive (benign, low grade/adenoma/mild dysplasia, moderate/intermediate dysplasia, severe/high-grade dysplasia/CIS) or invasive (Table 2a and b). Impact of margin status on recurrence The chance of recurrence in patients with negative margins was 16.3% (95% CI: 13.7%e19.1%). The I-squared statistic indicated that only 30.7% of the variance in the estimate was due to heterogeneity across the studies (Table 3). The pooled estimate indicated that the proportion of recurrence in benign non-invasive margins was 13.4% (95% CI: 7.5%e20.8%). The I-squared statistic indicated that 59.7% of the variance in the estimate was due to heterogeneity across the studies (Table 4). The chance of recurrence in invasive resection margin was 84.1% (95% CI: 73.0%e92.6%). The I-squared statistic indicated that there was no variation in the estimates (Table 5). Impact of gland histology on recurrence and survival On comparing benign versus malignant gland histology, the pooled odds ratio for the overall analysis was 16.08 (95% CI: 8.76e29.54). This was significantly different from 1.00 (p < 0.0001), indicating that the odds of recurrence were significantly higher in the invasive gland group compared to the non-invasive group. Isquared statistic indicated that only 20.4% of the variation across studies was due to heterogeneity (Table 6, Fig. 2). The study from Paye et al. (2000) was removed from the analysis as the odds ratio estimated for this study was extreme (1323.0). It only made a small contribution to the pooled odds ratio (weight ¼ 3.64), and was influential in determining the magnitude of the I-square statistic. It is also evident that the 5-year survival, where reported, in patients with benign disease (approaches 100% in most studies) is greater than those with malignant disease (31%e80%) (Table 7).
Statistical analysis Discussion Histology measures were reported as pooled percentages. A sine transformation was applied to these data to avoid confidence intervals below zero and above 100. The data were back transformed prior to analysis. The comparison of invasive versus non-invasive gland was reported as a pooled odds ratio. This was calculated from count data as presented in the papers and expressed as the odds of recurrence in the invasive gland group relative to that in the non-invasive group. A random effects meta-analysis model was applied to obtain the pooled estimates and heterogeneity was assessed using the I-square statistic. All analyses were completed using STATA v12.1 (STATA Corp., LP, TX, USA). Results A total of 82 abstracts were identified, of which 50 full articles were retrieved and analysed in-depth. 6 articles were rejected due to duplicity of data and a further 28 articles were rejected as they failed to report margin status. A total of 16 articles were included in the final analysis [4,6e8,14e25] (Table 1 e all 16 studies). A total of 1488 patients were included in this study, of which there were 1115 (75%) benign lesions resected and 373 (25%) malignant IPMNs resected. A total of 263 recurrences (18%) were noted of which 198 (75%) had a negative resection margin. Of the 263 patients with recurrence, 161 (61%) had invasive IPMN on initial histology and 28 (11%) had invasive carcinoma at the margin. Non-invasive histology was evident in 31% and only 8% had completely benign histology. A
The management of IPMN is ever evolving and little is understood regarding the impact of residual non-invasive margin status on the risk of future recurrence. This study provides evidence that the vast majority of recurrences occur in patients whose margins were negative on resection, however a positive transection margin for invasive malignancy places patients at a high risk of developing recurrence. It also proves that the odds of disease recurrence is higher in invasive versus non-invasive excised IPMN lesions regardless of the margin status. This is not surprising and suggests that recurrences are likely the result of a missed lesion at the time of initial resection. This meta-analysis does lend support to the theory that IPMN demonstrates a field defect in the pancreas, with the entire gland at risk of developing neoplastic disease. Some authors have suggested performing pre-operative endoscopic retrograde cholangiopancreatography followed by intra-operative pancreatography in order to detect all potentially malignant lesions [23], this however is not routine practice in most centres. Provided that clear transection margins can be achieved with partial resection, on the basis of this meta-analysis, there is no evidence to suggest that patients with invasive carcinoma should undergo a total pancreatectomy as first line treatment. Especially as this procedure is associated with severe co-morbidities such as exocrine dysfunction and brittle diabetes. From this data it is clear that patients with negative transection margins intra-operatively on frozen section do not warrant further resection, whereas those
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 4
Study (pub year) study time period
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998) Total
Total number of patients
Benign
Total invasive
Recurrence
Initial gland histology of those with recurrence Benign
low grade/ adenoma/ mild dysplasia
Moderate/ intermediate dysplasia
Severe dysplasia/cis
Initial margin of recurrence invasive
Negative
low grade/ adenoma/mild
intermediate grade/borderline
High grade/CIS
Invasive
118 78 35 100 130 208
82 78 22 75 130 126
36 0 13 25 0 82
20 6 7 13 22 44
0 0 0 1 0 8
2 0 0 0 5 0
4 2 0 0 8 0
2 4 0 0 9 3
12 0 7 12 0 33
17 2 7 9 16 38
1 2 0 2 6 2
0 1 0 1 0 0
0 1 0 0 0 0
2 0 0 1 0 4
191 104 24 41 127 82
191 104 15 31 58 45
0 0 9 10 69 37
32 10 7 10 19 17
na 4 1 0 0 0
na 0 0 0 0 0
na 0 0 0 0 0
3 6 0 0 0 1
0 0 6 10 19 16
31 7 1 4 17 9
1 3 6 1 0 0
0 0 0 3 0 0
0 0 0 0 0 0
0 0 0 2 2 8
62 113 45 30 1488
39 73 26 20 1115
23 40 19 10 373
7 31 15 3 263
0 5 2 0
3 0 0 0
0 0 0 0
0 0 0 0
4 26 13 3 161
2 25 10 3 198
1 0 2 0
0 0 0 0
0 0 0 0
0 6 3 0 28
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
PAN523_proof ■ 26 February 2016 ■ 4/8
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
Table 1 Total number of patients.
