Generalized development is a typical feature for MG patients who initially present with purely ocular symptoms, and 90% of OMG patients might progress to GMG within the first 2–3 years of disease course [1,2]. A lot of studies accounted several factors including onset age, AChR Ab titers, positive rate of RNS, thymoma and immunosuppressive therapy were associated with GMG conversion. As the only controlled factor, immunosuppressive treatments have been applied early for OMG patients intending to achieve the decline of generalization risk [2,4,12]. While some studies found small partial of OMG patients still converted to GMG even under sufficient immunosuppressive treatments [13]. Consequently, randomized controlled trials or clinical real-world studies are yet needed to clarify whether immunosuppressive therapy could avoid or delay GMG conversion. In the our reviewed study, all OMG patients received prednisone and/or other immunosuppressants.
Contrast to 90% of conversion rate without immunosuppressive treatments, only 38 patients (17.0%) of 223 OMG patients under sufficient immunosuppressive treatments converted to GMG in the present study. This result was consistent with previous studies which compared the conversion rate of the immunosuppressed group and the non-immunosuppressed group [6,7,13,14]. Yet the predictors for these patients who received early immunosuppressive therapy still remain unknown.
MG is a heterogeneous disease influenced by ethnicity, gender, age of onset, disease duration and other factors. In our research, age of onset, disease duration, positive facial nerve and axillary nerve RNS were deemed as potential predictors for GMG conversion through univariate logistic regression analysis except gender or thymic abnormalities.
Even males were more susceptible to OMG, demographic features of all 223 patients indicated no statistical difference of incidence between males and females, which was consisted with other previous studies considering age bias [15,16]. For the 38 generalized patients, gender disparity was not existed either. And our study also suggested gender was not a predictive factor for MG generalization by Kaplan-Meier curve analysis for the patients under immunosuppressant treatment.
According to previous studies, age of onset was considered to be a risk factor of GMG conversion. It was reported the conversion rate of juvenile-onset OMG patients fluctuated between 23% and 43% [3,17], while that of adult-onset OMG patients fluctuated between 31% and 49% [4,18]. Kamarajah and Wang suggested late-onset were positively related with generalization [19,20]. Mark et al [2] also reported patients were inclined to converted to GMG with increasing onset age. We declared age of onset could predict higher conversion risk even for patients under immunosuppressed therapy. As is reported in our research, median onset age was 37.1y of 38 patients who converted to GMG by the follow-up. Therefore, patients in our study were divided into juvenile-onset OMG and adult-onset OMG subgroups to analyze the discrepancy further. We found the number of juvenile-onset OMG patients was comparable to that of adult-onset OMG patients. And only positive rate of axillary nerve RNS was found significantly different between the two groups except gender, initial symptoms, positive rate of neostigmine test or positive rate of facial nerve RNS. The positive rate of axillary nerve RNS was higher in adult-onset OMG patients which may interpret the difference between the two age spectrums.
OMG patients may have abnormal RNS results [21], while the positive rate was not high. Whether positive RNS result could predict GMG conversion has not yet been completely elucidated. Previous reports indicated positive RNS could predict higher conversion rate [6,7]. Whereas other studies did not support this [1,4,6,22]. In the present study, in spite of immunotherapy, abnormality of facial nerve or axillary nerve RNS both turns out to be predictors of higher GMG conversion risk, and is demonstrated by Kaplan-Meier curve analysis. In addition, the number of patients under ulnar nerve RNS examination was limited, therefore the predictive value of that needs further study.
Some scholars indicated positive AChR-Ab, abnormal RNS tests and abnormal SFEMG tests in the early stages of disease could predict more GMG conversion [9,23,24]. While other researches pointed out these factors could not be predictors [8,25].Due to restrictions of earlier objective condition, we did not carry out the AChR-Ab titer detection. In further study, we will take it into consideration. What’s more, some researches pointed out thymic abnormality, especially thymoma could be a predictor. While, through multivariate logistic regression analysis, thymic abnormalities was excluded as a potential predictor for GMG conversion which might be related with early immunotherapy.
Reports deemed that OMG patients have lower conversion rates with longer disease duration [13,24,26]. In our study, disease duration came out of multivariate logistic regression analysis as the predictors of GMG conversion. But few focused on the factors influencing conversion interval [26,27]. Focus on the 38 generalized patients, our study analyzed these factors further and confirmed the conversion tempo was independent of gender, similarly with other studies [2]. Although abnormal facial nerve or axillary nerve RNS could indicate higher conversion risk, they had no effect on disease duration. While, with more older the age of onset, the earlier conversion would be. In summary, age of onset could be a key element in MG development both on the aspects of conversion risk and tempo.
Overall, onset age, disease duration and positive facial nerve RNS were indicated as the predictors of GMG conversion by multivariate logistic regression analysis.There was also a trend toward increased conversion rate with positive axillary nerve RNS by Kaplan-Meier curve analysis and univariate logistic regression analysis. It might be a predictor with enlarged sample size and prolonged follow-up time. The results we figured out gave us a hint that immunosuppressive agents likely reduced the conversion rate but they still could not avoid MG generalization for the above predictive factors.