Evidence from the study also suggests that spinal manipulative therapy may exacerbate pre-existing vertebral artery dissections.
The findings are published in the May 13 issue of Neurology.
Several previously published studies have reported cases in which spinal manipulative therapy of the neck has preceded stroke by minutes or weeks in patients who subsequently developed vertebral artery dissections, but these studies were not designed to establish causation. The current study teased out the variables that could explain the onset of stroke -- including addressing the question of whether patients with spontaneous cervical arterial dissection seek spinal manipulative therapy because of neck pain or whether spinal manipulative therapy either causes dissection or exacerbates a pre-existing dissection – and determined that treatment is an independent risk factor.
The researchers reported that patients under age 60 who had strokes or transient ischemic attacks from tears in the vertebral artery were six times more likely to have had spinal manipulative therapy in the 30 days prior to their stroke than patients who had strokes from other causes. The patients tended to be otherwise healthier than most stroke patients, lacking such risk factors as diabetes, high blood pressure and atherosclerosis.
"The number of chiropractic manipulations performed in the United States each year is in the millions, and the incidence of stroke from all causes is only 10 per 100,000, so we're not talking about large numbers of victims. But rare incidences do happen, and physicians and patients should be aware of spinal manipulation therapy as a rare but potentially causal factor in stroke," said the first author of the study, Wade S. Smith, MD, PhD, UCSF associate professor of neurology and director of the UCSF Neurovascular Service at UCSF Medical Center.
Given this link, he said, physicians presented with patients displaying the neurological symptoms of stroke or TIA – particularly younger patients - should be alert to the possibility that dissection is the cause, so that they can institute presumptive treatment to prevent further strokes. Likewise, he said, chiropractic practitioners performing spinal manipulative therapy should warn patients of the risk of cervical artery dissection. In fact, he says patients should be screened by a physician for symptoms of pre-existing vertebral arterial dissection, such as TIA, before beginning therapy, and be told that a significant increase in neck pain or neurological signs or symptoms within 30 days following spinal manipulative therapy warrants immediate medical evaluation.
In their paper, the researchers propose that since spinal manipulative therapy is a medical procedure, practitioners should carry out a formal consent process, in which risk would be disclosed. Smith noted that physicians are expected to disclose any risk to patients regarding procedures or drug therapies.
The researchers were not able to determine whether particular neck manipulations cause the rare cases of vertebral arterial dissection, though there is evidence in the scientific literature, says Smith, that rapid movements over short distances are safer than quick movements over long distances. (Scientists suspect that the vertebral artery is particularly vulnerable to mechanical dissection due to its horizontal course along the first cervical vertebra, where it can be compressed or placed under traction as the head is extended and rotated.)
Regardless, the tear in the artery causes the formation of a blood clot in the artery, either leading to the brain or in the brain itself, that blocks blood flow to the brain and leads to the often severe neurological impairment associated with stroke and TIA.
Approximately 16 to 19 percent of strokes in young patients are attributed to spontaneous cervical arterial dissection, often accompanied by neck or head pain. While most spontaneous cases do not have a clear cause, certain uncommon conditions, as well as trauma, can predispose people. However, it has been unclear whether minor trauma or simple, self-initiated head and neck motions can produce dissection.
In their study, the scientists reviewed all patients under age 60 at UCSF Medical Center and Stanford Medical Center stroke centers who developed cervical arterial dissection and stroke or TIA between 1995 and 2000. After excluding patients who either did not meet the criteria for the study or who declined to participate, the scientists examined the medical cases and conducted interviews with 51 dissection patients (mean age 41, +/- 10 years; 59 percent female) and 100 control patients (mean age 44 +/- 9 years; 58 percent female). Participants in the latter group, who had strokes not caused by dissection, were matched by gender and age, and were randomly selected.
Of this group, ten patients (seven dissection patients, three controls) were able to provide detailed information about the timing of spinal manipulative therapy and the timing of pain symptoms relative to stroke or TIA. Notably, the seven dissection patients received spinal manipulative therapy within 1.4 days of the neurovascular event, while the controls received it on average 8.4 days prior to stroke. Six of the dissection patients had their dissection in the vertebral artery, as opposed to other arteries leading to the brain. Of these seven patients, four had substantial increase or new and different pain immediately following spinal manipulative therapy – in contrast to the three controls who said they experienced relief of their pain -- and all four had vertebral artery dissections. Two of the dissection patients had a stroke within seconds of receiving spinal manipulative therapy. Both of these patients had vertebral dissections.
"These observations suggest that spinal manipulative therapy can directly produce dissection," says Smith. "It is highly improbably that a young patient will have a stroke and have had spinal manipulative therapy within seconds purely by chance, given the relatively low frequency of both events."
The researchers cited several limitations of their study, such as the fact that the data was collected retrospectively from a population of patients who responded to solicitation, leaving room for variables in patient recollection, variation in the diagnostic work-ups and patient self selection. However, they noted, while a prospective study could reduce the potential for such biases, it might be difficult to obtain a large enough sample given the rarity of the disease.
In an accompanying editorial in Neurology, neurologists Linda S. Williams, MD, and Jose Biller, MD, noted the limitations of the study but wrote that the six-fold risk determined by the scientists "must be taken seriously."
Moreover, they wrote, "In the absence of randomized controlled trial evidence demonstrating the effectiveness of cervical manipulation, the best current evidence suggests that the small risk of dissection and stroke outweighs the benefit of this treatment modality for patients with acute neck pain."
Co-authors of the study were S. Claiborne Johnston, MD, PhD, UCSF associate professor of neurology and director of the UCSF Stroke Service, UCSF Medical Center; E.J. Skalabrin, MD, of the Department of Neurology, University of Utah; M. Weaver, MS, of the Stanford University Center for Biomedical Ethics Department; P. Azari, St. George's University School of Medicine; G.W. Albers, MD, Department of Neurology, Stanford University and D.R. Gress, MD, of Lynchburg General Hospital.