J Manipulative Physiol Ther. 2009 (Sep); 32 (7): 586–591 ~ FULL TEXT
Gábor Ormos, MD, J.N. Mehrishi, PhD, FRCPath, Tibor Bakács, MD, DSc
Department of Rehabilitation,
Hungarian National Institute for Rheumatology and Physiotherapy (HNIRP),
OBJECTIVE: This case report discusses the treatment of 2 patients with cervicogenic headache (CHA) attending the Outpatient Clinic of the Hungarian National Institute for Rheumatology and Physiotherapy (Budapest, Hungary) and reviews the pathophysiology, therapeutic strategy, and problems associated with the treatment of CHA.
CLINICAL FEATURES: Patient 1 was a 27-year-old female who sustained a whiplash injury. A sharp, shooting headache developed, readily induced, and aggravated by just bending the neck backward or by turning her head. Magnetic resonance imaging revealed a disk protrusion at C4-C5 pressing the anterior cerebrospinal space. Patient 2 was a 62-year-old female who sustained a whiplash injury; her cervical movements became restricted, which precipitated headaches. Magnetic resonance imaging revealed a paramedian disk hernia between the C4 and C5 vertebrae that intruded into the right ventral cerebrospinal space.
INTERVENTION AND OUTCOME: After 4 weeks of manipulative therapy for patient 1, both active and passive range of motion returned to normal, and the high tumor necrosis factor-alpha (TNF-alpha) level (63 pg/mL) was substantially reduced (28 pg/mL). Patient 2 was started on manipulative therapy twice a week for 4 weeks; after 2 months, the patient became symptom-free, and high TNF-alpha level (72 pg/mL) was reduced greatly (35 pg/mL).
CONCLUSION: Two patients with whiplash injury and disk herniation developed CHA associated with very high TNF-alpha levels. After manipulative therapy, these patients became symptom-free, and their TNF-alpha levels decreased substantially.
KEYWORDS: Cervicogenic Headache, Tumor Necrosis Factor-α, Musculoskeletal Manipulations, Whiplash Injuries
From the Full-Text Article:
The incidence and severity of headache frequently may increase after head injury.  After whiplash injury (WI), cervicogenic headache (CHA) is relatively common. Cervicogenic headache arises from damaged structures around the neck, involving joints, ligaments, muscles, and cervical disks that have complex nerve endings. Opinions differ about the role of WI in the generation of CHA, and its treatment is also controversial. Neck support, neurosurgery, local anesthetic or steroid injections, and manipulative therapy (MT) among others are used.  Therefore, we thought that it would be worth reporting successful treatment of CHA from a new angle. Here, we present the first 2 patients with CHA found to have elevated blood tumor necrosis factor-α (TNF-α) levels, which were reduced after MT.
During the past decades, cytokines have been the subject of intense research for their potential usefulness to treat many disease states ranging from autoimmune disorders to cancer. In fact, the history of bench to bedside translation of research is an example of how it led to the identification of TNF-α as a key regulator of the loss of homeostatic immune-inflammatory responses in rheumatoid arthritis (RA) and a good therapeutic target.  Among the cytokines, TNF-α has proven to be a key ligand in triggering many intracellular processes, both physiologic and pathologic.  For example, TNF-α is a well-established mediator of inflammatory response and plays a critical role in a number pathologic condition, such as RA,  Crohn's disease,  and Jarisch-Herxheimer reaction.  Randomized clinical trials of anti-TNF reagents have demonstrated an acceptable safety profile and a marked clinical efficacy in cases of RA that have not responded adequately to conventional therapy. [3, 8, 9] Treatment of Crohn's disease [10-12] and Jarisch-Herxheimer reaction  with TNF-α antibodies have clearly demonstrated the central role of TNF-α in these pathologic conditions.
Tumor necrosis factor-α in the nervous system (eg, reference range, 0-5 pg/mL) is released after peripheral nerve injury. Tumor necrosis factor-α induces thrombus formation, intraneural edema, and a reduction of nerve conduction velocity. [13-15] Tumor necrosis factor-α has been suggested as the main candidate among substances potentially responsible for nerve root pain requiring TNF-α antagonists in patients with disk-related sciatica. 
It was thought that during WI any TNF-α that might be produced in large amounts could be harmful. We were curious to find a meaningful correlation between WI and an endogenously released TNF-α for possible institution of routine monitoring in our clinics. The purpose of this study was to report the treatment of 2 patients with CHA due to WI and the measurement of TNF-α levels before and after treatment.
The diagnoses and individualized treatment of many pathologic conditions have benefited from attempts to correlate meaningfully a patient's symptoms with any significant quantitative changes in the endogenously produced specific compounds in the blood. According to the diagnostic criteria described by Sjaastad et al, [16, 17] the 2 patients presented matched criteria specific for CHA.
