J Manipulative Physiol Ther 2006 (May); 29 (4): 330–335
Annabel L. Kier, DC, Peter W. McCarthy, PhD
Welsh Institute of Chiropractic, School of Applied Sciences, University of Glamorgan, Pontypridd, Rhondda Cynon Taff, Wales, UK
OBJECTIVE: To discuss the case of a patient with chronic headache. Although not in severe pain at time of consultation, signs and symptoms raised concern. The patient later had a cerebrovascular accident.
CLINICAL FEATURES: A 49-year-old man with non-traumatic chronic episodic head and neck pain presented for care. Examination and plain film radiographs were unremarkable, suggesting a mechanical origin for the symptoms; however, information in the case history raised concerns.
Working Diagnosis Several symptoms and signs were of concern. Although his symptoms might have been related to exacerbation of chronic insidious non-traumatic cervical facet irritation or be cervicogenic/tension type headaches and intermittent migraines, the difficulties in determining a diagnosis in this case included several suspicious features. From the history, the patient's apparent elevated blood pressure in conjunction with the history of monitoring for hypertension was of concern as was the family history of cardiovascular disease including stroke. Tinnitus experienced bilaterally and nausea without vomiting during a severe attack was also alarming.
These are all symptoms that separately are potential risk factors for cerebrovascular incidents, but when considered simultaneously could not be ignored.
Figure 1 – Demographics of the patient population at risk for CVA associated with cervical manipulation.
Figure 2 – Major signs and symptoms of vertebrobasilar ischemia (VBI); a mnemonic for remembering the signs and symptoms: The 5 D's And 3 N's.
In addition, he had reduced cervical range of motion in all directions during a severe attack, symptoms not readily expected in non-traumatic facet irritation, migraine, or cervicogenic/tension type headaches.
The examination was also suspect in that despite a typical fixation pattern noted in the cervico-thoracic spine, the moderate degree of joint restriction revealed by palpation did not correlate with that expected from such a chronic mechanical dysfunction. In addition, none of his head pain could be reproduced. The working diagnosis and associated concerns were explained to the patient. It was recommended that the patient discuss his case further with his general practitioner before initiating a treatment program involving cervical manipulation. The following week, he had a cerebrovascular accident leaving him hemiplegic and with speech impairment.
INTERVENTION AND OUTCOME: The patient was examined and not manipulated by the doctor of chiropractic but referred back to his general practitioner for a second opinion. The following week, the patient was admitted to hospital having had a cerebrovascular accident.
CONCLUSION: The possible indication of the prodrome to a stroke may lie in the case history rather than the examination findings and provocative testing.
From the Full-Text Article:
Many questions arose. Was this vascular incident due to a chronic underlying weakness or some vascular malformation that caused the recurring complaint of neck and head pain? If so, where was it located and was it the cause of the migraines, occipital head pain, neck pain, or a combination there of? Was it in progress at the time of consultation? If patients do present to clinicians with vascular incidents in progress that are mimicking mechanical neck and head pain, what signs and symptoms (single or in combination) can be highlighted as potential (amber flags), if not clear contraindicators (red flags) for cervical manipulation? What would have happened if cervical spinal manipulation was applied within such a short period before the CVA occurring?
Hindsight is marvelous, although in this instance, nothing can be changed to help the patient. The patient has kindly allowed access to relevant medical notes in order that others may hopefully learn from the event. The monitoring of hypertension had occurred for a considerable time with readings varying from 180/100 to 150/85 mm Hg. Although never treated for hypertension, ambulatory blood pressure monitoring had been recommended. The patient mentioned that he had just never got around to following up on his blood pressure. It was mentioned by the patient's partner that the family had actually not realized how frequent the headaches were occurring; it had been a recurring symptom on and off for years and had “become a part of normal life,” as had his use of NSAIDs. After the event, it was also realized that shopping included regular weekly purchases of NSAIDs, which were so commonplace that the family failed to recognize it and the patient failed to report the regular use although specifically questioned about it.
During the day before the event, the patient had spent a pleasant time with his family, had stopped to help someone change a car tire, had experienced a good night of sleep, and awoke refreshed. The morning had included non-exertional work on the farm and in the office. At approximately midday, the patient entered the house complaining of an acute and severe headache. A grayish appearance, clammy skin, and sweating was noticed by the partner, who subsequently called for assistance. A seizure occurred in the ambulance before reaching the hospital. After the appropriate scans and investigations, a diagnosis of large middle cerebral artery territory subarachnoid hemorrhage from an aneurysm was reached. The discharge condition when referred to a neurosurgical hospital was noted as dense right hemiparesis. Subsequently, surgical procedures were performed to clip the aneurysm, but unfortunately, further complications arose after a deep vein thrombosis from which an embolism settled in the lung tissues, nearly causing death. The family's recollection is that they were informed that the vascular weakness may have been congenital in origin and further weakened over time by the high blood pressure. Added to this, one cannot eliminate the effects of prolonged NSAID use.
