Displaying 1-10 letters out of 510 published
A radiological curiosity of hyperammonaemia!
We would like to draw attention to one important point in regard to hyperammonaemic encephalopathy which was not mentioned in the recent excellent article by Sutter and Kaplan discussing the imaging features of encephalopathy.
The cases of hyperammonaemic encephalopathy with neuroimaging features number less than 10 in the reported literature. As the authors stated they can develop cortical signal abnormalities.
However having recently had experience with three cases of hyperammonaemia we noted despite the very diffuse cortical signal change, the signal change didn't involve the perirolandic cortex, in each of the three cases. This was noted in all but one of the reported cases in the literature.
It is worth drawing attention to this, as this radiological feature may be a feature which is quite specific to hyperammonaemia. Clearly this is observation is based on a very small number of patients, but is worth bearing in mind when assessing patients with cortical signal change, which may spare the perirolandic cortex.
As an explanation for this process, it is possible that this is related to cortical cytoarchitecture (perineuronal nets-abundent in perirolandic cortex )/receptor characteristics leading to a differential sensitivity to the toxic insult of elevated ammonia.
 Karaarslan E, Arslan A. Perirolandic cortex of the normal brain. Low signal intensity on turbo FLAIR MR images. Radiology. 2003;227:538-541
Conflict of Interest:
New Insights on AD: Comment on The identification of cognitive subtypes in Alzheimer's disease dementia using latent class analysis
We read with interest the recent article published by Scheltens et. al.  The article helped us in understanding greater insights about AD dementia. As rightly stated by the authors that one magic bullet will never be found, but different therapeutic agents may benefit different subgroups of patients. The identification and importance of cognitive AD subtypes for making differentiated diagnoses will also help in the further cognitive rehabilitation of AD patients. Out of the three clusters classified by the authors, one was the memory indifferent group wherein all patients were diagnosed with probable AD with a MMSE score with mean 14. The review suggests how an MMSE score is not supportive of AD diagnosis.  We would like to know the view of the authors on how memory indifferent group a MMSE score. As the neuropsychological test battery used by the institute is dynamic, the use of MMSE in future can be avoided as it is not a sensitive measure for diagnosis of AD. There are other available measures which are found to be more sensitive than MMSE for cognitive screening like MoCA. Attention is one of the cognitive domains that are found to be impaired in people suffering from AD.  It's intriguing to know from the results of the paper which show how attention is comparatively less affected in these patients. Lastly, we would like to know the age of the patients as the article talks about the younger patients who were six times more likely to be classified in MOD- VISP, three times likely to be classified in SEV-DIFF and two times likely to be classifies in MILD-EXE. The information about the different age ranges would provide a better insight. We would like to thank the authors for carrying out this research as it provides greater understanding of the AD.
1. Scheltens, N. M., Galindo-Garre, F., Pijnenburg, Y. A., van der Vlies, A. E., Smits, L. L., Koene, T., & van der Flier, W. M. (2015). The identification of cognitive subtypes in Alzheimer's disease dementia using latent class analysis. Journal of Neurology, Neurosurgery & Psychiatry, jnnp-2014. 2. Roselli, F., Tartaglione, B., Federico, F., Lepore, V., Defazio, G., & Livrea, P. (2009). Rate of MMSE score change in Alzheimer's disease: influence of education and vascular risk factors. Clinical neurology and neurosurgery, 111(4), 327-330. 3. Galasko, D., Klauber, M. R., Hofstetter, C. R., Salmon, D. P., Lasker, B., & Thal, L. J. (1990). The Mini-Mental State Examination in the early diagnosis of Alzheimer's disease. Archives of Neurology, 47(1), 49- 52. 4. Neufang, S., Akhrif, A., Riedl, V., F?rstl, H., Kurz, A., Zimmer, C., & Wohlschl?ger, A. M. (2011). Disconnection of frontal and parietal areas contributes to impaired attention in very early Alzheimer's disease. Journal of Alzheimer's Disease, 25(2), 309-321. 5. Pendlebury, S. T., Markwick, A., de Jager, C. A., Zamboni, G., Wilcock, G. K., & Rothwell, P. M. (2011). Differences in cognitive profile between TIA, stroke and elderly memory research subjects: a comparison of the MMSE and MoCA. Cerebrovascular diseases (Basel, Switzerland), 34(1), 48-54. 6. Larner, A. J. (2012). Screening utility of the Montreal Cognitive Assessment (MoCA): in place of-or as well as-the MMSE?. International Psychogeriatrics, 24(3), 391. 7. Calderon, J., Perry, R. J., Erzinclioglu, S. W., Berrios, G. E., Dening, T., & Hodges, J. R. (2001). Perception, attention, and working memory are disproportionately impaired in dementia with Lewy bodies compared with Alzheimer's disease. Journal of Neurology, Neurosurgery & Psychiatry, 70(2), 157-164.
