Surgical management of portal hypertension has evolved significantly in terms of techniques and indications. Despite improvements in endoscopic therapy, 10%–15% of patients still require various forms of surgical intervention, either due to endoscopic failure or other reasons [1]. Splenectomy is the most frequently performed procedure as definitive management or as part of shunt surgery or devascularization. Although different surgical techniques have been explained in literatures, splenectomy in portal hypertension is very challenging due to extensive collaterals, thrombocytopenia, and enlarged spleen, resulting in significant intraoperative blood loss [2]. Early arterial ligation and late mobilization is the traditional method of splenectomy in portal hypertension [3]. The philosophical thought behind this technique is that the portal pressure reduces after arterial ligation, and the collateral flow decreases [3]. But splenic artery dissection can be dangerous during the initial stage of surgery as it increases the risk of damage to the splenic vein which may jeopardize the shunt. Early spleen mobilization offers good control of the hilum. We aim to compare the effect of early mobilization and delayed arterial ligation (EMDAL) with that of the conventional splenectomy technique in patients with portal hypertension.
During the study period from September 2011 to September 2022, 173 patients underwent surgical intervention for portal hypertension at our institution. Among these patients, 114 underwent the conventional method of splenectomy between 2011 and 2017 while 59 underwent splenectomy with the technique of early mobilization and late arterial ligation between 2018 and 2022. No selection criteria were used to select the patients for procedures. This analysis involved the comparison of two distinct populations across different time intervals. A prospectively maintained database was retrospectively analyzed, and patient demographics and perioperative and postoperative data were compared between the two groups. The study was approved by the Institute’s ethical committee (JIP/IEC/2020/09/47), and the requirement for obtaining informed consent was waived due to the retrospective nature of research.
A thorough history taking and physical examination were performed in all individuals with portal hypertension. Hemograms, liver function tests, kidney function tests, serology for hepatitis B and C virus, abdominal ultrasound (US) with Doppler imaging, and upper gastrointestinal endoscopy were performed for all patients whoever posted electively. CT scan was performed to plan for shunt surgery in selected patients. Pneumococcal, meningococcal, and H. influenza vaccinations were given to all patients before elective surgery and two weeks after the emergency surgery.
Splenectomy was carried out as a definitive surgery or part of other surgeries, such as devascularisation, proximal splenorenal shunt (PSRS), and unconventional shunt. An unconventional shunt was performed whenever technical and pathological factors hindered performing conventional PSRS. Splenoadrenal shunt, mesocaval shunt, interposition mesocaval shunt, interposition PSRS, Jejunal vein-IVC shunt, and left gastroepiploic vein-renal shunt were the unconventional shunts performed at our institute.
Position and incision
Under general anesthesia, patients were placed in a supine position and a wedge was placed under the left flank for elevation. The left trapdoor incision was commonly used for obtaining a better access and visualization. Left side skin flap was everted and fixed to the skin with a silk suture. Hebson retractor was applied to both the costal margins.
Portal pressure measurement
Portal pressure was assessed by inserting a 24-gauge cannula into a branch of the larger omental branch of the gastroepiploic vein using a cannulation technique. Precautionary measures were employed to avoid damage to the posterior wall of the vein during the procedure. A pressure transducer was attached to the cannula once it was confirmed that the blood was flowing freely, and it was connected to monitor and measure the pressure. The transducer was positioned such that it corresponded to the level of the cannulated vein, and it was zeroed to match the atmospheric pressure. The surgical table was adjusted to a flat position, and all pressure measurements were recorded in millimeters of Hg. Portal pressure was measured immediately after the laparotomy and after the splenectomy.
Splenic mobilization
Mobilization was started from the inferior pole of the spleen. Initially, the splenocolic ligament was dissected, and the spleen was mobilized medially to visualize the lateral and retroperitoneal attachments. Large collaterals were divided following suture ligation or clip application, and small collaterals linked to these ligaments were dissected with a Harmonic scalpel or Thunderbeat. After mobilizing the inferior pole, lateral and retroperitoneal attachments, the superior pole was dissected and the gastrosplenic ligament and short gastric vessels were divided. The rationale for approaching from the lower pole to the upper pole was to facilitate optimal exposure and preservation of the dissection plane within the perisplenic area, thereby minimizing the risk of capsular tear. The tail of the pancreas was mobilized with spleen and the splenopancreatic complex was delivered to the wound (Fig. 1). The fundamental principle underlying shunt surgery is the uncompromising avoidance of splenic vein injury. A significant factor contributing to such injuries is inadvertent dissection of the vein while it adheres to the operative field in the presence of an intact spleen. The EMDAL technique provides a valuable solution by enabling precise and controlled isolation of the splenic vein. This isolation is particularly crucial during extraction of the pancreatosplenic complex from the operative field.
The present study performed a comparative analysis between the EMDAL and conventional approaches, with an aim to elucidate the disparities and procedural outcomes. Descriptive statistics, including medians, ranges, and frequencies, were employed to characterize the dataset. For continuous variables, the student t-test was applied, while categorical data underwent assessment using the chi-square test. Statistical significance was determined at a 95% confidence interval, and a p-value of less than 0.05 was considered to denote statistical significance.
