Читать книгу Management of Complications in Oral and Maxillofacial Surgery - Группа авторов - Страница 71

 Etiology: failure to recognize risk factors, individual patient anatomy, poor surgical technique

 Management: monitoring, serial neurosensory examination, microneurosurgery, medical management

The incidence of neurological complications as a result of third molar surgery ranges from 0.4% to 11% [1, 2, 5, 6, 27]. Injury to the IAN is associated with spontaneous recovery in 96% of cases, and spontaneous recovery of LN injury is approximately 87% [27]. Sensory deficits that last longer than one year are likely to be permanent, and spontaneous recovery of sensation should begin within the first eight weeks following surgery [27]. According to the AAOMS White Paper on third molars, the incidence of IAN injury one to seven days postoperatively is 1–5%, while persistent alteration in sensation after six months ranges from 0% to 0.9% [1]. LN injury one day after surgery was reported in 0.4–1.5% of patients, with persistent sensory alteration at six months in 0–0.5% of patients [1]. The use of lingual retraction increased the incidence of temporary paresthesia; however, the incidence of persistent neurosensory deficit remained the same. In a study by Tay et al., 192 IANs in 170 patients were exposed during third molar surgery. Twenty percent reported paresthesia at one week follow‐up, and 6% had persistent paresthesia at one year [28, 29].

An increased risk of a persistent neurosensory deficit following an IAN injury is associated with increased age, female gender, complete bony impaction, horizontal tooth angulation, sectioning of the tooth multiple times, bone removal, surgeon experience, and duration of surgery [27]. Additionally, Rood et al. has described seven radiographic predictors of potential nerve injury [30, 31]. The most significant have been found to include diversion of the IAC, darkening of the roots, and interruption of the white line of the IAC. One in three patients with IAC diversion, and one in four patients with darkening of the root, or interruption of the white line of the IAC, exhibited impairment of sensation. These signs are highly sensitive, but not highly specific, for risk of nerve injury, and the absence of all radiographic risk factor signs has a strong negative predictive value [30]. Therefore, patients without any significant indicators of nerve injury are unlikely to have injury, patients with an injury are likely to have at least one of the predictors, and patients without injury commonly have predictors of injury radiographically. Other reported radiographic indicators such as deflection of the roots, narrowing of the roots, dark bifid root apices, and narrowing of the IAC were statistically unrelated to nerve injury [30].

LN injury is associated with increased age, female gender, distoangular tooth inclination, lingual tooth orientation, perforation of the lingual cortex, and presurgical pericoronitis (this tends to “scar” the LN closer to the surface mucosa increasing vulnerability to injury) [32, 33]. Often, flap reflection, tooth sectioning with extension through the tooth into the lingual plate, or lingual plate fracture may be causal factors in the injury [32, 33]. Due to the nerve's variable position, care must be taken when incisions are made and flaps reflected. Miloro et al. reported 10% of LNs positioned superior to the lingual crest, and 25% in direct contact with the bone of the lingual plate [28, 29, 34]. The mean vertical distance of the LN from the lingual crest is 2.75 mm, and the mean horizontal distance of the LN from the lingual plate is 2.53 mm [29]. Perforation of, or a defect in, the lingual plate may represent normal patient anatomy (50% of preoperative CBCTs may have an anatomic lingual plate defect [personal communication, Roger A. Meyer, Shahrokh C. Bagheri]), or may represent iatrogenic injury from a rotary instrument.

Injury to the LN or IAN due to local anesthetic injection occurs in approximately 1 in 785 000 cases, with 79% affecting the LN and 21% the IAN. The highest incidence is associated with 4% prilocaine (Citanest) or 4% articaine (Septocaine) solutions. The majority of cases, 85%, resolve within eight weeks and of the remaining 15%, one‐third will eventually resolve [31]. Unfortunately, patients with persistent paresthesia are not candidates for microneurosurgical repair since access in the pterygomandibular space is limited, and also, the nerve may not show a readily identifiable area of injury or neuroma formation to resect and repair due to the nature of the injury, which may be mechanical (from the needle itself or a barb on the needle tip) or chemical (from a concentration effect of the local anesthetic solution).

All patients who report paresthesia should be followed closely for resolution and appropriate objective neurosensory testing performed. The clinical neurosensory test should be performed to determine the degree of impairment, and to assess whether microneurosurgical intervention is indicated. There are three levels of testing according to Zuniga et al. [35]. Mechanoreceptive testing begins with level A testing. It is comprised of brush stroke directional discrimination and two‐point discrimination. It is important to test both normal (to establish a baseline) and abnormal areas, map out the area of impaired sensation by marking directly on the patient's skin, and consider photographs of the markings for future reference and comparison. Two‐point discrimination can be tested using a Boley gauge or the two non‐cotton ends of a cotton tip applicator. Testing should proceed in 2 mm increments of widening between the points until the patient can no longer discern two separate points. Normally, the IAN has a discriminatory threshold of 3–5 mm and the LN of approximately 2–4 mm [27]. Level B testing involves contact detection using von Frey hairs or Semmes–Weinstein monofilaments that deflect with a certain amount of pressure application, and normative values exist for these as well. Finally, level C testing of nociception includes pain (pinprick) and thermal discrimination, and typically these are recorded as all‐or‐none responses. Taste is not usually tested in LN injuries. The indications for microneurosurgical repair include the following: complete anesthesia beyond one to two months, profound nonresolving hypoesthesia (below a level of functional sensory recovery) after three months, early dysesthesia (may indicate neuroma formation), and a clinically observed Sunderland V nerve transection at the time of surgery [27, 36]. Referral to a surgeon proficient in microneurosurgery should be made if any of the above criteria are met, or if the surgeon is unfamiliar with nerve testing and possible treatment protocols [27]. In general, dysesthesia is managed with pharmacological therapy (e.g., gabapentin, pregabalin, amitriptyline), while hypoesthesia/anesthesia is managed with microneurosurgery. The details of microneurosurgical repair are beyond the scope of this chapter (Algorithm 2.5).