Suspected epidural morphine analgesia induced chronic urinary and bowel dysfunction in a cat

Abstract

A 12-year-old male castrated domestic shorthair developed chronic urinary retention, constipation and a decreased perineal reflex following a single lumbo-sacral epidural injection of morphine during general anesthesia. Similar adverse effects have been reported in humans following epidural analgesia, but this is the first reported case of both urinary and bowel dysfunction in a cat purportedly from an epidural. The cat was medically managed with manual bladder expressions, intermittent enemas, and various medications including bethanechol, cisapride and stool softeners. The cat continues to have long-term neurologic dysfunction 15 months post-onset. This case report describes a rare but serious potential risk of lumbo-sacral epidural injections in cats.

A 12-year-old male castrated, 3.8 kg domestic shorthair presented to the Veterinary Hospital of the University of Pennsylvania for a 9-month history of weight loss (1.3 kg) despite a normal appetite. Physical examination showed generalized muscle wasting and a grade II/VI parasternal systolic heart murmur. Previous evaluation ruled out diabetes mellitus, hyperthyroidism, and exocrine pancreatic insufficiency. Six years prior, the cat had an episode of heart failure likely due to endomyocarditis. Initial echocardiogram showed a mildly enlarged left atrium, but follow-up echocardiograms 1, 2 and 4 years later showed a normal heart. The most recent echocardiogram 9 months prior to its presentation for weight loss showed evidence of remodeling in the left ventricle, mild diastolic dysfunction and normal function of all valves. The cat has had no clinical signs of heart disease following its episode of heart failure and has not been on any cardiac medications. Abdominal ultrasound revealed diffuse thickening of the submucosal and muscularis layers of the small intestine and enlarged mesenteric lymph nodes. The owner opted for full-thickness intestinal biopsies via exploratory laparotomy.

Prior to surgery, the cat was sedated with dexmedetomidine (Dexdomitor; Pfizer) 4 μg/kg IM. Anesthesia was induced with propofol (Propoflo; Abbott) at 4.5 mg/kg IV and midazolam (Hospira) at 0.2 mg/kg IV. A previously described routine lumbo-sacral morphine epidural ¹ was chosen to provide peri-operative analgesia. The lumbo-sacral area was cleaned with antiseptic and a 22 gauge 1.5 inch needle was inserted just caudal to the dorsal spinous process of the L7 vertebrae. During needle placement, blood was seen in the spinal needle hub. Thus, the needle was removed and a new spinal needle was inserted through the ligamentum flavum. A 7 cc plastic Loss-of-Resistance syringe (Kimberly Clark) was used to detect the epidural space using the loss-of-resistance technique. When a small amount of saline (no more than 0.4 ml) could be injected without resistance signifying the correct placement of the needle in the epidural space, morphine (Baxter, 0.43 mg (0.43 ml)) with 0.86 ml of 0.9% saline was then slowly injected over 45–60 s. Anesthesia was maintained with a rebreathing coaxial pediatric circle system with isoflurane (Isoflo; Abbott) vaporizer settings ranging from 1.0% to 1.6% and oxygen flows ranging from 0.4 to 1 l/min. Fifteen minutes into anesthesia, the cat became hypotensive with Doppler systolic blood pressures of 85 mm Hg. The hypotension was unresponsive to two separate 5 ml/kg intravenous fluid boluses. Therefore, 1 h after start of anesthesia, a continuous rate infusion of phenylephrine (Baxter) at 2 μg/kg/min was started. ² The phenylephrine constant rate of infusion (CRI) was increased to 3 μg/kg/min 10 min later and the cat finally responded and maintained its systolic blood pressure at 110–130 mm Hg. The phenylephrine CRI was gradually decreased as the blood pressure held above 90 mm Hg and was eventually weaned off by the end of the procedure. The cat was administered 130 ml of intravenous isotonic fluids (Plasma-lyte A; Baxter) during anesthesia. Single full-thickness biopsies of the stomach, duodenum, jejunum and ileum were taken and the submucosa and mucosa were closed with 4-0 polydioxanone (PDS) in a simple interrupted pattern. The seromuscular layer of the stomach was closed with 4-0 PDS with an interrupted Lembert pattern. A marginal biopsy of the liver was also obtained using an encircling ligature with 3-0 PDS. Total anesthesia time was 2.5 h.

