Dear Editor,

Srikanth et al have provided vital information on Tuberculous meninigitis in patients above the age of 5o years in their article: "Clinicoradiological features of tuberculous meningitis in patients over 50 years of age J Neurol Neurosurg Psychiatry 2007; 78: 536-538" However, a few points are of note: In their data, since the diagnosis of TBM is based on several criteria, it is important to know how many had now many criteria for the diagnosis. The sensitivity and specificity of (CSF) Anti-tuberculous IgG antibody drops rapidly as we include patients with and without culture-proven active tuberculosis, and becomes as low as 60%* in patients with only clinically diagnosed tuberculosis. In the authors' series only 3 patients had culture-proven TBM. In all the other patients, ancilliary evidence has been used to get a diagnosis of TBM, including Mantoux test, response to therapy, and exposure to tuberculosis. By these criteria, it is very difficult to be sure of the diagnosis of TBM, which is a problem commonly faced in tropical countries. It would be useful to have some idea of what criteria, and how many, they used to diagnose Probable TB in individual patients. It would be acceptable if they had found culture-proven TBM patients having normal CT scans; however, since their series consists of a large number of patients without definitive proof of TBM, the assertion that normal CT scans can be associated with TBM remains questionable.

References

*J Clin Microbiol. 2001 Oct;39(10):3603-8. Links Clinical evaluation of anti-tuberculous glycolipid immunoglobulin G antibody assay for rapid serodiagnosis of pulmonary tuberculosis. Maekura R, Okuda Y, Nakagawa M, Hiraga T, Yokota S, Ito M, Yano I, Kohno H, Wada M, Abe C, Toyoda T, Kishimoto T, Ogura T. Toneyama National Hospital, Toneyama ity, Osa

Dear Editor,

The authors thank Dr Stanley for her interest in our manuscript. In contrast to the conventional neurophysiological techniques utilised in Dr Stanley's earlier studies, the more recent studies published in JNNP focused on changes in nerve excitability that occur during dialysis. These changes in excitability appear to link to serum potassium, and provide support for the hypothesis that potassium is a uremic neurotoxin.(1)

In terms of primacy, which was not mentioned in our JNNP paper, the effects of dialysis therapy on peripheral nerve function date as far back as the 1960's. For interested readers, the historical aspects of uraemic neuropathy and details of previous studies are included in a recent review of the area. (2)

Yours sincerely,

Matthew Kiernan

References

1. Bostock, H, Walters, R.J.L., Andersen, K.V., Taupe, D., Murray, N.M.F. & Kiernan, M.C. (2004). Has potassium been prematurely discarded as a contributing factor to the development of uraemic neuropathy? Nephrology, Dialysis, Transplantation 19:1054-1057.

2. Krishnan A.V., Kiernan M.C. (2007). Uremic neuropathy: Clinical features and new pathophysiological insights. Muscle Nerve 35:273-290.

Dear Editor,

We write in response to a recent review by Lies Bouwmeester(1) et al on the effectiveness of visual training for patients with brain damage. We do not agree with their comments on the study of Kasten et al (2). This remains (with its follow-up study) the only randomised placebo-controlled trial of Vision Restoration Therapy (VRT), and is the foundation of claims of effectiveness for this rehabilitation technique.

Bouwmeester's review includes only the retrochiasmal damage section of Kasten's study. It reports this trial to be of good quality. Specifically it states that it earned “a good methodological score on randomisation, blinding and comparability of the groups. We disagree that blinding or comparability was good.

In relation to blinding, allocation of patients to treatment or placebo groups was double-blind but there is no mention of blinding of visual field measurement. These measurements were largely automated but fixation monitoring required direct visualisation by the examiner. Doubt has been cast on the effectiveness of the additional automatic fixation monitoring procedure. The examiner therefore played an important role in visual field measurement and field testing was thus a potential source of bias. One should be mindful that the strongest criticism of this study has related to the adequacy of fixation control(3,4). In addition there is no comment in the paper regarding concealment; the researchers were probably aware of which group each patient was in from an early point in the study.

Regarding comparability of groups, Kasten’s Table 1 outlines the baseline characteristics of the two “post-chiasmatic injury groups. We stress that randomisation, particularly in small trials, does not guarantee that groups are comparable. We feel that in this study the groups differ in two important ways.

