Characteristics of patients who developed glucose intolerance in the early period after partial pancreatectomy

Abstract

Background

New-onset diabetes mellitus (DM) often develops after partial pancreatectomy. Little is known regarding how soon patients develop glucose intolerance after partial pancreatectomy. We investigated the incidence of and factors contributing to the development of glucose intolerance during hospitalization after partial pancreatectomy.

Patients and methods

We retrospectively analyzed the cases of 38 patients with normal glucose tolerance pre-surgery who underwent a partial pancreatectomy (pancreaticoduodenectomy, n = 23; distal pancreatectomy, n = 15). The patients’ glucose tolerance and insulin secretory/sensitivity values were determined by a normal meal tolerance test (NMTT) within 2 months post-surgery during their hospitalization.

Results

The post-surgery NMTT values revealed that 11 (28.9%) patients developed new-onset impaired glucose tolerance (the IGT group); the other 27 (71.1%) patients maintained normal glucose tolerance (the NGT group). The pre-operative hemoglobin A1c (HbA1c) levels were significantly higher in the IGT group (5.84%) versus the NGT group (5.58%, p = 0.034). There were no significant between-group differences in age, sex ratio, body mass index, the ratio of operative procedure (either pancreaticoduodenectomy or distal pancreatectomy), or post-operative insulin secretory values including the fasting/postprandial C-peptide index. The IGT group showed significantly higher insulin resistance assessed by the homeostasis model assessment of insulin resistance (HOMA-IR) versus the NGT group (1.52 ± 0.67 vs. 0.65 ± 0.42, p < 0.001).

Conclusion

After undergoing a partial pancreatectomy, approximately 30% of the patients developed glucose intolerance during the hospitalized period. Our findings indicate that pre-operative HbA1c and post-operative HOMA-IR values can be associated with developing glucose intolerance just after partial pancreatectomy.

Introduction

Patients who undergo a partial pancreatectomy often develop glucose intolerance due to the loss of the pancreatic endocrine parenchyma [1]. The incidence of diabetes after pancreatectomy was reported to increase with the passage of time [2]. The pre-operative factors that have been reported to contribute to the development of diabetes at the late period (> 1 year) after partial pancreatectomy are as follows: the progression of the underlying diabetogenic disease, such as chronic pancreatitis or pancreatic carcinoma, complicated before pancreatectomy [2], surgical methods including distal pancreatectomy [2.3] and extent resection of the pancreas [2.4–5], high body mass index (BMI) [4–6], female gender [4], high age [5], high level of preoperative hemoglobin A1c (HbA1c) [6, 7], low insulin sensitivity calculated by the homeostasis model assessment of insulin resistance (HOMA-IR), and histological fatty infiltration in the pancreas [7]. Few studies have investigated the incidence of the development of diabetes during the early postoperative period in patients who have undergone a partial pancreatectomy. We conducted the present study to determine the incidence of and the factors that contribute to the development of glucose intolerance in the early postoperative period during hospitalization among patients who have undergone a partial pancreatectomy.

Patients and methods

Figure 1 illustrates the patient enrollment process. We conducted a retrospective analysis of patients who underwent a partial pancreatectomy; i.e. a pancreaticoduodenectomy (PD) or a distal pancreatectomy (DP), at Nagasaki University Hospital during the ~ 4-year period from April 2014 to March 2018. We assessed the cases of all of the 154 patients who underwent a partial pancreatectomy during that period for study eligibility.

Fig. 1.

Patient selection flow chart. DM diabetes mellitus, IGT impaired glucose tolerance, NGT normal glucose tolerance, NMTT normal meal tolerance test

For the determination of each patient’s pre-operative glucose tolerance, the levels of fasting plasma glucose (FPG) and random postprandial plasma glucose (PPG) were used, in addition to the HbA1c level. During their hospitalization, the patients consumed the usual treatment foods including calorie-controlled and/or fat-restricted food. Each patient underwent a normal meal tolerance test (NMTT), using the treatment foods to estimate the postoperative glucose tolerance. NMTT is valuable as an insulin test in patients with type 2 diabetes [8]. The patients underwent the NMTT measuring plasma glucose and C-peptide within 2 months (mean 18.0 days) after undergoing their pancreatectomies for the evaluation of their residual capacity of insulin secretion in addition to their glucose tolerance. From the medical records, we obtained the patients’ clinical data including age, sex, height, weight, BMI, and the pre-operative levels of FPG, PPG, and HbA1c; the surgical method (PD or DP), the pathological findings of the resected pancreas, and the post-operative NMTT results.

Among the 154 patients eligible for the study, 105 patients were excluded because they had been diagnosed with diabetes or were suspected of having glucose intolerance (FPG > 110 mg/dL and/or PPG > 200 mg/dL and/or HbA1c > 6.5%) at the pre-operative screening. One patient diagnosed with pancreatic metastasis from other organs, six patients complicated with a diffuse pathological abnormality (including diffuse fiberization, fatty infiltration, or atrophy of the islets of Langerhans) in the resected pancreas, and four patients who did not undergo an NMTT after surgery were also excluded. We thus analyzed the cases of a final total of 38 patients.

