Seroconversion of anti-aquaporin-4 antibody in NMO spectrum disorder: a case report

Abstract

Dear Sirs,Anti-aquaporin-4 (AQP4) antibodies are known to be pres-ent both in patients with neuromyelitis optica (NMO) andNMO spectrum disorders (NMOsd), some limited forms ofNMO [1, 2]. The antibodies are considered to have animportant role in the pathogenesis of those disorders [3, 4].We had a patient with NMOsd who had acute longitudinallyextensive myelitis. Her serum anti-AQP4 antibody waspositive at the time myelitis developed, but was negative 8and 9 years before myelitis onset and seroconverted 3 yearsbefore the onset. Her case provides insight into when andhow anti-AQP4 antibody is produced.A 55-year-old Japanese woman developed numbnessand dysesthesia on April 10, 2010 which thereafter exten-ded symmetrically below the eighth cervical segment andcaused slight staggering gait in the dark. Spinal cord MRIT2-weighted images revealed longitudinally extensivemyelitis from the C4 to T2 vertebral segments with gado-linium-enhanced lesion from C7 to T1. She was referred tous, and admitted on May 18, 2010.Actually, she had visited us regularly for the follow-up ofanti-acetylcholine receptor (AChR) antibody-positivemyasthenia gravis (MG) since 1994. She had been treatedwith oral high-dose prednisolone from January 1995. InAugust 1995, she underwent extended thymectomy andhistological examination had revealed hyperplasia. Thosetreatmentshadrelievedallneurologicaldeficits by7 monthsafter surgery. From December 1998, the prednisolone dosewas maintained at 5 mg/day every second day with tempo-rary cessations. By May 2002, prednisolone therapy hadended, and she remained symptom-free.On her present admission, neurological examinationrevealed hypopallesthesia and hyperreflexia in her lowerlimbs, and mild weakness in her left adductor digiti miniand bilateral iliopsoas muscles. Cerebrospinal fluid exam-ination showed a normal cell count and protein levels witha negative oligoclonal band. Brain MRI detected noabnormality consistent with Barkhof’s criteria [5]. Anti-AQP4 antibodies, measured by enzyme-linked immuno-sorbent assays [6] in a serum sample obtained on May 18,2010 were positive. Our diagnosis was NMOsd.Serum samples had been stocked for MG researchbefore the onset of NMO. After receiving her informedconsent, we measured the presence of anti-AQP4 anti-bodies in her stocked serum samples. Figure 1 shows thelongitudinal change in the anti-AQP4 and anti-AChRantibodies before and after NMOsd onset. Anti-AQP4antibody remained negative 8 and 9 years before the onsetof longitudinally extensive myelitis and was seroconverted3 years before myelitis onset. Her anti-AQP4 antibody titrehad increased, and myelitis emerged.Recently, the case of a patient with NMO whose serumhad remained positive for anti-AQP4 antibody for morethan 10 years before onset was reported [7], but it is notclear whether all patients with NMO have anti-AQP4antibody ten or more years before onset. Our patient’s caseshows that seroconversion is possible and that elevation ofanti-AQP4 antibody does not cause NMO onset. Onset mayrequire long-term elevation of anti-AQP4 antibody or someother mechanism.We previously reported two patients with MG whopresented with NMO after thymectomy [8]. In that report,