Abstract
Headache and depression were studied in patients who had undergone operation for acoustic neuroma. A questionnaire with headache and Beck Depression Inventory scale were sent to 228 patients, of whom 192 (84%) responded. Preoperative headache was reported by 61 (32%) of the respondents (47 migraine and nine tension-type headache) and 122 (64%) respondents had postoperative headache (15 new migraine and four new tension-type headache). The new postoperative headache was chronic (≥3 months) in 86% and continued at the time of the survey in 55% and presented typically as severe short-lasting attacks provoked by physical stress, bending or coughing. Non-steroidal anti-inflammatory drugs were effective in most cases. Depression (usually mild) occurred in 24% of the respondents, being significantly more common in prolonged postoperative headache patients. The operation doubled the prevalence of headache (from 32% to 64%). Headache after acoustic neuroma operation appears to be a specific subgroup of postcraniotomy headache.
Introduction
Headache is a common aftermath of acoustic neuroma surgery. The prevalence of headache is reported to be from 23% to 34% at 3 months (1, 2) and from 16% to 29.5% at 1 year after surgery (1, 3, 4). Postoperative headache is a major problem and refractory to medication in one-tenth of patients (3–5). The aetiology and pathophysiology of this condition are unknown. Dural adhesions to nuchal muscles (6, 7), meningeal irritation due to bone dust after drilling (8, 9) and a combination of muscle tension and vascular and neuropathic mechanisms (10) have been suggested as explanations for pain. The prevalence of headache before the operation has been reported in some studies: 9% (11), one-fifth (2), one-third (12), 42% (10) and 49% (4). In one study preoperative headache ceased in 27% of patients, but was still a risk factor of postoperative headache (4).
The aims of the current study were to: (i) assess the prevalence and prognosis of headaches after acoustic neuroma surgery; (ii) classify pre- and postoperative headaches using the latest version of The International Classification of Headache Disorders (ICHD II) (13); (iii) identify the clinical characteristics of new-onset chronic headache after acoustic neuroma surgery; (iv) assess medical treatment of new-onset chronic headache after acoustic neuroma surgery; (v) assess the prevalence of depression after acoustic neuroma surgery; and (vi) identify possible risk factors of new-onset chronic headache after acoustic neuroma surgery in patients operated on at the Department of Neurosurgery, Helsinki University Hospital.
Patients and methods
All charts from 241 patients who had undergone operation for acoustic neuroma at the Helsinki University Hospital between January 1995 and September 2002 were reviewed. Six patients had died and seven were lost to follow-up. A diagnostic questionnaire on preoperative and postoperative headaches, pain and visual analogue scale (VAS) and other postoperative symptoms and the Beck Depression Inventory (BDI) I scale (14) were sent to all traceable patients (n = 228), of whom 192 (84%) responded. The severity of pain was assessed using the VAS (15). The characteristics of possible preoperative and postoperative headaches were assessed separately. The ease of differentiating between the two headaches types was also queried. Detailed analysis of the headaches included the location, quality, intensity and duration of pain, concomitant symptoms, and exacerbating and alleviating factors. Enquiries also covered medication for headache and its efficacy, and the duration of postoperative headache. The headache diagnoses were based on the latest version of the ICHD (13).
To control for confusing factors in the analysis of postoperative headache, a group of patients was formed with postoperative headache who met the following criteria: (i) no preoperative headache or easy differentiation between pre- and postoperative headaches; (ii) onset of headache within 1 week after the operation (which is an ICHD II criterion for postcraniotomy headache); (iii) duration of headache of at least 3 months (an ICHD II criterion for chronic headache); and (iv) no new postoperative migraine or tension-type headache (TTH). As craniotomy had been performed for acoustic neuroma surgery, the ICHD II criterion that craniotomy be performed for a reason other than head trauma was also met. This group was called the new postoperative headache group. Additionally, a subgroup of patients was formed from the new postoperative headache group who met the additional criteria: (v) intensity of pain severe, i.e. at least 60/100 using VAS after the first postoperative year, and (vi) headache was provoked or clearly exacerbated by physical stress. This group was called the new severe postoperative headache group.
The difference between the groups was tested using the non-parametric Mann–Whitney U-test. Binary logistic regression analysis was used to determine the probability of having postoperative headache.