During the study period, 173 patients with portal hypertension underwent splenectomy. Fifty-nine patients who underwent EMDAL were compared with those who underwent the conventional method of splenectomy (n = 114). Table 1 presents the demographic characteristics. Diagnosis of non-cirrhotic portal hypertension (NCPH) and chronic liver disease (CLD) was made based on the clinical, laboratory, radiological, and intraoperative findings. A total of 110 patients with NCPH and 4 patients with CLD underwent splenectomy by the conventional method, and 56 patients with NCPH and 3 patients with CLD underwent EMDAL.
Table 1 . Demographics of patients who underwent splenectomy
EALDEM technique (n = 114) | EMDAL technique (n = 59) | |
---|---|---|
Sex | ||
Female | 65 (57.0) | 37 (62.7) |
Male | 49 (43.0) | 22 (37.3) |
Diagnosis | ||
EHPVO | 54 (47.4) | 28 (47.5) |
NCPF | 56 (49.1) | 28 (47.5) |
CLD | 4 (3.5) | 3 (5.1) |
Surgery | ||
Splenectomy | 9 (7.9) | 5 (8.5) |
Splenectomy with devascularisation | 16 (14.0) | 3 (5.1) |
PSRS | 74 (64.9) | 35 (59.3) |
Unconventional shunt | 15 (13.2) | 16 (27.1) |
Values are presented as number (%).
EALDEM, early arterial ligation and late mobilization; EMDAL, early mobilization and delayed arterial ligation; EHPVO, extrahepatic portal vein obstruction; NCPF, non-cirrhotic portal hypertension; CLD, chronic liver disease; PSRS, proximal splenorenal shunt.
The types of surgical procedures in both groups were comparable. Median loss of blood was higher in the conventional group, 450 mL (range, 50–6,500 mL) compared to 200 mL (range, 50–1,200 mL) in the EMDAL method (p = 0.002), with a comparable duration of surgery in both groups. Patients who received a PSRS had significant less intraoperative bleed in the EMDAL group (p = 0.000). Clavien-Dindo grade I–II postoperative complications were comparable in both the conventional and EMDAL groups, at a rate of 37.9% and 33.9%, respectively. The most common complication in both groups was transient ascites, which was managed with diuretics in both groups. A higher percentage of Clavien-Dindo grade III/IV complications, such as post-operative intraabdominal bleeding, UGI bleeding, and left hydropneumothorax, was reported in the conventional group and one death was reported due to cardiac failure. Seven patients bled in the immediate post-operative period, with 5 cases of intraabdominal bleeding and two cases of upper GI bleeding in the conventional group. Six patients in the EMDAL group bled in the immediate post-operative period, of which 3 cases were of intraabdominal bleeding and 3 cases were of UGI bleeding. Two patients required exploration due to rebleeding in the conventional group, and other patients were managed conservatively in both groups (Table 2).
Table 2 . Intraoperative and perioperative outcomes
EALDEM (n = 114) | EMDAL (n = 59) | p-value | |
---|---|---|---|
Splenic vein diameter (mm) | 10.80 ± 4.29 | 10.23 ± 3.87 | 0.325 |
Pre-splenectomy pressure (mm Hg) | 28 (13–52) | 28 (14–43) | 0.732 |
Duration of surgery (h) | |||
Splenectomy | 3.5 (1–6) | 3.5 (4–6.5) | 0.042 |
Splenectomy with devascularisation | 3.5 (2–6) | 4 (2–6) | 0.875 |
PSRS | 5.5 (4–8) | 5.5 (3.5–7) | 0.694 |
Unconventional shunt | 5 (2–7) | 5.5 (4.5–9) | 0.572 |
Total | 5 (1–8) | 5.5 (2–9) | 0.476 |
Blood loss (mL) | |||
Splenectomy | 250 (50–6,500) | 350 (200–600) | 0.364 |
Splenectomy with devascularisation | 500 (200–800) | 400 (350–400) | 0.254 |
PSRS | 450 (50–2,000) | 200 (100–1,000) | 0.000 |
Unconventional shunt | 300 (150–1,200) | 275 (50–1,200) | 0.740 |
Total | 450 (50–6,500) | 200 (50–1,200) | 0.002 |
Clavien-dindo classification | |||
0 | 58 (50.9) | 36 (61.0) | 0.486 |
I | 34 (29.8) | 16 (27.1) | |
II | 7 (6.1) | 4 (6.8) | |
III | 12 (10.5) | 3 (5.1) | |
IV | 2 (1.8) | - | |
V | 1 (0.9) | - | |
Rebleed | 7 | 6 | 0.583 |
Ascites | 29 (25.4) | 17 (28.8) | 0.634 |
SSI | 8 (7.0) | 6 (10.2) | 0.471 |
Respiratory complication | 12 (10.5) | 7 (11.9) | 0.790 |
Blood transfusion | 0 (0–8) | 0 (0–2) |
Values are presented as mean ± standard deviation, mean (range), number (%), or number only.
EALDEM, early arterial ligation and late mobilization; EMDAL, early mobilization and delayed arterial ligation; PSRS, proximal splenorenal shunt; SSI, surgical site infection.