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Q3
a Study (publication year) study time period
Total no of patients
Non invasive gland
Invasive gland
Recurrence
Kim (2008) 1994e2004 White (2007) 1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer (2008) 1992e2005 Miller (2011) 1991e2010 Fujii (2010) 1991e2009 Azar (1996) 1980e1994 Paye (2000) 1991e1998 Nara (2009) 1986e2007 Nakagohri (2007) 1994e2006 Lubezky (2010) 1995e2008 Chari (2002) 1983e2002 Cuillerier (2000) 1980e1996 Sugiura (2002) 1983e1998
118 78 35 100 130 208
82 78 22 75 130 145
36 0 13 25 0 63
191 104 24 41 124 82 62 87 35 30
191 104 15 31 58 45 39 60 20 23
0 0 9 10 66 37 23 27 15 7
Invasive gland e no recurrence
Initial margin of recurrence -ve margins recurrence
-ve no recurrence
Non-inv. recurrence
Non.inv. margin no recurrence
Inv. Margins þve recurrence
Inv margins þve no recurrence
12 0 7 12 0 33
24 0 6 13 0 30
17 2 7 9 16 37
78 31 20 83 84 149
0 4 0 3 6 2
20 41 8 4 22 15
3 0 0 1 0 5
0 0 0 0 0 0
0 0 6 10 20 15 4 18 12 3
0 0 3 0 46 22 19 9 3 4
31 7 1 4 18 8 2 17 6 3
122 57 9 21 87 59 31 61 13 27
1 3 6 4 0 1 1 0 2 0
37 25 8 10 18 5 10 3 5 0
0 0 0 2 2 8 4 6 6 0
0 0 0 0 1 1 2 0 2 0
Recurrence Non-invasive gland
Invasive gland
20 6 7 13 22 44
8 6 0 1 22 11
32 10 7 10 20 17 7 23 14 3
32 10 1 0 0 2 3 5 2 0
b
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
PAN523_proof ■ 26 February 2016 ■ 5/8
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
Table 2 Invasive and non-invasive only.
5
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 6/8
6
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
Table 3 Percentage of recurrence in negative margins. Study
ES
[95% Conf.
Interval]
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998)
0.882 0.547 1.088 0.65 0.832 0.929 0.939 0.693 0.747 0.857 0.862 0.722 0.547 0.98 1.213 0.684
0.682 0.21 0.717 0.447 0.637 0.785 0.781 0.45 0.156 0.473 0.672 0.485 0.21 0.759 0.774 0.332
1.082 0.883 1.458 0.854 1.027 1.072 1.097 0.936 1.338 1.241 1.052 0.96 0.883 1.201 1.651 1.036
D þ L pooled ES
0.832
0.758
0.905
% Weight
Percent
Lower 95% CI
Upper 95% CI
18.2 7.3 26.8 10.2 16.3 20.1 20.5 11.5 13.3 17.3 17.5 12.5 7.3 22.1 32.5 11.2
11.2 1.1 12.3 4.9 9.8 14.6 14.5 5.0 0.6 5.5 10.9 5.8 1.1 13.7 14.2 2.7
26.5 18.3 44.4 17.2 24.1 26.1 27.2 20.3 38.5 33.8 25.2 21.3 18.3 31.9 54.0 24.5
100
16.3
13.7
19.1
100
% Weight
Percent
Lower 95% CI
Upper 95% CI
% Weight
1.2 9.7 2.9 43.7 22.4 13.8 3.7 12.0 43.3 29.9 1.3 21.0 12.2 6.7 31.1
1.1 3.0 2.4 13.5 9.4 2.2 0.1 2.9 20.1 10.3 1.2 1.1 0.5 5.1 5.9
10.1 19.9 22.7 76.8 39.0 32.9 11.9 26.1 68.2 54.7 11.2 56.0 35.6 46.7 65.0
7.5
20.8
100
% Weight
8.42 3.96 3.37 8.25 8.68 12.03 10.97 6.5 1.46 3.16 8.94 6.71 3.96 7.43 2.51 3.67
% Weight 8.42 3.96 3.37 8.25 8.68 12.03 10.97 6.5 1.46 3.16 8.94 6.71 3.96 7.43 2.51 3.67
I-squared (variation in ES attributable to heterogeneity) ¼ 30.7%.
Table 4 Percentage of recurrence in non-invasive margins. Study
ES
[95% Conf.
Interval]
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996)
0.22 0.635 0.34 1.444 0.986 0.76 0.389 0.708 1.437 1.158 0.231 0.952 0.713 0.524 1.183
0.208 0.346 0.313 0.752 0.622 0.298 0.075 0.344 0.931 0.652 0.218 0.211 0.148 0.456 0.49
0.648 0.924 0.993 2.137 1.35 1.222 0.703 1.072 1.943 1.664 0.681 1.692 1.279 1.504 1.876
D þ L pooled ES
0.75
0.555
0.946
100
13.4
% Weight
Percent
Lower 95% CI
Upper 95% CI
93.3 85.3 95.6 90.8 63.0 85.3 64.5 96.3 72.4
53.4 21.7 67.2 41.3 17.1 58.2 28.4 71.4 40.6
94.9 91.3 96.4 93.5 97.3 99.3 93.0 96.9 95.0
84.1
73.0
92.6
7.69 9.61 5.16 4.82 8.56 7.25 9.26 8.56 6.71 6.71 7.4 4.43 6.03 3 4.82
7.69 9.61 5.16 4.82 8.56 7.25 9.26 8.56 6.71 6.71 7.4 4.43 6.03 3 4.82
I-squared (variation in ES attributable to heterogeneity) ¼ 59.7%.
Table 5 Percentage of recurrence in invasive margins. Study
ES
[95% Conf.
Interval]
Kim (2008) 1994e2004 Wada (2005) 1998e2003 Schnelldorfer 2008 (1992e2005) Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996)
2.618 2.356 2.721 2.526 1.833 2.356 1.864 2.754 2.035
1.638 0.97 1.921 1.395 0.853 1.736 1.123 2.013 1.382
3.598 3.742 3.521 3.658 2.813 2.976 2.605 3.495 2.689
D þ L pooled ES
2.32
2.048
2.591
7.69 3.85 11.54 5.77 7.69 19.23 13.46 13.46 17.31 100
7.69 3.85 11.54 5.77 7.69 19.23 13.46 13.46 17.31 100
I-squared (variation in ES attributable to heterogeneity) ¼ 0.0%.
with margins positive for invasive malignancy should undergo further resection until a negative margin is achieved. It seems probable that only patients with invasive malignancy at the transection margin detected on histological resection postoperatively also warrant further resection. Most importantly,
patients with IPMN at the transection margin including evidence of carcinoma-in-situ or high-grade dysplasia do not benefit from further intervention. This in keeping with several other studies [8,16], however is in contrast to a meta-analysis by Leng et al., which reported that 53.85% of recurrence in non-invasive IPMNs
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 7/8
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
7
Fig. 2. Odds of recurrence in invasive versus non-invasive gland.