First, both groups were heterogeneous with regard to the mechanism of injury. Most notably there were no patients with traumatic brain injury at all in the control group. Also, only 22% (2 of 9) of patients in the treatment group had stroke, compared with 80% (8 of 10) in the placebo group. There is little reason to believe that the pathologies grouped here behave identically, as we are reminded by Robert McFazdean in his recent review of Vision Restoration Therapy(5). Group differences could (at least in part) result from differences in group constitution rather than treatment effects. Certainly one should treat with caution the authors statement that the study illustrates that patients who suffer from partial blindness due to trauma or stroke can benefit from VRT.

Second, the groups are not equivalent in lesion age. Spontaneous recovery is uncommon more than six months after brain injury, therefore for groups to be comparable the proportion of injuries less than six months old should be equivalent. Lesions in the treatment group were aged 6.8 +/- 11.4 months, compared with 7.2 +/- 6.3 months for the placebo group. A difference of 0.4 months between group means is not worrying but the spread of the groups is important. Assuming that the numbers quoted are mean +/- standard deviation (this information is not in the paper) the variance of the treatment group is over three times that of the placebo group. Since the treatment groups spread is larger than its mean and the lesion age cannot be negative, the distribution must be positively skew with a majority of values below the mean; a substantial proportion of lesions in this group must be well below the mean age (and below 6 months old). This is not true of the placebo group which has a lower spread and higher mean, and therefore contains more lesions more than 6 months old. Thus there is more inequality in the number of lesions less than 6 months old than is apparent from the means of the groups, there being more of the younger, potentially recoverable deficits in the treatment group.

Taken together these factors indicate a risk of significant bias, in both cases likely to favour a positive finding of the study. Taken with other criticisms of the study they cast further doubt on the validity of its findings. This is a difficult and controversial area but we consider that such methodological shortcomings may account for apparently irreconcilable differences between studies and should be brought to light. Detailed criticism of published papers along with further meticulously designed research will hopefully lay this matter to rest in the not too distant future.

Philip L Clatworthy

References

1. Bouwmeester L, Heutink J, Lucas C. The effect of visual training for patients with visual field defects due to brain damage. A systematic review. J Neurol Neurosurg Psychiatry 2006.

2. Kasten E, Wust S, Behrens-Baumann W, Sabel BA. Computer-based training for the treatment of partial blindness. Nat Med 1998;4(9):1083-1087.

3. Horton JC. Disappointing results from Nova Vision's visual restoration therapy. Br J Ophthalmol 2005;89(1):1-2.

4. Horton JC. Vision restoration therapy: confounded by eye movements. Br J Ophthalmol 2005;89(7):792-794.

5. McFadzean RM. NovaVision: vision restoration therapy. Curr Opin Ophthalmol 2006;17(6):498-503.

Competing interests: none

Dear Editor,

We consider your paper very interesting, especially comparing to our patient. He is a 65 year-old hypertensive, polivascular man presented with a serious symptomatic right internal carotid artery stenosis (90%) and a recent minor nucleo-basal stroke. He has been referred for cerebral angiography before carotid stenting procedure. At the beginning of the exam, when only 3 cc of contrast medium was been injected and catheter guide was in the aortic arc, patient presented generalized seizure and coma.

In the next 2 days the patient was unconscious . On the third day he was alert but confuse and unresponsive , he couldn’t look moving objects as for visual agnosia. After ten days his mental status improved with residual signs of cortical blindness. Motor or somatosensorial signs were never evident.

MRI was performed one day after angiography showing bilateral hyperintensity in the white matter of occipital and temporal lobes, in the cerebellum and basal ganglia, predominantly suggestive of vasogenic edema (diffusion-weighted images), but with limitate signs of possible cytotoxic in the occipital cortex, bilaterally (positive ADC). Electroencephalograms showed evidence of diffuse encephalopathy. Laboratory investigations were unremarkable. The patient was treated with corticosteroid therapy and clopidogrel.

After one month MRI showed remarkable improvement: altered signal intensity was present particularly in the occipital areas (where ADC signal had been previously altered). Because of MRI imaging we considered this acute encephalopathy as a type of reversible encephalopathy caused by a possible unknown neurovegetative reaction of cerebral arteries to angiography procedure. Your report also help us to consider in this patient a type of reversible encephalopathy due to cholesterol embolism.

Carmela Gerace, Enrico Cotroneo, Renato Gigli, Daniela Tessarolo, Fabio Massimo Corsi

Azienda ospedaliera S.Camillo-Forlanini

Figure 1