Conflict of Interest:
IgG subclass in combined central and peripheral demyelination associated with anti-neurofascin 155 antibodies
At paranodes of both central and peripheral nerves, neurofascin-155 (NF155) is expressed by the terminal loops of myelin and associates with the axonal cell adhesion molecules contactin-1 and contactin-associated protein-1. They are important in maintaining the integrity of axo-glial junction and forming barrier against lateral diffusion of nodal channels. Human IgG antibodies consist of four subclasses (IgG1-4) with different structural and functional characteristics. IgG4 are generally believed to be non-inflammatory antibodies because of their inability to activate complement. IgG4 autoantibodies to NF155 and CNTN1 were identified in patients with chronic inflammatory demyelinating polyneuropathy (CIDP) sharing common clinical features, including subacute symptom onset and poor response to intravenous immunoglobulins (IVIG) (1, 2). Anti-contactin -1 IgG4 autoantibodies appear to affect paranodal structure both in vivo and in vitro (1, 3), suggesting that they are pathogenic. Anti-NF155 antibodies were reported by a single Japanese group in a proportion of patients with combined central and peripheral demyelination (CCPD) (4, 5). It is unclear whether their autoantibodies belong to the IgG4 subclass, but we have identified three Japanese patients with CCPD presenting with anti-NF155 IgG4 antibodies (Miura and Yuki, unpublished results). Whereas their CCPD patients responded to IVIG, our patients did not always respond to IVIG (Fukami and Yuki, unpublished results). One possible explanation is that the IgG subclasses were different in these two cohorts. Our patients showed anti-NF155 IgG4 antibodies, which had a low affinity for Fc receptors and complement, and did not activate complement in vitro (Miura and Yuki, unpublished results). By contrast, the IgG subclass has not been examined in their studies. It is thus important to know whether the autoantibodies identified in their study belong to the IgG4 subclass, and the reason why there is discrepancy in the response to IVIG treatment between their study and ours if both patients demonstrated anti-NF155 IgG4 antibodies.
Referemces 1. Doppler K, Appeltshauser L, Wilhelmi K, Villmann C, Dib-Hajj SD, Waxman SG, et al. Destruction of paranodal architecture in inflammatory neuropathy with anti-contactin-1 autoantibodies. J Neurol Neurosurg Psychiatry. 2015 Feb 18. pii: jnnp-2014-309916. doi: 10.1136/jnnp-2014- 309916. 2. Querol L, Nogales-Gadea G, Rojas-Garcia R, Diaz-Manera J, Pardo J, Ortega-Moreno A, et al. Neurofascin IgG4 antibodies in CIDP associate with disabling tremor and poor response to IVIg. Neurology. 2014;82:879-86. 3. Labasque M, Hivert B, Nogales-Gadea G, Querol L, Illa I, Faivre- Sarrailh C. Specific contactin N-glycans are implicated in neurofascin binding and autoimmune targeting in peripheral neuropathies. J Biol Chem. 2014;289:7907-18. 4. Kawamura N, Yamasaki R, Yonekawa T, Matsushita T, Kusunoki S, Nagayama S, et al. Anti-neurofascin antibody in patients with combined central and peripheral demyelination. Neurology. 2013;81:714-22. 5. Ogata H, Matsuse D, Yamasaki R, Kawamura N, Matsushita T, Yonekawa T, et al. A nationwide survey of combined central and peripheral demyelination in Japan. J Neurol Neurosurg Psychiatry. 2015 Feb 11. pii: jnnp-2014-309831. doi: 10.1136/jnnp-2014-309831.