Table 6 Odds of recurrence in invasive versus non-invasive gland. Study
Odds ratio
[95% Conf.
Interval]
Kim (2008) 1994e2004 Raut (2006) 1990e2003 Wada (2005) 1998e2003 Schnelldorfer 2008 (1992e2005) Azar (1996) 1980e1994 Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998)
4.63 51.92 68.31 13.40 28.00 51.58 14.66 2.53 22.00 36.00 36.56
1.69 5.50 8.17 6.09 2.40 6.69 3.07 0.51 6.52 5.21 3.34
12.65 489.89 571.04 29.49 326.74 397.62 69.92 12.47 74.21 248.66 400.46
% Weight 15.03 5.70 6.21 17.87 4.94 6.58 9.51 9.25 12.64 7.13 5.15
D þ L pooled ES
16.08
8.76
29.54
100.00
I-squared (variation in ES attributable to heterogeneity) ¼ 39.3%.
occurred where the transected margin showed mild dysplasia [26]. In addition, in their analysis only 4.85% of recurrences occurred in non-invasive gland whereas our study found 32% of recurrences occurred in non-invasive glands commensurate with the field change associated with IPMNs. The vast majority of pancreatic resections performed for IPMN harbour no evidence of malignancy; therefore it stands to reason that the vast majority of recurrences are likely to be benign lesions. The literature lacks evidence of the lead time for lesions to develop from adenomas to carcinomas, nevertheless it can be argued that all patients with benign histology who develop recurrence should
be offered a re-resection, whereas those with invasive carcinoma should be treated with adjuvant chemotherapy, much like the standard practice for recurrence in adenocarcinomas of the pancreas post resection. Negative margins and non-invasive positive margins have a similar recurrence rate to each other, however margins with invasive histology had a higher risk of recurrence compared to benign glands, which is reflected in the 5-year survival of both these disease processes. There is evidence to suggest that the majority of recurrences occur in the remnant pancreas and not the resection margin [27], this adds weight to the argument that IPMN is usually
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 8/8
8
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8 Table 7 5-year survival in invasive and benign gland. Study
5 year survival benign
5 year survival malignant
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998)
98.20% 87% 74% 100% 81% 94% n/r 100% 95% n/r 100% n/r n/r n/r 97% 85%
65.30% N/A 34% 32% n/r 31% n/r n/r 50% n/r n/r n/r n/r n/r 44% 80%
n/r e not reported.
a multi focal disease and may explain why margin status, particularly benign/non invasive, has no bearing on the chance of recurrence. Conclusions Transection margins should be sent for frozen section when undertaking pancreatic resection for IPMN. In the event that the transection margin confirms invasive tumour, then further resections should be undertaken until carcinoma is no longer detected. Transection margins demonstrating persistent IPMN should also undergo further resection. Transection margins with non-invasive dysplasia do not require further resection. However, the risk of recurrence IPMN in the pancreatic remnant is high, and therefore all patients require further surveillance and should be offered re-resection in the event of recurrence in patients without evidence of invasive malignancy. Acknowledgements Dr Stuart Howell, Senior Statistician, Data Management and Analysis Centre, School of Population Health, The University of Adelaide. References [1] Ohhashi KM, Maruyama M. Four cases of “mucin-producing” cancer of the pancreas: a specific finding of the papilla of vater. Prog Dig Endosc 1982;20: 348e51. [2] Tanaka M, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012;12(3):183e97. [3] Hruban RH, et al. An illustrated consensus on the classification of pancreatic intraepithelial neoplasia and intraductal papillary mucinous neoplasms. Am J Surg Pathol 2004;28(8):977e87. [4] Raut CP, et al. Intraductal papillary mucinous neoplasms of the pancreas: effect of invasion and pancreatic margin status on recurrence and survival. Ann Surg Oncol 2006;13(4):582e94. [5] Nakagohri T, et al. Long-term surgical outcome of noninvasive and minimally invasive intraductal papillary mucinous adenocarcinoma of the pancreas. World J Surg 2002;26(9):1166e9. [6] He JCJLA, N.Makary MA, Hirose K, Choti MA, Schulick RD, Hruban RH, Pawlik TM, Wolfgang CL. Is it necessary to follow patients after resection of a benign pancreatic intraductal papillary mucinous neoplasm? J Am Coll Surg 2013;216:657e67. [7] White R, et al. Fate of the remnant pancreas after resection of noninvasive intraductal papillary mucinous neoplasm. J Am Coll Surg 2007;204(5): 987e93. discussion 993-5.
