gms | German Medical Science

A large number of pathological entities may develop in the petrous apex [Ref. 1]. Some are relatively aggressive but the common lesions, congenital cholesteatoma and cholesterol cysts, are not and can take years or decades before inflicting serious neurological deficits. In other words, many of these indolent lesions can be managed conservatively by watchful waiting.

Surgical access to the petrous apex can be extremely difficult and is not without risk of significant complications [Ref. 2]. A number of lateral and anterior approaches have been described, each offering specific advantages but sometimes accompanied by disadvantages, for example hearing loss [Ref. 2], [Ref. 3], [Ref. 4]. It is generally accepted that no one approach is suitable for all cases largely because of anatomical variations [Ref. 5], [Ref. 6].

At the European Skull Base Congress in 1999 we reported our experience with patients who had petrous apex disease [Ref. 7]. From this a management strategy for those with petrous apex cholesterol granulomas and cholesteatomas was defined which we have adhered to since. The aim of this study was to review the clinical status of our original cohort of patients and to re-consider our management strategy in the light of subsequent cases.

This was a retrospective analysis based on regular clinical reviews and interval scanning. The inclusion criteria were patients with cystic petrous apex lesions (congenital cholesteatoma and cholesterol cysts) treated between 1987 and 2004.

Eighteen patients were reported in 1999 of which 3 had radiographic anomalies that had been misreported as pathological lesions. Of the remainder, 8 patients had 9 cholesterol cysts and 7 patients had cholesteatomas. Since then, a further 10 patients with 11 cholesterol cysts and 9 patients with cholesteatomas have been added to our series. So, in total, we have 34 patients with 36 lesions.

In this cohort of 34 patients, 23 were male and 11 were female. The mean age at presentation was 42 years for those with cholesterol cysts and 52 years for those with cholesteatoma.

There were striking differences in the presenting symptoms between the two groups of patients. Overall, 58% of patients had a hearing loss. In the cholesteatoma group, 94% of patients had a hearing loss of which 75% were severe to profound. By contrast, 22% of the cholesterol cyst group had a hearing loss and only 6% (one patient) had a profound loss.

Facial nerve palsy was a prominent feature in the cholesteatoma group as 50% had some degree of weakness. By contrast, only 1 patient (6%) with a cholesterol cyst had a facial nerve palsy at the time of presentation.

No patient had strictly trigeminal symptoms in the cholesteatoma group even though a few had a non-descript headache. By contrast, 28% of patients with cholesterol cysts had either diminished facial sensation or classical trigeminal neuralgia. One patient with a cholesterol cyst also present with an abducens nerve palsy.

Vertigo was not a discriminating feature and was problematic in a third of patients in both group. Otorrhoea was a common feature in the cholesteatoma group (50%). This had developed either because the cholesteatoma had breached the surface either in the external auditory canal or the middle ear while in others surgeons had attempted mastoid surgery but underestimated the task.

The classic imaging characteristics of these lesions are well known and have been described previously. However, it is worth noting that variations in signal intensity on T2 weighted images of cholesterol cyst are common and reflect the constitution of the cyst, some areas being solid and organized while other parts are filled with fluid.

An algorithm of our management strategy is shown in Figure 1 [Fig. 1]. We advocate that all patients with epidermoid cyst should be considered for surgery. However, those with no cranial nerve deficits other than hearing loss and where there is no infective risk can be managed by conservatively by watchful waiting and interval scanning. This conservative approach is really only an option in the elderly.

Patients with cholesterol cysts pose a more difficult management problem. Those with minimal symptoms and no significant cranial nerve deficits may also be managed by watchful waiting and interval scanning. For those who have significant symptoms, the degree of hearing loss and whether other cranial nerve deficits are present become the determinant factors in management. We feel that patients with any degree of facial palsy or established trigeminal deficit would be better served by a resection in the first instance. Patients with serviceable hearing should be offered the option of a hearing preservation drainage procedure.

In cholesteatoma group, one patient was observed for 18 years without change of cyst size or symptoms. The rest of patients in this group all underwent resection of the cyst. Ten patients had a subtotal petrosectomy, three patients had resections using the type B infra-temporal fossa approach, one patient a transcochlear approach and another patient had a resection through a sub-occipital approach.

The mean follow up period for this group was 101 months (median: 103 months). Pre-operative otorrhoea resolved in all who presented with that symptom. Three patients sustained necrosis of their fat grafts, and in two patients, recurrent cholesteatoma developed that needed to be removed. Three patients acquired a temporary facial weakness following surgery all of which recovered to their pre-operative state or better. Two patients had complete facial palsies pre-operatively and had VII–XII anastomoses. They recovered to grade 3 and 4 House Brackmann function.

The mean follow-up period for the cholesterol cyst group was 62 months (median = 60.5 months). Of the 18 patients with 20 cholesterol cysts, 6 cysts have been managed conservatively with a follow-up period that ranged from 20–76 months. None have needed to be drained or resected. Six cysts have been drained using the infra-labyrinthine or transphenoidal routes. Hearing has been preserved in all these patients. One patient failed on clinical grounds because she developed facial spasm. Her cyst was resected and complete resolution of symptoms was achieved. The one patient who presented with an isolated VIth nerve palsy and normal hearing underwent a transphenoidal drainage procedure. She made a full recovery. There is radiographic evidence of residual or recurrent cyst contents in 4 patients but all remain asymptomatic and there is no evidence of increase in size of the cysts. 8 cysts have been resected using either the type B or translabyrinthine approaches. All have subsequently become asymptomatic and none have recurred. One patient in this group had a temporary facial weakness which recovered completely.

CT and MR techniques have transformed the diagnosis and management of petrous apex lesions [Ref. 8], [Ref. 9]. Once the diagnosis has been established, the management strategy for these two lesions should be different. That this is so is dictated by their very different natural histories. In former times it was believed that the management of choice for all petrous apex lesions was surgery [Ref. 10]. This is still largely correct for those patients with cholesteatoma as late onset infection can be life threatening at worst but many more patients develop significant cranial nerve palsies. However, for those patients with cholesterol cysts a watchful waiting policy can be safely adopted as long as there are no cranial nerve deficits.

In our experience, facial palsy and trigeminal nerve deficits were valid indications for surgical intervention in the cholesterol cyst group. The choice between resection and drainage is largely dependent on whether there is serviceable hearing. The majority drainage procedures in this series were performed through an infralabyrinthine approach (5 out of 6). While not universally satisfactory, one patient eventually requiring a resection, it has stabilized the disease process in the remainder.