Long-Term Assessment of Pancreatic Function after Pancreatectomy for Cystic Neoplasms

Abstract

Background:

With advances in cross-sectional imaging, pancreatic cysts are more frequently diagnosed and have become a common indication for pancreatectomy. The impact of pancreatectomy in these patients is important. The purpose of this study was to assess short-term outcomes, long-term nutritional status, quality of life (QOL), and pancreas function after pancreatectomy for cystic neoplasms.

Materials and Methods:

At a single institution, patients at least three years post-pancreatectomy for benign cystic neoplasms were identified. Using a validated questionnaire, short-term outcomes, long-term outcomes including endocrine and exocrine insufficiency, long-term nutritional status, and pre- and post-operative QOL were compared based upon operation and indication for resection.

Results:

Among 102 eligible patients, 70 had valid contact information and 51 (72.9%) agreed to participate. Median follow-up was 6 (4–8) years. Patients undergoing pancreatectomy for benign cysts had higher morbidity than a similar cohort resected for pancreatic adenocarcinoma (patients with at least 1 ≥grade 2 complication [49.0% vs 31.6%, P = 0.038]). After long-term follow up, pancreatectomy did not significantly affect perceived QOL. Half of patients had mild-moderate or severe malnourishment but pancreatic enzyme replacement was reported by only 4 (7.8%) patients. New-onset diabetes was present in 15 (29.4%) patients with median time to diagnosis of 6 (1–12) months post-resection.

Conclusion:

Pancreatectomy for benign cysts did not negatively impact patients’ perceived QOL. However, after long-term follow-up, malnutrition and pancreatic insufficiency occurred in a significant percentage and may be greater than previously estimated. Consideration of short and long-term outcomes should factor into preoperative counseling, especially in cysts with minimal risk of progression to malignancy.

INTRODUCTION

Pancreatic cysts are increasingly becoming a common imaging finding with a prevalence of 2% in the general population.¹ This estimate is likely conservative due to the widespread use and advancement of cross-sectional imaging. The majority of incidentally found cystic lesions are benign and require no further medical intervention.^2,3 However, preoperative characterization and prediction of the behavior of these lesions remains challenging as cystic lesions can vary widely in their risk of malignant transformation. Due to the aggressive nature of pancreatic ductal adenocarcinoma (PDAC) even at an early stage, pancreatectomy remains the cornerstone of treatment for cystic lesions with multiple high risk features including solid component, dilated main pancreatic duct, and concerning features on endoscopic ultrasound with fine needle aspiration (EUS-FNA).^4,5

Patient-centered outcomes, such as quality of life (QOL) are an important factor to consider following pancreatectomy. Prior studies examining long-term outcomes, such as overall survival and QOL in pancreatectomy patients have been done, but are focused largely on PDAC survivors.^6–11 Historically, pancreatectomy was reserved for malignant pathology, but lesions are now identified much earlier in the spectrum of disease. Patients, benefitting from detection of premalignant pathology, are now living longer after surgery, and the understanding of long-term outcomes including QOL, malnutrition, exocrine, and endocrine insufficiency are becoming increasingly relevant. Only a few studies have examined long-term outcomes in patients undergoing resection for benign cystic lesions but all have been centered on European populations.^12,13 The generalizability of those studies to the United States is consequently limited due to the inherent differences in patient populations and access to nationalized health care. Examination of patient outcomes following resection for benign disease in the United States is needed to direct pre-operative decision making and patient counseling.

Therefore, the objectives of this study were to evaluate the short-term morbidity of pancreatectomy for cystic neoplasms in comparison to pancreatectomy for pancreatic adenocarcinoma (PDAC) and to determine long-term pancreatic function following pancreatoduodenectomy and distal pancreatectomy for pancreatic cysts. We hypothesized that short-term morbidity would be higher in patients with lesions undergoing resection for cystic neoplasms due to the presence of a soft gland and a non-dilated duct and that long term pancreatic function, nutrition, and quality of life would be affected.

