Pancreaticoduodenectomy (PD) is a complex surgical procedure that is performed to treat peri-ampullary neoplasms, including tumors of the pancreas, duodenum, ampulla of Vater, and distal common bile duct [1-5]. While this surgical intervention offers potential curative benefits, it often leads to the loss of a significant portion of the pancreas, which can result in pancreatic insufficiency and endocrine dysfunction [3,6-8].

Understanding the impact of PD on the residual pancreatic volume and endocrine function is essential to optimize patient management and improve postoperative outcomes [6]. Moreover, the endocrine function of the pancreas has a profound effect on the patient’s quality of life [9-11]. In recent years, there has been growing interest in evaluating the long-term consequences of pancreatic resection, specifically in peri-ampullary neoplasms [1,3,12]. However, comprehensive studies focusing on pancreatic volume and endocrine function are limited [3,6,13,14]. Large-scale studies have rarely been conducted, because obtaining the volume of the pancreas is difficult.

Pancreatic volume assessment was conducted using various imaging modalities, such as computed tomography (CT) or magnetic resonance imaging, within a specified time frame postsurgery [3,6,15]. By comparing the pre-and postoperative pancreatic volumes, we could quantify the extent of tissue loss, and evaluate its impact on the endocrine function of the pancreas.

This retrospective study aimed to assess the pancreatic volume and endocrine function in patients who underwent PD for peri-ampullary neoplasms. By characterizing the extent of pancreatic tissue loss and its impact on endocrine function, we aimed to identify the potential predictors of postoperative pancreatic endocrine insufficiency, and develop strategies to optimize patient care.

Study design and participants

This retrospective study evaluated 353 consecutive patients with at least a 12-month follow-up period who underwent elective PD for peri-ampullary neoplasms at a single medical center between January 2011 and December 2020. Perioperative and postoperative outcomes, long-term endocrine function of the pancreas, and pancreatic volume changes 12 month postoperatively were evaluated. This study was approved by the Institutional Review Board of the Gachon University Gil Medical Center (No. GDIRB2020-306). This study was performed in accordance with the Declaration of Helsinki, and owing to its retrospective nature, the requirement for written informed consent was waived.

Variables, sources of data, and measurements

Data were collected on the following parameters for each patient: demographic characteristics, postoperative outcomes, including postoperative pancreatic fistula (POPF), adjuvant chemotherapy, length of hospitalization, HbA1c levels, and serial pancreatic volumes. POPF was defined according to the International Study Group on Pancreatic Fistula (ISGPF) criteria, which define a pancreatic fistula as a measurable volume of drainage fluid with an amylase concentration > 3 times the upper limit of normal after postoperative day 3. Three different grades of POPF were defined according to the clinical impact of POPF on the patient’s postoperative course [16].

Endocrine function evaluation

Endocrine function data was prospectively collected on the day before surgery, and 1 week and 3, 6, and 12 months after surgery. The endocrine function was evaluated by measuring the fasting blood glucose (FBG) and serum HbA1c levels. Glycemic control was classified as normal, impaired fasting glucose, or diabetes mellitus (DM), according to the American Diabetes Association guidelines [17]. Impaired fasting glucose was defined as FBG levels > 100 mg/dL, while DM was defined as FBG levels > 126 mg/dL [17]. Endocrine function impairment was defined as a deterioration in endocrine function control capacity, as shown by a change from normal before surgery to DM after surgery [17,18].

Pancreatic parenchymal volume changes measured using pancreas volumetry

The patients underwent pancreas-protocol 3-mm thick three-phase contrast-enhanced axial and coronal CTs preoperatively and at 1 week, and 3, 6, and 12 months postoperatively. The pancreatic volume was calculated using the AVIEW program (Version 1.1; Coreline Soft Co., Ltd.) (Fig. 1).

Fig 1. (A) AVIEW software (Version 1.1; Coreline Soft Co., Ltd.) for pancreas volumetry. (B) Pancreas volumetry and 3D reconstruction. (C) The pancreas was segmented manually to define the region of interest using the AVIEW software.