All statistical analyses were performed using SPSS 12.0.1 (SPSS Inc., Chicago, IL, USA). The study protocol was approved by the local ethics committee, and informed consent was obtained from all respondents.
Results
Ninety men (47%) and 102 women (53%), aged 23–79 years (mean 54 years), returned the questionnaires. The postoperative follow-up time was 1–9 years (mean 5 years). The surgical approach had been retrosigmoid in 179 (93%) patients, subtemporal in 12 (6%) and translabyrinthine in one patient. The largest diameter of the tumour varied from 3 to 55 mm (mean 20 mm, median 17 mm). Fifteen tumours were intracanalicular. The new postoperative headache group comprised 85 patients (Fig. 1). The new severe postoperative headache group consisted of 44 patients (25 male, 19 female), aged 31–76 years (mean 53 years), with a postoperative follow-up of 1–8 years (median 4 years). None of them was reoperated. Of these 44 patients, 11 had preoperative migraine (six with aura), but none had preoperative TTH.
Patient delineation.
Preoperative headache
The prevalence of pre- and postoperative headaches among the responders is shown in Fig. 1. Sixty-one (32%) patients reported preoperative headache. Of these, 47 (77%) met the International Headache Society (IHS) criteria for migraine [migraine without aura (MoA) 26 and migraine with aura (MA) 21] and nine (15%) for TTH. Five patients fulfilled the IHS criteria for chronic migraine. Seven patients had episodic (ETTH) and two had chronic tension-type headache (CTTH). One patient suffered from both migraine and TTH, and one patient had trigeminal neuralgia. Four patients had unclassified preoperative headache.
Twenty-three patients (of whom 87% were migraineurs) reported complete cessation of their preoperative headache after the operation, whereas 15 patients (of whom 73% were migraineurs) reported essential exacerbation of their preoperative headache after the operation.
Postoperative headache
Postoperative headache was reported by 122 (64%) patients. Of these, 39 had both pre- and postoperative headache, 89 had postoperative headache only. The new postoperative headache met the IHS diagnostic criteria for TTH in four patients, MoA in 12 patients and MA in three patients. No autonomic features accompanied the headache, and no patient with postoperative headache met the IHS diagnostic criteria for trigeminal autonomic cephalgias. The patients did not have temporomandibular disorders.
Twelve patients with pre- and postoperative headache were unable to differentiate between the two, whereas 27 patients found the differentiation easy. Of these 27 patients, 21 (77%) had preoperative migraine. In the 110 patients who had only postoperative headache or who could easily identify new postoperative headache from the previous one, the duration of headache was at least 3 months (i.e. met the ICHD II criterion for chronic headache) in 95 patients (86%) and at least 1 year in 82 patients (75%). The postoperative headache was ongoing during the survey in 61 patients (55%).
Clinical picture of new postoperative headache
Table 1 presents the intensity, temporal presentation, localization and provoking or exacerbating factors of headache in the new postoperative headache group and in the new severe postoperative headache group. Pain presented as attacks in 76% of patients in the new postoperative headache group and in 68% of patients in the new severe postoperative headache group. The duration of the attacks varied between 10 min and 24 h (median 1.5 h, mean 2.5 h) in the new postoperative headache group and between 10 min and 7 h (median 1.5 h, mean 2.0) in the new severe postoperative headache group. In patients in the new postoperative headache group whose pain presented as attacks and who had not had previous migraine, pain was accompanied by photo- and phonophobia in three, nausea in six and both in seven. In patients in the new postoperative headache group who had had previous migraine and whose new headache presented as attacks, pain was accompanied by photo- and phonophobia in two and by nausea in three. During the first postoperative year, the number of headache days/month varied between 1 and 30 (median 28, mean 20) in the new postoperative headache group, and between 1 and 30 (median 30, mean 24) in the new severe postoperative headache group. Gradual alleviation of headache after the first postoperative year was reported by 22 patients (26%) in the new postoperative headache group. Headache still continued at the time of this survey in 52 patients (61%) in the new postoperative headache group. The current frequency of attacks was 1–16/week (median 2) and the current duration of attacks was 10 min to 2 h.
Characteristics of postoperative headache in the new postoperative headache group (NPHG) and new severe postoperative headache group (NSPHG)
Of the patients whose headache was still present during the survey (52 patients in NPHG and 30 patients in NSPHG).