MATERIALS AND METHODS

Data Source

We analyzed data from a prospectively maintained Pancreas Surgery Registry in a high-volume academic pancreas center to identify patients undergoing pancreatectomy for cystic neoplasms. The database stores information including baseline demographics, comorbidities, perioperative characteristics, and graded complications within 60 days of surgery using the Accordion Severity Grading for Surgical Complications¹⁴ and the International Study Group of Pancreatic Fistula and International Study Group of Pancreatic Surgery definitions^15,16 for pancreatic fistula and delayed gastric emptying. A detailed definition of complications has been previously reported.¹⁷

Study Cohort

We identified patients undergoing pancreatoduodenectomy and distal pancreatectomy at a single institution between January 2005 and January 2014. All patients with benign cystic neoplasms on final pathology were included and cystic neoplasms with malignant features were excluded, including pancreatic adenocarcinoma, cholangiocarcinoma, ampullary carcinoma, duodenal adenomas, and tumors metastatic to the pancreas. Cystic lesions in the setting of neuroendocrine tumors and chronic pancreatitis were also excluded. Patients meeting inclusion criteria were contacted and consented by telephone. Permission for this study was obtained from an Institutional Review Board (H-35341). Surgeons in this study followed the consensus set forth by the International Association of Pancreatology, including the Sendai and Fukuoka guidelines for management of cystic neoplasms of the pancreas.^18–20

Short-Term Outcomes

All patients who underwent pancreatectomy for benign cysts were compared to a cohort of all patients who underwent pancreatectomy for PDAC within the same time period. Short-term outcomes were identified using the database as well as a comprehensive chart review when applicable. Outcomes included all complications within 60-days, length of stay (LOS), and overall mortality within 90-days.²¹ Short-term complications were graded using the Accordion Severity Grading for Surgical Complications.¹⁴ The Pancreatic Fistula Risk Score was calculated using an online medical calculator.^22,23

Long-Term Outcomes and Survey

Long-term outcomes were examined in patients who were resected for benign cysts. Patients who underwent resection for PDAC were excluded due to concerns regarding long term survival and confounding variables such as adjuvant therapy affecting QOL. Data on long-term outcomes were obtained through a telephone survey evaluating four sections of interest: (1) QOL, (2) nutritional status, (3) pancreatic endocrine, and (4) exocrine function. The survey was comprised of validated questions from the Subjective Global Assessment (SGA)²⁴ and the Functional Assessment of Cancer Therapy (FACT-Hep)²⁵ questionnaires. Questions about body mass index, gastrointestinal symptoms (nausea, vomiting, diarrhea, pain on eating), dietary intake, weight loss (recent or long-term), and day-to-day functional capacity (ability to perform activities of daily living) were derived from the SGA questionnaire. The FACT-Hep questionnaire provided validated questions concerning the following subcategories to assess QOL: physical well-being (PWB), social well-being (SWB), emotional well-being (EWB), functional well-being (FWB), and hepatobiliary specific subscale (HCS). The remaining variables including alcohol and tobacco intake, hospitalizations, and symptoms of exocrine and endocrine insufficiency were developed by the authors for this study. The Centers for Disease Control and Prevention (CDC) and National Institute of Health (NIH) definition of heavy drinking (15 drinks weekly or 5 or more on occasion for men and 8 drinks weekly or 4 or more on occasion for women) was used to report clinically significant alcohol intake.^26,27 Data from the questionnaire, including long-term nutritional status and incidence of pancreatic endocrine and exocrine insufficiency, was compared between pancreatoduodenectomy and distal pancreatectomy patients.

Statistical Analysis

Chi-square, Fisher’s exact, or McNemar’s test, when appropriate, was used to analyze categorical variables. The student’s t-test or Mann-Whitney test was used to evaluate continuous variables. All statistical analyses were performed using SPSS v25 (IBM Corp. Armonk NY, USA). Median values are reported with interquartile range (IQR) in the following format: median (IQR).