Conflict of Interest:
Sensory Ataxia and Anti-contactin-1 IgG4-Associated Paranodopathy
With interest, we read an excellent paper written by a German group, in which four patients with chronic inflammatory demyelinating polyneuropathy (CIDP) carried IgG autoantibodies against contactin-1 (CNTN1) expressed at the paranodes in the peripheral nerves. Human IgG antibodies consist of four subclasses (IgG1-4) with different structural and functional characteristics. IgG4 are generally believed to be non- inflammatory antibodies because they poorly bind to complement and Fc receptors Three of their patients had anti-CNTN1 IgG4 and sera from these three patients bound the paranodes in the mouse teased nerve fibers. On the other hand, the remaining patient with no anti-CNTN1 IgG4 antibodies in the serum did not bind the paranodes.
In a large cohort of Japanese patients with CIDP (n = 533) and Guillain-Barre syndrome (GBS) (n = 200), our group identified anti-CNTN1 IgG antibodies in 16 sera from patients with CIDP and five with GBS. IgG4 antibodies to CNTN1 were identified in 13 of 16 patients with CIDP, but none of those with GBS. IgG2 antibodies to CNTN1 were identified in three patients with CIDP and in five patients with GBS. We also blindly tested the sera on mouse teased fibers. Of interest, all the IgG4-positive CIDP sera strongly reacted against the paranodal domains but not for the IgG2-positive sera from patients with CIDP or GBS. These results are in line with those of the German group, suggesting that only the IgG4 antibodies are pathogenic.
We further tested whether these sera activate the complement pathway in vitro. None of the sera with anti-CNTN1 IgG4 antibodies induced the deposition of activated C3 components on enzyme-linked immunosorbent assay plates. These results support that anti-CNTN1 IgG4 antibodies do not fix complement.
The German group found out a typical clinical picture among the CIDP patients with anti-CNTN1 antibodies, namely acute onset of disease and severe motor symptoms, with three out of four patients manifesting action tremor. However, in our cohort we found out that presence of anti-CNTN1 IgG4 antibodies was significantly associated with the clinical sign of sensory ataxia. CNTN1 was widely expressed in large dorsal root ganglion neurons. Similarly, CIDP sera with anti-CNTN1 IgG4 antibodies stained large neurons in dorsal root ganglion sections and co-localized with CNTN1 staining in the soma and at the paranodes of sensory axons. These results support that the anti-CNTN1 autoantibodies induce the development of sensory ataxia. Nonetheless, the clinical feature of our patients contrasted with previous studies where none of the three German patients and only one of four Spanish patients with anti-CNTN1 IgG4 antibodies showed sensory ataxia.[1,3] The reason for this discrepancy is unclear, but this discrepancy should motivate international study groups to investigate the clinical feature of the anti-CNTN1 IgG4 antibodies among different countries, and their underlying mechanism.
References 1. Doppler K, Appeltshauser L, Wilhelmi K, et al. Destruction of paranodal architecture in inflammatory neuropathy with anti-contactin-1 autoantibodies. J Neurol Neurosurg Psychiatry 2015 Feb 18. pii:jnnp-2014- 309916. doi: 10.1136/jnnp-2014-309916. 2. Miura Y, Devaux N, Fukami Y, et al. Contactin 1 IgG4 associated to chronic inflammatory demyelinating polyneuropathy with sensory ataxia. Brain 2015 Mar 25. pii: awv054. 3. Querol L, Nogales-Gadea G, Rojas-Garcia R, et al. Antibodies to contactin-1 in chronic inflammatory demyelinating polyneuropathy. Ann Neurol 2013;73:370-80.
