Analysis of long-term outcomes after minimally invasive spleen-preserving distal pancreatectomy under the ‘Kimura-first’ strategy

Abstract

Introduction:

Spleen-preserving distal pancreatectomy (SPDP) can be carried out by the Kimura technique (KT) or Warshaw technique (WT). This study aimed to evaluate the long-term post-operative outcomes of the two minimally invasive SPDP methods under the ‘Kimura-first’ strategy with a particular focus on the haemodynamic changes in the splenogastric circulation.

Patients and Methods:

The electronic medical records and follow-up data of patients who underwent minimally invasive SPDP in our centre from March 2016 to July 2022 were reviewed. The haemodynamic changes in splenogastric circulation were monitored by post-operative computed tomography (CT) images, and the risks they caused were assessed by long-term follow-up.

Results:

A total of 112 patients (KT = 93 and WT = 19) were included in the study. The tumour size in the WT group was significantly larger than that in the KT group (P = 0.02). We also found less blood loss for patients who underwent KT (P = 0.02). The occurrence of gastric varices was significantly higher in the WT group (P = 0.022). There was no gastrointestinal bleeding in either group. There were two cases of splenic infarction in the WT group (11.1%), and the incidence was higher than that in the KT group (P = 0.026). The infarct area gradually decreased during periodic CT examinations and disappeared completely at the last review. The two groups of patients had similar results across the 15 items in three areas of the quality of life questionnaire.

Conclusions:

The ‘Kimura-first’ strategy, in which the WT is used as an alternative to the KT when the splenic vessels cannot be safely preserved, is feasible, and safe for minimally invasive SPDP.

INTRODUCTION

Distal pancreatectomy with splenectomy is used to treat benign and borderline malignant tumours located in the distal pancreas. However, as awareness of the haematological and immunological functions of the spleen has increased,[1,2] severe complications have been reported in the context of splenectomy.[3,4] Current studies suggest that the spleen should be preserved in treating these diseases.[5,6] Moreover, with the development of minimally invasive techniques, minimally invasive spleen-preserving distal pancreatectomy (MISPDP) has become the routine, preferred operation because of fewer post-operative complications, less blood loss and shorter hospital stays compared to open procedures.[7,8]

Spleen-preserving distal pancreatectomy (SPDP) can be carried out by the Kimura technique (KT),[9] which conserves the splenic artery and vein, or the Warshaw technique (WT), which ligates the splenic vessels while preserving the short gastric vessels and left gastroepiploic vessels.[10] The clinical benefits of performing either method remain unclear, and most of the current studies have reported no statistically significant difference in their perioperative results.[11,12,13] Therefore, the focus of studies on MISPDP is to assess the long-term outcomes, especially the haemodynamic changes in splenogastric circulation. Some studies have reported a higher incidence of gastric varices and splenic infarction after WT, which is the reason why KT should be the first choice.[12,14,15,16] Other studies have reported no gastrointestinal bleeding of the gastric varices during a long-term follow-up.[17] Therefore, they have urged that gastric varices should be regarded as a paraphysiologic phenomenon.[13,17,18,19] Patients with splenic infarction rarely need reoperation.[13,17,20,21] Moreover, WT is faster and easier than KT; therefore, some studies have recommended prioritising WT in cases, in which certain operations are difficult.[21]

Advances in computed tomography (CT) scanner technology and the development of three-dimensional (3D) visualisation software have helped tremendously in the diagnosis and treatment of diseases. Vascular reconstruction provides a visualisation of the 3D spatial relationship between vessels and adjacent organs and is widely used in the mesenteric vascular system.[22] The reconstruction of perigastric veins is more difficult due to the low concentration of contrast media. In recent studies, the haemodynamic changes in splenogastric circulation were mainly evaluated on cross-sectional CT images,[16,17] but none have been visually demonstrated by means of vascular reconstruction.

In a previous study, we proposed the ‘Kimura-first’ strategy, in which WT is used as an alternative to KT when the splenic vessels cannot be safely preserved. Moreover, we confirmed that this strategy was feasible and safe in terms of perioperative outcomes.[14]

Therefore, in this study, we aimed to verify the feasibility and safety of the ‘Kimura-first’ strategy by comparing the long-term outcomes of the two MISPDP methods. We paid special attention to the haemodynamic changes in the splenogastric circulation and visualised them by vascular reconstruction.