Due to the definitions of the NSPHG.
Medication response
Simple analgesics [non-steroidal anti-inflammatory drugs (NSAIDs), coxibs or paracetamol] provided significant pain relief in 67 patients (79%) in the new postoperative headache group, including 30 patients (35%) whose pain ceased completely with simple analgesics. Of these 52 patients, whose headache continued at the time of the survey, 48 (85%) had medication for their headache: 24 patients had medication on demand and 24 had regular medication (of whom four had additional medication on demand). NSAIDs or coxibs were used by 15 patients, paracetamol by eight and tramadol by one patient as on-demand medication. Ten patients had regular medication as monotherapy and 14 patients as combination medication. NSAIDs or coxibs were used on a daily basis by 13 patients, paracetamol by 11, antiepileptic medication by five (four of whom were on gabapentin and one on carbamazepine), tricyclic antidepressants by eight and muscle relaxants by four patients.
Patients' data were evaluated using the ICHD II criteria of medication overuse headache (MOH). Eleven patients fulfilled some of the criteria of MOH such as headache frequency (i.e. >15 days/month), use of simple analgesics ≥15 days/month and the quality of headache. However, 10 patients reported essential pain relief (measured with VAS) with simple analgesics, and in one patient, despite worsening the headache during analgesic use, the detoxification severely worsened the headache. These patients are therefore not typical representatives of MOH.
Effect of tumour size and location
The largest diameter of the tumour in the patients in the new postoperative headache group varied from 4 to 29 mm (mean 13 mm, median 11 mm, intracanalicular tumours 4). Tumour size was significantly smaller than in the group without chronic postoperative headache (mean 23 mm, median 22 mm). Multivariate analysis tested possible predictors: age, gender, existence of preoperative headache and tumour size. The analysis indicated that female gender (P = 0.04), lack of previous headache (P = 0.005) and small tumour size (P < 0.0001) predicted chronic postoperative headache. All patients in the new postoperative headache group had been operated on with the retrosigmoid approach.
Mood changes
Of the respondents, 145 (76%) had a Beck score <10 (absence of depression), whereas 37 (19%) had mild depression, six (3%) had moderate depression and four (2%) had severe depression. On the new postoperative headache group, 17 (33%) of 52 patients with continuing headache had depression (mild in 12, moderate in four and severe in one), whereas three (9%) of 33 patients without continuing headache had depression (mild in two and severe in one) (P = 0.015). Of those 10 patients who had moderate or severe depression, five had continuing new postoperative headache, four had no headache and one had had postoperative headache which had ceased by the survey. Of the 37 patients with mild depression, 12 (32%) had continuing new postoperative headache. Hence, in 50% of patients with moderate or severe depression and in 32% of patients with mild depression, headache may contribute to the presence of depression.
Discussion
The ICHD II defines chronic postcraniotomy headache as headache, which (i) is maximal in the area of craniotomy, (ii) follows craniotomy performed for a reason other than head trauma, (iii) develops within 7 days after craniotomy, and (iv) persists for at least 3 months after craniotomy (13). The references concerning chronic postcraniotomy headache in the latest IHS classification concern almost exclusively headache following acoustic neuroma surgery. The few exceptions are the paper on headache after retromastoid craniectomy for microvascular decompression with headache prevalence 29% at 1 month and 17% at 6 months (16), and the paper on headache after temporal lobectomy for intractable epilepsy with headache prevalence 12% at 1 year (17). Neither of those papers described the headache in detail.
The literature on the clinical characteristics of headache after acoustic neuroma surgery is scant. Pedrosa et al. have reported that headache after acoustic neuroma surgery is most often described as tension type, with episodic acute exacerbations mimicking migraine (11). In Vijyan's study also, the most common headache type for acoustic neuroma was of tension-type, the second most common being shooting pain near the surgical site or referred to the frontal area, and the third being throbbing pain (10). In one study the location of the headache was predominantly in the neck in 53%, on the top of the head in 34%, frontal in 5% and all over the head in 8% (1), whereas in another study the headache was occipital in 79%, frontal in 10% and located all over the head in 11% (2). In our study the location of chronic headache was frontal in 13%, occipital in 25% and both occipital and frontal in 61%. The headache was unilateral in 60% and bilateral in 40%. The maximal site of postcraniotomy headache after acoustic neuroma operation seems to extend beyond the surgical site. In Vijyan's study most of the patients (90%) described the headache as pressing (10). Physical stress (63%), emotional stress (46%) and bending down (29%) were the exacerbating factors of headache in one study (4), whereas head movements (83%), fatigue (58%) and exercise (40%) exacerbated the headache in another study (2). A recent postal survey of the members of the Acoustic Neuroma Association of the United States has reported that chronic headache after acoustic neuroma surgery occurs more than once daily in 46%, lasts 1–4 h in 43%, is of moderate intensity in 63% and is alleviated by NSAIDs in 61% (12).
