Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms

HPB

https://doi.org/10.1016/j.hpb.2018.09.015

ORIGINAL ARTICLE

Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms Rajesh Ramanathan1,2, Amr I. Al Abbas2, Travis Mason1, Luke G. Wolfe1 & Brian J. Kaplan1 1

Department of Surgery, Virginia Commonwealth University Medical Center, Richmond, VA, and 2Department of Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA, USA

Abstract Background: Distal pancreatic neuroendocrine tumors (PNET) and pancreatic cystic neoplasms (PCN) are often incidentally found in older adults, requiring careful consideration between operative management and watchful waiting. This study analyzes the short-term complications associated with distal pancreatectomy (DP) for PNET and PCN in older adults to inform clinical decision-making. Methods: Patients undergoing DP for PNET and PCN were analyzed using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database and the pancreatectomy procedure-targeted dataset. Associations between decade of age and 30-day outcomes were evaluated. Results: 1626 patients were analyzed from 2014 to 2015. 692 (42.6%) were younger than 60 years, 507 (31.2%) were sexagenarians, 342 (21.0%) were septuagenarians, and 85 (5.2%) were octogenarians. Minimally invasive approaches were used in 62.7%. While septuagenarians and octogenarians constituted 26.3% of the cohort, they were affected by 55.6% of reintubations, 66.7% of failures to wean, 82.4% of myocardial infarctions, and 57.1% of septic shock. Septuagenarians and octogenarians had longer hospital stays, as compared to those younger than 60 years. Conclusion: Septuagenarians and octogenarians are disproportionately affected by perioperative complications after DP for PNET and PCN. Careful patient selection and thorough counseling should be provided when surgery is considered. Received 30 January 2018; accepted 28 September 2018

Correspondence Brian J. Kaplan, 1200 E. Broad St, PO Box 980011, Richmond, VA 23298, USA. E-mail: brian.kaplan@ vcuhealth.org

Introduction The incidence of pancreatic neoplasms has increased dramatically in the past ten years, with the largest growth occurring among older age groups.1–4 A major contributor to this increase is the increased quality and utilization of cross-sectional imaging.4 Unlike localized pancreatic ductal adenocarcinoma (PDAC), which is highly aggressive with few meaningful alternatives to resection, pancreatic neuroendocrine tumors (PNET) and pancreatic cystic neoplasms (PCN) demonstrate more heterogeneous biology. This enables a role for surveillance and

Previous communication: Mini-oral presentation at the Americas HepatoPancreato-Biliary Association 2018 Annual Meeting in Miami, FL.

HPB 2018, -, 1–8

shared-decision making regarding resection. For localized lesions of the pancreatic body and tail, open or minimally invasive distal pancreatectomy (DP) can be performed. While DP generally has a lower morbidity profile than pancreatic head resections, it remains a complex operation with the potential for perioperative complications.5 The existing data on age-related risk of perioperative complications after DP remains unclear, with studies reporting both higher and lower perioperative complications in older adults.6–10 Having data on age-related risks of complications after DP from large, accurately-collected datasets can help better inform the discussion regarding resection for PNET and PCN in older adults. This study analyzes the American College of Surgeons-National Surgical Quality Improvement Project (ACS-NSQIP) dataset

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

2

and the pancreatectomy procedure-targeted dataset to assess agerelated short-term outcomes after DP for PNET and PCN.

tests were used to compare the means of continuous variables, and likelihood ratio chi-squared tests were used to compare the means of categorical variables. The non-

Methods Patient selection and characteristics The ACS-NSQIP is a surgical quality improvement and outcomes database comprising national, aggregated, patient-level data collected by trained data abstractors using chart review and administrative data. In addition, in 2014 ACS-NSQIP began recording pancreatectomy procedure-targeted variables and outcomes.11 Patient records of adults over the age of 18 years from the ACSNSQIP participant use file (PUF) and the pancreatectomy procedure-targeted PUF were merged for the years 2014–2015. Patients undergoing DP were identified using the Current Procedural Terminology (CPT) codes 48140, 48145, and 48146. Variables Patient variables assessed included: age, sex, body mass index (BMI), medical comorbidities, tobacco use, functional status, frailty, and malnutrition. Age was categorized into decades: below 60 years, sexagenarians (age 60–69 years), septuagenarians (age 70–79 years), and octogenarians (age 80 years and greater). ACSNSQIP categorizes functional status as independent, partially dependent or totally dependent. For this analysis, functional status was dichotomized as independent and non-independent. A previously validated, abbreviated 5-point modified frailty score (amFI) was calculated by assigning one point each for the following widely-captured conditions in the ACS-NSQIP dataset: non-independent functional status, diabetes, history of congestive heart failure (CHF), history of chronic obstructive pulmonary disease (COPD), and history of hypertension requiring medication.12 A composite surrogate of malnutrition was defined as preoperative albumin less than 3 g/dL and/or preoperative weight loss of greater than 10% of body weight. Pancreatectomyprocedure targeted variables analyzed were histology, minimally invasive surgery (MIS) approach, gland texture, duct size, transfusion requirement, and operative time. All outcomes within the ACS-NSQIP database are 30-day outcomes. The post-operative outcomes analyzed in this study included pneumonia, reintubation, failure to wean, renal insufficiency, renal failure, myocardial infarction (MI), cardiac arrest, stroke, sepsis, septic shock, venous thromboembolism (VTE), surgical site infection (SSI), wound dehiscence, organ space infection, hospital length of stay (HLOS), discharge destination, and death within 30 days. Pancreatectomy-specific outcomes assessed were pancreatic fistula and need for percutaneous drainage. Abbreviated definitions are provided in Table 1. Analysis Univariate modeling was used to examine the associations of decade with outcomes. Mann– Whitney U and Kruskal–Wallis

