The Pain of Maxillofacial Osteonecrosis (NICO)
©The Maxillofacial Center for Diagnostics & Research
Other Links NICO Clinical Page NICO Follow-up NICO Home Page Home Page		This Topic Pain or No Pain Types of NICO Pain Pain Mechanisms in NICO Nerve Damage References Table
The "Rule of 80s" for NICO (click on figure above to see enlargement; use back arrow to return).

Before preceding with a discussion of painful maxillofacial ischemic osteonecrosis, often referred to as neuralgia-inducing cavitational osteonecrosis (NICO), two important features of osteonecrosis should be emphasized, namely, that the disease may or may not produce pain and the intensity of symptoms is not related to the amount of bone destroyed. Recent dental literature has seldom discussed asymptomatic IO of the jaws, but the older literature, including the classic oral pathology textbook by G.V. Black, contains many examples of painless intramedullary "dry rot" or cavitation, usually under terms such as "bone caries" and "chronic osteitis" to distinguish it from osteomyelitis. Avascular bone associated with residual or unhealed extraction sockets, with or without pain, was also reported long ago and has recently been cited as a "red flag" or warning sign for medullary ischemia severe enough to prevent proper healing after surgery.

Clinical and clinicopathologic research indicates that medullary ischemia and occult inflammatory disease may be quite common in persons with chronic idiopathic facial pain. Investigators have found localized "zones of tenderness" in the affected alveolar quadrant of the majority of patients, and similar zones have been demonstrated on ipsilateral maxillary alveoli in persons with migrainous headache. Localized areas of elevated mucosal temperature and submucosal lymphocytic infiltration have also been reported in the painful quadrants of such patients, presumably from neurogenic inflammation in the extraosseous receptor fields of intraosseous alveolar nerves sensitized by chronic exposure to diseased marrow. Scintigraphy scans traditionally used for IO, ⁹⁹technetium-MDP (⁹⁹Tc-MDP) and ⁹⁹Tc Single Proton Emission Computed Tomography (SPECT) scans, also show localized areas of increased radioisotope uptake or "hot spots" in the painful quadrants of the majority of patients with idiopathic facial pain.

Neuropathic pain abatement with local anesthesia (the selective anesthesia test) or cold laser therapy seems to further confirm a relationship between the symptoms and some alveolar process, whether it be nerve damage or marrow disease. The permanent or long-term pain abatement resulting from conservative surgical removal of diseased marrow, without further damaging visible alveolar nerves, seems to speak toward the marrow disease as etiologically significant. In this light, it is important to know that ischemically damaged or chronically inflamed marrow is found quite routinely in biopsy samples taken from bone demonstrating hypersensitized surface zones, ⁹⁹Tc hot spots, or osteoporotic radiographic abnormalities. Moreover, dramatic pain abatement in a large proportion of NICO patients with the use of anticoagulants (without surgery or antibiotic therapy) confirms an association between the symptoms and prothrombotic tendencies in those patients, just as it does in painful IO of the hips.

Several investigators, some with large numbers of cases, have had good success treating these patients with decortication and curettage of diseased marrow (Table 1), so much so that it is advisable to rule out NICO prior to the designation of a truly idiopathic neuropathy and referral to a neurologist.

Prior to bone marrow biopsy the average NICO patient has been in pain for 6 years (up to 32 years), usually diagnosed as atypical facial neuralgia/pain, but also diagnosed as trigeminal neuralgia, chronic sinusitis, phantom toothache/pain, and various headaches, including migraine headache. In the hip the disease typically "progresses relentlessly" without treatment until the femoral head collapses, usually after 3-5 years. Many cases, however, stabilize and never go on to complete collapse and a significant minority undergo self-healing without treatment. Stabilized lesions without progressively increasing pain occur in the jaws as well, but are at constant risk of inflammatory events or vasoconstrictor use exacerbating the compromised marrow flow and pushing the disease over a threshold into more intense pain. We are used to thinking of irradiated bone as susceptible to osteoradionecrosis for the rest of the patient’s life, it is time to use similar thinking for all forms of ischemically damaged jaws.

The Pain of NICO The pain of ischemic osteonecrosis and its lesser counterpart, bone marrow edema, is quite variable and, as previously mentioned, it bears little relationship to the amount of bone destruction. Usually the pain involves a relatively broad area, it seldom occurs as the pin-point site of exquisite tenderness and pain which is the hallmark of the acute dental or periapical abscess. Nor are there obvious soft tissue signs of inflammation, although a small number of patients will have mild facial or alveolar mucosal edema and/or redness. Involved bone is typically tender to palpation, although this often is quite mild, and the clinician may have difficulty achieving complete anesthesia with routine local and block anesthetic procedures.