RESULTS

Cohort Characteristics

We identified 115 total patients who underwent pancreatectomy (either pancreatoduodenectomy or distal pancreatectomy) for benign cystic neoplasms. Thirteen patients were excluded only from long term analysis due to death prior to contact date (10), advanced Alzheimer’s disease (1), severe hearing loss (1), or completion of total pancreatectomy after initial pancreatoduodenectomy (1). This cohort was compared to a cohort of 131 patients who underwent pancreatectomy for PDAC (for short-term outcomes). Age was lower in the cohort of patients undergoing surgery for a cyst (61.45 vs 66.23, P=0.00199). Gender distribution was also significant (cyst: male 30.4%; PDAC: male 50.8%; P=0.00125). Charleson-Deyo co-morbidity score was not significantly different between the cohorts (0.62 vs 0.64, P=0.823). Only 8% received neoadjuvant radiation therapy. The median pancreatic duct size of patients that underwent pancreatoduodenectomy for cystic lesions was 3.0 (2.0–5.0) mm. The median pancreatic duct size for patients that underwent pancreatoduodenectomy for PDAC was 5.0 (4.0–6.0) mm.

Of the 102 eligible participants, 70 had valid contact information and 51 completed the survey for a response rate of 72.9% (Figure 1). Median age at the time of survey was 68 (63–77). General nutritional findings and medical history of participants are summarized in Table 1. Median follow-up was 6 (4–8) years post-pancreatectomy. The most common diagnoses for distal pancreatectomy patients were intraductal papillary mucinous neoplasm (20%), mucinous cystic neoplasm (33%), and serous cystadenomas (20%). The most common pathology for pancreatoduodenectomy patients was intraductal papillary mucinous neoplasm (71%) followed by serous cystadenoma (24%). Only 3/51 (5.8%) of patients, each with IPMN, had high grade dysplasia on final pathology.

Figure 1.

Flow diagram of patient selection

Table 1.

Nutritional and Medical History of Survey Participants

n=51	n (%) or median (interquartile range)
Age at follow-up survey	68 (63–77)
Body Mass Index at Surgery	27.8 (25.0–31.8)
Body Mass Index at Follow-up	26.7 (23.3–29.6)
Weight loss in last 6 months	15 (29%)
Time from operation until follow-up
3–4 years	18 (35%)
5–6 years	11 (22%)
7–8 years	13 (25%)
9–10 years	7 (14%)
11–12 years	2 (4%)
Dietary Intake
Normal intake	37 (73%)
Reduced intake	12 (24%)
Semisolid or liquid diet	0
Very poor intake or starvation	1 (2%)
Gastrointestinal symptoms last 2 weeks
Pain on eating	10 (20%)
Nausea	11 (22%)
Vomiting	1 (2%)
Diarrhea	13 (26%)
Functional Capacity
Normal	41 (80%)
Reduced capacity	6 (12%)
Unable to work	0
Ambulatory	3 (6%)
Alcohol abuse	15 (29%)
Current tobacco use	2 (4%)
At least 150 minutes exercise per week	30 (59%)
Hospitalized in last 2 years	21 (41%)
Enzyme Replacement	4 (8%)
Proton Pump Inhibitor	18 (35%)
Nutritional Supplement	3 (6%)
New Onset Diabetes Since Surgery	15 (29%)
Do you see yourself having nutritional problems?
Major malnutrition	0
Moderate malnutrition	5 (10%)
No nutritional problems	46 (90%)
Health status compared to others of same age?
Not as good	8 (16%)
Don’t know	3 (6%)
As good	12 (24%)
Better	28 (55%)

Short-Term Outcomes

Short-term 60-day outcomes between the 115 patients who underwent pancreatectomy for benign cystic were compared with the 131 patients who underwent pancreatectomy for an indication of PDAC during the same time period (shown in Table 2 stratified by pancreatoduodenectomy and distal pancreatectomy). Complications (≥ grade 2) for patients undergoing pancreaticoduodenectomy for cystic lesions were significantly greater than those undergoing pancreaticoduodenectomy for cancer. There was an increased frequency of all and clinically relevant pancreatic fistulas (P<0.001; P<0.001) and intra-abdominal abscesses (P=0.027). As expected, the Fistula Risk Score was significantly higher in the cohort undergoing pancreatoduodenectomy for cystic lesions compared to PDAC (4.64 vs 1.67, P<0.001).^22,23 The increase in complications did not translate to significant differences in length of stay or 90-day mortality (P=0.079; P=1). Complications, such as pancreatic fistula, wound infections, and intra-abdominal abscesses, in those undergoing distal pancreatectomy were not significant between patients undergoing resection for cystic neoplasms or cancer.