Conflict of Interest:
Nodopathy or paranodopathy: that is the question
In peripheral nerves, the domain organization of myelinated axons depends on specific axoglial contacts between the axonal membrane and Schwann cells at nodes, paranodes and juxtaparanodes. The term nodo- paranodopathy was originally proposed to characterize neuropathies with anti-ganglioside antibodies by a common pathological continuum starting with dysfunction/disruption at the nodes of Ranvier, a transitory nerve failure and resulting in axonal degeneration. Uncini and Kuwabara extended the concept to include neuropathies of different etiology (dysimmune, inflammatory, ischemic, nutritional and toxic) in which the involvement of the nodal region is determinant. As the pathogenic mechanisms are mainly focused on the node, the authors proposed to classify these neuropathies as nodopathies. We agree with the change from nodo-paranodopathy to "nodopathy" because the latter term is simple and easy to be kept in mind. However, as recently suggested by Doppler and colleagues, a new term "paranodopathy" should be used to define neuropathies associated with autoantibodies to paranodal cell adhesion molecules.
Cell adhesion molecules at the nodes and paranodes dictate the formation of the nodes of Ranvier and enable the rapid saltatory propagation. At paranodes, contactin-associated protein-1 associates with contactin-1 (CNTN1) on the axonal membrane and this complex interacts with a glial partner expressed by the terminal loops of myelin, neurofascin-155 (NF155). This ternary complex of glycoproteins is required for the formation and stability of the septate-like junctions which form an insulating barrier at paranodes. IgG4 autoantibodies to NF155 and CNTN1 have been documented in patients with chronic inflammatory demyelinating polyneuropathy, sharing common clinical features, including subacute symptom onset and poor response to intravenous immunoglobulin.[3-5] When introducing the term nodopathy/paranodopathy, Uncini and colleagues referred to patients with anti-ganglioside antibodies and conduction block, but nearly normal nerve conduction velocities. In contrast, both anti-CNTN1 and anti-NF155-mediated neuropathies are associated with a reduction of nerve conduction velocity.[3-5] The loss of paranodal axoglial junctions and the elongation of the nodal gap in these patients indicate that autoantibodies against CNTN1 could lead to paranode disruption and result in slow nerve conduction or conduction block. We agree with the new concept proposed by Doppler and colleagues that anti-CNTN1 and anti-NF155-mediated paranodopathies are a new subgroup of autoimmune peripheral neuropathy with distinct clinical phenotypes and damage to the paranodes.
References 1. Uncini A, Susuki K, Yuki N. Nodo-paranodopathy: beyond the demyelinating and axonal classification in anti-ganglioside antibody- mediated neuropathies. Clin Neurophysiol 2013;124:1298-34. 2. Uncini A, Kuwabara S. Nodopathies of the peripheral nerve: an emerging concept. J Neurol Neurosurg Psychiatry 2015 Feb 19. pii: jnnp-2014-310097. doi: 10.1136/jnnp-2014-310097. 3. Doppler K, Appeltshauser L, Wilhelmi K, et al. Destruction of paranodal architecture in inflammatory neuropathy with anti-contactin-1 autoantibodies. J Neurol Neurosurg Psychiatry 2015 Feb 18. pii: jnnp-2014- 309916. doi: 10.1136/jnnp-2014-309916. 4. Querol L, Nogales-Gadea G, Rojas-Garcia R, et al. Neurofascin IgG4 antibodies in CIDP associate with disabling tremor and poor response to IVIg. Neurology 2014;82:879-86. 5. Miura Y, Devaux N, Fukami Y, et al. Contactin 1 IgG4 associated to chronic inflammatory demyelinating polyneuropathy with sensory ataxia. Brain 2015 Mar 25. pii: awv054.
Conflict of Interest:
CNS involvement in V30M transthyretin amyloidosis: clinical, neuropathological and biochemical data from another center.
We read with great interest the recent study by Maia et al., which reports retrospectively central nervous system (CNS) involvement in patients with familial amyloid polyneuropathy associated with Val30Met mutation (ATTR Val30Met FAP)1. This complication was observed in 31% of ATTR Val30Met FAP patients with long survival due to liver transplantation (LT). CNS disorders occurred on average 14.6 years after the onset of TTR- FAP, which is beyond the average life expectancy (11 years) of non- transplanted patients2. The CNS symptoms were classified as negative (ischemia and aura) or positive (epileptic seizures). Magnetic resonance imaging (MRI) could not be performed in their population of patients; however, amyloid deposition from the meninges to the brain parenchyma was proposed as the neuropathological hallmark of those clinical manifestations, mainly based on historical necropsy studies1. Indeed, LT does not preclude amyloid production by the retinal pigment and choroid plexus, thus ocular and CNS involvement many years after LT is not an unexpected observation3. Moreover, atypical cases ATTR Val30Met FAP presenting with early CNS manifestations have been described4.