Patients often have difficulty describing the pain and outlining its area of involvement, and the pain may alternate between several maxillofacial sites or be referred some distance from the affected bone, especially to the ipsilateral neck, oropharynx, pre-auricular and periorbital regions, and calvarium. Most often the pain is a deep ache or sharp bone pain and there are two broad patterns of onset: 1) a slow, insidious increase in pain over months or years, often with intermittent periods of pain-free existence; 2) an abrupt onset, often occurring within hours or days of local trauma, infection or a dental procedure using local anesthetics with vasoconstrictors. Pressure and deep burning sensations are described by many. It is not unusual for the pain to prevent sleep or to awaken the patient, and it is often remarkably resistant to routine analgesics, although antibiotics may provide temporary relief.

The pain often varies over time, with intermittent symptoms becoming constant, with mild pain becoming more intense or changing character (for example, a deep, ill-defined ache becomes sharp and lancinating), and with additional areas of pain developing in at least a third of affected individuals. Spontaneous remission of disease symptoms may occur after years of pain.

When a triggering event is not obvious, the clinician should remember that the disease is one with multi-hit etiologic factors, and that the triggering event may have taken place weeks or months before the onset of pain. Estrogen-induced osteonecrosis and corticosteroid-induced osteonecrosis are especially prone to a long-delayed onset of symptoms, typically 4-6 months after beginning the medication.

A formal classification system probably cannot be established for a disease with such variable and varying symptoms, or one with such frequent intermixing of different pains in a single patient, but the following general pain types are those seen most frequently, based on our experience with many hundreds of NICO patients and as summarized by the Second National NICO Workshop:

Type I: Deep bone ache (mimics atypical facial neuralgia/pain). The most common NICO presentation is a moderately localized or full-quadrant sensation of deep aching, perhaps interspersed with episodes of lancinating pain and perhaps associated with a sensation of intramedullary pressure or burning. This symptom often begins insidiously and intermittently, but over a period of 6-12 months it becomes more constant. Occasional patients will have unchanging intermittent pain for years. In dentate individuals the pain is often described initially as toothache pain and may be exacerbated by endodontic therapy but, conversely, may temporarily diminish for 2-6 weeks after such therapy. The ache is characteristically resistant to NSAIDs, hydrocodone bitartrate, oxycodone hydrochloride, and anticonvulsants.

Type II: Sharp, non-lancinating pain (mimics phantom toothache). Deep, sharply intense, often debilitating and moderately well localized pain of a portion of the alveolar bone, the mandibular ramus or the zygoma, usually with a tenderness of the overlying mucosa and often diminished or eliminated temporarily by antibiotic therapy. This sharp pain is resistant to NSAIDs and is often non-responsive to hydrocodone bitartrate or oxycodone hydrochloride, but for a time usually responds moderately well to anticonvulsants (myelin stabilizers), such as phenytoin sodium (Dilantin), carbamazepine (Tegretol) and gabapentin (Neurontin).

Type III: Sharp, lancinating pain (mimics trigeminal neuralgia). Very intense, intermittent, debilitating "lightning burst" of pain lancinating to the ipsilateral periorbital or preauricular regions, oropharynx, or neck. Attacks last for only a few seconds but a series of attacks usually occurs for a 10-25 minute period. This trigeminal neuralgia-like pain may or may not be associated with an intraoral or facial trigger point and initially responds well to anticonvulsants, such as phenytoin sodium (Dilantin), carbamazepine (Tegretol) and gabapentin (Neurontin). It is unresponsive to analgesics but may temporarily improve with antibiotics. This pain might best be described as trigeminal neuralgia of alveolar origin.

Type IV: Intramedullary hypertension & myelopyrosis. The patient complains of an uncomfortable pressure phenomenon within a broad area of alveolar bone. The pressure phenomenon may be real, as poor outflow from diseased marrow results in backup pressures and intramedullary hypertension as much as four times greater than normal.¹ Because this is a fluid-associated phenomenon, the patient often can relate symptom changes to increasing or decreasing barometric pressures. This symptom is often accompanied by a deep burning sensation, myelopyrosis, usually involving most or all of a quadrant of alveolar bone. Neither of these symptoms is responsive to antibiotics or analgesics, although the myelopyrosis may diminish with hyperbaric therapy. Both symptoms almost always diminish or disappear with local anesthetic use.

