Analysis of Risk Factors of Pancreatic Injury during Elective Laparoscopic Splenectomy in Children.

INTRODUCTION: Laparoscopic splenectomy (LS) became the standard choice for splenectomy in children with benign hematological disease. There are few reports about pancreatic injury during LS. The purpose of this study is to spot on factors increasing the risk of pancreatic injury during LS in children. PATIENTS AND METHODS: A total of 140 children had LS for benign causes. Children were categorized into A and B groups. LigaSure™ was used to control pedicle in Group A, while endoscopic staplers were used in Group B. Preoperative levels of amylase, lipase, and lactate dehydrogenase (LDH) were obtained. The mean of pancreatic enzymes and LDH values was calculated on the 3 postoperative successive days.
RESULTS: A total of 71 boys and 69 girls had LS. The mean splenic size was 13.50 cm in Group A and 12.51 cm in Group B. The mean operative time in Group A was 41.91 min and in Group B was 56.36 min. The mean level of amylase was 42.99 IU/ml in Group A and 75.70 IU/ml in Group B (P = 0.001). The mean level of lipase was 37 IU/ml in Group A and 76.66 IU/ml in Group B (P = 0.001).
CONCLUSION: Pancreatic injury during LS is a rare complication usually presented on biochemical level. We believe that it is a hemostatic-dependent complication rather than splenic size or nature of disease.

INTRODUCTION

Laparoscopic splenectomy (LS) gained popularity since it had been introduced several decades ago and became the golden choice, especially among children with benign hematological disease.[1]

The advantages of bipolar-sealing devices and endoscopic staplers made the procedure easy and shortened the operative time.[23]

There are some complications that especially occur during LS in children such as bleeding from pedicle which is the main cause of conversion, splenosis, missed accessory spleens, injury to nearby structures, and ports-related complications.[4]

The aim of this study was to detect the magnitude of pancreatic injuries during LS in children with benign hematological diseases.

PATIENTS AND METHODS

During the period from March 2007 to December 2017, 140 children with benign hematological diseases had LS. The study had ethical committee approval. They were randomly categorized into two groups with closed-envelope method according to the method used to secure pedicle of the spleen. Group A included patients in whom bipolar-sealing devices (LigaSure™) were used to control the pedicle, while Group B included patients in whom endoscopic linear staplers were used to control pedicle. All patients had the routine preoperative investigations. In addition to these investigations, all patients had preoperative baseline of serum amylase, lipase, and lactate dehydrogenase (LDH). All have been vaccinated against pneumococci and capsulated organisms. Under general anesthesia, the patients were placed in right decubitus position with a 60° angle between the back of patient and operating table. The flank of patient was elevated. We used three-port technique in normal-sized spleens (with span ranged from 8 to 10 cm) and four ports in cases of splenomegaly. The ports' sites were at the umbilicus for telescope (10 mm), epigastric port (5 mm), and left midclavicular line port at or above the level of umbilicus (10 mm). The fourth port (5 mm) was used high in the epigastric position. Exploration for accessory spleens was the first step. Securing of lower splenic vessels with LigaSure™ was done[Figure 1]. The lesser sac was entered and securing short gastric vessels was performed. Then, the pedicle was dissected with the identification of the splenic artery and vein. In Group A, they were secured by LigaSure™ [Figure 2], while in Group B, they were secured using endoscopic linear stapler [Figure 3]. The splenic attachments were taken down using LigaSure™. The spleen was delivered inside a bag through the umbilical port. The umbilical port site was extended for 2 cm in cases of splenomegaly. During the postoperative period, all patients received broad-spectrum antibiotics. Oral intake was initiated as early as the patients tolerated. Serum amylase, lipase, as well as LDH were obtained in all cases on the 1st, 2nd, and 3rd postoperative days to detect pancreatic injury.

Figure 1

Level of amylase in both groups

Figure 2

Lipase levels in both groups

Figure 3

LDH levels in both groups

Statistical analysis

Statistical analysis was done using IBM SPSS version 24 (IBM, NY, USA). Descriptive statistics are presented as means ± standard deviation, medians (interquartile range), or percentages. Patients' demographic, preoperative, and postoperative data were compared between two groups using the Chi-square test for categorical variables and independent samples t-test for continuous variables.

RESULTS

We operated 140 children suffering from benign hematological diseases. The etiology for splenectomy was thalassemia in 40, spherocytosis in 15, and idiopathic thrombocytopenic purpura in 85 children. The mean age in Group A was 6.95 years; while in Group B, it was 7.17 years. Boys were 71. The longest splenic axis was 20 cm in both groups; the mean axis was 13.5 cm in Group A, while in Group B, it was 12.91 cm. The amylase, lipase, and LDH levels were within normal values in both groups preoperatively [Table 1].