Table 2.

Comparison of Short-Term (60-day) Complications

	All Procedures			Pancreatoduodenectomy			Distal Pancreatectomy
	Cyst (n=115)	PDAC (n=131)	p-value	Cyst (n=51)	PDAC (n=98)	p-value	Cyst (n=64)	PDAC (n=33)	p-value
Any ≥ Grade 2 complication	43 (37%)	38 (29%)	0.163	25 (49%)	31 (32%)	0.038	18 (28%)	7 (21%)	0.461
Postoperative pancreatic fistula	40 (35%)	18 (14%)	<0.001	24 (47%)	11 (11%)	<0.001	16 (25%)	7 (21%)	0.678
Clinically relevant postoperative pancreatic fistula	17 (15%)	6 (5%)	0.006	11 (22%)	2 (2%)	<0.001	6 (9%)	4 (12%)	0.731
Intra-abdominal abscess	13 (11%)	7 (5%)	0.088	9 (18%)	6 (6%)	0.027	4 (6%)	1 (3%)	0.659
Abdominal Fluid Collection	9 (8%)	14 (11%)	0.431	2 (4%)	12 (12%)	0.140	7(11%)	2 (6%)	0.713
Wound Infection	8 (7%)	6 (5%)	0.422	2 (4%)	6 (6%)	0.716	6 (9%)	0	.092
Delayed gastric emptying	19 (17%)	26 (20%)	0.501	10 (20%)	19 (19%)	0.974	9 (14%)	7 (21%)	0.369
Gastrointestinal Hemorrhage	2 (1.7%)	1 (0.8%)	0.600	2 (4%)	1 (1%)	0.270	0 (0%)	0 (0%)	1.0
Length of Stay, median (IQR)	7 (6–8)	7 (6–8)	0.345	7 (7–8.5)	7 (6–9)	0.1078	6 (5–8)	6(5–8)	0.518
90-day mortality	1 (0.9%)	3 (2.3%)	0.625	1 (2%)	3 (3%)	1.000	0 (0%)	0 (0%)	1.0

Note: PDAC = pancreatic adenocarcinoma, IQR = interquartile range

Long-Term Patient Reported Outcomes

Preoperatively, the majority of patients (58.8%) perceived a ‘high risk’ of malignant transformation of their lesion with 43.1% identifying malignant potential as the primary reason for deciding to undergo resection (versus symptomatic relief, or both symptomatic relief and malignant potential of lesion). Postoperatively, the majority of patients (62.7%) perceived ‘low risk’ of pancreatic malignancy in the future. At follow-up, 19 patients (37.3%) reported regular continued surveillance imaging. 90% of patients believed they had no nutritional problems, with the remaining 10% believing they had a moderate level of malnutrition. No patient believed they had a severe level of malnutrition. In terms of perceived health status, forty patients (78.4%) considered it to be as good as or better than other people of the same age, while 8 patients (15.7%) considered their health status to be worse than other people of the same age.

Among the cohort, 43 (84.3%) patients cited the risk of malignancy as the reason they chose to undergo surgery with the remainder citing symptom management. Thirty-nine patients (76.5%) reported a moderate-to-high risk of cancer prior to surgery while 32 (62.7%) perceived a negligible-to-low risk of malignancy after cyst resection (P≤0.001).

Long-Term Quality of Life

Thirteen patients had both pre-operative and post-operative QOL data available. Mean FACT-Hep QOL scores in these patients increased by 9.81 points post-pancreatectomy (95% CI −9.45 to 29.07, p=0.29). A complete FACT-Hep was obtained from all patients post-pancreatectomy. The overall median score for these patients was 160 (138–167.42) out of 180 total points. Table 3 shows a comparison between preoperative and postoperative quality of life. There was no difference in the QOL scores between subjects who had distal pancreatectomy (158.42 (139.75–164)) or pancreatoduodenectomy (162.67 (137–172)). Table 4 shows an in-depth analysis of the FACT-Hep QOL scores comparing patients who underwent pancreatoduodenectomy and distal pancreatectomy.