HPB 2018, -, 1–8

Table 1 Abbreviated definitions of ACS-NSQIP variables and out-

comes. Full definitions accessed at: https://www.facs.org/~/media/ files/quality%20programs/nsqip/ug12.ash and https://www.facs. org/~/media/files/quality%20programs/nsqip/pt_nsqip_puf_ userguide_2016.ashx Variable

Definition

Functional status

Ability to perform of activities of daily living (ADL) within 30 days prior to surgery Independent: Does not require assistance from another person for ADL Partially dependent: Requires some assistance from another person for ADL Totally dependent: Requires total assistance from another person for all ADL

Pneumonia

Radiologic findings, AND Clinical signs with microbiological confirmation OR Respiratory symptoms

Reintubation

Unplanned endotracheal or other breathing tube with ventilator support intraoperatively or within 30 days

Failure to wean

Ventilator requirement >48 h after surgery

Renal insufficiency

Creatinine increase >2 mg/dl from preoperative value, without need for dialysis

Renal failure

Renal insufficiency with need for dialysis

Cardiac arrest

Need for chest compressions or defibrillations

Myocardial infarction

Indicative ECG changes OR new troponin elevation >3× upper limit of reference range OR physician diagnosis

Stroke

Embolic, thrombotic or hemorrhagic vascular accident or stroke with dysfunction lasting >24 h

Sepsis

Two or more SIRS criteria AND positive culture or source, OR Operative confirmation of infarcted bowel, purulence, enteric contents

Septic shock

Sepsis AND organ and/or circulatory dysfunction

Surgical site infection

Infection involving only skin or subcutaneous tissue with purulent drainage OR microbiologic confirmation OR opening of wound with suspicion

Wound dehiscence

Loss of integrity of fascial closure or total breakdown of closure

Organ space infection

Infection involved organs or spaces other than the incision, confirmed with purulence OR microbiologic confirmation OR physician diagnosis

VTE

New thrombus treated with anticoagulation or vena cava filter

Reoperation

Unplanned return to the operating room within 30 days of surgery

Pancreatic fistula

Clinical diagnosis OR persistent drainage with one of the following: nihil per os and parenteral nutrition, drain continued beyond 7 days, percutaneous drainage, reoperation, or spontaneous wound drainage

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

3

parametric Nemenyi-Damico-Wolfe–Dunn test was used for pairwise comparisons of continuous variables between the age groups. For those postoperative outcomes that demonstrated a significant association with decade of age, multivariable models were used to assess whether decade of age was an independent contributor after adjusting for differences in baseline characteristics. Baseline characteristics with differences between decades at a significance of p < 0.10 on univariate analysis were included in the models. Binary logistic regression models were used for analyzing categorical outcomes, while ordinary least squares linear regression models were used for the continuous outcomes. All analyses were performed using JMP 12 (SAS Institute, Cary, NC). Institutional review board (IRB) approval IRB approval was waived for this study as the ACS-NSQIP is a de-identified publicly available database.

Results Between 2014 and 2015, 1626 patients underwent DP for PNET or PCN. 692 (42.6%) were younger than 60 years, 507 (31.2%) were sexagenarians, 342 (21.0%) were septuagenarians, and 85 (5.2%) were octogenarians. In comparison to those younger than 60 years, there were higher proportions of females in the older age groups. With increasing decade of age, BMI steadily decreased from 30.3 kg/m2 in those below 60 years to 29.4 kg/m2 among sexagenarians, 28.7 kg/m2 among septuagenarians, and 26.8 kg/m2 among octogenarians (p < 0.01) (Table 2). Higher prevalence of baseline comorbid conditions in older patients Older decades were associated with increased prevalence of diabetes, COPD, hypertension, CHF, and non-independent functional status as compared to those below 60 years. An exception to this was among octogenarians, whom had a decreased prevalence of diabetes. The median and mode amFI