Table 1

Demographic and preoperative data

	Group A (n=70)	Group B (n=70)	P
Age (years), mean	6.95±2.84	7.17±2.91	0.645
Sex
Males	38	33	0.398
Females	32	37
Splenic size (cm), mean	13.5±2.22	12.91±2.36	0.112
Diagnosis
Thalassemia	20	20
ITP	42	43
Spherocytosis	8	7
Biochemical markers (IU/L), mean
Amylase	32	35	0.451
Lipase	34	37	0.321
LDH	210	211	0.412

ITP: Idiopathic thrombocytopenic purpura, LDH: Lactate dehydrogenase

The mean operative time was 41.91 min in Group A, while it was 56.36 min in Group B. This represented the advantages of saving time in Group A as there was no need to exchange instruments during the procedure as the control of pedicle was achieved by bipolar sealing devices; however, in Group B, there was time consumed during loading and introduction of endoscopic stapler to control the pedicle. The mean amount of estimated blood loss in Group A was 66.14 ml, while in Group B, it was 97.5 ml. There were no intraoperative injuries of adjacent organs in either group. There were three cases converted to open approach in Group A. In Group B, five cases were converted to open approach. Six cases in Group A received blood transfusion while 13 cases received blood transfusion in Group B. Blood transfusion occurred intraoperatively [Table 2].

Table 2

Operative data

	Group A (n=70)	Group B (n=70)	P
Operative time (min), mean	41.91±6.43	56.36±7.47	0.001*
Estimated amount of blood loss (ml), mean	66.14±16.3	97.5±23.25	0.001*
Conversion (number of cases + percentage)	3 (4.2)	5 (7.1)
Intra operative blood transfusion (number of cases + percentage)	6 (8.5)	13 (18.5)

*0.001 is significant

During the postoperative period, oral intake was delayed in 17 cases of Group A on the 2nd postoperative day, while it was delayed in 22 cases in Group B on the 2nd postoperative day. The mean time of hospital stay in Group A was 2 days with standard deviation (SD) 1.08, while in Group B, it was 2.7 days with SD 1.01. The mean level of serum lipase measured on 3 successive postoperative days was 37 U/L with SD 27.01 in Group A, while it was 76.66 U/L with SD 39.6 in Group B. The mean level of serum amylase during the same period was 42.99 U/L with SD 27.01 in Group A, while it was 75.7 U/L with SD 42.11 in Group B. The mean serum level of LDH during these days was 315.5 U/L in Group A, while it was 345.49 U/L in Group B. Although the mean values of amylase, lipase, and LDH fall in the normal ranges between the two groups, these measures are still high in Group B. This may reflect some pancreatic affection in patients of Group B [Table 3].

Table 3

Early postoperative data

	Group A (n=70)	Group B (n=70)	P
Full oral intake, n (%)
1st day	53 (75.7)	48 (68.6)	0.346
2nd day	17 (24.3)	22 (31.4)
Hospital stay (days), mean	2±1.08	2.7±1.01	0.001*
Lipase (IU/L), mean	37±27.01	76.66±39.6	0.001*
Amylase (U/L), mean	42.99±27.01	75.7±42.11	0.001*
LDH level (U/L), mean	315.5	345.49	0.226

*0.001 is significant. LDH: Lactate dehydrogenase

Furthermore, the levels of amylase, lipase, or LDH did not affect by splenic size [Table 4].

Table 4

The relation between splenic size and pancreatic enzymes and lactate dehydrogenase levels in both groups

	Mean splenic size (cm)	Pancreatic enzymes and LDH	Mean levels (U/L)	P
Group A	≤10	Lipase	28.31	0.337
	>10		42.14
	≤10	Amylase	34.15	0.134
	>10		48.20
	≤10	LDH	293.65	0.188
	>10		328.50
Group B	≤10	Lipase	73.65	0.427
	>10		81.44
	≤10	Amylase	71.23	0.266
	>10		82.81
	≤10	LDH	340.77	0.346
	>10		348.26

LDH: Lactate dehydrogenase

Fourteen cases in Group A suffered from left-sided subphrenic collections diagnosed by ultrasound during the postoperative follow-up due to onset of fever. The mean volume was 77 ml. Ultrasound-guided aspiration was done for all cases, and the fluid was examined for its biochemical and microbial characters. The aspirate was bloody in 10 cases, while it was turbid serous fluid in four cases. The mean levels of amylase and lipase in collections of these cases were 1500 and 2000 U/L, respectively. In Group B, there were 18 patients with subphrenic collection also diagnosed by ultrasound. The mean volume was 130 ml. The aspirate was bloody in nine cases, while it was turbid serous fluid in the other nine cases. Amylase and lipase levels in the aspirated fluid were 2500 and 3000 U/L, respectively. A total of 31 patients suffering from collections responded to conservative measures which included ultrasound-guided drainage and follow-up. Only one patient in Group B needed surgical exploration and evacuation of collection after the failure of ultrasound drainage as the collected fluid was septate and thick [Table 5].

