Delayed spontaneous haematoma after minimally invasive prostatectomy

Background

Prostate cancer is the most common urological malignancy, and in the United Kingdom it is the second most common overall cancer. ¹ Radical prostatectomy is the standard surgical management for localised prostate cancer, ² and minimally invasive surgery has become the favoured approach in the United Kingdom. ³ Laparoscopic radical prostatectomy (LRP) and robotic-assisted radical prostatectomy (RARP) are widely performed in Europe, with meta-analysis data showing low re-intervention rates (2.7%) and a complication rate of just under 10%. ⁴ Post-operative haematoma formation is a recognised complication of minimally invasive prostatectomies, ⁵ with rates of 2–4% quoted.^3,6,7

However, the literature reports haematomas as a more immediate complication of the operation, with some patients requiring radiological or surgical re-intervention to manage the consequences. There is a paucity of literature on the development of late-onset haematomas, and to our knowledge there are no reports of a spontaneous haematoma occurring more than 30 days after the initial operation.^5,6

We present an unusual case of a man presenting with a late post-operative bleed that occurred spontaneously 5 weeks after surgery. A haematoma developed and was associated with complete disruption of the vesico-urethral anastomosis that imaging had shown to be intact 11 days post-operatively.

Case presentation

A 67-year-old man presented to a district general hospital with haematuria, nausea and general malaise, 5 weeks after a laparoscopic-assisted radical prostatectomy for a Gleason 7, T3 anterior prostate tumour. The pre-operative prostate specific antigen (PSA) level was 8.6. The primary surgery was uneventful, and there were no peri-operative complications. The vesico-urethral anastomosis was created with interrupted sutures. He was discharged home on the second post-operative day and attended for a routine cystogram, as is standard practice in our unit, 11 days after the operation. This did not show an anastomotic leak (Figure 1). The catheter was removed and he successfully voided urine with good immediate continence (0–1 pad/24 h). On presentation 5 weeks post-op, he stated that he had been in good health thereafter, and these were new, acute symptoms. However, he had experienced symptoms consistent with bladder spasms in the days preceding his re-admission to hospital.

OPEN IN VIEWER

Figure 1.

The original post-operative cystogram.

An ultrasound scan on the day of re-admission demonstrated an echogenic collection anterior to the bladder, tracking to the left iliac fossa. The impression was that this represented a post-surgical haematoma. Blood tests on admission revealed a haemoglobin level of 132 g/L, peripheral white cell count (WCC) of 10.4 × 10⁹/L, neutrophil count of 8.81 × 10⁹/L, lymphocyte count of 0.76 × 10⁹/L. He was pyrexial, with a temperature of 38.2°C, and had a mild acute kidney injury, with an eGFR of 56 mL/min/1.73 m². This was previously normal.

Treatment with intravenous amoxicillin (1 gramme, 8 hourly) and intravenous gentamicin (dose calculated based on weight, height and renal function, once daily) was commenced. A urinary catheter was re-inserted and intravenous fluids were given. A urine sample sent for culture and sensitivities did not grow any organisms, and blood cultures did not grow any organisms either. The gentamicin was stopped after 72 hours and switched to oral Clarithromycin (500 mg, twice daily).

A computerised tomography (CT) scan was performed 6 days after admission to the district general hospital. This demonstrated a persisting, unchanged, complex, loculated mixed-density collection around prostatic bed, suggesting both fluid and haematoma components (Figure 2). It also showed proximal and mid small bowel dilatation, without focal obstruction, suggesting an ileus. This was now 42 days after the primary surgery.

OPEN IN VIEWER

Figure 2.

CT scan demonstrating the suspected haematoma.

