3.1 Patient characteristics
One hundred twelve JMG patients (42 males, 70 females, male-to-female ratio 0.6) were
identified (Supplementary Table 1). The average age of onset was 8.5 years (range
0-20 years). The peak age of onset was biphasic: one peak was before the age of 5
and the other peak occurred during adolescence.
The most common symptoms were ptosis in 108 (96.4%) patients followed by diurnal symptom
change in 61 (54.5%), muscle weakness in 24 (21.4%), double vision in 23 (20.5%),
bulbar palsy in 18 (16.1%), respiratory symptoms in 10 (8.9%), head deviation in 3
(2.7%), ophthalmoplegia in 2 (1.8%), and facial palsy in 1 (0.9%) patient.
Clinical classification of the disease severity at onset was evaluated as class I
in 88 (78.6%), class II in 13 (11.64%), class III in 3 (2.7%), class IV in 4 (3.6%)
and class V in 4 (3.6%) patients. The age of onset for ocular type MG (n= 88) was
younger than that for systemic type (n= 24) (7.29 ± 6.17 vs. 13.07 ± 4.71, P=0.001).
Four patients had a family history of MG. Three patients had a family history of thyroid
diseases. Almost all patients were Asian, except one patient was Caucasian.
Serologic tests were performed in 107 (95.5%) patients. Peak anti-acetylcholine receptor
antibodies were ≥ 0.5 nmol/L in 60 (56%), 0.2-0.5 nmol/L in 11 (10%), and < 0.2 nmol/L
in 36 patients (34%). One seronegative patient had autoantibodies against MuSK and
presented with systemic type and bulbar palsy.
Repetitive stimulation tests were performed in 50 patients, and 35 patients exhibited
a typical decremental change. No correlation was observed between the anti-acetylcholine
receptor value and repetitive stimulation test results.
Two patients had presence of thymoma. The peak serology status of anti-acetylcholine
receptor antibodies was 49 and 5.47 each. After excluding them, 53 patients had done
chest computed tomography. 33 showed positive thymus hyperplasia, and 22 were negative.
No correlation was observed between the anti-acetylcholine receptor value and thymus
hyperplasia (Supplementary Table 2).
3.2 Outcome at two years and last outpatient clinic follow-up
After excluding patients without serum antibody data, one Caucasian, and one patient
positive for anti-MuSK antibodies, a total of 54 patients followed up for over two
years (range 24 to 258 months, mean 112 months) (Fig. 1). We compared the demographic
data of JMG patients with different levels of anti-acetylcholine receptor antibodies,
and no difference was observed (Table 1). In total, 42 patients (80%) were ocular
type and 12 patients (20%) were generalized type. Nine (9/54, 16.7%) achieved complete
remission without medication use at two years after diagnosis. Thirteen (24.1%) patients
achieved complete remission during later follow-up periods. Thirty-two (59.3%) patients
had improved symptoms but were maintained on oral medication, 5 (9.3%) patients were
maintained on a high dosage of immunosuppressant, and 4 (7.3%) patients had worse
symptoms. Fifty-two of 106 (49%) patients had incomplete follow-up over two years.
No gender, age, disease severity, and intervention differences were observed between
patients followed up for two years and those without.
3.3 Outcome versus anti-acetylcholine receptor antibody levels
We compared two-year outcomes among 54 JMG patients using receiver operating characteristic
curve (ROC) curves with anti-acetylcholine receptor antibodies at different levels.
ROC curves of primary and secondary outcomes are presented in Fig. 2 and 3. Analyses
revealed that a serum cut-off level of 0.2 nmol/L serves as an optimal predictive
value at two years after diagnosis (sensitivity: 80%, specificity: 66%, PPV: 92%,
NPV: 40%) (Table 2) and at last outpatient clinic (sensitivity: 83%, specificity:
57%, PPV: 85%, NPV: 53%) (Table 3). JMG patients with negative anti-acetylcholine receptor antibodies (< 0.2
nmol/L) were significantly more likely to achieve complete remission compared with
those with antibodies ≥ 0.2 nmol/L at two years (6/15 [40%] vs. 3/39 [7.7%], 95% confidence
interval [CI] 1.670 to 38.323) (Table 2, Fig. 4). In addition, JMG patients with anti-acetylcholine
receptor antibodies less than 0.2 nmol/L exhibited a higher rate of complete remission
compared with those with antibody > 0.2 nmol/L at last out-patient visit (8/15 [53.3%]
vs. 5/39 [12.8%], 95% CI 2.367 to 20.704) (Table 3, Fig. 4). After adjustment for age, gender, classification and thymectomy operation, the clinical outcome exhibited a significant difference if we used a serum level of
0.2 nmol/L as a cut-off (Table 4).
3.4 Clinical fluctuations
Four boys (aged 0.5 years, 1.5 years, 2.5 years and 7.3 years) with initial anti-acetylcholine
levels less than 0.2 nmol/L became positive later (0.25 nmol/L, 12.5 nmol/L, 2.19
nmol/L and 11.1 nmol/L). The occurrence of seroconversion was detected 11, 19, 107,
and 85 months after clinical onset, respectively. One of the 4 patients achieved complete
remission at 2 years but developed symptom relapse, and the other 3 were maintained
on oral medication throughout the study course.
3.5 Comorbidity
Comorbid diseases included Graves’ disease in 11 (9.8%) patients, Hashimoto disease
in 2 (1.8%), thymoma in 2 (1.8%), systemic lupus erythematous in 1 (0.9%), and rheumatic
arthritis in 1 (0.9%) patient.
We analyzed the characteristics of 13 patients with comorbid autoimmune thyroid disease.
Nine (69%) patients were females, and the mean onset age was older than JMG patients
without comorbid autoimmune thyroid disease (11.8 ± 5.8 years old vs. 8.0 ± 6.3 years
old, 95% CI 0.018 to 7.33). Ten (77%) patients had ocular type MG. Two had a family
history of hyperthyroidism, and two had a family history of MG. In 10 (77%) patients,
comorbid autoimmune thyroid diseases were diagnosed at or after the JMG diagnosis
was made. Patients with anti-acetylcholine antibody serum levels < 0.2 nmol/L were
less likely to have autoimmune thyroid disease (1/35 [2.9%] vs. 12/71 [16.9%], 95%
CI 0.018 to 1.161). No correlation found using a cut-off level of 0.5 nmol/L.
3.6 Pharmacologic and surgical treatment
Forty-six of 112 (41.1%) patients received both cholinesterase inhibitors and steroids,
35 (31.3%) patients received cholinesterase inhibitors only, 5 (4.5%) patients received
steroids only, and 8 (7.1%) patients did not receive any medication.
Eighteen of 112 (11%) patients received thymectomy, and 16 (16/18, 90%) were female.
Patients manifesting with initial systemic MG were more likely to require thymectomy
than those with initial ocular MG (12/24 [50%] vs. 6/88 [6.8%], 95% CI 4.31 to 43.49).
No difference in the primary and secondary outcomes was observed between patients
who received both or either cholinesterase inhibitors and steroids or who underwent
thymectomy.