[8] Miller JR, et al. Outcome of the pancreatic remnant following segmental pancreatectomy for non-invasive intraductal papillary mucinous neoplasm. HPB Oxf 2011;13(11):759e66. [9] Moher D, et al. Preferred reporting items for systematic reviews and metaanalyses: the PRISMA statement. Int J Surg 2010;8(5):336e41. [10] Stroup DF, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. J Am Med Assoc 2000;283(15):2008e12. [11] Adsay NVF, N.Furukawa T, Hruban RH, Klimstra DS, Kloppel G, et al. Intraductal neoplasm of the pancreas. In: Bosman F, Hruban RH, Theise ND, editors. WHO clasiification of tumours of digestive system F.T.C. Lyon: WHO Press; 2010. p. 304e13. [12] Andea A, Sarkar F, Adsay VN. Clinicopathological correlates of pancreatic intraepithelial neoplasia: a comparative analysis of 82 cases with and 152 cases without pancreatic ductal adenocarcinoma. Mod Pathol 2003;16(10): 996e1006. [13] Haugk B. Pancreatic intraepithelial neoplasia-can we detect early pancreatic cancer? Histopathology 2010;57(4):503e14. [14] Azar C, et al. Intraductal papillary mucinous tumours of the pancreas. Clinical and therapeutic issues in 32 patients. Gut 1996;39(3):457e64. [15] Sugiura H, et al. Clinicopathologic features and outcomes of intraductal papillary-mucinous tumors of the pancreas. Hepatogastroenterology 2002;49(43):263e7. [16] Nara S, et al. Clinical significance of frozen section analysis during resection of intraductal papillary mucinous neoplasm: should a positive pancreatic margin for adenoma or borderline lesion be resected additionally? J Am Coll Surg 2009;209(5):614e21. [17] Chari ST, et al. Study of recurrence after surgical resection of intraductal papillary mucinous neoplasm of the pancreas. Gastroenterology 2002;123(5): 1500e7. [18] Cuillerier E, et al. Outcome after surgical resection of intraductal papillary and mucinous tumors of the pancreas. Am J Gastroenterol 2000;95(2):441e5. [19] Paye F, et al. Intraductal papillary mucinous tumors of the pancreas: pancreatic resections guided by preoperative morphological assessment and intraoperative frozen section examination. Surgery 2000;127(5):536e44. [20] Schnelldorfer T, et al. Experience with 208 resections for intraductal papillary mucinous neoplasm of the pancreas. Arch Surg 2008;143(7):639e46. discussion 646. [21] Lubezky N, et al. Clinical presentation can predict disease course in patients with intraductal papillary mucinous neoplasm of the pancreas. World J Surg 2010;34(1):126e32. [22] Nakagohri T, et al. Surgical outcome of intraductal papillary mucinous neoplasms of the pancreas. Ann Surg Oncol 2007;14(11):3174e80. [23] Wada K, Kozarek RA, Traverso LW. Outcomes following resection of invasive and noninvasive intraductal papillary mucinous neoplasms of the pancreas. Am J Surg 2005;189(5):632e6. discussion 637. [24] Kim SC, et al. Intraductal papillary mucinous neoplasm of the pancreas: clinical characteristics and treatment outcomes of 118 consecutive patients from a single center. J Hepatobiliary Pancreat Surg 2008;15(2):183e8. [25] Fujii T, et al. Prognostic impact of pancreatic margin status in the intraductal papillary mucinous neoplasms of the pancreas. Surgery 2010;148(2):285e90. [26] Leng KM, et al. Impact of pancreatic margin status and lymph node metastases on recurrence after resection for invasive and noninvasive intraductal papillary mucinous neoplasms of the pancreas: a meta-analysis. Dig Surg 2012;29(3):213e25. [27] Sho M, et al. Pattern of recurrence after resection for intraductal papillary mucinous tumors of the pancreas. World J Surg 1998;22(8):874e8. Q2
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
PAN523_proof ■ 26 February 2016 ■ 1/8
Pancreatology xxx (2016) 1e8
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Review article
Q5
Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis
Q4
Neil Bhardwaj a, *, Ashley R. Dennison a, Guy J. Maddern b, Giuseppe Garcea a a b
Q1
University Hospitals Leicester, Department of HPB Surgery, Leicester, LE5 4PW, UK University of Adelaide, Discipline of Surgery, The Queen Elizabeth Hospital, Woodville, Adelaide, Australia
a r t i c l e i n f o
a b s t r a c t
Article history: Available online xxx
Introduction: IPMN is a relatively new clinical entity and surgeons are continuing to develop their understanding of this complex pathology. Little is known of the natural disease process post-resection of an IPMN, particularly the impact of gland histology and margin status on the chance of recurrence and survival in benign and invasive IPMN. Methods: An online search was conducted to evaluate and include those studies which reported on gland histology, margin status and disease recurrence in resected benign and malignant IPMN. A Meta analysis was then performed using a random effects model. Results: The chance of recurrence in non-invasive margin positive IPMN is similar to margin negative IPMN. The chance of recurrence is higher in invasive gland IPMN compared to non-invasive gland. The vast majority of recurrences occurred in patients with positive margins demonstrating invasion. Conclusion: All patients with intra- or post-operative evidence of invasive carcinoma at the resection margin should undergo further resection to achieve a negative margin. Patients with evidence of IPMN at the transaction margin (even with changes of high grade dysplasia/CIS) may not achieve any benefit from further resection. Patients with recurrence in benign/non-invasive IPMN should undergo re-resection, whereas patients with recurrence in invasive IPMN should not. Copyright © 2016, IAP and EPC. Published by Elsevier India, a division of Reed Elsevier India Pvt. Ltd. All rights reserved.
Keywords: Survival Resection IPMN Outcomes Resection margin
Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Definition of histology and margin status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of margin status on recurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of gland histology on recurrence and survival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
* Corresponding author. E-mail address: [email protected] (N. Bhardwaj). http://dx.doi.org/10.1016/j.pan.2016.02.008 1424-3903/Copyright © 2016, IAP and EPC. Published by Elsevier India, a division of Reed Elsevier India Pvt. Ltd. All rights reserved.
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 2/8
2
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
remnant transaction margin after a partial pancreatectomy for IPMN.
Introduction Intraductal papillary neoplasm (IPMN) of the pancreas was first described as a discreet entity in 1982 [1]. In 1996, the World Health Organisation revised pancreatic tumour classification and recognized IPMN as a separate pathological tumour to mucinous cystic tumours; these guidelines were recently updated in 2012 [2]. Morphologically, IPMNs are divided into three types; main-duct, branch-duct or combined, mainly based on imaging and/or histology. It is now well established that IPMNs exhibit a spectrum of neoplastic transformation similar to the adenoma-carcinoma sequence observed in colon cancer [3]. Histological grading of pancreatic cancer is graded from “hyperplasia” or adenoma (currently classified as “low-grade dysplasia”) to invasive carcinoma. The definition of “malignancy” has been variable, as there remains some debate as to whether to include “carcinoma in situ” (CIS) in the malignant category or whether that definition should be reserved for invasive carcinomas only. This makes it difficult to compare studies and may negatively impact a surgeon's ability to accurately predict patient prognosis. There has been increased understanding of IPMN pathology and behaviour with accepted guidelines and acknowledged risk factors. Non-invasive tumours have a favourable outcome compared to invasive tumours [4,5] and it would be reasonable to assume that IPMN (or even invasive carcinoma) at the pancreatic cut margin (on frozen section) would merit further resection to obtain clear margins. However, despite the recent advances in the understanding of IPMN, little is known regarding the potential risk of recurrence in patients with low grade, medium grade or high grade dysplasia at the pancreatic cut surface. There have also been reports of invasive carcinoma developing after resection of non-invasive IPMN [6e8]. The aim of this meta-analysis is to establish the future recurrence risk associated with resection of benign and malignant glands and in the various grades of dysplasia (low, medium and high) in the
Abstracts idenƟfied through database searching (n = 82 )
Methods Search strategy This was performed as per the guidelines issued in the PRISMA and MOOSE statement [9,10]. A systematic literature search of the Medline, Embase, Cochrane and Pubmed database was conducted over the last 20 years using the terms intraductal papillary mucinous neoplasm/IPMN/intraductal papillary mucinous tumour. The latest search was performed in 2014 and there were no language restrictions. Abstracts of all relevant clinical studies were reviewed and if relevant, the article was reviewed in detail and included in the analysis. Further articles were also identified from relevant reviews and cross-referenced with those already included to avoid duplication of studies (Fig. 1 e PRISMA flow diagram). Inclusion criteria All studies that reported on resection of primary IPMN, margin status, recurrence rate and detailed benign and malignant primary histological diagnoses were included. In addition, only studies reported in English with a minimum follow up of 2 years were included. There was insufficient detail in the studies examined to consistently record if the lesions resected were main-duct, branchduct or mixed type IPMN. Exclusion criteria Case reports and review series, in addition to all studies not reporting margin status, separate benign and malignant histology, follow up and recurrence rates were excluded. If more than one
AddiƟonal records idenƟfied through other sources (n = 10 )
Abstracts aŌer duplicates removed (n = 86)
Full text arƟcles screened (n = 50 )
Abstracts excluded as failure to meet criteria (n = 36)
Full-text arƟcles excluded as per exclusion criteria (n = 34 )
Studies included in qualitaƟve synthesis (n = 16 ) Fig. 1. PRISMA flow diagram.