Table 3.

Comparison of Pre-Operative and Post-Operative QOL Scores

FACT Scale/Subscale	Pre-Operative Median (IQR) QOL Score (n=13)	Postoperative Median (IQR) QOL Score (n=13)	P-Value
Physical Well-Being (0–28)	27 (26–28)	26 (24–27)	0.30
Social Well-Being (0–28)	24 (22–28)	26 (25–28)	0.49
Emotional Well-Being (0–28)	20 (18–22)	23 (22–24)	0.21
Functional Well-Being (0–28)	27 (22–28)	27 (24–28)	0.63
Hepatobiliary Cancer Subscale (0–72)	57 (48–64)	60.5 (56.75–67.25)	0.29
FACT-General (0–108)	96 (88–99)	102 (90–105)	0.59
Trial Outcome Index (0–128)	104 (96–119.25)	112 (100–121)	0.34
FACT-Hep Total (0–180)	148.63 (137.5–162.75)	163 (154.5–172.25)	0.29

Note: FACT = functional assessment of cancer therapy, IQR = interquartile range, QOL = quality of life

Table 4.

Comparison of Pancreatoduodenectomy and Distal Pancreatectomy QOL Scores

FACT Scale/Subscale (n=50)	Overall Median (IQR) QOL Score	Pancreatoduodenectomy Median (IQR) QOL Score	Distal Pancreatectomy Median (IQR) QOL Score	P-Value
Physical Well-Being (0–28)	26 (19.5–27.5)	26 (19–28)	25.5 (22–27)	0.57
Social Well-Being (0–28)	25.67 (22–28)	25.67 (24–28)	25.83 (20.13–28)	0.90
Emotional Well-Being (0–28)	22 (17.5–24)	22 (16–24)	22 (18–24)	0.77
Functional Well-Being (0–28)	26 (21–28)	26 (24–28)	26 (18.5–27)	0.30
Hepatobiliary Cancer Subscale (0–72)	61 (56–66)	62 (57–68)	59.5 (56–64.5)	0.37
FACT-General (0–108)	97.83 (80–103.83)	97.67 (80–104)	97.92 (81–102.75)	0.80
Trial Outcome Index (0–128)	111 (98–118)	113 (98–124)	110 (98.5–116.92)	0.56
FACT-Hep Total (0–180)	160 (138–167.42)	162.67 (137–172)	158.42 (139.75–164)	0.61

Note: FACT = functional assessment of cancer therapy, IQR = interquartile range, QOL = quality of life

Long-Term Nutritional Status

When assessing long-term nutritional status with the Subjective Global Assessment questionnaire, twenty-seven (52.9%) patients were well-nourished (SGA A Rating) and 23 (45.1%) had mild-moderate malnourishment (SGA B Rating). Only one (2.0%) patient was severely malnourished (SGA C Rating). Fifteen (29.4%) patients reported an average of 7.5 pounds of weight loss in the six months prior to follow-up. However, average weight loss at follow up was found to be 9.7 pounds and was significantly decreased from the time of surgery (P<0.001).

Long-Term Exocrine and Endocrine Pancreatic Function

Both exocrine and endocrine pancreatic functional outcomes status-post pancreatectomy for benign cystic neoplasms were assessed (Table 5). Four patients (7.8%) reported no evidence of subjective decline in exocrine pancreas function. However, 28 patients (55.0%) reported steatorrhea and 25 patients experienced early satiety. One fourth of patients had abnormal stool color, bloating after meals, a decrease in number of daily meals post-operatively, or loss of appetite. Despite at least half of patients being symptomatic, only 4 patients were taking pancreatic enzyme replacement (7.8%); 3 of which underwent a pancreatoduodenectomy. Additional patient-reported outcomes are reported in Figure 2.

Table 5.

