Placement of the needle at the L3 level for a left lumbar sympathetic block shows uptake of the dye by the left kidney via ureter. Needle was removed and then positioned lower at the left L4 level; patient had excellent relief of his sympathetic mediated pain. He was cautioned regarding blood in urine but ultimately had no sequelae from the initial needle placement in the left ureter and kidney; he subsequently returned for repeat blocks that were all performed at L4 level to avoid contact with the kidney. These images are from Dr. Anitescu’s personal library
The patient underwent anterograde pyelography which demonstrated extravasation at the upper third of the proximal ureter and concurrent formation of urinoma at the site with no passage to the distal ureter and bladder. The patient was taken to the operating room for exploration; the urinoma was drained and a ureteroureterostomy was performed; a nephrectomy was deemed unnecessary. The patient was seen 3 months later and had full resolution of his abdominal and lower back symptoms as well as full recovery from the ureteral injury.
Lumbar sympathectomy was first reported by a French surgeon, Leriche. In 1920s, the first percutaneous lumbar sympathetic block was first described by Kappis and Mandl, with several subsequent variations in technique . The goal of lumbar sympathetic block is to utilize local anesthetic and/or neurolytic agents to inhibit sympathetically mediated pain signal transmission at the level of the ganglion . While advances in imaging techniques have increased the safety of such blocks, neither fluoroscopy nor computed tomography has precluded ureteral injury . To date, there are seven case reports of such injury in the literature. Presentation of post-procedure ureteral injury is variable in both its symptoms and its time course, thereby warranting close follow-up and requiring a high index of suspicion for diagnosis.
Lumbar Sympathetic Block
Anatomy and physiology
The lumbar sympathetic ganglia are located in the retroperitoneum, anterolateral to the lumbar vertebral bodies at the attachment site of the psoas muscle. Variability exists in the number of ganglia as well as their location; cadaveric studies demonstrate three ganglia; however, up to five separate ganglia have been reported in the literature. The most common reported location of the ganglia is anterior to the L2/L3 intervertebral disc; as such, the ganglia can be blocked anywhere between L2 and L4 . Several surrounding structures make placement of a lumbar sympathetic block precarious. These structures include the somatic lumbar plexus, the intervertebral foramen with access to the epidural space, the subarachnoid space, and the spinal cord, as well as important vascular structures including the inferior vena cava anteriorly on the right and the abdominal aorta anteromedial on the left. Smaller vascular structures, such as the lumbar arteries and veins, also are present, as are the ureters . Complications may arise when such surrounding structures are compromised either by direct injury from the needle or injury secondary to injection of neurolytic agents.
A posterior oblique approach to the ganglia with the patient in the prone position is most commonly used. A lateral approach was also described by George Wallace in 1955 with the suggested advantage of increasing patient comfort; Wallace’s approach described a patient in the lateral position with needle entrance at the apex of the “lumbar triangle” or convergence of the border of the twelfth rib, the superior border of the iliac crest, and the paravertebral musculature at L2 . More recently, a transdiscal approach has been suggested with the aim of decreasing intra-psoas injection and thus decreasing genitofemoral neuritis . Approaches also vary in the number of needle placements utilized to achieve blockade. A study by Hong et al. demonstrated a significant difference in great toe temperature change with three injections (at L2, L3, and L4) as compared with single injection at L2 .
While the original technique was based solely on anatomic landmarks, complications arising secondary to the presence of local anatomic structures have led to the use of image guidance to improve safety profile. Imaging tools utilized include both ultrasound  and MRI , although fluoroscopy has been the most commonly used modality in pain management . X-ray fluoroscopy is currently the primary mode of imaging among pain physicians, although computer tomography (CT) fluoroscopy may offer improved safety as it allows visualization of visceral structures not readily seen on plain radiography [16, 17].
Local anesthetics and neurolytic agents
Sympathetic blockade with local anesthetic can be diagnostic and/or therapeutic. Repeated LSBs with local anesthetic have been performed  to treat CRPS with success. In one study, 86% of 29 patients in a case series with CRPS following knee surgery treated with LBS with .375% bupivacaine as the injectate demonstrated partial to complete relief of knee pain for variable durations .
Chemicals, thermal ablation, and radiofrequency ablation (RFA) have been used to extend the duration of the sympathetic blocking effect of LSB . Phenol and alcohol are the most common chemical agents used for blockade; however, case reports of agents such as botulinum toxin  and clonidine  also exist. Studies comparing phenol ablation to thermal ablation demonstrated increased sympathectomy in the phenol group; a similar study comparing phenol ablation to RFA revealed comparable levels of sympathectomy . The effects of chemical sympathectomy typically last about 3–6 months, whereas the effects of RFA last up to 1 year .
The incidence of neuralgia has been quoted between 6 and 40%. The most common neuralgia induced by LSB is in the genitofemoral distribution; however, symptoms affecting the lateral thigh have also been described.
Subarachnoid injection, post-dural puncture headache, paraplegia, aseptic meningitis, epidural injection, and subdural injection have all been described.
Complications arise from intravascular injection, intralymphatic injection, and bleeding secondary to puncture of vascular structures. Retroperitoneal hematoma has been reported as a complication secondary to vascular puncture in anticoagulated patients.
More unusual complications also have been reported, including pneumothorax, intervertebral disc penetration, allergic reactions to injectate, inability to ejaculate in men, and ureteral and/or renal injury.