During a later interview, it was noted that the patient was ambulatory around the house with the assistance of a stick, but in need of a wheel chair for any distance. During gait, the right lower limb was held extended with only very mild plantar or dorsiflexion of the foot and the right upper limb was held in a flexed position. Speech impediment was noted as was poor recollection of words. The family indicated that the patient experienced enhanced feelings of being overwhelmed and was becoming increasingly emotional.
As this was an intracranial hemorrhage and not a cervical or carotid artery dissection, it may be considered a less relevant case report with respect to cervical manipulation. However, the point of interest here though is that the patient did not present with overtly severe pain nor were there any neurological signs or symptoms. Except for the aspects already highlighted above from the history, the physical component of the case presented as a musculoskeletal disorder. It may even be plausible to consider that the hemorrhage had nothing to do with any of the symptoms. A significant counter point to this statement would be that the patient has not experienced headaches or neck pain since the stroke that occurred 3 years ago.
Patients who have vertebrobasilar stroke or stroke-like symptoms after neck manipulation have been found to demonstrate the demographic characteristics as seen in
A number of cases are documented where sustained, repetitive, or spontaneous head positioning during normal activities (ie, no external forces or trauma) have produced a brainstem ischemic accident, thereby suggesting that cervical arterial complications may occur as a consequence of apparently minor damage or minor trauma or daily activities. [3, 7, 8] There appears to be an increased risk of vascular injury to the cervical spine post-manipulation associated with, but not limited to, the following: congenital abnormalities, inherited disorders of collagen, elevated plasma homocysteine levels, previous vascular trauma, previous ischemic episodes and atherosclerosis subsequent to hypertension, smoking, oral contraceptives, and high alcohol consumption. [5, 9-10] It is therefore interesting to note that Hosoya et al  found atherosclerosis and hypertension to be rare in dissecting aneurysm of the vertebral artery. A curiosity that may originate is the fact that the study was performed on a Japanese population who, according to Beatriz et al,  carry a higher risk of hemorrhagic stroke than an equivalent population in the United States who have a increased risk for thromboembolic stroke. This highlights the importance of paying attention to a patient's ethnic background, diet, and lifestyle as well as presenting signs and symptoms.
Much research has been published over the recent years focusing upon blood flow in the cervical vascular structures in particular the vertebral artery. Questions have been raised concerning the validity and reliability of the vertebral artery insufficiency test, not to mention the potential traumatic effects of spinal manipulative therapy upon the vascular structures on the cervical spine. In light of this being outside the scope of this report, the reader is referred to other articles. [12–16]
Warning Signs and Symptoms
There are a number of signs and symptoms that require clinical consideration when contemplating treatment of the cervical spine ( Fig 2 ). These symptoms are relevant if they present singularly as well as together, but are particularly important when associated with sudden onset of severe neck and head pain. Dizziness is a known symptom of CVAs both intracranially and extracranially. However, it is important to emphasize that dizziness can also arise as a symptom secondary to cervical mechanoreceptor dysfunction,  which is not an uncommon symptom of whiplash-associated disorders and other conditions that are commonly managed by chiropractors. A variety of additional signs and symptoms may also be noted with CVAs, such as cerebellar signs, carotid bruits, and scintillating scotomata. [18, 19] Furthermore, if a dissection affects the sympathetic fibers associated with the internal carotid plexus, a Horner's syndrome might develop, whereas if the sympathetic fibers associated with the external carotid plexus were to be affected, it would impinge on the function of the facial sweat glands. 