Comparison Between Pancreatoduodenectomy and Distal Pancreatectomy Patients

	Pancreatoduodenectomy (n=21)	Distal pancreatectomy (n=30)	p-value
Gender
Female	13 (62%)	23 (77%)	0.255
Age*	68 (62–77)	68.5 (64.25–73.75)	0.4
Pathology			0.005
IPMN^	15 (71%)	6 (20%)
MCN	1 (5%)	10 (33%)
Serous cystadenoma	5 (24%)	6 (20%)
Cystadenoma NOS	0	2 (8%)
Lymphoepithelial cyst	0	2 (7%)
Pseudocyst	0	4 (13%)
Pseudopapillary tumor	1 (5%)	0
SGA			0.0651
Well-nourished (SGA A)	12 (57%)	15 (50%)
Mild-moderately malnourished (SGA B)	9 (43%)	14 (46.7%)
Severely malnourished (SGA C)	0 (0%)	1 (3%)
Pancreatic enzyme replacement†	3 (14%)	1 (4%)	0.08
New-onset diabetes	6 (29%)	9 (30%)	0.3

^*Mann-Whitney U Test

^{^}1 patient had moderate to severe dysplasia and 2 patients had focal high grade dysplasia. The remainder of patients had no, low, or moderate dysplasia.

^†Fisher’s Exact Test

Note: IMPN = intraductal papillary mucinous neoplasm, MCN = mucinous cystic neoplasm, SGA = subjective global assessment

Figure 2.

Pancreatic Functional Outcomes (Exocrine)

At long-term follow-up, 21 (41.2%) patients reported a diagnosis of diabetes, fifteen of which had new-onset diabetes mellitus (NODM) postoperatively (29.4%). There was no significant difference in NODM between pancreaticoduodenectomy and distal pancreatectomy patients (28.6% vs 30.0%, p=0.91). Median time at diagnosis of diabetes was 6 (1.25–12) months post-resection. Of those with NODM, 13 patients (86.67%) developed diabetes within one year of pancreatectomy. Insulin was required for glycemic control by 9 patients (60%), but no significant difference in incidence occurred between distal pancreatectomy and pancreatoduodenectomy (p=0.136).

DISCUSSION

An understanding of the true short- and long-term morbidity of pancreatectomy is necessary to determine the risk-benefit ratio that is vital to pre-operative counseling.^28,29 Of great importance to patient counseling is an understanding of patient centered outcomes. However, studies examining patient outcomes after resection for cystic neoplasms centered on European populations.^12,13 Therefore, we sought to assess the short and long-term risks of pancreatectomy for benign cystic neoplasms of the pancreas and patient perceived quality of life in the American population. With regard to short-term outcomes, our study revealed that compared to patients undergoing resection for cancer, peri-operative complications are greater in patients undergoing pancreaticoduodenectomy for cystic lesions. After long-term (6 year) follow up, our study, which is the only study to analyze long-term outcomes after distal pancreatectomy or pancreatoduodenectomy for benign cystic lesions in a U.S. population, revealed that half of patients exhibited symptoms of malnourishment and a third of patients developed diabetes post-operatively. We intentionally excluded patients with PDAC and chronic pancreatitis from our long-term outcome survey as the physiology of these diseases confounds our analysis of the effect of pancreatectomy on long-term nutritional status, quality of life, and pancreas function. Compared to patients undergoing pancreatectomy for cystic neoplasms, patients undergoing pancreatectomy for PDAC have a higher pre-operative incidence of recent weight loss and new onset diabetes.³⁰ The presence of these conditions pre-operatively impacts the prevalence of these conditions in the immediate and long-term post-operative periods.³¹ Preoperative exocrine pancreatic insufficiency is also more common in patients with malignancy and chronic pancreatitis—a population excluded from our analysis—secondary to the fibrosis of the pancreatic ducts and the fibrotic replacement of the parenchyma, respectively.³² In addition to the physiologic differences between patients undergoing pancreatectomy for cystic neoplasms and those undergoing pancreatectomy for cancer, another variable that significantly affects the long-term quality of life in PDAC patients is the receipt of adjuvant therapy.