Table 5

Postoperative collection content of amylase and lipase in both groups

	Group A	Group B	P
Volume (ml), mean	79.4	138.89	0.001*
Amylase (U/L), mean	1215.71	2547.22	0.001*
Lipase (U/L), mean	1921.43	2913.89	0.001*

*0.001 is significant

The levels of postoperative amylase and lipase in both groups are shown in Figures 1 and 2.

DISCUSSION

The advances in minimally invasive surgeries lead to significant improvement in many conventional operations. Furthermore, these advances shorten the operative time and reduce the rates of associated complications.

During LS, there are several techniques that used to control the rich blood supply of spleen as endoscopic staplers, clips, bipolar-sealing devices, suture ligature, and ultrasonic shears.[5]

There are several complications that may occur during LS in children with benign hematological diseases. Bleeding is the most frequent complication and considered the main cause of conversion to open procedure.[6]

According to available data reporting injury to adjacent structures during LS, there are few clinical trials that document these complications.

Apart from general and common complications of LS, Chand (2001) was the first researcher that spotted on pancreatic injuries during LS.[7]

In a series, the author noticed that out of 94 patients who underwent LS using surgical staplers for pedicle control, 15 patients developed pancreatic injuries. The author classified these injuries into minor or major. Furthermore, the author included states of hyperamylasemia and hyperlipasemia. To explain these complications, the author assumed that both splenomegaly and use of staplers stood behind the occurrence of pancreatic injuries.[7]

During the current study, cases of splenomegaly did not represent an obvious cause of pancreatic injuries in both groups of study.

On the other hand, the use of surgical staplers during pedicle control in Group B was associated with the onset of elevated levels of amylase, lipase, and LDH. Furthermore, in Group B, there was an increased incidence of left subphrenic collections.

There were some main limitations in the study of Quand; the first, it was a single group of patients with only one method of pedicle control which was staplers, and the second was the absence of records of preoperative estimations of either amylase or lipase. Furthermore, there were no records of LDH during postoperative periods in complicated cases.

Ultrasound percutaneous aspirations in cases presented later with left subphrenic collections showed shooting levels of both amylase and lipase. This was the same finding as Quand observed.

Casaccia et al. reported that the incidence of pancreatic complication was not rare after LS and assumed that it ranged from 4.5% to 16% in patients with benign hematological diseases. This was relatively high. The main concern about this study was the use of endoscopic staplers to control the pedicle.[8]

The use of LigaSure™ during this study did not require long segment of skeletonizing splenic vessels to be sealed. In contrast to the use of endoscopic staplers, there should be enough segment of skeletonized splenic vessels for the staplers to be fitted and fired safely. Hence, the proximity of the tail of pancreas to splenic hilum represented the great chance to its injury while applying jaws of staplers.

In a study of Khirallah et al., they showed that there were no cases of pancreatic injury during LS in children with benign hematological diseases. They used LigaSure™ only to control hilum.[9]

In this study, we did not report cases with pancreatic fistula as a complication of LS in both groups. This was established according to the definition of International Group of Study on Pancreatic Fistula.[10]

Tsutsumi et al. reported that the incidence of pancreatic fistula after LS was 22%. It is a life-threatening morbidity and assumed that prophylaxis should be attempted during the procedure. They limited their study to only adults with hypersplenism secondary to liver cirrhosis. They used surgical staplers to control the pedicle. They came to conclusion to use fibrin glue sheets at the site of transected vessels during LS.[11]

CONCLUSION

LS in children with benign hematological diseases is a safe procedure. Although pancreatic complications represented a little common morbidity in literature with a range from 4% to 22%, also all these studies were either on adults or using staplers during the procedure. This made us believe that the occurrence of pancreatic injury is a hemostatic-dependent procedure rather than operative time or splenic size. In our study, we found the use of BSD represented a safe hemostatic tool with less incidence of pancreatic complications during LS.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.