After this CT scan, he was transferred to the regional tertiary urology centre. Blood tests on admission revealed a WCC of 12.7 × 10⁹/L, haemoglobin of 113 g/dL and normal renal function. A nasogastric tube was inserted to manage his ileus, and he was again commenced on intravenous amoxicillin (1 gram, 8 hourly) and intravenous gentamicin. Both his nausea and pyrexial episodes persisted, and his inflammatory markers continued to deteriorate despite antibiotic therapy. Two days after transfer, his WCC was 18.5 × 10⁹/L and C-reactive protein was 154 mg/L.

A repeat cystogram was performed which demonstrated a probable anastomotic breakdown with a moderately sized contained cavity at the prostatic bed (Figure 3). There was no extravasation of contrast into the retroperitoneal space, and the appearances were an improvement compared to the previous CT scan.

OPEN IN VIEWER

Figure 3.

Repeat cystogram demonstrating the anastomotic leak.

Over the following days, there was both a clinical and biochemical improvement, as he remained apyrexial and haemodynamically stable. Two weeks after his initial presentation, he had a WCC of 6.4 × 10⁹/L and C-reactive protein of 44, and no longer required a nasogastric tube, with his ileus resolving and gut function returning to his normal baseline.

He was discharged with his catheter still in place with a 2-week course of oral Solifenacin (5 mg, once daily) in view of his history of bladder spasms.

Outcome and follow-up

A further cystogram was performed 2 weeks after discharge. This showed a persistent anastomotic leak, although with slightly improved appearances. The urinary catheter remained in place and a further cystogram was arranged 2 weeks later, which showed the leak had healed and there was no contrast extravasation (Figure 4). The catheter was removed and he successfully passed urine. This was 82 days after his original operation and 45 days after presenting unwell to hospital.

OPEN IN VIEWER

Figure 4.

Final cystogram showing resolution of the anastomotic leak.

He has been reviewed in the outpatient clinic since this discharge from hospital and has remained clinically well, with an undetectable PSA. His function is good, requiring only one pad a day for mild incontinence. There will be a further review in the outpatient clinic in 6 months, and the plan is to continue on Solifenacin for at least the next 2–3 months.

Discussion

We could find no literature on similar presentations, and the fact that it has not happened previously in the operating surgeon’s career (with approximately 2000 minimally invasive prostatectomies) makes it difficult to know with certainty the cause of the bleed that led to haematoma formation and disruption of the anastomosis. As the patient reported symptoms consistent with bladder spasms in the days preceding his re-admission, one hypothesis is that he had such a significant spasm that it actually compromised the integrity of the relatively new anastomosis, causing a leak and then subsequent haematoma to form. There are no confirmed reports of this happening in previous literature; however, anecdotal experience suggests that those men with greater symptoms of bladder spasm post-operatively are more likely to have early anastomotic disruption. We are unaware of any patient presenting with an anastomotic disruption associated with haematoma formation so long after surgery, particularly when the initial cystogram had shown no leak and the patient had been so well.

Studies of post-operative bleeding complications after minimally invasive prostatectomies ⁷ report on the possible requirement of re-intervention, either through angioembolisation or a return to the operating theatre. This patient was conservatively managed, largely due to the delayed nature of presentation and the likelihood it would be difficult to drain an organised haematoma. The overall recovery, and lack of discernible early impact of this complication on his function, suggests that conservative management of a delayed-onset haematoma is a reasonable approach to management.

A systematic analysis of the three techniques for prostatectomy (open, laparoscopic and robotic) did show that there were fewer peri- or post-operative complications with RARP than either open or laparoscopic RP, including estimated blood loss. ⁸ However, given the rarity of the complication discussed, it is not clear if the surgical technique was a contributing factor.

Although rare, this presentation nevertheless had significant consequences for the patient in question. He spent 15 days as a hospital inpatient after his re-admission, being treated with intravenous antibiotics for the majority of that time, and in total had a urinary catheter re-inserted for almost 6 weeks. It is reassuring that it has not had any obvious detrimental impact on his post-operative function. However, in future, we will consider early administration of medication in men complaining of over-active bladder symptoms in the post-operative period.