Short-term morbidity of pancreatic resection, especially pancreatoduodenectomy, is greater for patients undergoing resection of pancreatic cysts than it is for patients undergoing resection for cancer. Van der Gaag et al.¹³ assessed long term outcomes in patients resected for cystic neoplasms (n=65) with a reported complication rate of 20%; however, the classification system and time frame after surgery were not provided. In contrast, Falconi et al.¹² demonstrated a 52% complication rate using the Dindo-Clavien classification in their cross-sectional study of patients resected for benign tumors. Heerkens et al.³³ similarly reported a 39% incidence of severe complications in their study examining quality of life after resection for malignancy. While there is significant variability in complication occurrence, this likely represents the variation in indication for surgery with an increased risk of complications in benign resections due to the well-established risk factors for postoperative pancreatic fistula of small duct and soft pancreas relative to resection for PDAC in which a more firm pancreas with a larger duct would be less prone to short-term morbidities. These risk factors are a major component of the Pancreatic Fistula Risk Score; a prospectively validated score that accurately predicts clinically relevant postoperative pancreatic fistula based on intraoperative findings.^22,23 In the present study, the complication rate of 49% is consistent with the inclusion of patients resected for benign disease. Of note, short-term morbidity was greater in patients undergoing pancreaticoduodenectomy but decision to resect should be based upon current consensus guidelines as well as a careful and detailed discussion with the patient about risks and benefits of an operation. While it is generally known that misuse of alcohol can put patients at risk for surgical complications, the 29% incidence of alcohol abuse in the cohort of patients who underwent pancreas resection for benign cystic lesions corresponds to the statistic put forth by the Substance Abuse and Mental Health Services Administration (SAMHSA) in which 26.9% of adults nationally had engaged in binge drinking in 2015.^34–36 This consistency between our own study’s investigation and national agency estimates suggest that the short-term morbidities highlighted in our results are not primarily or in great part due to alcohol abuse.

Our findings expand upon the existing literature in terms of quality of life after pancreas resection. In a recent cross-sectional study by Massachusetts General Hospital, Fong et al.⁷ examined quality of life of 305 patients who underwent pancreatoduodenectomy for benign or malignant lesions. Median follow up was 9.1 years with a higher quality of life noted when compared to sex and age adjusted healthy individuals. No difference was noted when accounting for resection due to benign versus malignant pathology. However, the study only encompasses patients who underwent pancreatoduodenectomy and does not include pre-operative baseline testing. Huang et al.³⁷ also compared quality of life in post-pancreatoduodenectomy patients (n=192) to a cohort of healthy laparoscopic cholecystectomy patients (n=31) and discovered no difference. In a comparative study, Van der Gaag et al.¹³, examined quality of life following resection for benign cystic neoplasms (n=65) and found no difference in physical or mental quality of life scores when compared to healthy references. However, this study encompasses patients cared for under nationalized health care and is therefore, limited in terms of generalizability to the US due to the fundamental differences between the health care systems and patient populations. The current study provides the only focused examination of perceived quality of life in a US population undergoing distal pancreatectomy or pancreaticoduodenectomy for benign disease.

An important long-term concern in patients after pancreatic resection is nutritional status. Frequently, post-operative nutritional status returns to pre-operative levels with dietary counseling. However, long term information on nutritional status is less well characterized as the majority of studies examine pre-operative nutritional status.³¹ Post-operative nutritional assessments usually center on weight loss, changes in BMI, and exocrine insufficiency. However, weight loss and BMI do not provide an accurate assessment. BMI does not account for tissue types (fat and muscle) and the change in composition that occurs over time (i.e. muscle wasting).³⁸ Patients may therefore benefit from routine screenings with a knowledgeable provider or a multi-disciplinary team post-resection may improve identification of nutritional and vitamin deficiencies and raise patient awareness. Despite the lack of nutritional testing preoperatively, the current study provides unique insight into patient perceptions and the prevalence of malnutrition post resection. Other symptoms that patients may experience after pancreaticoduodenectomy for benign or malignant lesions are gastroesophageal reflux disease and peptic ulcer disease. These processes are related to functional changes in the duodenum and pancreas.³⁹ In order to minimize symptoms and complications related to these processes, it is our practice to discharge all pancreaticoduodenectomy patients on proton-pump inhibitors post-operatively.