Based on their post-pancreatectomy NMTT results, we classified the 38 patients into two groups: those with normal glucose tolerance (NGT) and those with impaired glucose tolerance (IGT). The NGT group was the patients who fulfilled both FPG < 110 mg/dL and 2-h PPG < 200 mg/dL during the NMTT. The IGT group was the patients with either FPG ≥ 110 mg/dL or 2-h PPG ≥ 200 mg/dL. We calculated the C-peptide index (which reflects the residual β-cell function) for each patient as serum C-peptide (ng/mL)/plasma glucose (mg/dL) × 100 [9]. As an index of insulin resistance, the HOMA-IR was calculated by the following formula: FPG (mg/dL) × fasting serum insulin (μU/mL)/405 [10]. Between-group differences were analyzed by paired t-tests for continuous variables and by the χ² test for categorical variables, using IBM SPSS Statistics software. P-values < 0.05 were considered significant.

Results

Among the 38 patients, 27 (71.1%) patients comprised the NGT group, and the other 11 (28.9%) patients were the IGT group (Fig. 1). As summarized in Table 1, the mean age and sex distribution were not significantly different between the NGT and IGT groups. The preoperative HbA1c values (%) were significantly higher in the IGT group compared to the NGT group (5.84 vs. 5.58, respectively; p = 0.034). There were no significant differences in body weight or BMI at hospitalization between the groups. In the NGT group, 16 patients had undergone a PD and 11 patients had undergone a DP. In the IGT group, seven patients had undergone a PD and four patients had undergone a DP. There was no significant difference in the ratio of PD and DP between the groups (p = 0.647).

Table 1.

The clinical characteristics of the patients who developed IGT after partial pancreatectomy and the patients who did not

	NGT (n = 27)	IGT (n = 11)	p value
Age, years	66.4 ± 15.5	72.8 ± 6.6	0.262
Males:females	14:13	6:5	0.880
Preoperative HbA1c, %	5.58 ± 0.35	5.84 ± 0.26	0.034
BW at admission, kg	53.3 ± 10.0	57.4 ± 11.0	0.252
BMI at admission, kg/m2	20.9 ± 2.7	22.3 ± 2.1	0.112
Operative procedure, PD:DP	16:11	7:4	0.647

BMI body mass index, BW body weight, IGT impaired glucose tolerance, NGT normal glucose tolerance, PD pancreaticoduodenectomy, DP distal pancreatectomy

As shown in Table 2, the means of the number of postoperative days at which the NMTT was carried out were not significantly different between the groups. The plasma glucose levels at fasting and 2-h postprandial during the NMTT were significantly higher in the IGT group compared to the NGT group. The fasting serum insulin levels (µIU/mL) were significantly higher in the IGT group than the NGT group (5.11 ± 2.19 vs. 2.82 ± 1.61, p = 0.003), but there were no significant between-group differences in the serum C-peptide levels at fasting or 2-h postprandial of the NMTT.

Table 2.

Results of the normal meal tolerance test (NMTT) after partial pancreatectomy

	NGT (n = 27)	IGT (n = 11)	p-value
Postoperative period until NMTT, days	24.2 ± 12.9	18.5 ± 4.1	0.147
Fasting plasma glucose, mg/dL	90.8 ± 11.0	119.9 ± 10.4	< 0.001
Postprandial plasma glucose, mg/dL	121.3 ± 26.5	163.6 ± 25.4	< 0.001
Fasting serum C-peptide, ng/mL	1.26 ± 0.72	1.38 ± 0.36	0.600
Fasting serum insulin, μU/mL	2.82 ± 1.61	5.11 ± 2.19	0.003
Fasting CPI	1.36 ± 0.70	1.14 ± 0.26	0.347
Postprandial serum C-peptide, ng/mL	3.58 ± 1.99	3.74 ± 2.05	0.774
Postprandial CPI	2.89 ± 1.50	2.23 ± 1.09	0.243
HOMA-IR	0.65 ± 0.42	1.52 ± 0.67	<0.001

CPI C-peptide index, HOMA-IR homeostasis model assessment of insulin resistance

Our analysis of the patients’ post-operative insulin resistance and insulin secretory capacity calculated using the HOMA-IR and C-peptide index revealed that the HOMA-IR values of the IGT group were significantly higher than those of the NGT group (1.52 ± 0.67 vs. 0.65 ± 0.42, p < 0.001), demonstrating that the IGT group had significantly higher insulin resistance compared to the NGT group. In contrast, there were no significant differences in the C-peptide index at fasting or 2-h postprandial during the NMTT between the groups, demonstrating that the insulin secretory capacity of the IGT group was comparable to that of the NGT group.

There were no significant differences between the surgical method groups (PD, n = 23; DP, n = 15) in age, sex ratio, BMI, or the glucose metabolism ratio. Only the number of post-operative hospitalization days was significantly different (longer) in the PD group compared to the DP group (25.7 ± 12.9 vs. 18.3 ± 6.3, p = 0.042) (Table 3).

Table 3.