Table 2 Baseline characteristics by decade of age

All patients (n [ 1626)

Group A <60 years (n [ 692)

Group B 60–69 years (n [ 507)

Group C 70–79 years (n [ 342)

Group D ‡80 years (n [ 85)

P-Valuea

Mean Age, years (SD)

59.8 (14.1)

46.4 (10.0)

64.7 (2.7)

74.1 (2.7)

82.4 (2.4)

<0.001

Male Gender, n (%)

644 (39.6)

201 (29.1)

241 (47.5)

165 (48.3)

37 (43.5)

<0.001

Mean BMI, kg/m2 (SD)

29.5 (6.6)

30.3 (7.2)

29.4 (6.3)

28.7 (5.6)

26.8 (5.0)

<0.001

Malnutrition, n (%)

371 (23.6)

163 (23.6)

118 (23.3)

69 (20.2)

21 (24.7)

0.614

Diabetes, n (%)

341 (21.0)

95 (13.7)

137 (27.0)

97 (28.4)

12 (14.1)

<0.001

COPD, n (%)

50 (3.1)

10 (1.5)

17 (3.4)

19 (5.6)

4 (4.7)

0.003

Hypertension, n (%)

800 (49.2)

206 (29.8)

305 (60.2)

227 (66.4)

62 (72.9)

<0.001

CHF, n (%)

8 (0.5)

0 (0)

6 (1.2)

2 (0.6)

0 (0)

0.031

Non-independent, n (%)

16 (1.0)

1 (0.1)

6 (1.2)

9 (2.6)

0 (0)

0.001

amFI, median (IQR)

1 (0–1)

0 (0–1)

1 (0–2)

1 (0–2)

1 (1)

<0.001

PNET, n (%)

742 (45.6)

320 (46.3)

273 (53.9)

113 (33.0)

36 (42.4)

<0.001

PCN, n (%)

844 (54.4)

372 (53.8)

234 (46.2)

229 (67.0)

49 (57.7)

<0.001

MIS, n (%)

1014 (62.4)

443 (64.0)

292 (57.6)

223 (65.2)

56 (65.9)

0.062

Soft Gland, n (%)

488 (30.0)

214 (73.5)

153 (74.3)

98 (68.5)

23 (88.5)

0.185

Small Duct, n (%)

258 (15.9)

113 (68.5)

73 (50.3)

64 (54.7)

8 (34.8)

<0.001

Transfusion, n (%)

138 (8.5)

48 (6.9)

44 (8.7)

38 (11.1)

8 (9.4)

0.152

Median operative time, min (IQR)

197 (144–265)

210 (150–285)

188 (142–259)

184 (139–235)

191 (137.5–248)

<0.001

Male gender: A vs. B, p < 0.01; A vs. C, p < 0.01; A vs. D, p < 0.01. BMI: A vs. C, p = 0.03; A vs. D, p < 0.01; B vs. D, p < 0.01. Diabetes: A vs. B, p < 0.01; A vs. C, p < 0.01; B vs. D, p < 0.01; C vs. D, p < 0.01. COPD: A vs. B, p = 0.03; A vs. C, p < 0.01. Hypertension: A vs. B, p < 0.01; A vs. C, p < 0.01; A vs. D, p < 0.01; B vs. D, p = 0.02. CHF: A vs. B, p < 0.01; A vs. C, p = 0.04. Non-independent: A vs. B, p = 0.02; A vs. C, p < 0.01; C vs. D, p = 0.04. amFI: A vs. B, p < 0.01; A vs. C, p < 0.01; A vs. D, p < 0.01. Small duct: A vs. B, p < 0.01; A vs. C, p = 0.02; A vs. D, p < 0.01. Median operative time: A vs. B, p < 0.01; A vs. C, p < 0.01. a Pairwise analysis of statistically significant (p < 0.05) differences between groups.

HPB 2018, -, 1–8

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

4

was 1 amongst all the older decades, indicating that most older patients had only one of the five measured conditions. 77.6% of octogenarians, 72.2% of septuagenarians, 65.3% of sexagenarians, and 34.1% of those below 60 years were affected by at least one of the five comorbidities, as defined by ACS-NSQIP (Table 2). Shorter operative time and smaller ducts in older decades Operative times were longest in those below 60 years (median 210 min) as compared to the older decades (p < 0.01). Despite a higher BMI among those below 60 years, there was no correlation between BMI and operative time using linear regression (r2 = 0.02, p = 0.15). Additionally, there were higher proportions of patients with small ducts (<2 mm) in the older age groups, with the greatest statistical difference between those below age 60