Anesthetic recovery was normal. Methadone (Bioniche, 0.2 mg/kg IV) was administered during recovery and the cat received intravenous fluids at 2 m/kg/h. The urinary bladder was expressed in the immediate postoperative period. The cat was also administered two doses of buprenorphine (Bedford, 0.01 mg/kg IV) for pain management overnight. Although the cat was not observed to have urinated or defecated in his litter pan, urine spots were noted on the bedding the day after surgery on which he was discharged. Histopathology of his biopsy samples showed chronic fibrosing lymphocytic jejunitis and small cell lymphoma in the ileum, which was subsequently managed with prednisolone (PrednisTab; Butler, 1.3 mg/kg PO q 24 h) and chlorambucil (Leukeran; GlaxoSmithKline, 0.5 mg/kg PO every other day).

Four days following discharge, the cat returned for urinary incontinence and no contemporaneous bowel movements. On physical examination, a large, easily expressible bladder, and a fecal-distended descending colon were palpated. Neurological examination revealed a markedly decreased perineal reflex and anal tone. Normal sensation was noted to the tail, perianal and preputial region, but tail tone was flaccid with no appreciable movement. The remainder of the examination was normal. The neuroanatomic localization was S1–S3 spinal cord segments involving the pelvic and pudendal nerves.

The cat was managed in-hospital with enemas and frequent manual expressions of the urinary bladder. He was started on bethanecol (compounded, 0.3 mg/kg PO q 12 h) which was subsequently discontinued due to persistent vomiting. Cisapride (compounded) was administered at the maximum dose tolerated without vomiting (1.6 mg/kg PO in the morning and 1.3 mg/kg PO in the afternoon). Lactulose (0.5 ml/kg PO q 8 h) was administered as a stool softener.

Three weeks postoperatively, the cat continued to have a decreased perineal reflex but had improved tone and appreciable voluntary movement of the tail. Three months postoperatively, the cat had normal tail movement, anal tone, and perianal sensation but decreased perineal reflex persisted. Over the next year, the cat's urinary retention and constipation decreased with continued medical management. One year postoperatively, the cat was able to void larger volumes of urine voluntarily with some urine retention and dribbling. With medications, the cat was eventually able to have a bowel movement once every 2–4 days, although intermittent enemas approximately every 6 weeks were still needed. The cat did develop a urinary tract infection 6 and 10 months postoperatively, likely secondary to urine retention and immunosuppression from prednisolone and chlorambucil.

The micturition reflex is regulated by a combination of autonomic and somatic nervous control. Important for urine retention, the sympathetic nervous system relaxes the detrusor muscle and contracts the trigone. ³ The preganglionic fibers of the hypogastric nerve exit the L2–L5 spinal cord segments in cats, synapse at the caudal mesenteric ganglion, then divide into the hypogastric nerves which course to the pelvic plexus. Important for urine voiding, the parasympathetic preganglionic axons of the pelvic nerve originate from S1–S3 spinal cord segments, which branch to innervate the urogenital organs, rectum and descending colon. Stretch receptors on the bladder wall send afferent signals via the pelvic nerve to the spinal cord that ascend to the brain. From here, descending parasympathetic efferent signals cause the detrusor muscle to contract and the internal and external sphincter to relax causing micturition. ⁴ Voluntary control of micturition is mediated through the pudendal nerve which has contributions from S1–S3 spinal cord segments and innervates the striated muscle of the urethra. The pudendal nerve also divides into the caudal rectal nerve which innervates the external anal sphincter and the perineal nerves which innervate the skin of the anus and perineum. ⁵

Similarly, the act of defecation is also regulated by a combination of autonomic and somatic nervous control. The sympathetic nervous system prevents defecation by sending inhibitory signals to the descending colon and rectum while sending excitatory signals to the internal anal sphincter. The parasympathetic nervous system acts through the pelvic nerves to facilitate defecation through contraction of the descending colon, rectum and internal anal sphincter. ⁶ Rectal distension with feces stretches the puborectalis muscle causing a reflex relaxation of the internal anal sphincter. ⁷