MATERIALS AND METHODS

Study design

Patients who underwent minimally invasive (including laparoscopic and robotic) SPDP in our centre from March 2016 to June 2022 were included in this study. Patients were divided into two groups according to whether the splenic vessels were removed: (1) KT and (2) WT. The study protocol was approved by the ethics committee of our hospital. Written informed consent was obtained from the patients for publication of this manuscript and any accompanying images.

Data collection

Patients with a pathologic diagnosis of benign or low-grade malignancy were included in the study, excluding those diagnosed with cancer. The information was collected by reviewing the electronic medical record system. Baseline information collected included sex, age, body mass index (BMI), diabetes, hypertension, smoking history, operation time, estimated amount of blood loss (EBL), pathological diagnosis, size and location of the mass, length of stay after surgery and post-operative complications. The imaging data before and after surgery were carefully reviewed and compared to assess the presence of splenic infarction, splenomegaly or perigastric varices. Vascular reconstruction was performed in patients who showed perigastric varices on cross-sectional CT images.

Definitions

Clinically relevant post-operative pancreatic fistula (CR-POPF) was defined according to the 2016 update of the definition and grading of POPF by the International Study Group of Pancreatic Surgery.[23] The former ‘grade A post-operative pancreatic fistula,’ which has been redefined as a ‘biochemical leak,’ was excluded. Perigastric varices were diagnosed when tortuous veins along the gastric wall were larger than 5 mm in diameter and separated from the splenic hilum.[24]

Follow-up

Patients underwent CT review 1 month and 3 months after surgery. If there were no specific positive radiographic findings, a CT review was recommended every 6 months to 1 year. If splenic infarction, splenomegaly and perigastric varices were found, CT review was recommended every 3 months.

Three to 6 months after discharge, patients were asked whether they had the following symptoms by telephone or outpatient visit: (1) Symptoms of gastrointestinal bleeding (haematemesis and melena), (2) Frequent colds and fevers, (3) Left upper abdominal pain, (4) New-onset diabetes, (5) Any reason for readmission and (6) Any other discomfort. The European Organisation for Research and Treatment of Cancer, quality of life (QOL) Questionnaire-Core 30 (EORTC QLQ-C30) was used to evaluate the QOL of the patients. Our questionnaire was sent to the patients through an Internet link.

Operative techniques

We followed the ‘Kimura-first’ strategy, switching to the WT once it was determined that the splenic arteries and veins could not be safely preserved. The two techniques were performed as previously described.[14]

Statistical analysis

Percentages and frequencies are used to represent the categorical variables. The normality of continuous variables was verified using the Shapiro–Wilk test. The continuous variables are reported using means and standard deviations or median and interquartile range. Continuous variables were compared by independent-sample t-tests or non-parametric Mann–Whitney U-tests according to whether they followed a normal distribution. The Chi-square test or Fisher’s exact test was used for analysing the categorical variables. The QLQ-C30 questionnaire is divided into three areas: functional area, general health area and symptom area. It consists of 15 items, and the scores for each item are linearly transformed. Higher scores in functional and general health areas indicate a better QOL, and higher scores in symptomatic areas indicate worse QOL. Data were considered statistically significant at P < 0.05.

RESULTS

Kimura versus Warshaw

Characteristics

A total of 112 patients (KT = 93, WT = 19) were included in the study, with similar basic nutritional statuses (BMI: KT: 22.95 ± 3.74, vs. WT: 23.41 ± 3.3, P = 0.62) and underlying disease histories (including hypertension, diabetes and smoking history) between the surgery groups. The tumour size in the WT group was significantly larger than that in the KT group (KT = 3.5 [2.0–5.0] vs. WT = 5.0 [3.5–6.0], P = 0.02). There was no significant difference in the pathological diagnosis between the two groups [Table 1].

Table 1.

The characteristics and perioperative outcome of patients underwent Kimura or Warshaw