years and octogenarians (68.5% vs. 34.8%, p < 0.01). There were no statistically significant differences in the use of MIS (57–65%), gland texture, or transfusion requirements (Table 3). Higher incidence of complications and prolonged HLOS amongst older decades, but less need for percutaneous drainage There were seven mortalities (0.4%), and 308 (18.9%) patients experienced postoperative complications. Overall, the frequency of each individual complication was low; of the 1626 patients, there were 18 reintubations, 18 failures to wean, 8 renal failures, and 17 MI. Octogenarians had greater incidence of reintubation than those <60 years and sexagenarians (4.7% vs. 0.6% and 0.8%, p < 0.01). Compared to those <60 years, septuagenarians and octogenarians had higher rates of failure to wean (1.1% vs. 2.3% and 4.7%, p < 0.01). Similarly, septuagenarians and

Table 3 Postoperative complications with pairwise analysis by decade of age

All patients (n [ 1626)

Group A Group B Group C Group D P-Valuea <60 years (n [ 692) 60–69 years (n [ 507) 70–79 years (n [ 342) ‡80 years (n [ 85)

Pneumonia, n (%)

34 (2.1)

12 (1.73)

10 (1.97)

8 (2.34)

4 (4.71)

0.333

Reintubation, n (%)

18 (1.1)

4 (0.6)

4 (0.8)

6 (1.8)

4 (4.7)

0.003

Failure to wean, n (%)

18 (1.1)

2 (0.3)

4 (0.8)

8 (2.3)

4 (4.7)

<0.001

Renal insufficiency, n (%)

8 (0.5)

3 (0.43)

2 (0.39)

2 (0.58)

1 (1.18)

0.796

Renal Failure, n (%)

8 (0.5)

2 (0.3)

4 (0.8)

0 (0)

2 (2.4)

0.027

Cardiac Arrest, n (%)

6 (0.4)

4 (0.58)

2 (0.39)

0 (0)

0 (0)

0.491

MI, n (%)

17 (1)

1 (0.1)

2 (0.4)

10 (2.9)

4 (4.7)

<0.001

Stroke, n (%)

2 (0.1)

0 (0)

2 (0.39)

0 (0)

0 (0)

0.22

Sepsis, n (%)

87 (5.4)

33 (4.77)

36 (7.1)

16 (4.68)

2 (2.35)

0.148

Septic Shock, n (%)

7 (0.4)

3 (0.4)

0 (0)

2 (0.6)

2 (2.4)

0.021

VTE, n (%)

37 (2.3)

14 (2.02)

9 (1.78)

10 (2.92)

4 (4.71)

0.299

Dehiscence, n (%)

9 (0.6)

4 (0.58)

5 (0.99)

0 (0)

0 (0)

0.25

Organ space infection, 167 (10.3) n (%)

77 (11.13)

59 (11.64)

27 (7.89)

4 (4.71)

0.088

SSI, n (%)

43 (2.6)

21 (3.03)

12 (2.37)

8 (2.34)

2 (2.35)

0.87

Mortality, n (%)

7 (0.4)

2 (0.29)

3 (0.59)

2 (0.58)

0 (0)

0.756

Discharge home, n (%) 52 (3.2)

27 (8.54)

14 (6.01)

7 (4.58)

4 (10.26)

0.318

Readmission, n (%)

267 (16.4)

136 (19.7)

76 (15.0)

43 (12.6)

12 (14.1)

0.018

Reoperation, n (%)

37 (2.3)

14 (2.0)

13 (2.6)

4 (1.2)

6 (7.1)

0.007

HLOS, days (SD)

6.2 (5.2)

5.8 (4.1)

6.0 (3.4)

7.1 (7.7)

7.8 (8.7)

0.012

Pancreatic fistula, n (%)

348 (21.4)

152 (22.19)

110 (21.87)

73 (21.47)

13 (15.92)

0.539

Perc. Drainage, n (%)

212 (13.0)

102 (14.7)

74 (14.6)

32 (9.4)

4 (4.7)

<0.001

Reintubation: A vs. D, p < 0.01; B vs. D, p = 0.02. Failure to wean: A vs. C, p < 0.01; A vs. D, p < 0.01; B vs. D, p = 0.02. MI: A vs. C, p < 0.01; A vs. D, p < 0.01; B vs. C, p < 0.01; B vs. D, p < 0.01. Readmission: A vs. B, p = 0.04; A vs. C, p < 0.01. Reoperation: A vs. D, p = 0.02; C vs. D, p < 0.01. HLOS: A vs. C, p < 0.01; A vs. D, p < 0.01. Perc. Drainage: A vs. C, p = 0.01; A vs. D, p < 0.01; B vs. C, p = 0.02; B vs. D, p < 0.01. a Pairwise analysis of statistically significant (p < 0.05) differences between groups.

