Editor—The review on the topic of ‘chronic adhesive arachnoiditis’ (CAA) from obstetric epidurals by Rice and colleagues1 was apparently triggered by a series of articles that appeared in one of the London tabloids, fostered by some of the members of the Arachnoiditis Trust. These articles were unreasonable to many of us that remember the statistics of maternal deaths in the 1970s in the UK,2 when general anaesthesia was the predominant form of analgesia; aspiration of gastric contents and difficulty with tracheal intubation were the main culprits. I also feel that it is the right of women in labour to ask for pain relief, and anaesthetists ought to provide it for them. But we cannot deny that neuroaxial anaesthesia produces morbidity and that neurological deficits are probably one of the most serious. Unfortunately, the authors of the review lost the opportunity to assess the subject of neurological deficit and arachnoiditis (ARC) after epidural anaesthesia. Instead of being impartial, they attempted to prove that adhesive arachnoiditis does not happen as frequently as the patrons of the ‘Trust’ claimed it did and, when it does occur, they dismissed it as irrelevant.
Allow me to say for the record, that I do not belong to the Arachnoiditis Trust and I do not agree with their attempt to ban epidural anaesthesia for women in labour. Properly executed, epidural analgesia is, at the present time, the safest approach. However, by focusing mostly on the old concept of CAA, the authors of the review failed to recognize that ARC is an integral feature in most injuries to the intrathecal neural structures resulting in a variety of neurological deficits occurring after spinal interventions.34 These causes include: myelograms; spinal or epidural anaesthesia; invasive pain relief procedures; infections and blood entering the cerebrospinal fluid (CSF) from epidural blood patches; haematomas; trauma; or spinal operations. The arachnoid is now recognized as an active organ that responds to any invasion by initiating an inflammatory response proportional to the degree of injury. This reaction lasts ∼2 months; if not treated, it may progress into a chronic proliferative phase in which scarring, fibrosis and adhesions become permanent.4 These two phases are distinctly identified in radiological images with ‘enhanced’ or oedematous nerve roots,5 located in the anterior half of the dural sac with the appearance of ‘stars’ (Fig. 1) in the inflammatory phase and ‘clumped’ nerve roots346 forming bizarre patterns adhering to each other and to the dural sac, in the chronic proliferative phase (Fig. 2).
Concerning the recognition of symptoms typical of ARC, Rice and colleagues1 listed in Table 1 vague symptoms described in a list of publications before 1992; no mention was made specifically of the severe, continuous, burning pain in the lower back and extremities, accompanied by dysaesthesia and muscle spasms, as well as bladder, bowel and/or sexual dysfunction, which are all frequent manifestations of this disease. No reference was made to the burning characteristic implying neuropathic pain, as this concept had not yet been understood in that period. The injury may be traumatic from needle tips,78 catheters,9 dural tears,10 etc. or chemical from such substances as lidocaine 5%,11 2-chloroprocaine 3%,12 Myodil,13 neurolytics,14 blood,15 etc.
There were also some inconsistencies present in this review. There were nine previous publications describing similar lesions in the literature, before the 1909 article by Victor Horsley was published.16 The diagram on p. 110 of Rice's paper1 was taken (with permission) from a 1972 publication; much has been found since then about the anatomy of the spinal meninges; the arachnoid has indeed two layers with the CSF in between them, but the spinal cord does not occupy 90% of the dural sac as represented in their Figure 1. The authors quoted Long17 stating that ‘less than 1000 cases of arachnoiditis have been reported in the 50 years prior to 1992’; this is incorrect as it did not include Eastern European, Latin American or Asian publications. In 14 yr, I have obtained the medical history, examined and reviewed all the radiological studies in 374 patients with confirmed ARC; however, the incidence of a disease cannot be determined by the number of cases reported in the literature, nor by insurance claims or by mail surveys, especially in a disease with a high degree of iatrogenicity, which hampers accurate reporting.1718
The authors cited a 1964 reference (no. 108) noting that local anaesthetics cross the dura; more recently, elegant studies done by Bernards and colleagues19 have found that this passing is selective and that they are not transported by the arachnoid villi at the dural cuff,20 nor through the radicular arteries.21 Current understanding suggests that an arachnoiditic process is part of most post-interventional neurological deficits such as the cases of cauda equina syndrome after spinal or epidural anaesthesia.822 The most common radiological finding is clumped nerve roots (Fig. 2), but the ‘empty sac’ syndrome, deformity of the dural sac, intrathecal calcification, and fibrosis are also forms of arachnoiditis,861723 as well as some cases of syringomyelia caused by needle puncture,7 and postlaminectomy pseudomeningoceles.2 Cauda equina lesions may be recognized by obtaining coronal views in MRI.622
In essence, the authors searched for the subject ‘adhesive arachnoiditis’ and found the citations to the old cases of constrictive pachymeningitis after repeated intrathecal injections of dyes or steroids, infection or multiple surgical procedures. With the diagnostic tools at hand, specifically MRI, and CAT scan postmyelogram (Figs 1 and 2), the diagnosis of ARC can be made promptly. If it is in the inflammatory phase, treatment may prevent ARC from progressing into the chronic proliferative stage.46 By emphasizing old concepts, they missed the chance to impartially analyse the subject and give an important message that would advance everyone's knowledge.
Editor—We would like to thank Dr Aldrete for his letter regarding our review article,1 supported by a large volume of his own work.
We would like to make it clear that, whilst our concern about CAA after obstetric epidurals was initially aroused owing to articles by the Arachnoiditis Trust, it is because we realize that: (i) epidural intervention may have an effect on the arachnoid mater, and (ii) that a link between CAA and obstetric epidurals as claimed by the Arachnoiditis Trust would have devastating clinical implications for the women and the practice of obstetric regional analgesia and anaesthesia, that we constructed our review. We do not dismiss CAA as irrelevant in any way.
As Aldrete correctly points out, the remit of our review was to find a link between CAA and obstetric epidurals. We would like to reassure him that we undertook this task with an open mind and reject his accusation of our partiality. We conducted a thorough and impartial review of all the evidence published in peer-reviewed journals.
We are not the only reviewers who have failed to find a link between CAA or indeed arachnoiditis and obstetric epidurals.24–26 However, we did report on the few specific cases of CAA directly related to epidural anaesthesia from 1983 to 2000 in Table 2 of our review.1 We were interested to read of Aldrete's personal series of 374 patients with arachnoiditis and wondered how many were related to obstetric epidurals; unfortunately, this has not been reported or published.
Table 1 in our review referred to a summary of the symptoms reported in CAA and not early changes of arachnoiditis as it was referred to by Aldrete. We would also disagree with Aldrete in his interpretation of Reynolds' report8 that ‘an arachnoiditic process’ was involved in the conus damage during spinal anaesthesia. Direct trauma, as evidenced by a syrinx in the conus, was the cause. We have confirmed this with Prof. Reynolds.
We are glad that Aldrete agrees with us that in the light of current evidence, we should not withhold regional analgesia from women in labour, and also that he supports our conclusion that a full clinical examination and MRI investigation will help in the detection and diagnosis of arachnoiditis in obstetrics.
1Isle of Wight, UK 2Poole, UK 3Basingstoke, UK