Selective dorsal rhizotomy

39.1

History

Despite clear advantages over complex repair-strategies of irreversible structural changes, selective dorsal rhizotomy (SDR) is still not recognized as an early prophylactic surgical option in most current guidelines. This striking reluctance can be understood by examining the fraught history of SDR. Initial attempts at using SRD to treat spasticity date to the early 20th century, yet due to an allegedly high rate of intraoperative complications or early return of spasticity , it was not adapted as a viable therapy. This commonly misinterpreted early history of SDR urgently needs correction. Intradural spinal surgery started with Victor Horsley in 1887 , reporting the removal of an intradural tumor, in a previously paraplegic British army captain, who regained his ability to walk after surgery. This new approach gained traction and was used to interrupt the transmission of adverse sensations by cutting dorsal spinal nerve roots. Rhizotomy can be attributed to two separate individuals, one in London and the other in New York in 1888. Charles Loomis Dana ( Fig. 39.1 ) proposed an SDR for a 43-year-old patient with brachial plexus neuralgia secondary to a shrapnel and stretch injury. Attempted therapies, including nerve stretching, amputation of the arm, and neuroma excision, were unsuccessful. Robert Abbe ( Fig. 39.2 ) performed a C6–C8 posterior spinal rhizotomy ( Fig. 39.3 ) at St. Luke’s Hospital in New York the same year with good results . During this procedure, nerve root localization and function were determined by electric stimulation.

On December 24, 1888, William H. Bennet performed a dorsal rhizotomy at St. Georges Hospital in London on a 45-year-old man with severe leg pain due to tibial osteomyelitis. Despite an amputation, the pain remained, and the patient subsequently underwent lumbar and sacral rhizotomies on the left side. He experienced good pain relief but unfortunately suffered and died from an intracranial hemorrhage some days later. Both reports encouraged multiple, rarely cited attempts, to use rhizotomy for pain relief in neuralgia and sciatica , in tumors and in so-called gastric crises . These experimental therapeutic attempts achieved expectedly inconsistent results and were the primary reasons that SDR was not accepted as a viable procedure.

Based on the scientific work of Charles Scott Sherrington , Otfrid Foerster ( Fig. 39.4 ) was the first to apply SDR specifically for the treatment of spasticity , again followed by many other rarely cited surgeons . In 1911 he published on 81 patients from his surgical series of 107 patients, 54 of them suffering from spasticity due to CP . Almost all of these patients benefited from surgery . Perioperative morbidity and mortality remained within the historically appropriate benchmarks. Unfortunately, the side effects and unreliable results of dorsal root sections for other reasons than spasticity caused SDR to vanish from the surgical theaters.

In 1954 SDR for cerebral palsy (CP) reemerged due to the contributions of Gros et al. in Montpellier. He modified the previously rather quantitative approach to root sectioning, introducing a complex algorithm of electromyography (EMG) data with functional aspects in order to preserve specific muscle groups relating to individual functional patterns .

The access to the roots L2 through S2 was traditionally achieved via a multilevel laminectomy. The actual approach to the roots is still under debate. The shift from approaching the roots at their medullar origin rather than at their dural exit, as proposed by Groves as early as 1911 and supported by Foerster , finally provided the basis for limiting the access in the first instance to a moderately extended laminectomy T11 to L2:

…It has usually been advised in dealing with the lumbar and sacral roots, to approach them at their point of exit from the dural cavity. But this necessitates the removal of all the lumbar laminae. These, on account of their size, their depth, and the thickness of the muscles covering them, are the most difficult to remove. It seems to me that it is much simpler and easier to attack the lower roots at the point where they all lie close together at their origin from the spinal cord. This necessitates a smaller removal of laminae, because the roots lie crowded close together instead off apart…

Gros adopted and modified this approach, further reducing the surgical access to three levels with the patient in a stable lateral position. Rootlet sections were limited to 80% per root ( Fig. 39.5 ). Fasano , using a similar access in prone position, questioned the importance of the quantity of sectioned rootlets and replaced this approach by a concept of tailored “functional SDR,” basing the section of rootlets exclusively on pathologic EMG results. Peacock and Arens returned to serial laminectomies L2 to S1, out of concern for the proper protection of the roots S2 to S5 through the limited exposure used by Fasano. Park finally showed that a correctly placed single-level laminectomy with the patient in prone position provides a sufficient overview of the conus including the entire group of sacral rootlets. This until now most limited access, together with the additional systematic section of L1 rootlets to improve hip mobility, was established in 1991 and modified by the author in 2008.

39.2

Patient selection

Earlier in history, the indication for SDR in CP was not tied to the patient’s ability to walk. Although Foerster already discusses a limitation, excluding “slight spasms” and “complete paralysis,” CP affected children, showing the full range of spasticity-induced handicaps, were submitted to surgery in the early stage of SDR. Later, SDR was increasingly focused on improving walking patterns, and the indication was restricted to Gross Motor Function Classification System (GMFCS) Grades 2 and 3 in children . Currently, some centers make SDR available again for selected severely handicapped patients performing on GMFCS Grades 4 and 5, based on the promising first experiences of Park in St. Louis, Dekopov, Tomskĭ, Shabalov, and Salova at the Burdenko Neurosurgical Clinic in Moscow and Vloeberghs in Nottingham.

In our actual setting the evaluation of a child with CP is carried out by a multidisciplinary team, consisting of an orthopedic surgeon, a physiotherapist, a neuropediatrician, a neurosurgeon, and a technician for assistive devices. The group is experienced in almost all major current therapy options, including Botox injections, any kind of orthopedic surgeries and minimal invasive procedures, physiotherapy and the optimization of orthoses, walkers or wheelchairs. Although requiring a higher organizational input, I do believe in the advantage of complementary over competitive expertise. With everybody in the same room at the same patient a tailored therapy concept can be established on the spot, coordinating the input of all experts and avoiding the common confusion of families and caretakers, traveling from expert to expert and discussing contradictory statements. On top and most preciously the team gains quickly experience as patients come back and the efficacy of previous decisions can be reviewed.

SDR in this setting is offered to children between 2 and 18 years of age with a diagnosed spastic diplegia, spastic quadriplegia, or spastic hemiplegia and independent mobility with or without assistive device. MRI is mandatory to exclude severe damage to the basal ganglia and to confirm CP in patients with hemiplegia by characteristic features such as periventricular leukomalacia or stroke-related tissue loss. This offer is extended to selected adults between 19 and 50 years of age with only mild orthopedic deformities, if they are close to walking independently without assistive device. Functional improvement by other therapy options—that is, Botox-injections—supports the indication.

In the case of concurring orthopedic surgical indications for muscle release or tendon release procedures, we recommend the reduction of spasticity by SDR first. After major orthopedic surgery we recommend to postpone SDR for 1 year to regain lost muscle strength.

SDR is also offered to selected children performing on GMFCS Grades 4 and 5 as an alternative to intrathecal baclofen therapy. The target of a surgical reduction of spasticity in these cases is an improvement of quality of life regarding pain, sleep, and positioning. We do not promise functional gains, but we are frequently impressed by an improvement of upper extremity function, speech, and cognitive performance.

SDR is not recommended to patients who suffered congenital brain infection, hydrocephalus unrelated to premature birth, or severe head trauma. SDR does not influence dystonia or severe scoliosis and should not be offered to concerned patients.

39.3