These 2 cases describe the first attempts at obtaining evidence for the significantly high levels of the cytokine TNF-α produced in the blood after “WI-induced CHA” that apparently benefited from MT and was associated with significantly reduced TNF-α levels. In many bone-related pathologic conditions, proinflammatory substances secreted by the nucleus pulposus are thought to be involved. When preexisting or concomitant mechanical injury to a nerve root occurs, for example, due to the compressive force exerted by a damaged protruding cervical intervertebral disk, the proinflammatory substances can cause nerve root pain. Plasma levels of cytokines are increased after nerve compression. Endoneural injections of TNF-α cause thermal hyperalgesia and mechanical allodynia, edema of the nerve root, damage to Schwann cells, and activation of macrophages. Endogenous TNF-α causes pain-related behavior in models of nerve dysfunction.  Consistent with this, selective inhibition of TNF-α prevents nucleus pulposus-induced thrombus formation, intraneural edema, and reduction of nerve conduction velocity. These findings have implications for future pharmacologic treatment strategies of sciatica. 
The therapeutic goal of the MT was to relieve the hypomobility. Nerve blocks can be used diagnostically and therapeutically. Diagnostically, the nerves in question are the greater and lesser occipital nerve as well as the spinal root.  Because of constant mechanical irritation, increased TNF-α levels may originate from any of the injured or dysfunctional anatomical sites in CHA cases.  Although the general consensus is that CHA arises from the upper cervical segments, recent studies suggest that disk herniation at the lower cervical segments may also be involved in the etiology of CHA. The affected cervical levels, singled out by magnetic resonance, anesthetic blockades and x-ray examinations indicate mainly the C4-5, C5-6, and C6-7 levels. [22, 23] In this context, it should also be noted that perhaps the raised level of TNF-α was only one of several contributing etiologic factors.
Martelletti  found that the cytokine pattern of CHA patients is similar to cluster headache, “biased toward an inflammatory status.” He investigated the levels of interleukin-1 β and TNF-α in the sera of CHA patients and found them higher compared to patients with migraine without aura and healthy subjects. Increase of cerebrospinal fluid TNF-α levels was also found in almost all the new patients with daily persistent headache and suggested a role for TNF-α in the pathogenesis of this condition.  It seems that TNF-α, interleukin-1 β, and interleukin-10 may be involved in the pathogenesis of migraine attacks. 
To the best of our knowledge, the cases presented here are the first WI patients with CHA in whom very high TNF-α levels (63 and 72 pg/mL compared to the reference range of 0-5 pg/mL) were demonstrated. These findings acquire greater significance, especially, because other rheumatologic or immunologic conditions associated with increased TNF-α levels were excluded. After successful restoration of the spinal segmental function by MT, which resulted in the reduction of the pathologic mechanic irritation and intervertebral compressive force respectively, TNF-α levels were substantially reduced (28 and 35 pg/mL). These values are still much higher than the reference range (0-5 pg/mL) suggesting that some structural damage may well have remained, albeit without significant clinical symptoms. This finding is compatible with the concept that endogenously released molecules and factors that are pain producing and TNF-α additionally may have a special role in the pathophysiology of CHA.
In retrospect, one may wonder why diskectomy was recommended in these patients with headache, if, as it turned out, it was manageable using MT. Is it fortuitous that an unnecessary procedure was avoided simply by that the patients refused it. We rather believe that neurosurgery could be postponed in our cases because neurologic symptoms and instability had not yet became apparent and MT was worth trying before opting for diskectomy.
At the Manual Medicine Outpatient Clinic of the National Institute for Rheumatology and Physiotherapy (Budapest, Hungary), many headache patients are treated, and their number is increasing. The diagnosis and treatment involve MRI, x-rays, followed by physical therapy, injections, and occasionally surgery that is rather stressful and does not always resolve the problems.
We believe that this preliminary finding may be useful for some clinicians while we hope to collect further data on such patients.
The data on the cases of WI-triggered CHA here presented are consistent with the experimental studies concluding that biomaterials, particularly TNF-α, may play a role in the pathophysiology of spinal disk symptoms and some categories of headaches. What this preliminary note adds are the following:
(i) increases awareness about the raised levels of TNF-α,
(ii) suggests the need to discover any meaningful correlation between any raised TNF-α level and the clinical state, and
(iii) suggests the need to develop a possible monitoring blood test for levels of TNF-α that might then require appropriate attention.
More important, the recommended diskectomy might well be postponed and eventually perhaps avoided by the administration of MT. These are the first WI patients with CHA in whom very high TNF-α levels (63 and 72 pg/mL) were substantially reduced (28 and 35 pg/mL) after successful restoration of the spinal segmental function and reduction of the intervertebral compressive force by MT. Further data are, however, required to confirm the pathophysiologic role of TNF-α in CHA.
Manipulative therapy resolved clinical symptoms and reduced
TNF-α levels in these 2 patients.
Endogenously released biomaterials that may play a role in
the pathophysiology of spinal disk symptoms and in some
categories of headaches.
Diskectomy in similar cases might possibly be postponed or
eventually avoided through a trial of manipulative therapy.