In our center, a similar experience with de novo CNS dysfunction in long survivors with ATTR Val30Met FAP has been identified. From 83 patients with a disease duration longer than 10 years (mean 15.6, SD 5.1), in whom 91.5% (76/83) were submitted to LT (mean duration after LT of 10.8 years), 16.9% (14/83) had some CNS manifestation. The mean interval between disease presentation and CNS manifestation was 16.9 years (SD 5.5). In the group of affected patients, 64% (9/14) had a transitory ischemic accident-like episode, 7% (1/14) ischemic stroke, 7% (1/14) brain hematoma, 42.8% (6/14) epileptic seizures, 7% (1/14) a psychosis-like event and 92.8% (13/14) cognitive impairment (11 with mild cognitive impairment criteria and 2 with minor cognitive impairment) although no patients had criteria for dementia. Our patients underwent brain CT scan, EEG, neurovascular workup (blood tests, EKG, transcranial doppler, carotid ultrasound and echocardiogram) and formal neuropsychological evaluation including Mini Mental State Examination (MMSE), tests of attention, memory and executive functions. Brain MRI (without gadolinium) was performed in two patients with seizures, which excluded ischemic or hemorrhagic lesions. Neuropsychological assessments disclosed attention and initiative deficits as the most critical findings, suggesting dysexecutive (frontal) involvement. A 40 year-old male, in whom LT was performed at the age of 26 (three years after the disease onset), had an acute right parieto-occipital hematoma, two years after his initial symptoms of epileptic seizures. During surgical decompression, a cerebral biopsy was done that showed TTR amyloid deposition in meningeal vessels and cerebral parenchyma with no beta A4 amyloid deposition (Figure 1). This observation was in agreement with the neuropathological findings described by Maia et al1. In general, our results support the detailed report by Maia et al1. In our population, seizures were somewhat more frequent, and MRI could be performed in two of those patients, excluding vascular changes. Furthermore, neuropsychological testing disclosed mild executive impairment in almost all patients, which raises the hypothesis of a predominant frontal amyloid deposition. Many questions remain open and should be addressed in future studies, in particular the risk of dementia and seizures in this group of patients. This has critical implications regarding medical decision towards LT in TTR-FAP patients.
1. Maia LF, Magalh?es R, Freitas J et al. CNS involvement in V30M transthyretin amyloidosis: clinical, neuropathological and biochemical findings. J Neurol Neurosurg Psychiatry 2015; 86:159-67. 2. Coutinho P, Martins da Silva, A, Lopes Lima J, Resende Barbosa A. Forty years of experience with type I amyloid neuropathy. Review of 483 cases. In: Glenner G, Costa P and de Freitas A. (eds). Amyloid and Amyloidosis 1980. Amsterdam: Execerpta Medica, pp. 88-98. 3. Sandgren O, Kjellgren D, Suhr O. Ocular manifestations in liver transplant recipients with familial amyloid polyneuropathy. Acta Ophthalmol 2008; 86:520-524. 4. Herrick MK, DeBruyne K, Horoupian DS, Skare J, Vanefsky MA, Ong T. Massive leptomeningeal amyloidosis associated with a Val30Met transthyretin gene. Neurology 1996; 47:988-992.
Conflict of Interest:
Cerebellar ataxia: lessons from research
I read with interest the review by Tada et al. , which, based on the current knowledge on the cerebellar systems, propose an interesting framework to interpret cerebellar ataxias in clinical settings. The Authors classify cerebellar ataxias into two main categories: those with a loss of Purkinje cells, translating into a malformation of internal models; and those with a disturbance of afferent systems, causing a mis- selection of internal models. From a clinical point of view, the first type of deficit, as observed in SCA31 and SCA6, results in decomposition and dysmetria. In addition, the loss of Purkinje cells may initially affect just the quality, and not the quantity, of internal models, and new internal models could be reconstructed using the residual circuits. This could explain why patients with pure cerebellar ataxia initially respond well to rehabilitation. Concerning the second category, a further classification differentiates the disturbances of the corticopontocerebellar system, such as in MSA-C and SCA2, from those of the spinocerebellar system, such as in SCA1, SCA3, FRDA and AOA1. In all of them there is dysmetria and dysrhythmia but, while the pontocerebellar disturbance affects the internal model selection for well-trained limb movements, such as writing or playing an instrument; disturbances of the spinocerebellar system affect actions requiring continuous feedback from the periphery, such as stance and gait. Finally, a third type of ataxia, such as the DRPLA, involves the efferent system (deep nuclei), resulting in a deficit of the overall cerebellar function.