Type V: Headache. Ipsilateral and bilateral headache pain, presumably referred, of the frontal, temporal or occipital regions, including migrainous headache (usually without aura). The headache temporarily disappears or dramatically diminishes with a positive hyperesthesia/anesthesia test. Cervical and shoulder aching may occur along with the headache or at times when there is no headache. Alveolar bone may or may not be painful.   

There is much that is not yet understood about the pain of osteonecrosis, although it is clear that symptoms result largely from the fluid dynamics associated with reduced marrow outflow, elevated intramedullary pressures, ischemia and recurring microinfarction. The reduced outflow phenomenon is so unique that it has led in long bones to the development of a "stress test" whereby saline injected into the femoral head causes an immediate increase in pressure and pain. Conversely, more than half of all painful ischemic hips exhibit dramatic and rapid, often instant pain relief when decortication or core decompression allow pressure release, as do most NICO patients after alveolar decortication and curettage. The relief experienced with anticoagulants also substantiates an association with an unusual intramedullary fluid environment.

Although nerve damage is not required to explain the pain of osteonecrosis, nerves traversing ischemically damaged or inflamed marrow are bound to be affected by the toxins, immunoglobulins and inflammatory mediators enveloping them, and the vasa nervorum are influenced by the same ischemia, increased pressures and microinfarctions which affect the other tissues of the bone, perhaps engendering a necrotizing angiopathic neuropathy similar to that described by Dyck as a painful consequence of peripheral vascular disease in diabetics. When the maxillofacial region is involved, moreover, there are significant compounding factors related to the presence of major branches of the very complex trigeminal nerve, the nerve with the largest ganglion and the most extensive innervating zone of all peripheral nerves in humans. No other marrow spaces contains an equivalent nerve.

Presumably, this has profound influences on the nature and intensity of associated pain, as does the considerably heightened frequency of trauma and infection. Progressive sensitization and degeneration of alveolar nerves and specialized dental sensory presso- and mechanoreceptors, even deafferentation and brainstem neuroplastic alterations after tooth-pulp removal, can go far to explain why the regions innervated by these nerves account for the vast majority of neuralgias in humans. Conversely, while the marrow disease influences alveolar nerves, the nerves in turn can adversely influence the marrow disease through neurogenic inflammation. Prolonged and enhanced neuropeptide release, especially of substance P or calcitonin-gene-related peptide, neural control of vasoactivity, and the local mitogenic and trophic effects occurring at sensory nerve endings within injured and healing tissues, may evoke additional tissue injury instead of helping to maintain tissue integrity and strengthen the repair process.

Damaged axons frequently undergo wallerian degeneration and die back to their cell bodies, perhaps leaving a small, frayed proximal stump or a neurovascular bundle represented by a string of degenerative fibrosis with no residual neural tissue. However, non-wallerian myelin degeneration also occurs and, in fact, seems to represent the majority of damage found in residual nerves from NICO biopsy samples. The presence of normally myelinated, viable nerve fibers with areas of missing myelin may help to explain the unusually and varied symptoms in a fashion similar to the traumatic neuroma with its partially unmyelinated fibers.

Remyelinization of damaged nerves may also contribute to a neuropathy by producing internodes of varying lengths with subsequently altered conduction velocities and, of course, toxins from damaged marrow may be transported centrally as far as the nucleus caudalis, potentially causing degenerative changes at distant sites. The latter phenomenon may, in fact, explain the minor lymphocytic infiltration, fibrosis and myelin sheath abnormalities common to gasserian ganglia and associated sensory roots in postmortem examinations. Other ganglia do not show these signs of inflammatory damage.

Elevated levels of circulating anti-peripheral nerve myelin (anti-PNM) antibodies have been found in NICO patients, suggesting chronic exposure of the peripheral myelin to the immune system. The sera of healthy humans normally show none of these antibodies but some NICO patients have had levels as high as or higher than those found in the Guillain-Barré syndrome. A potential cause of nerve damage in cases of jawbone osteonecrosis is emerging from the work of Haley at the University of Kentucky. Using standard and well established neurotoxicity assays, his evaluation of a large number of NICO tissue samples has identified extreme neurotoxicity in almost every sample. It is not at this time known whether the toxins are being generated by microorganisms entrapped in osteonecrotic lesions, by the necrotic debris itself, or by an inflammatory response to the debris or microorganisms.

All of these factors are local alveolar phenomena but we must be ever mindful of the powerful influences of higher brain centers on the perception and modulation of local symptoms.  

Table 1: Listing of all published follow-up studies of NICO patients. 
References can be found in NICO Follow-up Page.

    * overlapping patient pools, i.e. some patients probably reported in multiple papers.