Variable	Kimura (n=93), n (%)	Warshaw (n=19), n (%)	P
Age (years), mean±SD	43.42±13.5	49.21±13.10	0.09
Sex
Female	70 (75.3)	16 (84.2)	0.59
Male	23 (24.7)	3 (15.8)
BMI, mean±SD	22.95±3.74	23.41±3.34	0.62
History of diabetes	13 (14.0)	4 (21.1)	0.66
History of hypertension	10 (10.8)	4 (22.1)	0.39
History of smoking	7 (7.5)	1 (5.3)	1.0
Tumour size (cm), median (IQR)	3.5 (2.0–5.0)	5.0 (3.5–6.0)	0.02
Location
Body and tail	48 (51.6)	13 (68.4)	0.48
Body	21 (22.6)	3 (15.8)
Tail	21 (22.6)	2 (10.5)
Others	3 (3.2)	1 (5.3)
EBL (mL), median (IQR)	50.0 (50.0–100.0)	100 (50–350)	0.02
Operation time (min), mean±SD	235.40±62.56	251.60±62.83	0.37
Pathological diagnosis
SCN	26 (28.0)	6 (31.6)	0.27
SPN	19 (20.4)	5 (26.3)
NET	15 (16.1)	2 (10.5)
MCN	13 (13.9)	6 (31.6)
PC	8 (8.6)	0
IPMN	3 (3.2)	0
Others	9 (9.7)	0
Morbidity	25 (26.9)	4 (21.1)	0.81
CR-POPF	12 (12.9)	1 (5.3)	0.58
Reoperation	1	0	0.84
Readmission	8 (8.6)	1 (5.3)	0.98
LOSAS (days), median (IQR)	7.0 (5–7)	7.0 (6.0–8.0)	0.24

BMI: Body mass index, SCN: Serous cystic neoplasm, MCN: Mucinous cystic neoplasm, SPN: Solid pseudopapillary neoplasm, NET: Neuroendocrine tumour, PC: Pancreatic cyst, IPMN: Intraductal papillary mucinous neoplasm, EBL: Estimated blood loss, LOSAS: Length of stay after surgery, CR-POPF: Clinically relevant post-operative pancreatic fistula, SD: Standard deviation, IQR: Interquartile range

Perioperative outcome

There was no significant difference in the operation time between the two groups (KT = 236.21 ± 62.2 vs. WT = 258.29 ± 59.4, P = 0.223); however, we observed less blood loss for patients who underwent KT (KT = 50.0 [50.0–100.0] vs. WT = 100 [50–350], P = 0.02). Eight patients in the KT group were readmitted, three for abdominal infection and five due to a grade B pancreatic fistula. One patient in the WT group was readmitted for incisional hernia (P = 0.98). The incidence of post-operative complications (including pancreatic fistula, abdominal infection, pneumonia, etc.,) was similar in both groups. There were 12 patients in the KT group and 1 in the WT group with a grade B pancreatic fistula (P = 0.58). One splenectomy was necessary for post-operative splenic vein haemorrhage in the KT group. The length of hospital stay after surgery was similar in the two groups (KT = 7.0 [5–7] vs. WT = 7.5 [6.0–8.3], P = 0.189) [Table 1].

Changes in splenogastric circulation

Post-operative imaging data were available for 109 patients (KT: 91 vs. WT: 18). No patients in the KT group but two in the WT group had a splenic infarction (P = 0.026). One of these patients underwent periodic CT examinations, which showed gradual reduction of the low-density area and complete absorption at the last review [Table 2 and Figure 1]. Fifteen patients in the KT group and two in the WT group had splenomegaly [P = 0.83, Table 2 and Figure 2]. The occurrence of gastric varices was significantly higher in the WT group [P = 0.022, Table 3 and Figure 3]. None of our patients were diagnosed with perigastric varices by endoscopy. In the KT group, two patients developed low-grade obliterations of the splenic veins (18.2%). In the vascular reconstruction images, more tortuous and dilated veins and collateral veins had formed at 2 months after surgery than at 1 week after surgery [Figure 4].

Table 2.

Comparison of the changes of splenogastric circulation between the Kimura and Warshaw

Variable	Kimura (n=91), n (%)	Warshaw (n=18), n (%)	P
Splenic infarction	0	2 (11.1)	0.026
Splenomegaly	15 (16.5)	2 (11.1)	0.83

Figure 1.

Computed tomography images of the patient with a splenic infarction at regular review, each review showed a reduction in infarct size (red arrow) compared to the previous one. (a) 1 month after surgery, there is a low-density area obviously, (b) 2 months after surgery, (c) 4 months after surgery, (d) 9 months after surgery, (e) 2 years after surgery, low-density area was almost invisible, (f) 3 years after surgery, low-density area was completely absorbed

Figure 2.

Computed tomography images of the patient with splenomegaly. (a) before surgery, (b) 1 month after surgery, the spleen is significantly enlarged

Table 3.

Comparison of the changes of splenogastric circulation between the Kimura and Warshaw

Variable	Kimura (n=11), n (%)	Warshaw (n=8), n (%)	P
Perigastric varices	2 (18.2)	6 (75.0)	0.022

Figure 3.

(a) Perigastric varices (red arrow) in Kimura technique,(b)Perigastric varices (red arrow) in Warshaw technique

Figure 4.