This classification has the great value of building a bridge between basic research and clinical knowledge, and it's simple enough to be implemented in everyday clinical practice, but it deserves some comments.
First, as already acknowledged by the Authors, the pathology of degenerative ataxias is much more complex than the proposed schematization, as there is usually a combined degeneration of various systems and the very initial pathological alterations in many cases are still controversial .
Second, the proposed classification is based on the classical view that the cerebellar cortex has a uniform structure and that a unique neural computation (the so called "cerebellar transform") is performed throughout it . This implies that the functional role of the different cerebellar areas is dictated by its input and output connections and that, once known these connections, the impact of a lesion can be predicted relatively easily. Nowadays, it looks that this traditional concept requires sophistication. In fact, as underlined in a recent review by Cerminara et al. , cytoarchitectural and physiological variations have been identified in the cerebellar cortex, supporting the non-orthodox hypothesis that the cerebellar cortex, and therefore its information processing, is not uniform. In this case, classifying ataxias just on the basis of the anatomical regions affected could not be enough to predict the functional deficit. Interestingly, it seems that, the configuration of Purkinje cells death in neurodegeneration is not random, being probably genetically determined . Therefore, a deeper knowledge of these patterns and at the same time of the cerebellar cortex computations could lead to a more accurate phenotypic classification of ataxias.
Third, Tada et al.  focused their observations on motor symptoms. As the cerebellum is nowadays known to have also non-motor roles, a similar classification regarding cognitive functions in ataxias would be of interest, even if, again, the complexity of neurodegeneration would make this task quite arduous .
Finally, formally exploring the motor differences between these categories, with an accurate registration of movement kinematics and a motor learning task, would be of interest. This would require testing subjects in an early stage of disease, to evaluate if motor performance and motor learning vary across the different neurodegeneration patterns and with disease progression. This approach could finally lead to development of new rating scales but also to an improvement of actual neurorehabilitation approaches.
1. Tada M, Nishiwaza M, Onodera O. Journal of Neurology Neurosurgery and Psychiatry 2015; 0:1-7. doi: 10.1136/jnnp-2013-307225
2. Fratkin JD and Vig PJS. Handbook of Clinical Neurology 2012; 103:111-125.
3. D'Angelo E, Casali S. Frontiers in Neural Circuits 2013; 10:116. doi: 10.3389/fncir.2012.00116. eCollection 2012.
4. Cerminara NL, Lang EJ, Sillitoe RV, Apps R. Nature Reviews Neuroscience 2015; 16:79-93.
5. Burk K. Cerebellum 2007; 6(3):280-6.
Conflict of Interest:
Azathioprine and cyclosporine for myasthenia gravis in pregnancy
We read the recently published UK guideline for treating myasthenia gravis (MG) patients in pregnancy with great interest . As stated by the workgroup in their article, few studies are available concerning drug safety in pregnancy in these patients. In our experience, most drugs indeed carry a very small risk in the treatment of MG during pregnancy. However, based on additional literature we would advise some caution regarding use of azathioprine and cyclosporine during pregnancy.
Cleary and others described pregnancy outcomes in 476 women using AZT during early pregnancy, mostly for inflammatory bowel disease . A significant increase in atrioventricular septum defects was found (OR 3.18), pre-term birth, low birth weight and small stature were also more significant. A trend towards an increase in all birth defects was found (OR 1.41, 95% CI 0.98-2.04) was found in women with IBD and azathioprine, compared to IBD patients without azathioprine.