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 3/8
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
study was reported from the same institute then the most recent publication was included. Definition of histology and margin status Resected IPMNs were either classified as malignant or benign and similarly transection margin status was classified into malignant (invasive carcinoma), non-invasive benign or negative. Benign tumours were further classified according to the latest guidelines into adenoma, also known as mild dysplasia, moderate dysplasia (previously known as borderline) or high grade dysplasia, often referred to as carcinoma-in-situ (CIS) [2]. However, according to the latest WHO guidelines of 2010, CIS should be abandoned as a term in favour of high grade dysplasia [11]. Some studies referred to Pancreatic intra-epithelial neoplasms (PanIN) when describing resection margins. PanINs are pathologically distinct entities to IPMNs and are recognised pre-cursers to pancreatic adenocarcinoma [12,13]. Although they are a distinct histological diagnosis, they do share certain common characteristics with IPMNs; being also a precursor for pancreatic adenocarcinoma. They are usually small (<5 mm) intraductal non-invasive lesions formed by metaplasia and proliferation of ductal epithelium. They are similarly graded into mild, moderate or severe dysplasia (PanIN-1, PanIN-2 and PanIN-3, respectively) [3]. For the purposes of this metaanalysis these lesions, when described, in the remnant margin were categorised mild, moderate, severe similar to their IPMN counterpart. Recurrence Recurrence was defined as pathological evidence of recurrence on repeat resection or radiological evidence of recurrence on surveillance imaging. All recurrences: at the resected margin, distant to the margin but in the gland and/or distant to the gland were grouped under one category. No recurrences occurred distant to the gland without evidence of intra-pancreatic recurrence.
3
further refinement of the data was required due to the small numbers and therefore histological status of the gland and the cut margin were grouped together and defined as either non-invasive (benign, low grade/adenoma/mild dysplasia, moderate/intermediate dysplasia, severe/high-grade dysplasia/CIS) or invasive (Table 2a and b). Impact of margin status on recurrence The chance of recurrence in patients with negative margins was 16.3% (95% CI: 13.7%e19.1%). The I-squared statistic indicated that only 30.7% of the variance in the estimate was due to heterogeneity across the studies (Table 3). The pooled estimate indicated that the proportion of recurrence in benign non-invasive margins was 13.4% (95% CI: 7.5%e20.8%). The I-squared statistic indicated that 59.7% of the variance in the estimate was due to heterogeneity across the studies (Table 4). The chance of recurrence in invasive resection margin was 84.1% (95% CI: 73.0%e92.6%). The I-squared statistic indicated that there was no variation in the estimates (Table 5). Impact of gland histology on recurrence and survival On comparing benign versus malignant gland histology, the pooled odds ratio for the overall analysis was 16.08 (95% CI: 8.76e29.54). This was significantly different from 1.00 (p < 0.0001), indicating that the odds of recurrence were significantly higher in the invasive gland group compared to the non-invasive group. Isquared statistic indicated that only 20.4% of the variation across studies was due to heterogeneity (Table 6, Fig. 2). The study from Paye et al. (2000) was removed from the analysis as the odds ratio estimated for this study was extreme (1323.0). It only made a small contribution to the pooled odds ratio (weight ¼ 3.64), and was influential in determining the magnitude of the I-square statistic. It is also evident that the 5-year survival, where reported, in patients with benign disease (approaches 100% in most studies) is greater than those with malignant disease (31%e80%) (Table 7).
