Endoscopic management of splenic pseudocysts associated with acute and chronic pancreatitis

Introduction

Pancreatic pseudocysts are usually located in peri-pancreatic area but can occasionally occur at atypical locations like liver, spleen, mediastinum, pelvis, and kidney [1-7]. Pseudocysts in these atypical locations are rare and pose difficult diagnostic and therapeutic challenge. Splenic pseudocysts (SP) are rare and have been previously reported as case reports with most of them being treated with surgery [7]. Literature supporting endoscopic drainage of SP is scant [1-7].

In this study, we describe clinical and radiological characteristics of SP in 11 patients (4 of whom had been earlier published as case reports or their data had been included in a paper published earlier) [4,5,7] as well as our experience with endoscopic drainage and clinical outcome in these patients.

Patients and methods

We performed a retrospective analysis of patients with SP seen at our unit from January 2002 to June 2015. Clinical records were reviewed to identify patient symptoms and imaging findings. Patients were referred to us for endoscopic drainage and were treated by attempted endoscopic transpapillary drainage. All patients were symptomatic, and had SP documented on contrast-enhanced computed tomography (CECT) scan. Endoscopic ultrasound (EUS) has been available at our institution for the past 6 years and patients seen during that time frame also underwent EUS examination (EG-3670 URK radial echoendoscope or EG-3870 UTK linear echoendoscope Pentax Inc., Tokyo, Japan; GF-UE160-AL5 radial echoendoscope or GF-UCT 180 linear echoendoscope; Olympus Pvt. Ltd., Gurgaon, India). All patients provided informed consent prior to endoscopic treatment and protocol was approved by institutional ethics committee.

Patients were treated with attempted endoscopic transpapillary drainage with a nasopancreatic drain (NPD) or pancreatic stent. Intravenous ciprofloxacin or ofloxacin was administered for antibiotic prophylaxis and hyoscine butyl bromide was used to inhibit duodenal peristalsis. Endoscopic retrograde pancreatography (ERP) was performed by standard technique using a TJF 145 or TJF 160 or TJF-Q180V (Olympus Pvt. Ltd., Gurgaon, India) side-viewing duodenoscope under conscious sedation using intravenous midazolam. Pancreatic duct (PD) disruption was defined by free extravasation of contrast outside the ductal system after contrast injection of main PD or dorsal duct (in patients with pancreatic divisum). PD disruption was defined as complete when main duct upstream to disruption was not visualized and as partial when the main duct was visualized upstream from site of disruption. After confirming disruption, a 5/7Fr stent or 5-Fr NPD was placed across papilla into the PD. An attempt was made to place endoprosthesis across disruption.

Following ERP, patients with NPD placement, were instructed to empty drainage bag and record daily drain output. They were advised to report back if there was no drainage for 24 h or color of the output changed to bilious, indicating displacement of NPD into duodenum. When blockage was suspected (no output for 24 h), NPD was flushed with sterile saline and flow established by suction using a disposable syringe. The patients with stent placement were asked to report back if there was worsening of abdominal pain or onset of new symptoms. All these patients were followed up every 2 weeks for: 1) clinical evaluation, and 2) abdominal ultrasound. CECT of abdomen was done when there was complete clinical recovery along with complete resolution of pseudocysts on abdominal ultrasound.

The patients who did not respond to transpapillary drainage or had worsening of abdominal pain or had new onset/persistent fever with leukocytosis underwent repeat imaging. Patients with the same or increased size of pseudocyst underwent percutaneous radiological drainage or EUS-guided transmural single time complete aspiration with 19 G needle or surgery after an interdisciplinary consultation involving a gastroenterologist and a surgeon.

Therapeutic success was defined as symptomatic improvement with radiological resolution of all pseudocysts on CECT of abdomen and therapeutic failure was defined as need for surgical intervention at any time. Following resolution, the stent/NPD was removed and a repeat pancreatogram obtained to document healing of ductal disruption. Thereafter, these patients were regularly followed up in our clinic.

Results Eleven patients with SP (all male; mean age ± SD: 40.5±8.8 years; age range: 28-60 years) were studied (Table 1). Seven patients had chronic pancreatitis and 4 patients had acute pancreatitis. Patients with acute pancreatitis presented 8 to 32 weeks after the onset of acute attack. All patients had abdominal pain and 2/11 (18%) patients had also fever. Five (45%) patients complained of left upper quadrant discomfort/heaviness. The size of SP ranged from 2.5 to 15 cm (median 6 cm).

Table 1 Profile of patients with splenic pseudocysts

SP were well demonstrated on CECT (Fig. 1). Magnetic resonance cholangiopancreatography (MRCP) was done in 3 patients (Fig. 2) and ductal communication was noted in 1 (33%) patient. EUS was performed in 5 patients with good visualization of SP and echogenic contents were noted in 3 patients.