For cyclosporine, a meta-analysis of multiple studies shows a non- significant increase of malformations (OR 3.83, 95% CI 0.75-19.6) and borderline significant prematurity (OR 1.52, 95% CI 1.00-2.32) . Although inconclusive, the limited data available suggests at least some caution. Overall, prematurity and low birth weight are reported more frequently in women exposed to cyclosporine, although it is unclear if these are related to cyclosporine or maternal disease .
In our opinion, both azathioprine and cyclosporine cannot be considered completely safe. Although both are unlikely to be major teratogens, we consider both to be relatively contra-indicated. Our recommendation would be
1) to discuss the small risks associated with use of azathioprine and cyclosporine and
2) to discontinue both immunosuppressants from planned conception and during the first trimester in patients with stable disease and no or mild weakness.
1. Norwood F, Dhanjal M, Hill M, et al. Myasthenia in pregnancy: best practice guidelines from a U.K. multispecialty working group. J Neurol Neurosurg Psychiatry. 2014 May;85(5):538-43.
2. Cleary BJ, Kallen B. Early pregnancy azathioprine use and pregnancy outcomes. Birth Defects Res A Clin Mol Teratol. 2009 Jul;85(7):647-54.
3. Bar Oz B, Hackman R, Einarson T, et al. Pregnancy outcome after cyclosporine therapy during pregnancy: a meta-analysis. Transplantation. 2001 Apr 27;71(8):1051-5.
4. Paziana K, Del Monaco M, Cardonick E, et al. Ciclosporin use during pregnancy. Drug Saf. 2013 May;36(5):279-94.
Conflict of Interest:
Re:How accurate is the record linkage?
We thank Dr. Voci for his comments on our article. As described in the paper, our data comprise Hospital Episode Statistics (HES) obtained from the English national Health and Social Care Information Centre (HSCIC) and mortality record abstracts obtained from the Office for National Statistics. All data from both sources were anonymised to the same standard by encryption of personal identifiers before being supplied to the Oxford study team for record linkage studies.
The main data item used in the algorithm for linkage between successive HES records for the same person was "HES ID", an encrypted unique personal identifier generated by HSCIC. HES ID was present for 100% of the HIV HES records, 100% of the reference cohort, and 100% of the multiple sclerosis HES records. For the linkage of HES records to death records, probabilistic matching was used, based on encrypted NHS number, encrypted postcode, encrypted date of birth, and sex.
To illustrate the quality of the matching, and further to Dr. Voci's suggestion, we have followed up the HIV cohort for tuberculosis, a disease known to be associated with HIV. The rates in the HIV cohort were very substantially elevated compared with those in the reference cohort comprising people without HIV: there were 539 observed and 33.8 expected cases of tuberculosis in the HIV cohort; and 4763 observed and 5268.2 expected cases of tuberculosis in the reference cohort. The rate ratio, standardised for age and other factors as described in our original article,2 was 17.6 (95% CI 16.1 to 19.3). Restricting the outcomes to death records only, the standardised rate ratio was also very high: there were 34 observed and 1.4 expected cases of tuberculosis in the HIV cohort, and 496 observed and 528.6 expected cases of tuberculosis in the reference cohort. The standardised rate ratio was 25.5 (95% CI 17.5 to 36.17), i.e. there was a 25-fold increased death rate from tuberculosis in the HIV cohort. We have confidence in the dataset and the methods used to match and link records in it.
1. Voci C. How accurate is the record linkage? J Neurol Neurosurg Psychiatry
2. Gold J, Goldacre R, Maruszak H, et al. HIV and lower risk of multiple sclerosis: beginning to unravel a mystery using a record-linked database study. J Neurol Neurosurg Psychiatry 2014 Aug 4 [Epub ahead of print] doi: 10.1136/jnnp-2014-307932
Conflict of Interest:
How accurate is the record linkage?
This is an interesting study. However, I am somewhat concerned whether the record linkage could be appropriately performed. In most healthcare information systems, HIV patients characteristics are protected. Did the authors have enough information on the patients to perform probabilistic matching? It would be of interest if the authors could show that the matching is adequate by showing that there is not a general reduced risk of developing other autoimmune diseases and confirming known HIV associations, for example with TB.
Conflict of Interest:
Visit the full archive of podcasts for JNNP here >>
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of JNNP.
View free sample issue >>
Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.