The percentage change in total pancreatic volume at 12 months after surgery was calculated as follows [6]: = [(pancreatic volume at 12 months–preoperative pancreatic volume)/preoperative pancreatic volume] × 100

The percentage change in the remnant pancreatic volume at 12 months after surgery was calculated as follows [6]: = [(pancreatic volume at 12 months–pancreatic volume at 1 week)/pancreatic volume at 1 week] × 100

Main outcomes

The primary outcome measured in this study was the percentage change in the remnant pancreatic volume at 12 months after surgery. The secondary outcomes were endocrine function, overall morbidity, and clinically relevant POPF rates.

Operation methods

The operations were performed by three specialized hepato-biliary-pancreatic surgeons. Double-layer duct-to-mucosa pancreaticojejunostomy (PJ) technique was performed in all cases. External or internal PJ stent was inserted in all PJ anastomosis, and the type of stent was decided by the surgeon at the time of operation.

Statistical analyses

The patient demographics and perioperative and postoperative outcomes were summarized using descriptive statistics. Continuous variables are presented as the mean ± standard deviation (for normally distributed variables), or the median and range (for non-normally distributed variables). Pancreatic volume was calculated using CT at various time points. The percentage changes in the total pancreatic volume and remnant pancreatic volume at 12 months after surgery were calculated using the formulae provided in the study. Univariate analyses were conducted to identify the potential predictors of new-onset DM after PD. Hazard ratios and p-values were calculated to assess the significance of identified risk factors. All analyses were performed using IBM SPSS Statistics (Version 22.0; IBM Corp.).

Patient demographics

Table 1 lists the patient demographics. The mean age was 65.4 years, and the sex ratio was 1.38 (males, n = 205, 58.1%). The proportion of prediagnosed DM was 31.4% (n = 111). The origins of peri-ampullary neoplasms were as follow: the pancreas (n = 173, 49.0%), common bile duct (n = 96, 27.2%), ampulla of Vater (n = 65, 18.4%), and duodenum (n = 19, 5.4%). No significant differences were observed between the ‘No DM after PD’ and ‘New-onset DM after PD’ groups in age, sex, body mass index, American Society of Anesthesiologist classification, or origin of neoplasm.

Table 1 . Patient demographics

	Total (n = 353)	No DM after PD (n = 178)	New-onset DM after PD (n = 64)	p-valuea)
Age (yr)	65.4 ± 10.3	65.0 ± 10.5	63.5 ± 11.0	0.327
Sex				0.461
Male	205 (58.1)	99 (55.6)	39 (60.9)
Female	148 (41.9)	79 (44.4)	25 (39.1)
Preoperative body mass index (kg/m2)	23.2 ± 3.3	23.1 ± 3.1	23.2 ± 3.1	0.893
ASA classification				0.109
1	34 (9.6)	25 (14.0)	7 (10.9)
2	296 (83.9)	146 (82.0)	50 (78.1)
3	23 (6.5)	7 (3.9)	7 (10.9)
Preoperative DM	111 (31.4)	0 (0)	0 (0)
Preoperative hypertension	164 (46.5)	107 (60.1)	38 (59.4)	0.918
Origin of neoplasm				0.812
Pancreas	173 (49.0)	81 (45.5)	33 (51.6)
Common bile duct	96 (27.2)	44 (24.7)	12 (18.8)
Ampulla of Vater	65 (18.4)	38 (21.3)	12 (18.8)
Duodenum	19 (5.4)	11 (6.2)	5 (3.1)

Values are expressed as means ± standard deviations or number (%).

ASA, American Society of Anesthesiologist; DM, diabetes mellitus; PD, pancreaticoduodenectomy.

^a)p-value was calculated through comparison between the ‘No DM after PD’ and ‘New-onset DM after PD’ groups.

Perioperative and postoperative outcomes

Table 2 lists the perioperative and postoperative outcomes. The mean operative time was 561 minutes. The clinically relevant POPF rate (grades B and C) was 22.4% (n = 79). A total of 318 patients (90.1%) were diagnosed with peri-ampullary malignant neoplasms, and 220 (62.3%) received adjuvant chemotherapy after PD. Chronic pancreatitis was observed in 10 patients (2.8%). The mean postoperative hospital stay was 20.1 days. No significant differences were observed in operation time, length of postoperative hospital stay, PJ stent, POPF, and proportion of malignancy between the ‘No DM after PD’ and ‘New-onset DM after PD’ groups.

