• 
• 

Three Part Question

In [an adult patient presenting with a cluster headache] does [intranasal lidocaine] improve [pain relief]?

Clinical Scenario

A 37 year-old man, who is known to suffer from cluster headaches, presents to the Emergency Department with a severe unilateral headache associated with lacrimation, rhinorrhoea and restlessness. He has already taken his own triptan and has been put on oxygen therapy on arrival. You remember being told that intranasal lidocaine can help in cluster headaches and you wonder what the evidence is for this therapy.

Search Strategy

Medline using NHS Evidence 1950–14 June 2013

Embase using NHS Evidence 1980–14 June 2013

(((exp CLUSTER HEADACHE/) OR ((cluster AND headache).ti,ab)) AND ((exp LIDOCAINE/) OR (lidocaine.ti,ab) OR (lignocaine.ti,ab))) [Limit to: (Languages English)]—Medline 40 records, Embase 241

The Cochrane Library : Issue 6 of 12, June 2013

MeSH descriptor: [Cluster Headache] explode all trees AND MeSH descriptor: [Lidocaine] explode all trees 0 unique results

Search Outcome

Four papers were found to be relevant

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Kittrelle et al 1985 U.S.A	Five patients, four male and one female, aged between 24 -70 years, experiencing frequent cluster headaches. Headaches were induced using nitroglycerin. Patients then received 1ml of 4% lidocaine into the nostril ipsilateral to the pain.	Case series	Pain relief at 3 minutes as a percentage	Four patients experienced a 75% or greater reduction in intensity of their headache within 3 minutes.	Small study of only five patients. No control arm (placebo or non-placebo). Uses an experimental model for cluster headache (not a clinical trial). Does not use other established treatments for cluster headache (sumatriptan and oxygen).
			Resolution of autonomic symptoms	Autonomic symptoms were reported to have improved in these patients but incomplete details were provided.
Hardebo et al 1987 Sweden	22 male and two female patients, aged between 24 – 68 years, experiencing frequent cluster headaches. Five patients received cocaine (0.3 ml 10% cocaine), 12 patients received lidocaine (0.5-0.8ml 4% lidocaine), seven patients received both treatments. 11 patients received xylometazoline hydrochloride (0.4-0.8ml of 0.1% solution) either alone or with the anaesthetics. All medication was self- administered using a nasal dropper.	Non-randomised clinical trial.	Pain reduction as a percentage	In the group treated with cocaine pain was reduced by 50% or more in 3/5 patients. In the lidocaine group pain was reduced by 50% or more in 4/12 patients. In the group combining lidocaine and cocaine pain was reduced by 50% or more in 3/7 patients In the patients that used xylometazoline no pain reduction was experienced by any patient.	Treatments not randomised. Small study without power calculation. No control group and no placebo. Not blinded. No statistical analysis attempted. Incomplete data provided. Outcomes unclear (how pain was scored and at what time). Commonly used treatments (oxygen and sumatriptin) not used.
			Reduction of autonomic symptoms	Autonomic features were reduced in 4/5 patients when cocaine was used and in 6/12 patients using lidocaine. In the group combining lidocaine and cocaine autonomic symptoms were reduced in 5/7 patients. In the patients that used xylometazoline no reduction was seen.
Robbins 1995 U.S.A	32 male patients with episodic cluster headache, aged between 23-63 yrs. Patients were instructed to use four sprays of 4% lidocaine in the ipsilateral nostril for two consecutive cluster headaches. Patients were permitted to use usual abortive agents such as oxygen after ten minutes.	Case series	Pain relief expressed as percentage	46% of patients experienced no relief, 27% reported 20-40 percent relief, 27% reported 40-60 percent relief.	Route of application different to that used in other trials this may change efficacy No control arm. Small sample.
Costa et al 2000 Italy	Nine male patients, aged 31-56 years, experiencing frequent cluster headaches had headaches induced using nitroglycerin. Once a headache attack became established patients were treated with cotton swab previously immersed in a 10% solution of cocaine hydrochloride (1 ml, mean amount of application 40-50 mg), or 10% lidocaine (1 ml), or saline. Swabs were introduced under anterior rhinoscopy into the nostrils, placed in the region corresponding to the sphenopalatine fossa of both sides, and left there for at least 5 min.	Crossover trial	Pain score on a visuo-analogic scale ranging from 0 to 10.	Pain intensity decreased to 3.5 (from 5) for cocaine and 4 (from 5) for lidocaine after 5 min. In the case of saline, pain intensity further increased (from 5 to 7.1) after intranasal application (P < 0.001 vs. both drugs). Complete cessation of pain occurred after 31.3 (+/-13.1) min for cocaine, 37.0 (+/-7.8 min) for lidocaine, and 59.3 (+/-12.3 min) for saline (P <0.01 saline vs. both drugs). There was no significant difference between cocaine and lidocaine at all times, although a trend towards a better effect of cocaine was observed after 5 min (P =0.07 )	Small study of nine patients. No power calculation. No description of how blinding or randomisation were achieved. Uses an experimental model for cluster headache (not clinical trial. The method of giving medication (via anterior rhinoscopy) may be difficult to replicate in clinical practice. Does not use other established treatments for cluster headache (sumatriptin and oxygen).

Comment(s)

In the first half of the 20th century topical cocaine was the treatment of choice for cluster headache, with the local anaesthetic lidocaine being subsequently introduced as a non-addictive alternative. It has been proposed that lidocaine may be effective in cluster headaches by blocking neural transmission at the sphenopalatine ganglion, activation of which has been implicated in the pathophysiology of attacks. All studies identified reported improved pain outcomes when intranasal lidocaine was used. However, the quality of studies was poor and this provides insufficient evidence to support its use in routine clinical practice. All had small sample sizes and did not use standard treatments for cluster headache (oxygen and triptans). Only one study (Costa et al) used randomisation, blinding and placebo. The method of administering lidocaine varied between trials and in the case of Costa et al this would not be easily reproducible in the ED. Two studies used an experimental model of inducing headaches with nitroglycerin that may respond differently to spontaneous cluster headache. The nature of cluster headache means that trials are problematical. Attacks are of short duration and variable in nature, making it difficult to interpret whether the disappearance or reduction of pain is as a result of treatment or due to the natural history of the condition. Also as some treatments (triptans and 100% oxygen) have a good evidence base it is questionably unethical to perform trials that do not use them on the control arm.

Clinical Bottom Line

While all trials showed an improvement in pain levels when intranasal lidocaine was used, the included studies were of poor quality and provide insufficient evidence to support the routine use of intranasal lidocaine in routine clinical practice

References

1. Kittrelle JP, Grouse DS, Seybold ME. Local Anesthetic Abortive Agents. Archives of Neurology 1985; 42: 496-497.
2. Hardebo JE, Elner A. Nerves and Vessels in the Pterygopalatine Fossa and Symptoms of Cluster Headache. Headache 1987; 27: 528-532
3. Robbins L, Intranasal Lidocaine for Cluster Headache. Headache 1995;35:83-84.
4. Costa A, Pucci E, Antonaci F, et al. The effect of intranasal cocaine and lidocaine on nitroglycerin- induced attacks in cluster headache. Cephalalgia 2000;20:85-91.