Demographics of the patients who underwent a partial pancreatectomy

	PD (n = 23)	DP (n = 15)	p-value
Age, years	68.3 ± 12.3	69.8 ± 15.5	0.752
Sex, male:female	12:11	8:7	0.782
Post-operative hospitalization days, n	25.7 ± 12.9	18.3 ± 6.3	0.042
Glucose metabolism, NGT: IGT	16:7	11:4	0.647

IGT impaired glucose tolerance, NGT normal glucose tolerance

Discussion

The incidence of the development of diabetes after PD has been reported to be 18%–39%, and the reported incidence after DP is 5%–42% [2]. Various risk factors for the development of glucose intolerance in the late period after a pancreatectomy have been investigated [2–6]. A recent study indicated that the preoperative HbA1c value, the preoperative HOMA-IR value, and the presence of histologically confirmed pancreatic fatty infiltration were predictive factors of the development of glucose intolerance at > 1 year after pancreatectomy [7]. The complication of chronic pancreatitis was also reported to contribute to the development of new-onset diabetes in patients who showed normal glucose tolerance pre-pancreatectomy before undergoing a PD [11] or DP [12]. Another study showed that female sex, higher BMI, and the resected pancreas volume were independent risk factors for the development of glucose intolerance in patients who have undergone a DP [4].

The reported incidence rates of the development of new-onset diabetes mellitus (DM) caused by partial pancreatectomy are inconsistent; the following points should be considered as the reasons for the inconsistency. The surgical procedures (PD, DP, and other) and the observation period for the incidence of DM differ among the relevant studies, and a standard method of DM screening after partial pancreatectomies has not been established. Burkhart et al. demonstrated that patients not complicated with DM preoperatively were most likely to develop new-onset DM when they underwent a DP compared to a PD [2]. Those authors speculated that glucose intolerance might be directly related to the proportion of resected beta cells because greater amounts of beta cells are resected in a DP compared to a PD for the following reasons: (1) the volume of pancreas resected in a DP is larger than that resected in a PD, and (2) the numbers of beta cells are generally unevenly distributed in the pancreas, with increasing density toward the pancreas tail.

It was reported that existing chronic pancreatitis [11, 12] and fatty infiltration in the resected pancreas [7] posed an increased risk of the development of DM after partial pancreatectomy. However, we have found no report of a study recruiting only patients who did not have such a diffuse lesion in the resected pancreas for investigations of whether a PD or a DP is more likely to be followed by the development of DM.

In the present study, we excluded the patients whose resected pancreas specimen contained a diffuse pathological abnormality such as fiberization, fatty infiltration, or atrophy of Langerhans islets. After the exclusion of those patients, we observed no significant difference in the incidence of the development of glucose intolerance within 2 months after the operation between the patients underwent a PD and those who underwent a DP. This result suggests that glucose intolerance after a partial pancreatectomy cannot be explained only by a decrease in the volume of beta cells.

Ishibashi et al. reported that the values of both preoperative HbA1c and HOMA-IR predicted the development of glucose intolerance at > 1 year after pancreatectomy [7]. Although we evaluated the glucose tolerance of the patients during the early period (mean 18.0 days) after pancreatectomy, our results are consistent with that previous study; i.e., postoperative glucose intolerance was associated with higher preoperative HbA1c values. We could not evaluate the HOMA-IR values in the present patients during the preoperative period. It was reported that insulin sensitivity did not change shortly after PD in non-diabetes patients partly due to the reduction of pancreatic glucagon [13]. We speculate that our patients’ values of insulin resistance calculated by the HOMA-IR were hardly changed between before and after their pancreatectomies because the glucagon level could be decreased or the insulin action in the insulin-sensitive organs such as muscle and liver were maintained through the hospitalization period.

This study had several limitations. First, the sample size was small (n = 38). We need to perform an additional analysis comparing among 4 groups (NGT with PD, IGT with PD, NGT with DP, and IGT with DP) to evaluate the effect of the difference in a surgical procedure on the incidence of developing glucose intolerance when the number of participants will increase. Second, the number of postoperative days at which the NMTT was carried out differed among the patients because this number was not fixed in the study protocol. In addition, the inflammation accompanying the patients’ surgeries might have influenced the patients’ glucose tolerance. Third, the NMTT is considered to be less accurate for the evaluation of glucose tolerance compared to the 75-g oral glucose tolerance test.

Conclusion

Approximately 30% of the present patients without preoperative diabetes developed new-onset glucose intolerance during the early period after undergoing a partial pancreatectomy, irrespective of their operative procedure. In the early postoperative period after a partial pancreatectomy, a patient’s insulin resistance might be more closely involved in glucose intolerance compared to the residual capacity of insulin secretion.

Compliance with ethical standards

Conflicts of interest

All authors declare no conflict of interest.

Research involving human participants

The study was approved by the Ethics Committee of Nagasaki University Hospital (Date of approval: Feb 18, 2019, Approval number: #19021813). The summary of this study objective and research details was acknowledged on our website because of the retrospective design. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) in accordance with the Helsinki Declaration of 1964 and later versions.