An epidural injection in dogs and cats is typically given at the lumbo-sacral junction between the internal and external laminae of the dura mater surrounding the spinal cord. The subarachnoid space between the arachnoid and the pia mater contains the cerebrospinal fluid and is used for subarachnoid injections, also known as intrathecal injections. ⁸

In this case, possible etiologies of the clinical signs include traumatic injury to the sacral spinal nerves, accidental subarachnoid injection, spinal ischemia, or neoplasia. Magnetic resonance imaging was recommended but the owner declined further diagnostic evaluation. Due to the acute onset of signs following the surgery, traumatic injury during the epidural analgesia is considered most likely. Direct injury to the preganglionic axons of the pelvic nerve would disrupt the parasympathetic regulation of micturition and defecation causing urinary retention and constipation. Damage to the pudendal nerve would prevent voluntary control of micturition, decrease anal tone and perineal reflexes.

Another consideration includes the accidental puncture of the dura mater during needle placement resulting in an intrathecal injection of high dose morphine intended for an epidural injection. In humans, lack of spontaneous bowel movements, decreased perineal reflexes, and chronic urinary retention requiring self catheterization have been reported following intrathecal analgesic procedures. ^9,10 However, this is unlikely in this case as correct epidural placement with the second needle was determined by the loss-of-resistance technique, and no cerebrospinal fluid was observed.

Vascular injury to the spinal nerves should also be considered due to the prolonged general anesthesia and hypotensive episode. In one study that looked at 339 cats with spinal cord disease, only 1/15 cats was seen with a single vascular lesion in the lumbo-sacral area whereas others were affected in multiple spinal cord segments. ¹¹ It is possible that the cat was more susceptible to an ischemic or embolic event at the sacral spinal cord due to the mild trauma the epidural needle likely caused.

Excessive vasoconstriction causing inadequate spinal blood flow in response to the phenylephrine CRI is another possible etiology of the cat's clinical signs. Given that low spinal blood flow could occur in the absence of signs of poor perfusion elsewhere in the body, it is difficult to speculate on the adequacy of this cat's spinal blood flow. However, the authors are not aware of published reports of phenylephrine at 2–3 μg/kg/min causing neurologic sequelae secondary to low spinal blood flow in cats, nor have the authors observed this complication at their teaching hospital in which phenylephrine is administered frequently for hypotension in anesthetized cats.

Opioid mediated urinary retention is reported at the highest rates when administered epidurally. ¹² Morphine has been seen to have supraspinal, spinal and peripheral mechanisms causing blockage of parasympathetic detrusor muscle contractions and internal sphincter relaxation. ⁴ Rates of urinary retention following opioid epidurals are as high as 10–15% in human literature ¹² and less than 11% in veterinary literature. ⁷ In human studies, bladder function has been seen to resolve within 16 h as the effects of the opioid dissipate. ¹³ It is possible that initial opioid mediated urinary retention and lack of manual bladder expressions during the early stages of urinary retention postoperatively may have led to excess stretch of the bladder detrusor muscle causing bladder atony. However, opioid effects do not account for the cat's concurrent constipation following the morphine epidural, nor does it explain the chronicity of the clinical signs. The constipation may be a result of the intestinal lymphoma but is unlikely in this case due to the acute onset of concurrent urinary retention postoperatively.

Neoplasia most commonly affects the lumbosacral segments of the spinal cord in cats. ^11,14 However, the lack of progression of some neurologic deficits and improvement in others observed through regular exams performed within 15 months postoperatively makes neoplasia unlikely in this case.

To the best of the authors’ knowledge, this is the first reported case of urinary and fecal dysfunction following epidural analgesia in a cat. Postoperative urinary retention in dogs and cats has been reported following epidural analgesia with morphine and bupivicaine ^15,16 and numerous accounts of urinary retention and or fecal incontinence following both epidural and spinal analgesia have been reported in humans. ^9,10,17−21 Although rare, urinary and bowel dysfunction remains a serious potential risk following epidural or spinal analgesia in humans and animals.