We suggest that headache after acoustic neuroma operation should be classified as a specific subgroup of postcraniotomy headache. Postcraniotomy headache subsides spontaneously in most cases, but chronic postoperative headache severely affects the everyday life of patients, and this study has therefore concentrated on this problem. This was why the new severe postoperative headache group, representing the most severely affected subgroup of our headache patients, was formed soley for the purposes of this study. In our patients, chronic headache after acoustic neuroma surgery typically presented as severe attacks provoked by physical stress, bending or coughing. Headache was either uni- or bilateral, and the maximum pain was not necessarily at the site of the operation, which is the only feature that conflicts with the IHS criteria of postcraniotomy headache.
Weiss et al. (18) have reported post-traumatic migraine in 35 patients without previous headache, and the activation of the common headache pathway has been suggested to explain this (19). Similarly, craniotomy, especially when the retrosigmoid approach is used (5, 8), can activate the headache pathway and initiate headache with some migrainous features. However, in only 15 of our patients did the new postoperative headache meet the IHS criteria of migraine headache. In patients whose headache after acoustic neuroma surgery presented as attacks, the attacks were shorter than migraine headache and the accompanying photo- and phonophobia, and nausea was less common. To avoid confusion with migraine, only those patients with preoperative migraine who could easily differentiate between new headache and previous headaches were included in the new postoperative headache group, and patients with new postoperative migraine were excluded. In fact, patients with preoperative migraine had phono- and photophobia associated with their postoperative headache less often than those without previous migraine. The short duration of the attacks and their consistent abrupt presentation after certain provocations, such as lifting a heavy object or bending, were the clearest features in their differentiation from migraine. Interestingly, preoperative migraine ceased after the operation in 20 patients. The operation also exacerbated the migraine in 11 preoperative migraineurs and provoked new migraine in 15 patients. In most cases, postoperative headache was too intense to be TTH. Furthermore, TTH is not aggravated by physical activity.
The mechanisms of headache after acoustic neuroma surgery are not yet understood. The hypothesis of dural adhesions to nuchal muscles has led to a shift from craniectomy to craniotomy and to late cranioplasty in previously craniectomized patients. This, unfortunately, has not abolished the problem of chronic postcraniotomy headache (2, 8, 9). A small study with short follow-up has suggested that chronic pain after acoustic neuroma surgery is a mild incisional pain, although tenderness in the incision area (present in 46% of the patients) was unrelated to the presence or duration of postoperative pain (20). In a study of possible pathophysiological mechanisms of headache after acoustic neuroma surgery, abnormal activation of nuchal muscles or vestibular imbalance did not explain headache, whereas occipital neuralgia explained it in a few patients (21). There may be several mechanisms, with variable contributions from case to case. In patients with continuous headache, neuropathic mechanisms may contribute to pain, whereas in patients whose headache presents as attacks, the activation of the trigeminal pain systems may dominate. The diagnosis of neuropathy requires clinical examination and cutaneous sensory testing of the pain area (22). Unilateral headache, contralateral to the operation, was reported in the charts of eight patients, which argues against a cutaneous neuropathic mechanism as explanation of pain in these patients.
The dura of the posterior fossa has rich sensory innervation, which is thought capable of generating pain (23–25). A prominent feature of headache after acoustic neuroma surgery is its occurrence or worsening through normally innocuous activities producing an increase in intracranial pressure, such as bending or coughing. This suggests an exaggerated intracranial mechanosensitivity. Animal studies have shown sensitization of polymodal nociceptors of dura to mechanical stimuli by inflammation (26, 27). Sensitization may originate from perioperative manipulation of dura or prolonged chemical meningitis in response to bone dust (2, 9, 21). Schaller et al. have reported an association of laboratory-proven aseptic meningitis with chronic postoperative headache (2). In their patients with postoperative headache, gadolinium magnetic resonance images showed dural enhancement. The dural origin of the headache also fits well with the variable location of chronic headache after acoustic neuroma surgery.