Figure 1 Computed tomography showing a large splenic pseudocyst

Figure 2 Magnetic resonance imaging of the abdomen showing a large splenic pseudocyst

All 11 patients underwent attempted endoscopic transpapillary drainage through the major (n=10) or minor papilla (n=1). Ten patients had single partial disruption of main PD at tail region of pancreas (Fig. 3) whereas one patient with acute pancreatitis had complete disruption at neck region. A bridging endoprosthesis was placed in all 10 patients with partial disruption (Fig. 4) whereas non bridging stent was placed in patient with complete disruption. A 5 Fr stent was placed in 6 patients, 5 Fr NPD in 4 patients, and 7 Fr stent in 1 patient, respectively. None had stricture or calculus in PD.

Figure 3 Endoscopic retrograde cholangiopancreatography showing a partial disruption at tail end. Guide wire negotiated across disruption

Figure 4 Endoscopic retrograde cholangiopancreatography showing bridging transpapillary stent placed across disruption

Outcome in patients with chronic pancreatitis (n=7)

Five (71%) patients had marked improvement in symptoms following transpapillary drainage. In these patients, resolution of SP as well as associated other pseudocysts and pleural effusion was observed within 8 weeks (median 6 weeks). Another 2 patients who had fever, echogenic contents on EUS and large size of pseudocyst (9 and 10 cm respectively) continued to be febrile with no relief in pain or decrease in size of pseudocyst. Therefore, these two patients underwent single time EUS-guided complete aspiration of the splenic pseudocyst 48 and 96 h of transpapillary drainage (Figs. 5, 6). Following aspiration of purulent material, both these patients had improvement and became afebrile within 48 h. One of these patients had uneventful recovery and was discharged. The other patient had massive gastrointestinal bleed on the 7th day of endoscopic transmural drainage and computed tomographic (CT) angiography revealed a 2.3 cm splenic artery pseudoaneurysm. As patient had hemodynamic compromise, emergency surgery was performed and patient had an uneventful post operative course.

Figure 5 Endoscopic ultrasound-guided transmural aspiration of splenic pseudocyst. Echogenic contents can be noted

Figure 6 The splenic pseudocyst has been completely emptied

On follow up, one patient with large SP (15 cm) developed chronic gastric volvulus after healing, possibly because of peri-gastric adhesions. There has been no recurrence of pseudocysts in these successfully treated patients over a follow up period of 14 months to 13 years.

Discussion

In spite of close proximity of the spleen to the pancreatic tail, SP are rare complication of both acute as well as chronic pancreatitis with majority of them occurring in adult patients with alcohol-induced pancreatitis [1-10]. A large series of 500 patients with chronic pancreatitis, reported only 5 (1%) patients with SP [8]. Similarly, another study of 238 patients with pseudocysts, reported splenic involvement in 5.9% patients only [9]. SP have been very rarely reported in acute pancreatitis [11]. A study of 100 patients with acute pancreatitis who were studied with CECT reported no patient with SP [12]. The SP result either from dissection of the splenic parenchyma by pseudocyst or tracking of pancreatic fluid into the spleen along the course of splenic vessels [10].

SP usually do not have any specific signs and symptoms with majority of patients presenting with abdominal pain. Presence of left upper quadrant heaviness or left pleural effusion may suggest splenic involvement as was seen in 45% and 55% of our patients [8-10,13]. Because of lack of specific symptoms and signs, CECT and MRCP play an important role in diagnosis and follow up of patients. MRCP can also delineate ductal anatomy and also occasionally demonstrate communication of SP with PD. EUS is also a useful investigational modality for evaluating pseudocysts because by determining solid necrotic debris it helps in planning optimal therapy [8].

Because of rarity, ideal management of SP is not clear and surgery in the form of splenectomy with or without distal pancreatectomy, cyst resection or drainage has been commonly used as therapeutic modality [5]. There are few reports of successful treatment by percutaneous radiological drainage but risk of external pancreatic fistula especially in patients with PD abnormality limits its use [5,9,15].

While endoscopic transmural and/or transpapillary drainage has been successfully used in patients with abdominal pseudocysts, there is paucity of endoscopic data in patients with SP [2,4,5,7]. We have previously published case reports of successful resolution of SP with endoscopic transpapillary drainage using NPD or stent or combination of EUS-guided transmural aspiration with transpapillary drainage [2,4,5,7]. In this study also, 8/11 (73%) patients with SP who underwent endoscopic intervention had resolution of all the pseudocysts within 8 weeks.

Of eleven patients, 10 patients had partial disruption which could be bridged in all and 8 (80%) patients had successful outcome. It has been earlier demonstrated that patients with partial disruption respond better to endoscopic treatment than those with complete disruption [5]. In our study, 3 patients with echogenic contents in SP on EUS did not respond to transpapillary drainage alone. One of these could be successfully treated with combined transpapillary and transmural drainage, whereas other two required surgery. This observation is in accordance with previously published reports that have shown that endoscopic transpapillary drainage alone has poor results in fluid collections with solid debris [5,14].

In conclusion, SP is rare complication of both acute as well as chronic pancreatitis and endoscopic transpapillary drainage is safe and effective modality for its treatment especially when it is non-infected with clear contents and is associated with partial ductal disruption that can be bridged by an endoprosthesis.