Head and Neck Pain
Symptoms related to CVA frequently include unilateral, unprompted posterior cervical pain of acute onset. These symptoms may precede the onset of a stroke by several days and must be considered important warning signs. [18–20] A significant positive association has been shown between carotid artery dissection and migraine particularly with respect to young females. [9, 18] Furthermore, unlike migraines (with or without aura), the pain associated with internal carotid artery dissection is often described as a constant ache, rather than being pulsatile.  Therefore, a diagnosis of CVA may indeed be delayed due to a preexisting history of migraine headache.  It should also be noted that although women report headaches twice as often as men, chronic headache in males is an independent predictor for stroke and therefore consequently a potential indicator of an underlying disease process.  One form of headache worth noting for its association with CVA is thunderclap headache.  The term thunderclap headache is ascribed to a headache of instantaneous onset, sudden and unexpected as a “clap of thunder” such as that described by patients experiencing a subarachnoid hemorrhage. Thunderclap headache has been associated with a migraine variant know as “crash migraine,” spontaneous intracranial hypotension, and unruptured subarachnoid aneurysm, among other conditions. Often, negative findings of both CT scan and lumbar puncture divert from the potential indication of more sinister intracranial and extracranial vascular pathology such as acute aneurysmal expansion and therefore an increased risk of future rupture. Due to this complication, it has been suggested that cerebral angiography should be performed to either verify or rule out the presence of an unruptured aneurysm in cases of sudden onset severe headache. [23, 24]
In short, as with other disease processes, the presentation of CVA can be quite variable making the process of reaching an accurate diagnosis difficult. Symptoms associated with CVAs can range from one extreme to the other, from sudden severe neck and head pain with associated neurological signs and symptoms through no change in preexisting symptoms such as headache, to no pain at all.  The latter was confirmed by Hosoya et al,  who suggested that intracranial vertebrobasilar artery dissection is probably more frequent than previously considered, following their incidental findings of asymptomatic brainstem dissections.
The Clinician's Dilemma
The clinician's problems with respect to cervical manipulation and possible CVA are 2-fold. First, the literature is unclear and without consensus on the estimated risk factors related to cervical manipulation. Second, as the actual physical presentation of CVAs can be complex and diverse, arriving at a diagnosis can consequently become difficult. In addition, there is the possibility that a patient unknowingly reports their self-medication incorrectly, potentially misleading the clinician, as occurred in the case reported here.
Arterial dissection after cervical manipulation is commonly estimated as an incident in every 1 to 2 million cervical manipulations; however, the range extends from 1:400000 to 1:5.85 million.  Although the medical literature does not support a clear causal relationship between chiropractic manipulative therapy and internal carotid artery dissection,  publications concerning such events after cervical spine manipulation continue to be reported. The suggestion of underreporting of such events was highlighted by Depeyron et al  who ascertained the incidence of vertebrobasilar accident after cervical manipulation to be 30 times higher than previously considered. In contrast, Breen9 emphasized there is a lack of available high-quality data to provide the denominator required for any calculation of either the risk or the odds ratios of stroke after manipulation and thereby to assess causation. This problem would benefit from interprofessional collaboration between researchers from all the manipulative disciplines and neurologists. Indeed, recently, it has been suggested that spinal manipulation should remain under very strict medical control in an attempt to minimize risks of complications.  This proposal could only be successful should superior skills be developed in assessment and triaging, a challenge exemplified in a case report where a male patient developed a large infarct in the left cerebellar hemisphere after being treated twice within 24 hours by two medically qualified doctors initially using mobilization and subsequently manipulation. 
The difficulty in detecting and diagnosing a CVA in progress is not restricted to the chiropractic profession. Recently, Latronico et al28 described a 24-year-old mother presenting at an emergency department, complaining of a sudden onset of neck pain and paresthesia in both arms. Not having any recollection of trauma, manipulation (specifically chiropractic), or abrupt head movement, she was treated with analgesics and sent home. After a second presentation at the emergency room 24 hours later, she was admitted and diagnosed as tetraparetic due to spinal cord infarct. It is notable that the CVA in progress was not apparent even to emergency room staff during the first observation. The report does not detail the initial examination findings; however, it would be interesting to know whether CVA was considered as one of the initial differential diagnoses.
Although it would appear prudent to avoid cervical manipulation in patients with pre existing cerebrovascular disease, Rubinstein and Haldeman  described the use of rotational cervical manipulation in treating a patient who had previously undergone a traumatically induced dissection of the internal carotid artery. As the patient responded well, it was proposed that spinal manipulative therapy could be included in the treatment plan in selected cases with full and complete informed consent. In light of the above, a recent Cochrane Collaboration  report stated that neither manipulation nor mobilization was superior when comparing manual treatment for mechanical neck disorders. Indeed, most benefit appears to be achieved when used in a multimodal setting combined with exercises. Furthermore, a potentially important difference between the outcome of cervical manipulation and mobilization is that the former is more likely to result in objectively minor adverse reactions, with less patient satisfaction and poorer outcomes. 
It is imperative to avoid overstating the risk of CVA after cervical spinal manipulation, yet it is equally crucial to recognize it as a direct or indirect clinical possibility. Clinical apathy can arise from assuming that as there is no conclusive proof for CVAs as a consequence of cervical manipulation. It remains essential for both the patient's health and the profession's future that we remain vigilant—vigilant not only regarding knowledge concerning the potential presence of CVAs in progress, their underlying causative, and/or contributing factors, but arguably more importantly, with respect to the clinical diversity with which they may present. Thoughts linger concerning the possible legal consequences, which may subsequently have occurred, should spinal manipulation have taken place the day of consultation for this particular patient.