Artigo Acesso aberto Produção Nacional Revisado por pares

Are Antibody Panels Under‐Utilized in Movement Disorders Diagnosis? No

2021; Wiley; Volume: 8; Issue: 3 Linguagem: Inglês

10.1002/mdc3.13172

ISSN

2330-1619

Autores

Francisco Cardoso,

Tópico(s)

Neurological and metabolic disorders

Resumo

The use of panels to diagnose presumably auto-immune movement disorders must be seen in the broader context of development of new diagnostic tests. As it has been clearly established in other fields of medicine, the indiscriminate use of new diagnostic tests increases the frequency of false alarms, overdiagnosis and overtreatment in a substantial number of individuals.1, 2 It is important to emphasize that the aim of this article is to tackle not the role of specific antibody-testing to study a clinical syndrome, but the use of panels. The latter involves testing of a multitude of antibodies regardless of the clinical suspicion. As it will be detailed in following paragraphs, panels lack specificity and sensitivity; panels are not cost-effective; and, perhaps more importantly, they go against the foundations of clinical medicine. Therefore, the only possible answer to the question of the title of the article is NO. Sensitivity is defined as the proportion of individuals with a disease who have a positive result for a given test. Specificity shows the proportion of individuals without a disease who have a negative result for a given test. Kim and Jeon rightly state that “the value of a test depends on how much its result can increase the probability of a condition compared to the pre-test value”.3 The crucial problem is that sensitivity and specificity are currently used based on post-test probability. This means that the numbers are derived from retrospective analysis of the performance of the test after the diagnosis has been defined. In this scenario, there is a bias of artificially inflating the value of the test. In clinical practice, the situation is different: as we have in front of us a patient whose diagnosis needs to be established, it is necessary to learn the pre-test (prospective) probability. This figure will inform us the likelihood of the occurrence of the suspected diagnosis. The pre-test probability is better dealt with Bayesian statistics, which, unfortunately are much less used in the literature. It is important to keep this limitation of the literature in mind when judging the diagnostic value of one given test. Needless to say, the greater the number of tests ordered, the greater the risk of biases if one employs post-test probability. To illustrate these problems in a more concrete way, I will discuss specific situations. Dahm et al.4 studied the seroprevalence of nine brain defined antigens in 1703 healthy controls and 2533 individuals with neurological diseases. Taking, for instance, the example of anti-GAD65, they have found that 9 (0.5%) controls and 13 (0.5%) patients tested positive for this antibody. Obviously, the test was not able to discriminate healthy individuals from subjects with neurological diseases. The argument brandished by the supporters of the use of panels is the lack of specific clinical features in the group of patients. Yet this means that tests, including antibodies, are only helpful if ordered in the context of well-defined clinical features. Later in this article, I will return to this issue. Another study illustrates even better the lack of sensitivity and specificity of panels. Lang and Prüss5 described the results of neuronal surface antibodies in the serum and spinal fluid of an impressive 743,299 samples. Sixty one of 26,430 (0.23%) healthy controls tested positive in serum. The positivity in 67,069 disease controls (who lacked any specific clinical findings) was just 1.5% (989). When the authors studied patients with clinical features loosely suggestive of an underlying autoimmune condition, the positivity grew to 21% (1015 of 4832). This is certainly an improvement but, yet it is still a disappointing sensitivity. In contrast, the positivity among patients with a high suspicion of an autoimmune condition jumped to 69.2% (2203 of 3185). Although the sensitivity is far from ideal, it is obvious that unless there is a good clinical characterization of patients, panels are of limited usefulness. One concrete example of lack of specificity of antineuronal antibodies is Sydenham chorea (SC), a well-established auto-immune condition related to infection with β-hemolytic Streptococcus.6 Several years ago, we reported that the tiers of anti-basal ganglia antibodies on ELISA were significantly higher than controls in 95% of individuals with SC.7 This number suggested a high specificity and sensitivity. However, subsequent studies showed that the anti-basal ganglia antibodies are not specific of SC.8, 9 The study of Brilot and colleagues is particularly illustrative of the problems with specificity and sensitivity of anti-basal ganglia antibodies. It is true that the titer of these antibodies in subjects with SC discriminate this group from healthy controls, other neurological conditions, Tourette syndrome and patients who met criteria for PANDAS. However, as all studied subjects had circulating IgG against basal ganglia, the specificity is really low. Moreover, among the patients with SC, the statistical difference was accounted for by approximately 50% of the individuals. This demonstrates that the sensitivity of anti-basal ganglia is low as well. The conclusion is that the diagnosis of SC remains exclusively based on clinical grounds.6 Opsoclonus-Myoclonus-Ataxia syndrome (OMAS) is an even better example of the limitations of the use of panels of anti-neuronal antibodies. This condition is characterized by a variable combination of opsoclonus with cerebellar ataxia and/or myoclonus. Armangué and colleagues10 reported on a large cohort of 114 individuals with OMAS. Forty-five of these subjects had an underlying neoplasm, defining a paraneoplastic form of OMAS (P-OMAS) while the remaining individuals had idiopathic OMAS (I-OMAS). Both groups underwent an investigation with comprehensive panel of antineuronal antibodies. Disappointingly, just 13 (11.4%) of the subjects with P-OMAS tested positive for onconeural antibodies: anti-Ri, anti-Ma2, anti-Zic4, and anti-CRMP5. Similar results were found in the I-OMAS group, with positivity in 12 (10.45%) of individuals: anti-GlyR, anti-GABAB-R, anti-NMDAR, and anti-DPPX. The conclusion is