Statistical analysis Discussion Histology measures were reported as pooled percentages. A sine transformation was applied to these data to avoid confidence intervals below zero and above 100. The data were back transformed prior to analysis. The comparison of invasive versus non-invasive gland was reported as a pooled odds ratio. This was calculated from count data as presented in the papers and expressed as the odds of recurrence in the invasive gland group relative to that in the non-invasive group. A random effects meta-analysis model was applied to obtain the pooled estimates and heterogeneity was assessed using the I-square statistic. All analyses were completed using STATA v12.1 (STATA Corp., LP, TX, USA). Results A total of 82 abstracts were identified, of which 50 full articles were retrieved and analysed in-depth. 6 articles were rejected due to duplicity of data and a further 28 articles were rejected as they failed to report margin status. A total of 16 articles were included in the final analysis [4,6e8,14e25] (Table 1 e all 16 studies). A total of 1488 patients were included in this study, of which there were 1115 (75%) benign lesions resected and 373 (25%) malignant IPMNs resected. A total of 263 recurrences (18%) were noted of which 198 (75%) had a negative resection margin. Of the 263 patients with recurrence, 161 (61%) had invasive IPMN on initial histology and 28 (11%) had invasive carcinoma at the margin. Non-invasive histology was evident in 31% and only 8% had completely benign histology. A
The management of IPMN is ever evolving and little is understood regarding the impact of residual non-invasive margin status on the risk of future recurrence. This study provides evidence that the vast majority of recurrences occur in patients whose margins were negative on resection, however a positive transection margin for invasive malignancy places patients at a high risk of developing recurrence. It also proves that the odds of disease recurrence is higher in invasive versus non-invasive excised IPMN lesions regardless of the margin status. This is not surprising and suggests that recurrences are likely the result of a missed lesion at the time of initial resection. This meta-analysis does lend support to the theory that IPMN demonstrates a field defect in the pancreas, with the entire gland at risk of developing neoplastic disease. Some authors have suggested performing pre-operative endoscopic retrograde cholangiopancreatography followed by intra-operative pancreatography in order to detect all potentially malignant lesions [23], this however is not routine practice in most centres. Provided that clear transection margins can be achieved with partial resection, on the basis of this meta-analysis, there is no evidence to suggest that patients with invasive carcinoma should undergo a total pancreatectomy as first line treatment. Especially as this procedure is associated with severe co-morbidities such as exocrine dysfunction and brittle diabetes. From this data it is clear that patients with negative transection margins intra-operatively on frozen section do not warrant further resection, whereas those
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 4
Study (pub year) study time period
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998) Total
Total number of patients
Benign
Total invasive
Recurrence
Initial gland histology of those with recurrence Benign
low grade/ adenoma/ mild dysplasia
Moderate/ intermediate dysplasia
Severe dysplasia/cis
Initial margin of recurrence invasive
Negative
low grade/ adenoma/mild
intermediate grade/borderline
High grade/CIS
Invasive
118 78 35 100 130 208
82 78 22 75 130 126
36 0 13 25 0 82
20 6 7 13 22 44
0 0 0 1 0 8
2 0 0 0 5 0
4 2 0 0 8 0
2 4 0 0 9 3
12 0 7 12 0 33
17 2 7 9 16 38
1 2 0 2 6 2
0 1 0 1 0 0
0 1 0 0 0 0
2 0 0 1 0 4
191 104 24 41 127 82
191 104 15 31 58 45
0 0 9 10 69 37
32 10 7 10 19 17
na 4 1 0 0 0
na 0 0 0 0 0
na 0 0 0 0 0
3 6 0 0 0 1
0 0 6 10 19 16
31 7 1 4 17 9
1 3 6 1 0 0
0 0 0 3 0 0
0 0 0 0 0 0
0 0 0 2 2 8
62 113 45 30 1488
39 73 26 20 1115
23 40 19 10 373
7 31 15 3 263
0 5 2 0
3 0 0 0
0 0 0 0
0 0 0 0
4 26 13 3 161
2 25 10 3 198
1 0 2 0
0 0 0 0
0 0 0 0
0 6 3 0 28
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
PAN523_proof ■ 26 February 2016 ■ 4/8
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
Table 1 Total number of patients.
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Q3
a Study (publication year) study time period
Total no of patients
Non invasive gland
Invasive gland
Recurrence
Kim (2008) 1994e2004 White (2007) 1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer (2008) 1992e2005 Miller (2011) 1991e2010 Fujii (2010) 1991e2009 Azar (1996) 1980e1994 Paye (2000) 1991e1998 Nara (2009) 1986e2007 Nakagohri (2007) 1994e2006 Lubezky (2010) 1995e2008 Chari (2002) 1983e2002 Cuillerier (2000) 1980e1996 Sugiura (2002) 1983e1998
118 78 35 100 130 208
82 78 22 75 130 145
36 0 13 25 0 63
191 104 24 41 124 82 62 87 35 30
191 104 15 31 58 45 39 60 20 23
0 0 9 10 66 37 23 27 15 7
Invasive gland e no recurrence
Initial margin of recurrence -ve margins recurrence
-ve no recurrence
Non-inv. recurrence
Non.inv. margin no recurrence
Inv. Margins þve recurrence
Inv margins þve no recurrence
12 0 7 12 0 33
24 0 6 13 0 30
17 2 7 9 16 37
78 31 20 83 84 149
0 4 0 3 6 2
20 41 8 4 22 15
3 0 0 1 0 5
0 0 0 0 0 0
0 0 6 10 20 15 4 18 12 3
0 0 3 0 46 22 19 9 3 4
31 7 1 4 18 8 2 17 6 3
122 57 9 21 87 59 31 61 13 27
1 3 6 4 0 1 1 0 2 0
37 25 8 10 18 5 10 3 5 0
0 0 0 2 2 8 4 6 6 0
0 0 0 0 1 1 2 0 2 0
Recurrence Non-invasive gland
Invasive gland
20 6 7 13 22 44
8 6 0 1 22 11
32 10 7 10 20 17 7 23 14 3
32 10 1 0 0 2 3 5 2 0
b
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
PAN523_proof ■ 26 February 2016 ■ 5/8
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
Table 2 Invasive and non-invasive only.
5
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 6/8
6
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
Table 3 Percentage of recurrence in negative margins. Study
ES
[95% Conf.
Interval]
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998)
0.882 0.547 1.088 0.65 0.832 0.929 0.939 0.693 0.747 0.857 0.862 0.722 0.547 0.98 1.213 0.684
0.682 0.21 0.717 0.447 0.637 0.785 0.781 0.45 0.156 0.473 0.672 0.485 0.21 0.759 0.774 0.332
1.082 0.883 1.458 0.854 1.027 1.072 1.097 0.936 1.338 1.241 1.052 0.96 0.883 1.201 1.651 1.036
D þ L pooled ES
0.832
0.758
0.905
% Weight
Percent
Lower 95% CI
Upper 95% CI
18.2 7.3 26.8 10.2 16.3 20.1 20.5 11.5 13.3 17.3 17.5 12.5 7.3 22.1 32.5 11.2
11.2 1.1 12.3 4.9 9.8 14.6 14.5 5.0 0.6 5.5 10.9 5.8 1.1 13.7 14.2 2.7
26.5 18.3 44.4 17.2 24.1 26.1 27.2 20.3 38.5 33.8 25.2 21.3 18.3 31.9 54.0 24.5
100
16.3
13.7
19.1
100
% Weight
Percent
Lower 95% CI
Upper 95% CI
% Weight
1.2 9.7 2.9 43.7 22.4 13.8 3.7 12.0 43.3 29.9 1.3 21.0 12.2 6.7 31.1
1.1 3.0 2.4 13.5 9.4 2.2 0.1 2.9 20.1 10.3 1.2 1.1 0.5 5.1 5.9
10.1 19.9 22.7 76.8 39.0 32.9 11.9 26.1 68.2 54.7 11.2 56.0 35.6 46.7 65.0
7.5
20.8
100
% Weight
8.42 3.96 3.37 8.25 8.68 12.03 10.97 6.5 1.46 3.16 8.94 6.71 3.96 7.43 2.51 3.67
% Weight 8.42 3.96 3.37 8.25 8.68 12.03 10.97 6.5 1.46 3.16 8.94 6.71 3.96 7.43 2.51 3.67
I-squared (variation in ES attributable to heterogeneity) ¼ 30.7%.