HPB 2018, -, 1–8

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

5

octogenarians had higher incidence of MI, as compared to those <60 years and sexagenarians (2.9% and 4.7% vs. 0.1% and 0.4%, p < 0.01) (Table 3). Septuagenarians and octogenarians had significantly longer HLOS as compared to those <60 years and sexagenarians (7.7 and 8.7 vs. 5.8 and 6.0 days, p = 0.01). Notably, readmission was higher amongst those <60 years as compared to sexagenarians and septuagenarians (19.7% vs. 15.0% and 12.6%, p = 0.02), but not significantly different from octogenarians. There was no statistical difference in rates of discharge home versus facility (Table 3). Of the 1626 patients, 212 (13.0%) required percutaneous drainage. Septuagenarians and octogenarians had decreased need for percutaneous drainage as compared to those <60 years and sexagenarians (9.4% and 4.7% vs. 14.7% and 14.6%, p < 0.01). There were no differences in the development of sepsis or septic shock, however reoperations were more common among octogenarians as compared to those <60 years and septuagenarians (7.1% vs. 2.0% and 1.2%, p < 0.01) (Table 3). The overall incidence of pancreatic fistula was 21.4% with no significant differences among the age groups. Postoperative complications disproportionately affected older decades Although the incidences of reintubation, failure to wean, renal failure, and MI were low, the distribution of these complications were skewed towards older decades. Septuagenarians and octogenarians constituted 26.3% of the whole cohort, but were responsible for 55.6% of reintubations, 66.7% of failures to wean, 82.4% of MI, and 57.1% of septic shock (Fig. 1).

<60 years

60-69 years

Multivariable analysis Multivariable modeling was used to investigate the independent contribution of decade of age to each of the statistically significant post-operative complications. The following baseline characteristics and operative factors with significance of p < 0.10 on univariate analysis were included: sex, log-transformed BMI, diabetes, COPD, hypertension, CHF, functional status, MIS, operative time, small duct, and histology. In the multivariable model, decade of age was independently associated with decreased need for percutaneous drainage. Compared to those less than age 60 years, sexagenarians had 1.61 lower odds (95% CI: 1.01–3.50), septuagenarians had 1.60 lower odds (95% CI: 1.03–3.92), and octogenarians had 3.58 lower odds of requiring drainage (95% CI: 1.02–25.31). Although the multivariable model displayed statistically significant associations between decade of age and MI, renal failure, and reoperation, the confidence intervals were very large indicating that these statistical relationships were not valid. The unreliably large confidence intervals are likely related to the low frequency of those events. Decade of age was not associated with reintubation, failure to wean, septic shock, and readmission.

Discussion Unlike in PDAC, where in the absence of prohibitive comorbidities all technically resectable lesions benefit from resection, the management of PNET and PCN is more nuanced. Furthermore, most patients with PNET and PCN are asymptomatic from their disease. This study therefore focuses on patients with PNET and PCN, in order to help guide clinical and shared decision-

70-79 years

≥ 80 years

100%

80%

60%

40%

20%

0%

Figure 1 Contribution to postoperative complications by decade. Older decades disproportionately affected by complications

HPB 2018, -, 1–8

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

6

making regarding DP. This study identifies associations between older ages and increased perioperative complications. Specifically analyzing the association between age and postoperative outcomes reveals that despite the low overall complication rate after DP, the cardiopulmonary and renal complications disproportionately affect septuagenarians and octogenarians. DP has long been recognized to be a lower-risk operation with fewer physiologic derangements than PD.13–19 A retrospective series of laparoscopic DP in Norway reported that patients over age 70 years demonstrated lower postoperative morbidity compared to those below 70 years, concluding that age did not place patient at higher risk of systemic postoperative complications.10 Upon closer examination, however, the Norwegian cohort had mean BMI of 25 kg/m2, with only 40% having baseline hypertension and 18% having diabetes. By contrast, the septuagenarians in the ACS-NSQIP dataset had a mean BMI of 28.7 kg/m2, 66% incidence of hypertension, and 28% incidence of diabetes, suggesting that outcomes of older adults from Northern Europe may not reflect those of a North American cohort. This study affirms the low overall incidence of postoperative morbidity after DP, but does illustrate that older age poses a higher risk for complications. The higher baseline incidence of CHF, COPD and hypertension in older adults likely predisposes them to have the few MI, renal failures and respiratory complications noted in the dataset. It is also important to acknowledge the strict definitions used for the comorbidities by ACS-NSQIP, especially with regards to time of diagnosis relative to the operation date. These stringent criteria likely result in underrepresentation of the true incidence of comorbidities affecting older adults. Thus, while our data suggests that age is an independent risk factor for perioperative complications, these conclusions are tempered by the potential underestimation of comorbidities, and the low event rate. The low event rates of the cardiopulmonary and renal perioperative complications limit the statistical validity of the multivariable analysis. A meta-analysis of both PD and DP highlighted increased incidence of pneumonia and length of stay in the elderly, however noted no increase in pancreatectomy-specific complications such as POPF, delayed gastric emptying or bleeding.6 In retrospective series of DP alone, older age has been associated with no increased need for percutaneous drainage and decreased POPF risk, presumably due to the fatty involution of the pancreas and decrease in exocrine function.7–9 In ACS-NSQIP, we also noted a lower need for percutaneous drainage and similar POPF rate. This suggests that older adults are not at increased risk for technical complications, however are at increased risk of systemic complications as illustrated by the increased rate of septic shock and reoperation. Limitations in the dataset do not allow more granular analysis of the indications for reoperation, as the higher rate of reoperations could be associated with the lower percutaneous intervention rate among octogenarians. In this study, we also found a shorter operative time by greater than 20 min amongst older adults. Limitations of the dataset do