Table 2 . Perioperative and postoperative patient outcomes

	Total (n = 353)	No DM after PD (n = 178)	New-onset DM after PD (n = 64)	p-valuea)
Operation time (min)	561.0 ± 101.0	560.0 ± 100.9	580.3 ± 112.7	0.185
Length of postoperative hospital stay (day)	20.1 ± 18.0	20.6 ± 22.4	21.6 ± 16.8	0.739
Pancreaticojejunostomy stent				0.563
External stent	285 (80.7)	145 (81.4)	50 (78.1)
Internal stent	68 (19.3)	33 (18.6)	14 (21.9)
Postoperative pancreatic fistula	177 (50.1)	94 (52.8)	34 (53.2)	0.181
Grade A	98 (27.8)	56 (31.5)	14 (21.9)
Grade B	77 (21.8)	38 (21.3)	20 (31.3)
Grade C	2 (0.6)	0 (0)	0 (0)
Grade B or C	79 (22.4)	38 (21.3)	20 (31.3)
Adjuvant chemotherapy	220 (62.3)	99 (55.6)	41 (64.0)	0.390
Gemcitabine	150 (42.5)	60 (33.7)	30 (46.8)
5-fluorouracil, leucovorin	51 (14.4)	28 (15.7)	6 (9.3)
5-fluorouracil, leucovorin, irinotecan, oxaliplatin	7 (2.0)	5 (2.8)	2 (3.1)
Others	12 (3.4)	6 (3.3)	3 (4.7)
Malignancy	318 (90.1)	161 (90.4)	55 (85.9)	0.317
Combined chronic pancreatitis	10 (2.8)	5 (2.8)	2 (3.1)	0.345

Values are expressed as mean ± standard deviation or number (%).

DM, diabetes mellitus; PD, pancreaticoduodenectomy.

^a)p-value was calculated through comparison between the ‘No DM after PD’ and ‘New-onset DM after PD’ groups.

Endocrine function of pancreas and pancreatic parenchymal volume changes measured using pancreatic volumetry

The 1-week, and 3-, 6-, and 12-month postoperative HbA1c levels were 6.8%, 6.8%, 7.1%, and 7.6%, respectively. The 1-week, and 3-, 6-, and 12-month postoperative proportions of patients with DM diagnosed before PD combined with the patients with new-onset DM after PD were 39.7%, 42.8%, 43.9%, and 49.6%, respectively (Table 3). The preoperative and postoperative 1 week, 3-, 6-, and 12-month mean pancreatic volumes were 82.3, 38.7, 28.1, 24.9, and 25.5 mL, respectively (Table 4). The percentage change in total pancreatic volume 12 months after PD was –53.4%, while the percentage change in remnant pancreatic volume 12 months after PD was –37.1%.

Table 3 . Pancreatic endocrine function

Total (n = 353)	Preoperative	Postoperative 1 wk	Postoperative 3 mon	Postoperative 6 mon	Postoperative 12 mon
HbA1c level (%)	7.3 ± 1.9	6.8 ± 1.4	6.8 ± 1.6	7.1 ± 1.7	7.6 ± 2.0
New-onset DM after surgery (n)	-	29	11	4	20
DM diagnosed before surgery plus new-onset DM after surgery (accumulated)	111 (31.4)	140 (39.7)	151 (42.8)	155 (43.9)	175 (49.6)

Values are expressed as mean ± standard deviation or number (%).

DM, diabetes mellitus; -, not available.

Table 4 . Pancreatic volumetry

	Value
Pancreas volume (mL)
Preoperative	82.3 ± 30.3
1-wk postoperative	38.7 ± 17.4
3-mon postoperative	28.1 ± 14.3
6-mon postoperative	24.9 ± 10.4
12-mon postoperative	25.5 ± 11.1
Percentage volume changes (12-mon after surgery)
Changes in total pancreatic volume	–53.4 ± 14.6
Changes in remnant pancreatic volume	–37.1 ± 17.1

Values are expressed as mean ± standard deviation.

Risk factor analysis for new-onset diabetes mellitus in patients after pancreaticoduodenectomy

Table 5 presents the univariate risk factor analyses for new-onset DM after PD. In the univariate analyses, no significant difference was observed between the ‘No DM after PD’ and ‘New-onset DM after PD’ groups in age, body mass index, operation time, sex ratio, adjuvant chemotherapy, origin of neoplasm, POPF grade B or C, PJ stent type, remnant pancreatic volume at 12 months after surgery, or changes in total and remnant pancreatic volumes at 12 months after surgery.

