Cerebellar syndrome after occipital nerve block: A case report

Abstract

Background

Occipital nerve blocks are commonly used in the treatment of different types of refractory headaches. The procedure is considered safe, and serious complications have rarely been described.

Case presentation

We report a serious complication of occipital nerve blockade secondary to the penetration of local anesthetic and non-steroidal anti-inflammatory drugs into the posterior fossa in a patient affected by type I Arnold Chiari malformation.

Conclusions

This case reminds that a proper injection technique is mandatory to avoid potentially severe complications when performing occipital nerve blocks.

Background

Occipital nerve blocks (ONB) are used in the treatment of migraines, cervicogenic headache, cluster headache, and post-traumatic headaches. ONB are considered safe and are rarely associated with complications (1). However, possible complications include mild local adverse reactions, such as cutaneous hematomas, local hair loss, myonecrosis, anaphylaxis, and intra-arterial injections leading to local anesthetic toxicity. Serious adverse events have been reported with subarachnoid injections that culminate in brain stem toxicity, especially in cases of previous posterior fossa surgeries (2,3).

Case presentation

An 18-year-old woman with medically-refractory headache of suspected cervicogenic origin was considered for ONB with non-steroidal anti-inflammatory drugs (NSAID) and local anesthetic by her general practitioner (GP). She was also known to have type I Arnold Chiari Malformation (CM-I).

In the GP office, a combination of metamizole (500 mg/ml), piroxicam (20 mg/3 ml), and xylocaine (30 mg/3 ml) was injected in the occipital region, with the patient in a sitting position and the neck in antero-right flexion. Immediately after the injection, the patient presented with nausea, vomiting, vertigo, and loss of muscle tone. She was subsequently admitted to our emergency department where a neurological examination on admission showed the inability to maintain sitting or standing position, hypoesthesia in the Ramsay-Hunt area, severe dysphagia, left hypoacusis, and left-sided proportional strength deficit without pyramidal signs. The patient also complained of nausea and vomiting.

Cerebral CT findings confirmed a CM-I and showed a small linear hyper-density of the right cerebellar amygdala and the presence of air micro-bubbles in the right cerebellar hemisphere (Figure 1 (a) and (b)). Subsequent brain MRI confirmed a linear lesion in the right cerebellar amygdala (Figure 1 (b) and (c)). Neck X-ray and cervical CT results were unremarkable.

Figure 1.

Neuroimaging: CT ((a) and (b)) and MRI ((c) and (d)) findings.

The clinical picture spontaneously and gradually improved to a complete recovery in 15 days and the patient returned home with no deficits or limitations.

An informed consent for the publication of this case report was provided by the patient.

Discussion

We report the case of a serious complication of ONB, which to our knowledge was previously undescribed. The injection of local anesthetic and NSAID, a drug combination not recommended by current guidelines (4), penetrated into the cranio-spinal cavity leading to a cerebellar lesion. Our case should warn nonspecialized doctors to perform ONB with extreme caution and highlights that a patient’s incorrect neck position might increase the risk of needle penetration into the cranial cavity. According to current practice guidelines, ONB can be performed using a landmark-based technique with the patient in a sitting or prone position with the head slightly flexed forward (4,5). The nerve can be usually found by palpation of the occipital artery and is typically located just medial to the artery at the lateral third of an ideal line going from the occipital protuberance to the mastoid process. In this case described, an erroneous injection technique was certainly applied, with an abnormal/excessive antero-lateral neck flexion position and a site of injection lower to the ideal mentioned line.

Moreover, treatment of this kind outside the hospital setting, where immediate life support and intensive care treatments are unavailable, can seriously harm the patients in cases of complication. In our case, the procedure was performed in a private medical practice setting; hence, the outcome of the complication might have been catastrophic in the case of wider brainstem toxicity.

The presence of a CM-I malformation might be seen as an incidental anatomical predisposing factor: the intracranial diffusion of lidocaine and NSAID caused a toxic-ischemic damage directly in the cerebellar tonsils, due to their pathological downward position in relation to the foramen magnum. The poor technique is witnessed by neuroimaging findings, showing the signs of penetration of the needle directly into the posterior fossa towards the cerebellar tonsils (see Figure 1). Although in CM-I patients, disorganized cranio-cervical ligaments (6) with poorly arranged collagen bands and interspersed adipose tissue have been described (7), we do not retain the alteration of these ligaments directly related to the complication observed in our patient. Direct intravascular injection of the local anesthetic and NSAID may have also contributed to the clinical features.

Skull base abnormalities have been associated with complications of ONB (2,3) and are therefore considered a contra-indication for performing ONB. In the cases described, even without a traumatic penetration into the cranial cavity, the simple injection made in proximity of the craniotomy/skull defect was indeed associated with the local anesthetic penetration leading to brainstem toxicity.

It should also be noted that CM-I, although asymptomatic, might sometimes present with head and neck pain, along with an occipital headache felt at the base of the skull. This can typically be triggered by coughing, sneezing, or other Valsalva maneuvers. Therefore, before treating a cervicogenic headache with ONB, CM-I should be cautiously considered because of possible clinical overlap. In the case of our patient, a post-hoc medical history did not reveal all the ICHS-3 criteria (8) for a headache caused by CM-I (lack of brainstem, cerebellar or lower cranial nerves-related symptoms, lack of temporal relation to the CM-I, variable duration with attacks lasting up to 24 hours).

Our case highlights that poor ONB technique might be dangerous and harmful for patients and should serve as a warning for non-specialized doctors. In this respect, performing ONB under ultrasound guidance (9), which allows precise and safe location of the anatomical sites of nerve blockade, might reduce the risk of side effects. Our case also suggests the importance of considering the presence of CM-I before any ONB procedure is performed, given the possible clinical overlap with other forms of headache that can be treated with ONB. As recently advocated, the skull integrity should be carefully assessed before any ONB procedure (10) and previous posterior fossa craniotomies should be regarded as an absolute contraindication for intervention.

Clinical implications

Occipital nerve block should be performed with extreme caution by non-specialized doctors.

Occipital nerve block should be performed according to current practice guidelines using a landmark-based technique, injecting at the lateral third of an ideal line going from the occipital protuberance to the mastoid process. Ultrasound guidance might be considered to minimize side effects in selected cases.

Previous posterior fossa surgeries should be considered as an absolute contraindication to perform occipital nerve blocks.

Arnold Chiari malformation should be considered in the differential diagnosis of cervicogenic headaches.