HPB 2018, -, 1–8

not provide insight as to why, but it could be related to higher risk lesions in the younger populations, or a deliberate approach to minimize anesthetic exposure in older patients. Similarly, older adults had marginally, but statistically significant, longer hospital stays, and lower rates of readmission with no difference in discharge to a facility. This suggests that in older adults, surgeons may be electing to appropriately prolong hospital stay. The ACS-NSQIP dataset does not have information on the use of enhanced recovery after surgery (ERAS) pathways, which may influence the relationships between HLOS and readmission. PNET are the second most common neoplasm of the pancreas, and rates of detection of PNET are increasing. This is attributed primarily to the increased utilization and improved resolution of cross-sectional imaging, rather than a true increase in the disease incidence.20 The increased detection of incidental PNET poses a dilemma for clinicians who aim to balance the potential for tumor-related morbidity with expected operative morbidity. In fact, an early study that adopted an aggressive surgical approach for incidental, nonfunctioning PNET found that nearly half of all patients experienced a complication after DP, reaffirming that DP too carries a significant risk for postoperative complications.5 Numerous studies since have shown that management of small PNET with imaging surveillance is safe, and does not result in worse with tumor-related survival.21–26 Current PNET management guidelines from various societies recognize the role of nonoperative surveillance, however none incorporate age into the decision analysis.27–29 Similar to PNET, there has been an increase in the detection of PCN. The overall prevalence of incidental pancreatic cysts has been reported as 13.5%, however in those over 70 years, the cumulative prevalence was 40.2%.6 Management of PCN is area with evolving guidelines that incorporate symptoms, imaging and endoscopic criteria, and fluid and molecular analyses to predict risk of malignancy.22,30–32 Imaging surveillance is an acceptable strategy for select PCN, with reported crossover rates to resection of approximately 6.5%.22 A strength of this analysis is the large patient numbers and the granular procedure-specific data afforded by the pancreatectomy procedure-targeted dataset. The ACS-NSQIP dataset offers several additional advantages over pure administrative data registries. ACS-NSQIP is a nationally validated registry and uses a combination of administrative data and trained NSQIP abstractors to collect data according to strict predefined definitions. This overcomes the suboptimal validity of pure administrative databases, which suffer from over-reliance on accuracy of medical coders and physician documentation.33,34 However, there are several limitations related to the datasets used. Inherent is a selection bias in that all the analyzed patients have undergone DP, thereby precluding an intent-to-treat analysis. Secondly, the decision to resect a PNET often involves a period of observation to document growth velocity, which alters malignant risk and increases the urgency to resect. ACS-NSQIP data do not allow us to speculate on the period of preoperative observation or growth velocity. PCN is an umbrella term inclusive of pathologies with a range of

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

7

malignant potential from MCN to SCN. Imaging, endoscopic analysis, fluid studies and molecular characterization provide additional data about the likely pathology and malignant potential – this is not captured in the ACS-NSQIP PUF or the pancreatectomy PUF. ACS-NSQIP reliably captures data on the pathologic diagnosis of the specimen, but does not as reliably capture the preoperative diagnosis, thereby limiting our insight into the preoperative factors guiding the surgeon’s decision to embark on resection. The ACS-NSQIP dataset does not contain information on overall or disease-free survival, limiting survival analysis. Additionally, it does not provide insight into the rate of discovery of invasive cancer in patients undergoing DP for presumed noninvasive PNET and PCN. These factors limit the ability of this analysis to speculate on the potential benefits of resection versus observation alone. This study does, however, provide meaningful data on the age-related risks of complications, which can be used to guide the discussion of risks of resection.