A minimum follow-up of 12 months after the surgery showed UGI bleeding in 14 patients in the conventional group and 8 patients in the EMDAL group, all were managed with endovascular ligation. Intraabdominal bleeding was not reported in any of the cases.
Conventional splenectomy is described as early splenic hilar dissection and ligation of the splenic artery to reduce the blood flow to the spleen and congestion and pressure along the spleno-systemic collaterals. Watanabe et al. [4] performed partial splenic artery embolization prior to splenectomy to reduce bleeding from the splenic vessel during hilar resection. In patients with splenic hilar bleeding, hemorrhage usually occurred as a result of the initial dissection causing splenic vessel injury [2]. Splenomegaly itself restricts adequate work area, causing the capsule or small veins of collateral circulation to tear and ooze; thus, contaminating the operative field and making the surgical procedure more challenging. Also, it is important to avoid bleeding during the procedure, as it is very difficult to control massive hemorrhage from a major vessel or capsule fracture [5]. Therefore, in our view, complete mobilisation of the spleen and complete division and sufficient elevation of the upper pole of the spleen are necessary to achieve control of the splenic hilum before performing dissection in that area. Thus, we performed EMDAL in 59 patients. Dissection between the artery and vein was performed at a later stage.
On comparing the conventional method of splenectomy and EMDAL method, the mean blood loss was less in the EMDAL group. Intraoperative blood transfusion was not significantly different in the two groups, which was determined by the preoperative haemoglobin level, in view of an emergency splenectomy, and associated GI bleeding. Zhu et al. [5] performed 135 open splenectomy procedures for portal hypertension using the conventional method; and the mean operative time was 156 ± 36 minutes with median blood loss was 633 ± 340 mL, and the mean requirement of packed red blood cells was 4.4 ± 2.1.
Clavien-Dindo grade III/IV complications were reported in the conventional group, where two cases required reexploration and one case died due to cardiac failure. One of the patients had diaphragmatic injury leading to left sided pneumothorax. Early mobilization, especially upper pole mobilization, will help to reduce the diaphragmatic injury and avoid thoracolaparotomy. None of the patients in the EMDAL group underwent reexploration.
Laparoscopic splenectomy is performed for portal hypertension, owing to advances in laparoscopic devices and energy sources, such as harmonic scalpel, Ligasure vessel sealing system, and endoscopic vascular stapler, and improvements in surgical procedures [6,7]. A total of 89 shunt surgeries were conducted following conventional splenectomy, while 51 shunt surgeries were performed using the EMDAL technique. Despite the existence of reports detailing laparoscopic splenectomy and devascularization for portal hypertension, the practice of laparoscopic shunt surgery has remained uncommon. Notably, we did not undertake shunt surgeries through a minimally invasive approach. Chen et al. [2] reviewed multiple literatures comparing laparoscopic and open splenectomies and concluded that the laparoscopic procedure was superior to the open approach, in terms of blood loss, hospital stay, complication rate, and liver functional improvement, but it was difficult to derive a statistical conclusion due to lack of high-quality evidence. Kawanaka et al. [8] and Zhu et al. [5] described early partial mobilisation of the spleen with late hilar vascular control using an endovascular stapler. Kawanaka et al. [8] stated that uncontrollable bleeding usually occurred during division of the splenic hilar pedicles and during dissection of the upper pole of the spleen.
Although vessels within the splenic hilum can be dissected and divided individually with vascular clips, in some cases, manipulation is quite tedious and not encouraged. The splenic vessels should be controlled and divided en masse with a linear cutting stapler. But in laparoscopic splenectomies, high rate of pancreatic leak was reported compared to that in the open approach due to endovascular stapler application without identifying the pancreatic tail [2,9]. Laparoscopic splenectomy, when performed using a meticulous surgical technique, is not an inherent risk factor for pancreatic leak [10]. Also, the benefit of autotransfusion of the spleen after splenic artery ligation is lacking with the endovascular stapler method [11]. We can suggest the EMDAL method for laparoscopic splenectomy to reduce the conversion rate, diaphragmatic and colonic injury, even in splenectomy for other indications. This technique holds particular significance in hematological disorders, such as significant thrombocytopenia and other cytopathies. By employing this technique, dissection can be effectively minimized; thus, offering valuable assistance in managing these conditions.
Early mobilisation and delayed arterial ligation will aid in splenic hilum control and haemorrhage control. Complete spleen mobilisation, particularly of the upper pole, minimises diaphragmatic damage and provides adequate working space for splenectomy with less intraoperative bleeding; thus, avoiding thoraco-laparotomy. Late hilar dissection, performed after the hilum has been properly controlled, assists in locating the pancreatic tail and lowers the risk of pancreatic fistula.
The authors thank all residents who maintained the portal hypertension data over 10 years in the Department of Surgical Gastroenterology, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India.
None.
No potential conflict of interest relevant to this article was reported.
Conceptualization: HSL, BP. Data curation: All authors. Methodology: All authors. Visualization: HSL, BP, KR. Writing - original draft: All authors. Writing - review & editing: All authors.