As a significant proportion of patients develop postoperative headache, the possibilities of preventing headache should be examined. Various solutions, such as avoiding craniectomy (7), performing craniotomy instead of craniectomy (28) or closing the bone defect with meta-acrylate plate (1), have been suggested in the neurosurgical literature. After early enthusiastic reports the results have been less convincing, and variations in neurosurgical techniques have not solved the problem of postoperative headache. The efficacy of pre- or perioperative pharmacological interventions in prevention of chronic postoperative headache has not been studied widely. One study on venlafaxine in the prevention of postmastectomy pain has shown promising results (29). Several studies have shown a favourable effect of a single preoperative dose of gabapentin on perioperative and early postoperative pain, but none of these studies has looked at the reduction of chronic pain (30). Because of the rarity of acoustic neuroma, such studies are difficult in this patient group. An attractive approach would be to test whether pharmacological control of peri- and postoperative inflammation could prevent sensitization of dura and its sensory innervation and hence reduce prolonged postoperative headache.
Our results showing good efficacy of simple analgesics is in accordance with previous literature (1, 3, 8). The good pain relief provided by simple analgesics suggests them as first-line treatment for these patients. In severe cases they can be used regularly, whereas in less severe cases medication on demand is sufficient. Sumatriptan as an abortive medication for a headache attack has been tested in a small open, non-randomized study, showing alleviation of pain in one-third of patients (21), which, however, is also the proportion of placebo responders in triptan studies (31). Four of our patients used triptans for migraine, but it provided relief from postcraniotomy headache for none of them. Tricyclic antidepressants and gabapentin were the most commonly prescribed medications for the most refractory patients in our series. As these agents are neuromodulators and have proven efficacy in migraine prophylaxis (32, 33) and neuropathic pain (34), their clinical effect on pain does not refer directly to any specific mechanism.
MOH must be considered in the differential diagnosis of postcraniotomy headache. In this study 11 patients were found with use of analgesics and presence of bilateral headache on ≥15 days/month. However, 10 of these patients reported marked relief with simple analgesics, which speaks against MOH in these patients. One patient reported increased headache during the use of simple analgesics, but after the cessation the daily headache worsened severely. This patient was followed up in the pain clinic and was treated with paracetamol and gabapentin. The frequency and intensity of the headache gradually diminshed and the medication was slowly tapered off. Finally, the headache disappeared completely. In Vijyan's study, the prognosis for complete recovery was poor, as only 24% reported complete relief of headache, but prognosis for improvement of the severity of pain was better: 32% of patients noted gradual relief of headache in the course of years (10). Levo et al. (4) have reported that if headache after acoustic neuroma operation is present at 1 year after the operation, it remains chronic, whereas in another series the headache slowly alleviated over 2 years (9). In one-quarter of our patients, gradual relief of headache continued also after the first year. Headache intensity, frequency and duration of headache attacks tend to diminish with time. On the other hand, the patients in the new postoperative headache group with continuing headache had been followed for 1–9 years (median 4 years), which suggests that, without fulfilling criteria of MOH, the headache remains chronic in many patients.
Conclusions
Headache after acoustic neuroma operation is relatively common and is (for unknown reason) associated with small tumour size (1, 4, 5). We suggest that this headache should be classified as a specific subgroup of postcraniotomy headache, and that in the 3rd edition of the ICHD, the diagnosis of postcraniotomy headache be revised so as not to exclude bilateral headaches or headaches contralatral to operational side or headaches with maximal intensity other than in the area of the craniotomy. The typical features are aggravation of headache by physical stress, bending and coughing. Headache presents more commonly as attacks, but can also be continuous. We assume that sensitization of the sensory innervation of dura contributes to pain. According to our experience, simple analgesics (NSAIDs, coxibs, paracetamol) are the most useful drugs for these patients as the first-line agents. Some patients find additional pain relief from tricyclic antidepressants and gabapentinoids. The mechanisms of headache after acoustic neuroma operation need further studies to become fully explained.