that the sensitivity and sensibility of panels of antibodies for the diagnosis of OMAS is low. Other authors have also found the same low positivity for different anti-neuronal antibodies, leading them to conclude that “as no specific biomarker is yet clearly linked to OMAS, the diagnosis is still essentially based on the clinical presentation of the patients”.11 Cost has always been a critical issue in medical care. Nowadays its importance has grown, having become a charged political issue all over the world regardless of the political and economic model of the countries. As clinicians do not remain isolated in an ivory tower, we need to be involved with economic aspects of practicing medicine. Table 1 brings costs of panels in different countries of the world. The conclusion is that panels of anti-neuronal antibodies are quite expensive. Although there are no published studies of the cost-effectiveness of panels, they certainly add a substantial economic burden on the clinical care of patients. There are also other aspects that need to be highlighted. One of them is the variability of the components of the panels. As there is no clear scientific reason behind this variability, it is likely related to economic interests of laboratories. Another interesting aspect is the difficulty in determining the price of these panels. During preparation of the talk that resulted in this article, it was not easy to unearth the cost of ordering anti-neuronal antibodies panels. This was particularly notable in countries with strong private health care system. One wonders if doctors are purposefully left in the dark about pricing in order not to have the chance to reflect on the economic burden related to ordering panels. The time-honored process of making a diagnosis was established in the late nineteenth century. It involves listening to patients, examining them, building a hypothesis and checking on it. The process is time consuming and, not infrequently, emotionally taxing. However, it is a crucial step to the creation and consolidation of a bond between clinician and patient. This bond is what will allow the doctor to have the trust of the patient and console her or him even in face of a tragic outcome. In contrast, panels offer the allure of bypassing this process. Not only this creates a fracture of the bond between patient and physician, but also it often leads to overdetection and overdiagnosis.2 These two unintended consequences are harmful to patients and health system because they often result in visits to health centers, ordering of even more tests, prescription of unnecessary drugs and other therapeutic measures, and a psychological toll on patients and care givers. Sadly, these problems are not restricted to the field of movement disorders. As extensively discussed elsewhere, medicine in general is under threat by the indiscriminate use of diagnostic tests.1 Another reason for the popularity of diagnostic panels is the siren song of being able to emulate Oedipus. One must not forget that after deciphering the riddle of the sphynx, he paid a tragic price for his unbridled thirst for knowledge.12 Such a Promethean risky desire to steal the fire of knowledge from the gods is commendable in those living in academia. In ordinary clinical practice, however, there are not rare moments when we must humbly accept the fact that we simply do not know. Other areas in neurology have dealt with the issue of the role of antibodies panels to establish diagnosis of presumably auto-immune conditions. In a recent comprehensive assessment of the diagnosis of seizures secondary to auto-immune process, the Oxford group decreed that “in the absence of other features of autoimmune encephalitis, antibody-positivity is of doubtful clinical significance”.13 Such a conclusion is based on the same reasons as we have discussed in this article: the sensitivity and specificity of these tests just become helpful if there is a proper characterization of the clinical features of the patients suggesting an underlying specific auto-immune etiology. Even in the field of movement disorders, there are other voices in this same direction. After studying a cohort of patients with movement disorders, a Japanese group concluded that “cell-surface antibodies should be determined based on individual phenotype”.14 More recently, in an astute editorial Dale and Rostásy remarked that “as clinicians test phenotypes with lower pretest probability (…) the chance of a false-positive result in serum becomes as likely as a true-positive result”.15 Of course, it is not the case of being against ordering antibodies testing. But as Dr. Bettina Balint and Professor Hans-Michael Meinck stated, “antibodies themselves are not disease-specific per se”.16 Therefore, all the arguments reviewed in this article support the notion that from scientific, clinical and economic points of view, ordering antibodies must be done exclusively in the context of well-defined clinical hypotheses. The indiscriminate use of antibodies panels is simply against all available evidence. This article is based on a talk given at the 2020 MDS Virtual Congress. Many thanks to colleagues throughout the world who helped me gathering information on panels as discussing some of the points described in this article: Joseph Jankovic (Houston, TX), Kapil Sethi (Augusta, GA), Steven Frucht (New York, NY), Tiago Mestre (Ottawa, Canada), Andrew Lees (London, UK), Werner Poewe (Innsbruck, Austria), Carlo Colosimo (Terni, Italy), Esther Cubo (Burgos, Spain), Joaquim Ferreira (Lisbon, Portugal), Hirohisa Watanabe (Nagoya, Japan), Yoshikazu Ugawa (Fukushima, Japan), Beomseok Jeon (Seoul, Korea), Marcelo Merello (Buenos Aires, Argentina), and Mayela Rodríguez-Violante (Mexico City, Mexico). (1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the first draft, B. Review and Critique. F.C.: 1A, 1B, 1C, 3A, 3B This manuscript refers to a critical reflection on a specific issue (namely, the role of use of antibody panels to diagnosis movement disorders). As such, it did not involve patients, thus not requiring ethical committee approval or patient consent. The author confirms that he has read the Journal's position on issues involved in ethical publication and affirms that this work is consistent with those guidelines. No specific funding was received for this work and the author has no conflicts of interest to declare. The author has no additional disclosures to report.

Referência(s)