(a) Before surgery, the mass located in the tail of the pancreas (red arrow), (b) 1 week after surgery, (white arrow) right gastric veins; (red arrow) left gastric veins; (yellow arrow) right gastroepiploic veins; no obvious varicose veins formation, (c) 2 months after surgery, (white arrow) right gastric veins; (red arrow) left gastric veins; (yellow arrow) right gastroepiploic veins;the veins were obviously tortuous anddilated, compare to B

Follow-up results

A total of 73 patients had a follow-up after surgery (KT: 59 vs. WT: 14). The mean follow-up time was 29.4 months (KT = 27.0 (10–43) vs. WT = 34.5 (6.8–51.5), P = 0.599). There was no gastrointestinal bleeding in either group. There were four cases of newly developed diabetes in the KT group and one in the WT group (P = 1.0).

Quality of life

A total of 39 patients completed the QOL questionnaire (KT = 30, WT = 9). The EORTC QLQ-C30 is divided into three blocks. Higher scores in functional areas and general health areas indicate better QOL and higher scores in symptom areas indicate poor QOL. There were no significant differences between the two groups in any of the three areas. Patients had the lowest scores for emotional functioning, and fatigue was the most common symptom. Details on the responses to the questionnaire are listed in Table 4.

Table 4.

Comparison of the quality of life between the Kimura and Warshaw

Items	Kimura (n=30)	Warshaw (n=9)	P
Physical functioning	86.8	86.3	0.957
Role functioning	93.9	100.0	0.196
Emotional functioning	79.0	73.9	0.576
Cognitive functioning	84.4	74.1	0.183
Social functioning	88.9	74.8	0.085
Global functioning	82.0	77.9	0.462
Fatigue	30.7	34.2	0.850
Nausea and vomiting	2.2	0	0.432
Pain	17.2	14.8	0.750
Polypnoea	11.7	11.1	0.945
Insomnia	26.3	24.1	0.867
Appetite loss	11.1	11.1	0.962
Constipation	18.6	11.1	0.459
Diarrhoea	8.3	18.5	0.235
Financial difficulties	13.3	25.0	0.246

DISCUSSION

In this study, we reviewed the perioperative and long-term outcomes of 112 patients who underwent minimally invasive distal pancreatectomy from 2016 to 2022. The two groups of patients were comparable due to their similar personality characteristics and basic physiological conditions. Interestingly, WT is generally considered to be associated with a shorter operation time.[13] However, our results showed that the operation time of the WT group was longer than that of the KT group (KT = 235.40 ± 62.56 min, WT = 251.60 ± 62.83 min, P = 0.37). We believe that there are two reasons for this result. First, the size of the tumour in the WT group was significantly larger than that in the KT group (KT 3.5 [2.0–5.0] vs. WT = 5.0 [3.5–6.0], P = 0.02). Second, WT is regarded as an alternative procedure to KT in our institution when the tumour applies long-segment compression to the splenic vessels or there is uncontrolled intraoperative bleeding.[14] Therefore, WT is usually applied for the relatively difficult subsequent operation, which can explain why the WT group, despite undergoing a simpler procedure, had a longer operation time and more EBL than the KT group. Moreover, no significant difference was observed in terms of CR-POPF (grade B) (KT = 12.9% vs. WT = 5.3%, P = 0.58) and overall morbidity (KT = 26.9% vs. WT = 21.1%, P = 0.81) between the two procedures. A recent meta-analysis by Li et al. enrolled 7 studies with 579 patients and demonstrated similar post-operative results to the current study.[12]

WT requires ligation of the splenic vessels and retention of the left gastroepiploic vessels and the short gastric vessels to maintain the splenic blood supply. In KT, dissection of the splenic vein from the pancreas requires extensive manipulation that may lead to vascular endothelial injury, and thrombosis and inflammation are more likely due to the anatomical and haemodynamic characteristics of the vessels. Some studies have reported low-grade obliterations of the splenic veins in 17.2% to 74% of patients after laparoscopic KT.[25,26]

In summary, both KT and WT will cause changes in the splenogastric circulation to varying degrees, which is the main purpose of this study. Kohan et al. reported that 63% of patients experienced a certain degree of splenic hypoperfusion at 1 week after WT; however, 83% of patients had normal perfusion at 6 months post-operatively.[27] In the present study, the incidence of splenic infarction in the WT group was significantly higher than that in the KT group (P = 0.026). Therefore, we followed up with one patient for 4 years. During this period, the infarct area gradually decreased in size according to periodic CT examinations and disappeared completely at the last review [Figure 1]. This is consistent with the view of previous studies, showing that splenic infarcts only require conservative treatment and rarely require secondary surgery.[13,17,20,21]