4. Kimura W, Nagai H, Kuroda A, Muto T, Esaki Y. (1995 Dec) Analysis of small cystic lesions of the pancreas. Int J Pancreatol 18:197–206. 5. Haynes AB, Deshpande V, Ingkakul T, Vagefi PA, Szymonifka J, Thayer SP et al. (2011 May) Implications of incidentally discovered, nonfunctioning pancreatic endocrine tumors: short-term and long-term patient outcomes. Arch Surg 146:534–538. 6. Sukharamwala P, Thoens J, Szuchmacher M, Smith J, DeVito P. (2012 Oct) Advanced age is a risk factor for post-operative complications and mortality after a pancreaticoduodenectomy: a meta-analysis and systematic review. HPB 14:649–657. 7. Paye F, Micelli Lupinacci R, Bachellier P, Boher J-M, Delpero J-R. (2015 Feb) Distal pancreatectomy for pancreatic carcinoma in the era of multimodal treatment. Br J Surg 102:229–236. 8. Ecker BL, McMillan MT, Allegrini V, Bassi C, Beane JD, Beckman RM et al. (2017 Aug 29) Risk factors and mitigation strategies for pancreatic fistula after distal pancreatectomy: analysis of 2026 resections from the international, multi-institutional distal pancreatectomy study group. Ann Surg. https://doi.org/10.1097/SLA.0000000000002491 [Epub ahead of print]. 9. Yoshioka R, Saiura A, Koga R, Seki M, Kishi Y, Morimura R et al. (2010 Jan) Risk factors for clinical pancreatic fistula after distal pancrea-

Conclusions

tectomy: analysis of consecutive 100 patients. World J Surg 34:

This study adds data on the perioperative risks for those undergoing DP specifically for PNET and PCN. We report that despite a low overall rate of postoperative complications, the majority of those complications affect septuagenarians and octogenarians. In contrast to DP for PDAC or benign disease, the long-term outcomes and management strategies differ for those with PNET and PCN. These data can thus help inform the decision-making process of resection versus observation in older adults from the perspective of short-term perioperative risk factors and complications.

121 –125. 10. Sahakyan MA, Edwin B, Kazaryan AM, Barkhatov L, Buanes T, Ignjatovic D et al. (2017 Jan) Perioperative outcomes and survival in elderly patients undergoing laparoscopic distal pancreatectomy. J Hepatobiliary Pancreat Sci 24:42–48. 11. ACS National Surgical Quality Improvement Program [Internet]. [cited 2018 Nov 28]. Available from: https://www.facs.org/quality-programs/ acs-nsqip. 12. Subramaniam S, Aalberg JJ, Soriano RP, Divino CM. (2018 Feb) New 5factor modified frailty index using american College of surgeons NSQIP data. J Am Coll Surg 226:173–181.e8. 13. Yamada S, Shimada M, Utsunomiya T, Morine Y, Imura S, Ikemoto T et al. (2012 Sep) Surgical results of pancreatoduodenectomy in elderly

Funding

patients. Surg Today 42:857–862.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

14. Scurtu R, Bachellier P, Oussoultzoglou E, Rosso E, Maroni R, Jaeck D. (2006 Jun) Outcome after pancreaticoduodenectomy for cancer in elderly patients. J Gastrointest Surg 10:813–822.

Author contributions

15. Ito Y, Kenmochi T, Irino T, Egawa T, Hayashi S, Nagashima A et al.

RR contributed to conception, data acquisition, analysis, interpretation drafting and revision. AIA contributed to interpretation, drafting and revision.

(2011 Sep 11) The impact of surgical outcome after pancreaticoduodenectomy in elderly patients. World J Surg Oncol 9:102.

TM contributed to analysis, interpretation and revision. LGW contributed to

16. Bathe OF, Levi D, Caldera H, Franceschi D, Raez L, Patel A et al. (2000

analysis and interpretation. BJK contributed to conception, drafting and

Mar) Radical resection of periampullary tumors in the elderly: evaluation

revision.

of long-term results. World J Surg 24:353–358. 17. Barbas AS, Turley RS, Ceppa EP, Reddy SK, Blazer DG, 3rd, Clary BM

Conflicts of interest

et al. (2012 Feb) Comparison of outcomes and the use of multimodality

None declared.

therapy in young and elderly people undergoing surgical resection of pancreatic cancer. J Am Geriatr Soc 60:344–350. 18. Ballarin R, Spaggiari M, Di Benedetto F, Montalti R, Masetti M, De

References 1. Cancer Statistics [Internet]. National Cancer Institute. [cited 2018 Nov 28]. Available from: https://www.cancer.gov/about-cancer/understan

patients. J Gastrointest Surg 13:341–348. 19. Ramanathan R, Mason T, Wolfe LG, Kaplan BJ. (2018 Feb 5) Predictors of

ding/statistics. 2. Berger NA, Savvides P, Koroukian SM, Kahana EF, Deimling GT, Rose JH et al. (2006) Cancer in the elderly. Trans Am Clin Climatol

short-term readmission after pancreaticoduodenectomy. J Gastrointest Surg 22:998–1006. 20. Kuo EJ, Salem RR. (2013 Sep) Population-level analysis of pancreatic

Assoc 117:147–155. discussion 155-156. 3. Sial SH, Catalano MF. (2001 Jun) Gastrointestinal tract cancer in the elderly. Gastroenterol Clin North Am 30:565–590.