Table 5 . Risk factor analysis for new-onset DM in patient after PD

Variable	No DM after PD (n = 178)	New-onset DM after PD (n = 64)	Univariate analysis
			HR (95% CI)	p-value
Age (yr)	65.0 ± 10.5	63.5 ± 11.0	0.987 (0.961–1.013)	0.326
Body mass index (kg/m2)	23.1 ± 3.1	23.2 ± 3.1	1.006 (0.919–1.101)	0.893
Operation time (min)	560.0 ± 100.9	580.3 ± 112.7	1.002 (0.999–1.005)	0.186
Male sex	99 (55.6)	39 (60.9)	0.803 (0.449–1.439)	0.461
POPF grade B or C	38 (21.3)	20 (31.3)	1.675 (0.884–3.172)	0.114
Adjuvant chemotherapy	99 (55.6)	41 (64.1)	1.422 (0.788–2.566)	0.242
Pancreas origin	81 (45.5)	33 (51.6)	1.275 (0.719–2.259)	0.406
Malignancy	161 (90.4)	55 (85.9)	0.645 (0.272–1.531)	0.320
External stent	145 (81.4)	50 (78.1)	1.230 (0.609–2.485)	0.563
Preoperative pancreatic volume (mL)	81.2 ± 24.6	79.5 ± 27.6	0.997 (0.984–1.011)	0.694
Remnant pancreatic volume 12-mon after surgery (mL)	25.8 ± 9.6	23.9 ± 8.6	0.978 (0.939–1.019)	0.296
Change in total pancreatic volume 12-mon after surgery (%)	50.4 ± 14.9	53.8 ± 14.0	0.984 (0.960–1.009)	0.200
Change in remnant pancreatic volume 12-mon after surgery (%)	38.4 ± 17.1	38.3 ± 16.6	1.000 (0.982–1.019)	0.968

Values are expressed as mean ± standard deviation or number (%).

DM, diabetes mellitus; PD, pancreaticoduodenectomy; POPF, postoperative pancreatic fistula; HR, hazard ratio; CI, confidence interval.

This retrospective study, evaluating pancreatic volume through pancreatic volumetry and assessing endocrine function in patients undergoing PD for peri-ampullary neoplasms, revealed several significant findings. The focus of this study was on examining the long-term pancreatic functional outcomes, encompassing pancreatic volumetry, in patients who had undergone PD for peri-ampullary neoplasms. Studies with a 12-month follow-up, specifically evaluating endocrine function and pancreatic volume in these patients, are scarce, due to the challenges associated with accurately measuring pancreatic volume [3,6,9].

Pancreatic parenchymal atrophy after pancreatectomy is a common phenomenon, and the mechanisms for the atrophy of the pancreatic parenchyma are as follow [3,6,9]. First, surgical procedures can disrupt the blood supply to the remaining pancreatic tissue. The pancreas relies on a robust vascular network for its function, and any compromise in blood flow can contribute to tissue atrophy. Second, when a significant portion of the pancreas is removed, the remaining tissue may undergo functional changes and adaptation, including a decrease in enzyme production and hormonal regulation. This can contribute to overall atrophy. Third, the surgical procedure itself can lead to changes in the structure of the remaining pancreas. Fibrotic tissue formation and alterations in the architecture of the organ may occur in response to the surgery. While atrophy is a common consequence of pancreatectomy, the degree of atrophy and its impact on overall health can vary, depending on the extent of the surgery, and the underlying condition being treated.

During the 1-year period following PD, approximately 20% of patients experienced new-onset DM. The remnant pancreatic volume also steadily decreased by approximately 37% one year after surgery, with the total volume decreasing by approximately 53%, accounting for the resected pancreas volume. Immediately after PD, most patients reduced their food intake, and some temporarily discontinued DM medications. One month after PD, patients generally tended to regain their presurgery levels of food intake. In subgroup analyses conducted in the current study, no significant differences were observed between the early-onset DM (n = 29) and late-onset DM (n = 20) groups in preoperative and postoperative outcomes.

