Lumbar Sympathectomy for Lower Extremity Ischemic Ulcers

Anatomy and Physiology

Preganglionic sympathetic fibers that provide outflow to the lower extremities arise from cell bodies in the lateral gray substance of spinal cord segments T10 to L3. These fibers emerge by way of the anterior spinal roots and reach the sympathetic chain through the white rami communicantes. After synapsing with ganglionic cells, the postganglionic fibers pass through the gray rami communicantes to join somatic nerves. These nerves are then distributed to the peripheral blood vessels and sweat glands. Sympathetic innervation of the leg passes through the L1–L4 ganglia. The sympathetic trunks enter the abdomen as paired structures lying on the anterolateral region of the vertebral bodies. This places the left and right trunks behind the aorta and vena cava, respectively ( Figure 1 ).

The number of lumbar ganglia varies from three to five. The L1 and L2 ganglia can be fused and lie behind the diaphragmatic crura. Because some preganglionic fibers bypass the ganglia or cross over to the chain on the contralateral side, complete sympathetic denervation of an extremity requires division of preganglionic fibers at their origin, as well as removal of the appropriate ganglia and intercommunicating fibers. Although removal of L2 and L3 ganglia is probably sufficient for most clinical indications, the L4 ganglion should also be removed because of the possibility of collateral innervation. The L1 ganglion should be left intact to avoid the complication of retrograde ejaculation in men, although this usually occurs only with bilateral L1 ganglionectomy.

The effect of sympathectomy on lower extremity blood flow has been the subject of considerable research. Resting skin blood flow is regulated primarily by the demands of thermal homeostasis. To effect cooling, large volumes of blood are diverted through cutaneous arteriovenous anastomoses (AVAs). These are large-diameter (50-μm) shunts between arterioles and venules that bypass nutrient capillaries. These AVAs have little or no intrinsic resting muscle tone but depend on sympathetic vasoactivity to control their diameter. In contrast, skin arterioles exhibit a relatively high baseline myogenic tone, which is little influenced by sympathetic denervation. Thus, sympathectomy would be expected to have little effect on skin capillary blood flow but should dilate AVAs.

Direct measurement of the blood flow partitioned through AVAs versus capillaries using radioactive microspheres in both normal and ischemic canine hind limbs has shown that the eightfold increase in total extremity blood flow after sympathectomy is entirely accounted for by increased AVA flow. This increased AVA flow, rather increased nutrient skin capillary flow, accounts for increased skin temperature. Using isotope-clearance techniques, skin capillary blood flow in humans has been shown to be unchanged after sympathectomy, despite increased temperature and total extremity blood flow. Increased transcutaneous oxygen tension (TcP o ₂ ) in ischemic limbs after sympathectomy is also caused by the increased flow through AVAs, not an increase in nutrient vascular supply to the dermal tissues.

There is little evidence that sympathetic nerves have any vasoconstrictive influence on larger arteries, and even so, most significant collateral vessels are proximal to the area denervated by the standard L2 to L4 sympathectomy. The transient increase in collateral blood flow after sympathectomy is likely a result of decreased resistance from the maximal dilation of AVAs. In ischemic canine hind limbs following femoral artery ligation, vasodilation after sympathectomy results from dilation of AVAs, rather than skin or muscle arterioles, despite a measurable oxygen debt in the extremity. Thus, there is little if any evidence that sympathectomy results in an increase in nutrient blood flow to skin. Furthermore, sympathectomy causes no increase in blood flow to exercising skeletal muscle, which is regulated by local metabolic factors.

Clinical Results

A small number of randomized trials have been conducted to compare the effects of lumbar sympathectomy with conservative treatment in patients with unreconstructable critical limb ischemia. None have shown beneficial effects of sympathectomy on objective endpoints. A systematic review of four clinical trials and four observational studies found that sympathectomy did not show significant differences in grade of intermittent claudication, mortality, and amputation compared with medical management.

In contrast, noncontrolled clinical studies have reported a benefit of lumbar sympathectomy in approximately 50% of patients, but the benefits related to subjective endpoints such as rest pain or patient-reported walking distance. Relief of rest pain and complete ulcer healing for 6 months after sympathectomy have been reported in 28% to 73% of patients, illustrating the large variation in results. Unfortunately, increased skin temperature is often erroneously cited as evidence of the beneficial effect of sympathectomy in these patients. Most studies have concluded that the best results of lumbar sympathectomy are obtained in patients with only mild disease (ankle-to-brachial index >0.3) and small, superficial ulcers, the very patients who are likely to heal with local wound care alone, making it difficult to differentiate the true effect of lumbar sympathectomy from a placebo effect. In fact, placebo treatments of patients with small cutaneous ulcers and rest pain in randomized drug trials have had a beneficial result in more than 50% of patients, an outcome comparable to most observational studies of lumbar sympathectomy.

These results notwithstanding, some continue to recommend lumbar sympathectomy for patients with distal atheroembolization causing severe pain and local, unreconstructable ischemia. In such patients, a beneficial effect of sympathectomy may be related to potential interruption of pain-amplification networks, such as postulated in the treatment of posttraumatic pain syndrome. Intermittent sympathetic nerve blocks, as a diagnostic or even therapeutic modality, may be advantageous in such patients before considering permanent sympathectomy. Unfortunately, no controlled trials exist to justify sympathectomy for this application. Finally, lumbar sympathectomy has been recommended by some to improve graft patency if used in conjunction with proximal arterial bypass. In this area, randomized clinical trials have clearly shown that adjunctive sympathectomy has no benefit for improving graft patency.

Despite the lack of evidence supporting clinical benefit of lumbar sympathectomy in patients with unreconstructable critical limb ischemia, the procedure is still routinely performed for this indication, particularly in Europe. One study showed that 75% of practitioners in Ireland and the United Kingdom feel that chemical lumbar sympathectomy in particular plays a role in their current practice, although this is likely an overestimate caused by selection bias.