HPB 2018, -, 1–8

Ruvo N et al. (2009 Feb) Do not deny pancreatic resection to elderly

neuroendocrine tumors 2 cm or less in size. Ann Surg Oncol 20: 2815–2821.

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015

HPB

8

21. Bettini R, Partelli S, Boninsegna L, Capelli P, Crippa S, Pederzoli P et al.

28. Kunz PL, Reidy-Lagunes D, Anthony LB, Bertino EM, Brendtro K,

(2011 Jul) Tumor size correlates with malignancy in nonfunctioning

Chan JA et al. (2013 May) Consensus guidelines for the management

pancreatic endocrine tumor. Surgery 150:75–82.

and treatment of neuroendocrine tumors. Pancreas 42:557–577.

22. Gaujoux S, Partelli S, Maire F, D’Onofrio M, Larroque B, Tamburrino D

29. Oberg K, Knigge U, Kwekkeboom D, Perren A. (2012 Oct) Neuroen-

et al. (2013 Dec) Observational study of natural history of small sporadic

docrine gastro-entero-pancreatic tumors: ESMO Clinical Practice

nonfunctioning pancreatic neuroendocrine tumors. J Clin Endocrinol

Guidelines for diagnosis, treatment and follow-up. Ann Oncol

Metab 98:4784–4789.

23(Suppl. 7):vii124–vii130.

23. Lee LC, Grant CS, Salomao DR, Fletcher JG, Takahashi N, Fidler JL

30. Allen PJ, D’Angelica M, Gonen M, Jaques DP, Coit DG, Jarnagin WR

et al. (2012 Dec) Small, nonfunctioning, asymptomatic pancreatic

et al. (2006 Oct) A selective approach to the resection of cystic lesions

neuroendocrine tumors (PNETs): role for nonoperative management.

of the pancreas: results from 539 consecutive patients. Ann Surg 244: 572–582.

Surgery 152:965–974. 24. Rosenberg AM, Friedmann P, Del Rivero J, Libutti SK, Laird AM. (2016

31. Tanaka M, Chari S, Adsay V, Fernandez-del Castillo C, Falconi M,

Jan) Resection versus expectant management of small incidentally

Shimizu M et al. (2006) International consensus guidelines for man-

discovered nonfunctional pancreatic neuroendocrine tumors. Surgery

agement of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology 6:17–32.

159:302–309. 25. Sadot E, Reidy-Lagunes DL, Tang LH, Do RKG, Gonen M,

32. Singhi AD, McGrath K, Brand RE, Khalid A, Zeh HJ, Chennat JS et al.

D’Angelica MI et al. (2016 Apr) Observation versus resection for small

(2017 Sep 28) Preoperative next-generation sequencing of pancreatic

asymptomatic pancreatic neuroendocrine tumors: a matched case-

cyst fluid is highly accurate in cyst classification and detection of

control study. Ann Surg Oncol 23:1361–1370. 26. Uribe Galeano C, Fabregat Prous J, Busquets Barenys J, Pelaez

advanced neoplasia. Gut 67:2131–2141. 33. Ramanathan R, Leavell P, Stockslager G, Mays C, Harvey D, Duane TM.

Serra N, Secanella Medayo L, Ramos Rubio E et al. (2017 Feb)

(2014 Oct) Validity of international classification of diseases, ninth

Nonfunctioning, small, incidental pancreatic neuroendocrine tumors:

revision, clinical modification (ICD-9-CM) screening for sepsis in sur-

results of a nonoperative approach cohort. Cir Esp 95:83–88.

gical mortalities. Surg Infect 15:513–516.

27. Falconi M, Eriksson B, Kaltsas G, Bartsch DK, Capdevila J, Caplin M et al.

34. Ramanathan R, Leavell P, Stockslager G, Mays C, Harvey D, Duane TM.

(2016) ENETS consensus guidelines update for the management of patients

(2013 Jun) Validity of agency for healthcare research and quality patient

with functional pancreatic neuroendocrine tumors and non-functional

safety indicators at an academic medical center. Am Surg 79:578–582.

pancreatic neuroendocrine tumors. Neuroendocrinology 103:153–171.

HPB 2018, -, 1–8

© 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ramanathan R, et al., Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms, HPB (2018), https://doi.org/10.1016/j.hpb.2018.09.015