Previous studies have suggested that post-operative splenomegaly is a manifestation of abnormal splenogastric circulation.[20] In the current study, the incidence of splenomegaly in the two groups was similar (KT = 16.5% vs. WT = 11.1%, P = 0.83), but the probability of perigastric varices following WT was significantly higher than that following KT (P = 0.022). For perigastric varices, the main concern is the risk of gastrointestinal bleeding. However, none of the 112 patients who were followed up had gastrointestinal bleeding. Ferrone et al. reported that all patients with gastric varices had no gastrointestinal bleeding during a follow-up of up to 21 years.[17] Butturini et al. obtained similar results that gastric varices may be interpreted as a paraphysiologic phenomenon.[19] The current study is the first to show the process of perigastric varix and collateral circulation formation after Warshaw surgery by means of vascular reconstruction [Figure 4]. Although the splenic vessels were severed, the intact gastroepiploic vessels, short gastric vessels and collateral vessels could maintain the blood supply to the spleen. Therefore, although no serious complications were caused by the changes in gastrosplenic circulation, WT surgery resulted in a higher probability of perigastric varices and splenic infarction.

In conclusion, the Warshaw operation had a more substantial effect on splenogastric circulation than the KT. Therefore, our centre carefully proposed the KT as the first choice for minimally invasive SPDP. Moreover, neither splenic infarction nor perigastric varicose veins showed serious complications, such as gastrointestinal bleeding or the need for reoperation, in the long-term follow-up. Therefore, the ‘Kimura-first’ strategy, in which the Warsaw technique is used as an alternative when the splenic vessels cannot be safely preserved, is feasible and safe for performing MISPDP.

In the past, the spleen was often removed in distal pancreatectomy. Currently, surgeons tend to preserve the spleen, given its important immunological and haematological functions. Di Sabatino et al. reported that amongst the four diagnostic methods for splenic dysfunction, the detection of Howell–Jolly bodies in peripheral blood is the most suitable method for the clinical evaluation of splenic function.[1] Other studies have evaluated splenic function in this way, but their results have been inconsistent. The frequency of peripheral blood samples showing Howell–Jolly bodies was 0% and 48% in patients undergoing WT, respectively.[27,28] The cause of the discrepancy between the two studies is not clear. Moreover, the specificity and sensitivity of this test are still debated, and quantitative diagnosis is also inaccurate.[1] Therefore, we carefully propose the idea that although the effects of the two surgical methods on splenic function are different, the differences are not large enough to be reflected in the existing detection methods. Taking the important haematological and immunological functions of the spleen into account, this difference should thus affect the QOL of the patients.

Consequently, in this study, we paid special attention to the difference in post-operative QOL between the two groups. To our knowledge, there are no studies comparing the QOL after KT and WT. Zhang et al. reported that laparoscopic SPDP resulted in a better QOL than laparoscopic distal pancreatosplenectomy.[29] A recent multicentre randomised clinical trial by van Hilst et al. reported that minimally invasive (robot-assisted or laparoscopic) and open distal pancreatectomy resulted in a similar QOL.[30] In the present study, the two groups achieved similar results across the 15 items in the three areas of the questionnaire. In terms of specific scores, surgery had the greatest impact on emotional functioning, and fatigue was the most common symptom. Therefore, from the perspective of post-operative QOL, WT is feasible as an alternative method when KT cannot be safely implemented.

The limitations of single-centre retrospective studies are unavoidable. We followed the ‘Kimura-first’ strategy; therefore, the Warshaw group yielded limited data, which may have led to statistical bias. This may partly be due to the high rate of splenic vascular preservation. Our average follow-up period was <3 years, and the conclusion still needs to be verified by longer follow-up of a large sample in multiple centres.

CONCLUSIONS

Although WT surgery resulted in a higher probability of perigastric varices and splenic infarction, no serious complications were caused by the changes in the gastrosplenic circulation. Therefore, in terms of long-term outcome, the ‘Kimura-first’ strategy, in which the WT is performed as an alternative when the splenic vessels cannot be safely preserved, is feasible and safe for performing MISPDP.

Financial support and sponsorship

This study was financially supported by the Medical Minimally Invasive Centre Programme of Fujian Province, China.

Conflicts of interest

There are no conflicts of interest.