Table 4 Percentage of recurrence in non-invasive margins. Study
ES
[95% Conf.
Interval]
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996)
0.22 0.635 0.34 1.444 0.986 0.76 0.389 0.708 1.437 1.158 0.231 0.952 0.713 0.524 1.183
0.208 0.346 0.313 0.752 0.622 0.298 0.075 0.344 0.931 0.652 0.218 0.211 0.148 0.456 0.49
0.648 0.924 0.993 2.137 1.35 1.222 0.703 1.072 1.943 1.664 0.681 1.692 1.279 1.504 1.876
D þ L pooled ES
0.75
0.555
0.946
100
13.4
% Weight
Percent
Lower 95% CI
Upper 95% CI
93.3 85.3 95.6 90.8 63.0 85.3 64.5 96.3 72.4
53.4 21.7 67.2 41.3 17.1 58.2 28.4 71.4 40.6
94.9 91.3 96.4 93.5 97.3 99.3 93.0 96.9 95.0
84.1
73.0
92.6
7.69 9.61 5.16 4.82 8.56 7.25 9.26 8.56 6.71 6.71 7.4 4.43 6.03 3 4.82
7.69 9.61 5.16 4.82 8.56 7.25 9.26 8.56 6.71 6.71 7.4 4.43 6.03 3 4.82
I-squared (variation in ES attributable to heterogeneity) ¼ 59.7%.
Table 5 Percentage of recurrence in invasive margins. Study
ES
[95% Conf.
Interval]
Kim (2008) 1994e2004 Wada (2005) 1998e2003 Schnelldorfer 2008 (1992e2005) Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996)
2.618 2.356 2.721 2.526 1.833 2.356 1.864 2.754 2.035
1.638 0.97 1.921 1.395 0.853 1.736 1.123 2.013 1.382
3.598 3.742 3.521 3.658 2.813 2.976 2.605 3.495 2.689
D þ L pooled ES
2.32
2.048
2.591
7.69 3.85 11.54 5.77 7.69 19.23 13.46 13.46 17.31 100
7.69 3.85 11.54 5.77 7.69 19.23 13.46 13.46 17.31 100
I-squared (variation in ES attributable to heterogeneity) ¼ 0.0%.
with margins positive for invasive malignancy should undergo further resection until a negative margin is achieved. It seems probable that only patients with invasive malignancy at the transection margin detected on histological resection postoperatively also warrant further resection. Most importantly,
patients with IPMN at the transection margin including evidence of carcinoma-in-situ or high-grade dysplasia do not benefit from further intervention. This in keeping with several other studies [8,16], however is in contrast to a meta-analysis by Leng et al., which reported that 53.85% of recurrence in non-invasive IPMNs
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 7/8
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8
7
Fig. 2. Odds of recurrence in invasive versus non-invasive gland.
Table 6 Odds of recurrence in invasive versus non-invasive gland. Study
Odds ratio
[95% Conf.
Interval]
Kim (2008) 1994e2004 Raut (2006) 1990e2003 Wada (2005) 1998e2003 Schnelldorfer 2008 (1992e2005) Azar (1996) 1980e1994 Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998)
4.63 51.92 68.31 13.40 28.00 51.58 14.66 2.53 22.00 36.00 36.56
1.69 5.50 8.17 6.09 2.40 6.69 3.07 0.51 6.52 5.21 3.34
12.65 489.89 571.04 29.49 326.74 397.62 69.92 12.47 74.21 248.66 400.46
% Weight 15.03 5.70 6.21 17.87 4.94 6.58 9.51 9.25 12.64 7.13 5.15
D þ L pooled ES
16.08
8.76
29.54
100.00
I-squared (variation in ES attributable to heterogeneity) ¼ 39.3%.
occurred where the transected margin showed mild dysplasia [26]. In addition, in their analysis only 4.85% of recurrences occurred in non-invasive gland whereas our study found 32% of recurrences occurred in non-invasive glands commensurate with the field change associated with IPMNs. The vast majority of pancreatic resections performed for IPMN harbour no evidence of malignancy; therefore it stands to reason that the vast majority of recurrences are likely to be benign lesions. The literature lacks evidence of the lead time for lesions to develop from adenomas to carcinomas, nevertheless it can be argued that all patients with benign histology who develop recurrence should
be offered a re-resection, whereas those with invasive carcinoma should be treated with adjuvant chemotherapy, much like the standard practice for recurrence in adenocarcinomas of the pancreas post resection. Negative margins and non-invasive positive margins have a similar recurrence rate to each other, however margins with invasive histology had a higher risk of recurrence compared to benign glands, which is reflected in the 5-year survival of both these disease processes. There is evidence to suggest that the majority of recurrences occur in the remnant pancreas and not the resection margin [27], this adds weight to the argument that IPMN is usually
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
PAN523_proof ■ 26 February 2016 ■ 8/8
8
N. Bhardwaj et al. / Pancreatology xxx (2016) 1e8 Table 7 5-year survival in invasive and benign gland. Study
5 year survival benign
5 year survival malignant
Kim (2008) 1994e2004 White (2007)1983e2000 Raut (2006) 1990e2003 Wada (2005) 1998e2003 He (2013) 1995e2010 Schnelldorfer 2008 (1992e2005) Miller 2011 (1991e2010) Fujii 2010 (1991e2009) Azar (1996) 1980e1994 Paye 2000 (1991e1998) Nara 2009 (1986e2007) Nakagohri 2007 (1994e2006) Lubezky 2010 (1995e2008) Chari 2002 (1983e2002) Cuillerier 2000 (1980e1996) Sugiura 2002 (1983e1998)
98.20% 87% 74% 100% 81% 94% n/r 100% 95% n/r 100% n/r n/r n/r 97% 85%
65.30% N/A 34% 32% n/r 31% n/r n/r 50% n/r n/r n/r n/r n/r 44% 80%
n/r e not reported.