In addition to examining quality of life, our study assessed long term pancreatic exocrine and endocrine insufficiency in the only US study to focus solely on pancreatectomy for cystic lesions. Fong et al.⁷ previously assessed 119 patients post-pancreatoduodenectomy (for cystic or malignant lesions) for exocrine and endocrine insufficiency and found an incidence rate of 50% and 11% at a median follow up of 9.1 years. Falconi et al.¹² similarly assessed exocrine and endocrine insufficiency among Italian patients undergoing resection for benign disease with a 33% and 18% incidence at 5 years. Insufficiency in this study was determined by abnormal 72-hour fecal chymotrypsin and oral glucose tolerance tests. Fujii et al.⁴⁰ reported that 47% and 9% of Japanese patients developed exocrine insufficiency or new onset diabetes mellitus following pylorus preserving pancreatoduodenectomy in 55 patients after a median follow up of 42.5 months.⁴⁰ In a more recent study, Roeyen et al.¹⁰ examined 78 patients undergoing pancreatoduodenectomy for oncologic indication and reported an incidence of 20% and 40% respectively. While each study has merits, the conflicting incidences likely arise from variable diagnostic testing and indications for surgery. Discordant from the previous studies, we report only 8% of patients were taking enzymes which likely reflects the prohibitive cost. Even though a minority reported requiring supplementation, more than half of patients described symptoms indicating exocrine insufficiency. The 29% incidence of new-onset diabetes within our study remains consistent with the literature; however, due to limited sample size we were not able to demonstrate a difference in the incidence of diabetes following a pancreatoduodenectomy compared to distal pancreatectomy. This rate of new-onset diabetes should be an important consideration for surgeons prior to resection of benign cystic lesions who are weighing risk of malignant potential of lesions versus complications from pancreatectomy.

Despite the strengths, this study must be interpreted in the context of several limitations. It is retrospective in nature and can be subject to recall bias. Sample size was small and was limited by patient participation and validity of contact information. The study yielded a lower than expected overall response rate. However, the impact of this is unclear as 32 patients were lost to follow-up due to invalid or outdated contact information. Over 70% (51/70) of patients with valid contact information were successfully contacted and consented to participate in the study. Additionally, long-term outcomes data was obtained through telephone interviews to increase participation but may limit patient disclosure of sensitive subject matter. Data supports the possibility of patients under-reporting symptom severity when contacted for follow-up directly by a provider as opposed to automated or anonymous web-based survey administration.^41,42. Pre-operative questionnaires were not performed which limited the ability to draw conclusions about changes in patient perceptions over time. While this study provides valuable data on patients’ perceived long-term QOL and functional outcomes specifically for the resection of benign cystic lesions, this study does not have a control group to compare QOL and pancreas function between patients with cystic lesions who undergo resection compared to those who pursue surveillance. This is a limitation of our database which only follows patients who undergo resection. This limitation highlights an area where further research is required as the natural history of cystic neoplasms with regard to their effect on QOL/pancreas function has not been evaluated. The study was strengthened by the utilization of multiple validated assessment tools including the Subjective Global Assessment and FACT-Hep. Acknowledging these limitations, this study provides a comprehensive and patient-centered evaluation of short- and long-term outcomes after pancreatectomy specifically for pancreatic cysts and provides important data to clinicians in pre-operative counseling as well as post-operative monitoring.

CONCLUSION

Short-term morbidity of pancreatic resection, especially pancreatoduodenectomy, is greater for patients undergoing resection of pancreatic cysts than it is for patients undergoing resection for cancer. In addition, a significant portion of these patients will experience long term complications including diabetes, exocrine insufficiency, and malnutrition. Regular long-term follow up and screening for pancreas insufficiency following pancreatectomy is required to ensure appropriate diagnosis and treatment. In addition to the assessment of malignant transformation, this data should be an integral part of the preoperative discussion and analysis of the risk-benefit ratio of pancreas resection versus continued observation of cystic neoplasms.