Jung et al. [9] reported a retrospective study including 122 patients who completed 12 months of follow-up using abdominal CTs and a quality-of-life questionnaire to evaluate remnant pancreatic atrophy following PD. This study reported that malignancy and adjuvant chemo-radiotherapy were associated with the severity of the atrophy. At 12 months postoperatively, quality-of-life scores and nutritional indices were mostly not associated with atrophy, but stool elastase levels decreased significantly, and the incidence of new-onset DM was higher in the severe atrophy group [9]. The development of DM in patients after PD was anticipated to be influenced by the cumulative effects of various factors, including pancreatic volume status, alterations in diet, and chemoradiotherapy. In the present study, risk factor analyses were conducted for new-onset DM in patients after PD. However, malignancy, adjuvant chemotherapy, and pancreatic atrophic changes following PD were found to be irrelevant in this study. This is attributed to the inherent difficulty in controlling variables, such as the postsurgery dietary intake of each patient, their level of physical activity, the type and dosage of anticancer drugs administered, family history of diabetes, and variations in skeletal muscle mass that can impact blood sugar control. Consequently, obtaining statistically significant results in the risk factor analysis was deemed unattainable, suggesting inherent limitations in the study design.

Shin et al. [3] reported a prospective randomized controlled study of 213 patients with long-term morphological changes in the pancreas who underwent PD using pancreatic volumetry according to the pancreatic drainage method. The median percentage change in remnant pancreatic volume over 12 months was 45.2% (range, 3.2%−82.8%) in the external stent group, and 46.6% (range, 6.2%−75.1%) in the internal stent group, a difference that was not statistically significant (p = 0.221) [3]. Factors associated with pancreatic exocrine or endocrine function, including stool elastase levels (p = 0.571) and the new-onset DM rate (p = 0.179), were also comparable between the external and internal stent groups [3]. Since there is not much difference in the degree of change in pancreatic volume depending on external or internal PJ stent, it can be inferred that there will be no difference in new-onset DM incidence. External PJ stent was performed in 285 (80.7%) patients, while internal PJ stent was performed in 68 (19.3%) patients in the current study. No significant difference was observed in PJ stent type between the ‘No DM after PD’ and ‘New-onset DM after PD’ groups in the current study.

The results of this study demonstrate that PD has a significant impact on both pancreatic volumes and endocrine functions. The percentage change in the remnant pancreatic volume 12 months after surgery was –37.1%, indicating a substantial reduction in the size of the remaining pancreas. The percentage of patients diagnosed with new-onset DM after surgery increased over time, with 49.6% being diagnosed at the 12-month follow-up. These findings highlight the potential long-term effects of PD on pancreatic volumes and endocrine functions. Few studies have investigated the causal relationship between pancreatic volume loss and incidence of DM [7,13,14,19].

The present study has several limitations. Firstly, the results are susceptible to biases and limitations inherent in a retrospective design. Secondly, pancreas volume is calculated by manually delineating the region of interest, introducing a potential for error, compared to the actual pancreas volume [20]. To mitigate this, two surgeons and two researchers independently confirmed the region of interest. Thirdly, the current study lacks data pertaining to the exocrine function of the pancreas. Nevertheless, it stands out as one of the few studies that encompasses a large number of patients, tracking changes in pancreatic volume and endocrine function over a 12-month period post-PD.

In conclusion, this retrospective study offers valuable insights into the assessment of pancreatic volumes and endocrine functions in patients who underwent PD for peri-ampullary neoplasms. The current study did not identify a causal relationship between pancreatic endocrine dysfunction and atrophic changes in the pancreas following PD. Large-scale prospective studies are necessary to explore the mechanisms, associations, and consequential effects of decreased pancreatic volume and endocrine dysfunction in the pancreas following pancreatectomy.

The abstract of this paper was published in European Journal of Surgical Oncology under the title ‘Long-term changes in pancreatic volume and endocrine function after pancreaticoduodenectomy for peri-ampullary neoplasms using pancreas-volumetry: A retrospective single center study’ (https://doi.org/10.1016/j.ejso.2022.11.442).

This study was supported by the Research Program of the Korean Association for Hepato-Biliary-Pancreatic Surgery (KAHPBS-23-03), Gachon University Gil Medical Center (No. FRD2021-20), and Gachon University Research Fund of 2023 (GCU-202308820001).

No potential conflict of interest relevant to this article was reported.

Conceptualization: DHL. Data curation: All authors. Methodology: All authors. Writing - original draft: DHL, JY. Writing - review & editing: All authors.