In general, nocardia bacteria are opportunistic bacteria that mainly infect immunocompromised patients. Although MG is an autoimmune disease that requires immunosuppressive treatment, nocardia infection is rarely observed in these patients. We performed the literature review by searching the MEDLINE database using the key words “nocardia,” “nocardiosis, “and “myasthenia gravis.” We found that up until our search, only 7 case reports of MG patients with nocardiosis had been reported[3-9]. To improve understanding of the disease process, we summarized the clinical characteristics of these patients, and these are listed in table 1. Seven out of 8 patients had a good prognosis. All patients were treated with glucocorticoids and other immunosuppressants. Three patients had thymoma, and 1 patient had Good’s syndrome. There was no clear relationship between nocardia infection and MG or thymoma. However, patients with Good’s syndrome, who usually have thymoma and immunodeficiency, easily acquire opportunistic infections. Although our patient had low IgM levels, he had no thymoma and normal levels of lymphocyte subgroups, and a diagnosis of Good’s syndrome was therefore not considered. Repeated immunoglobulin tests showed that he had low levels of IgM and normal levels of IgG, IgA and lymphocyte subgroups. Selective immunoglobulin M deficiency (sIgMD) was considered in our patient and is a rare primary immunodeficiency disease characterized by low levels of IgM (below two standard deviations of the mean) in association with infection and normal levels of IgG and IgA[10]. In addition, our patient had the history of diabetes for many years. Therefore, immunosuppressant use, sIgMD and diabetes were the predisposing factors for nocardia infection in our patient. We concluded the data in table 1. The following factors may be preconditions for nocardia infection in MG patients. The primary factor was the use of immunomodulator and/or immunosuppressive agents, consistent with numerous previous studies[11, 12]. The second factor was age. We found that MG patients with nocardia infection were 49-79 years old, indicating that older patients with MG might be more susceptible to nocardiosis. This is consistent with the fact that older MG patients have a higher incidence of tumours, such as thymoma, and low immunity. The last factor is gender, as supported by our finding that 7 out of 8 MG patients with nocardia infection were male patients. In summary, older male MG patients being treated with immunomodulator and/or immunosuppressive agents might represent a population that is susceptible to nocardia infection.
The clinical manifestations of nocardia infections are very heterogeneous and nonspecific. Lung, brain and skin are the most commonly affected sites[13]. All 8 patients had lung lesions. The infection also involved the muscles, heart, and kidneys. In addition to lung, brain and skin lesions, our patient also had ocular lesions. To the best of our knowledge, this is the first case report of an MG patient with disseminated nocardiosis with ocular lesions.
Ocular tissue is an unusual site for disseminated nocardiosis, and ocular infection is generally diagnosed as local nocardiosis, with presents a keratitis or endophthalmitis resulting from ocular trauma or surgery [14]. Occasionally, ocular infection might also be caused by haematogenous dissemination via the choroidal circulation [15]. The ocular nocardia infection that occurred in our patient may have been caused by haematogenous spread because the patient did not have eye trauma or a history of surgery and he had no abnormal signs in his eyeball. The prognosis of ocular nocardiosis is generally poor. Blindness is a common consequence, and ophthalmectomy is performed in approximately 30% of these patients. For these reasons, regular ophthalmologic screening should be performed in patients with suspected disseminated nocardiosis [15, 16]. In our patient, the ocular lesion was located behind the left eyeball, which finally led to retinal detachment. After he received suitable treatment, the ocular lesion completely vanished, but his vision was not restored.
Due to the paucity of trials, there are no formal guidelines to direct drug choice and treatment duration in nocardiosis. Most clinicians agree that CNS nocardiosis warrants a long course of treatment, with 12 months commonly recommended [17]. Empirical treatment of disseminated nocardiosis usually involves three antibiotics, including imipenem or ceftriaxone, TMP-SMX, and amikacin. TMP-SMX is thought to be the cornerstone of treatment for nocardia infections and is also the drug of choice for cerebral nocardiosis due to its good penetration into the CNS. Other drugs, including meropenem, cefotaxime, minocycline, moxifloxacin, levofloxacin, linezolid, tigecycline, and amoxicillin/clavulanic acid, are also used for the treatment of these patients [12]. MG patients should be treated with the proper antibiotics because some antibiotics can aggravate the disease. Patients with nocardiosis often have an underlying autoimmune disease or are receiving immunosuppressive treatment. Therefore, a combination of antibiotics is recommended in the beginning, and a single drug can be maintained after the clinical symptoms are relieved [12]. Immunosuppressive therapy will increase the risk of infection and the difficulty of treating infection in patients with MG. The use of immunosuppressants in MG patients with infections is an important issue. By reviewing the literature[18] and combining our findings with our own clinical practice experience, we cautiously suggest that if the infection can be controlled, immunosuppressive therapy can be continued in MG patients. However, when an infection is hard to control with administration of the proper antibiotics and becomes life-threatening, physicians should reduce the dose of immunosuppressants or even stop it. We stopped the use of azathioprine and continued a tapered dose of methylprednisolone in our patient when he developed leukocytopenia.
Because of the nonspecific manifestations of nocardiosis, most patients with nocardia infection are not diagnosed in the early stage of the disease, and it normally takes from 42 days to 12 months after the appearance of symptoms to achieve a clear diagnosis [19, 20]; this causes a substantial physical and emotional burden in these patients. The prognosis of nocardiosis depends on the location and severity of the infection as well as the overall condition of the patient. The curable rate of pulmonary nocardiosis is approximately 90% with timely treatment, while the one-year mortality rate is high in patients with CNS involvement. The mortality rate is 10-fold higher in patients with solid organ transplantation from nocardiosis than in those without [11, 12]. None of the eight patients included in Table 1 had a history of organ transplantation. Although 4 of the 8 patients had CNS lesions, these lesions were relatively small, the symptoms were mild, and the diagnosis and treatment were timely, and this may have been the reasons for their good prognosis.
In conclusion, nocardiosis is a life-threatening infectious disease, and diagnosis in the early stage and appropriate antibiotic therapy are crucial to prognosis. Ocular involvement is rare in disseminated nocardiosis, and patients who are suspected to have disseminated nocardiosis should receive ophthalmologic screening. Neurological physicians should be aware of nocardia infection in MG patients.