a multi focal disease and may explain why margin status, particularly benign/non invasive, has no bearing on the chance of recurrence. Conclusions Transection margins should be sent for frozen section when undertaking pancreatic resection for IPMN. In the event that the transection margin confirms invasive tumour, then further resections should be undertaken until carcinoma is no longer detected. Transection margins demonstrating persistent IPMN should also undergo further resection. Transection margins with non-invasive dysplasia do not require further resection. However, the risk of recurrence IPMN in the pancreatic remnant is high, and therefore all patients require further surveillance and should be offered re-resection in the event of recurrence in patients without evidence of invasive malignancy. Acknowledgements Dr Stuart Howell, Senior Statistician, Data Management and Analysis Centre, School of Population Health, The University of Adelaide. References [1] Ohhashi KM, Maruyama M. Four cases of “mucin-producing” cancer of the pancreas: a specific finding of the papilla of vater. Prog Dig Endosc 1982;20: 348e51. [2] Tanaka M, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012;12(3):183e97. [3] Hruban RH, et al. An illustrated consensus on the classification of pancreatic intraepithelial neoplasia and intraductal papillary mucinous neoplasms. Am J Surg Pathol 2004;28(8):977e87. [4] Raut CP, et al. Intraductal papillary mucinous neoplasms of the pancreas: effect of invasion and pancreatic margin status on recurrence and survival. Ann Surg Oncol 2006;13(4):582e94. [5] Nakagohri T, et al. Long-term surgical outcome of noninvasive and minimally invasive intraductal papillary mucinous adenocarcinoma of the pancreas. World J Surg 2002;26(9):1166e9. [6] He JCJLA, N.Makary MA, Hirose K, Choti MA, Schulick RD, Hruban RH, Pawlik TM, Wolfgang CL. Is it necessary to follow patients after resection of a benign pancreatic intraductal papillary mucinous neoplasm? J Am Coll Surg 2013;216:657e67. [7] White R, et al. Fate of the remnant pancreas after resection of noninvasive intraductal papillary mucinous neoplasm. J Am Coll Surg 2007;204(5): 987e93. discussion 993-5.
[8] Miller JR, et al. Outcome of the pancreatic remnant following segmental pancreatectomy for non-invasive intraductal papillary mucinous neoplasm. HPB Oxf 2011;13(11):759e66. [9] Moher D, et al. Preferred reporting items for systematic reviews and metaanalyses: the PRISMA statement. Int J Surg 2010;8(5):336e41. [10] Stroup DF, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. J Am Med Assoc 2000;283(15):2008e12. [11] Adsay NVF, N.Furukawa T, Hruban RH, Klimstra DS, Kloppel G, et al. Intraductal neoplasm of the pancreas. In: Bosman F, Hruban RH, Theise ND, editors. WHO clasiification of tumours of digestive system F.T.C. Lyon: WHO Press; 2010. p. 304e13. [12] Andea A, Sarkar F, Adsay VN. Clinicopathological correlates of pancreatic intraepithelial neoplasia: a comparative analysis of 82 cases with and 152 cases without pancreatic ductal adenocarcinoma. Mod Pathol 2003;16(10): 996e1006. [13] Haugk B. Pancreatic intraepithelial neoplasia-can we detect early pancreatic cancer? Histopathology 2010;57(4):503e14. [14] Azar C, et al. Intraductal papillary mucinous tumours of the pancreas. Clinical and therapeutic issues in 32 patients. Gut 1996;39(3):457e64. [15] Sugiura H, et al. Clinicopathologic features and outcomes of intraductal papillary-mucinous tumors of the pancreas. Hepatogastroenterology 2002;49(43):263e7. [16] Nara S, et al. Clinical significance of frozen section analysis during resection of intraductal papillary mucinous neoplasm: should a positive pancreatic margin for adenoma or borderline lesion be resected additionally? J Am Coll Surg 2009;209(5):614e21. [17] Chari ST, et al. Study of recurrence after surgical resection of intraductal papillary mucinous neoplasm of the pancreas. Gastroenterology 2002;123(5): 1500e7. [18] Cuillerier E, et al. Outcome after surgical resection of intraductal papillary and mucinous tumors of the pancreas. Am J Gastroenterol 2000;95(2):441e5. [19] Paye F, et al. Intraductal papillary mucinous tumors of the pancreas: pancreatic resections guided by preoperative morphological assessment and intraoperative frozen section examination. Surgery 2000;127(5):536e44. [20] Schnelldorfer T, et al. Experience with 208 resections for intraductal papillary mucinous neoplasm of the pancreas. Arch Surg 2008;143(7):639e46. discussion 646. [21] Lubezky N, et al. Clinical presentation can predict disease course in patients with intraductal papillary mucinous neoplasm of the pancreas. World J Surg 2010;34(1):126e32. [22] Nakagohri T, et al. Surgical outcome of intraductal papillary mucinous neoplasms of the pancreas. Ann Surg Oncol 2007;14(11):3174e80. [23] Wada K, Kozarek RA, Traverso LW. Outcomes following resection of invasive and noninvasive intraductal papillary mucinous neoplasms of the pancreas. Am J Surg 2005;189(5):632e6. discussion 637. [24] Kim SC, et al. Intraductal papillary mucinous neoplasm of the pancreas: clinical characteristics and treatment outcomes of 118 consecutive patients from a single center. J Hepatobiliary Pancreat Surg 2008;15(2):183e8. [25] Fujii T, et al. Prognostic impact of pancreatic margin status in the intraductal papillary mucinous neoplasms of the pancreas. Surgery 2010;148(2):285e90. [26] Leng KM, et al. Impact of pancreatic margin status and lymph node metastases on recurrence after resection for invasive and noninvasive intraductal papillary mucinous neoplasms of the pancreas: a meta-analysis. Dig Surg 2012;29(3):213e25. [27] Sho M, et al. Pattern of recurrence after resection for intraductal papillary mucinous tumors of the pancreas. World J Surg 1998;22(8):874e8. Q2
Please cite this article in press as: Bhardwaj N, et al., Management implications of resection margin histology in patients undergoing resection for IPMN: A meta-analysis, Pancreatology (2016), http://dx.doi.org/10.1016/j.pan.2016.02.008
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130