Ocular versus generalized myasthenia gravis: a continuum associated with acetylcholine receptor antibody titers

The aim of this study was to evaluate clinical and serological differences between the ocular myasthenia gravis (oMG) and generalized MG (gMG). This study is a retrospective chart review, in which data was collected from patients fulﬁlling 2 of 3 diagnostic MG criteria (positive antibodies, evidence of neuromuscular transmission defect on neurophysiological examination, positive effect of pyridostigmine treatment). 350 patients were included and data concerning demographics and MG medical history were collected . Patients with oMG accounted for 15.7 % of the included patients. The two subgroups differed signiﬁcantly in oMG having a later age at onset, lower AChR antibody-titers, longer doctor-to-diagnosis delay and less intensive MG treatment. Additionally, patients with oMG were faster at reaching a well-controlled disease state. Thymus pathology, number of antibody-positive (95.9 % of gMG and 94.5 % of oMG), sex, number of other autoimmune diseases and delay before drug stability did not differ between oMG and gMG. In conclusion, oMG is presumably a milder form of gMG characterized by lower AChR antibody-titers, a milder phenotype, and a quicker response to a less aggressive treatment. But otherwise, oMG and gMG show very


Highlights
• More than 90% of both patient groups, oMG and gMG were antibody-positive • AChR antibody titer above 100 nmol/L seems to predict generalization • oMG patients responded better to treatment, achieving quicker symptom satisfaction • 1/3 (350) of all MG patients in Denmark was included in this chart review • No difference in thymus pathology was found between oMG and gMG patients

Introduction
Myasthenia Gravis (MG) is an autoimmune disease leading to activity-induced weakness in skeletal muscles [1].Worldwide, the prevalence of MG is 40-180 patients per million people [1].
MG can be divided into ocular MG (oMG), experiencing ptosis and/or diplopia exclusively, and generalized MG (gMG) with additional weakness in muscles of extremities, neck, face and pharynx.
It has generally been reported that 15%-20% of MG patients have a pure ocular presentation [1][2][3].Some clinicians view oMG and gMG as two separate disease entities and in support of this, only half of oMG have been reported to carry pathogenic MG-specific antibodies versus more than 90% of gMG patients [2,[4][5][6][7].The aim of this study is to compare characteristics of oMG and gMG systematically.

Study design
This retrospective chart review collected data from patients diagnosed with MG and followed at the Copenhagen Neuromuscular Center (CNMC), Denmark, hence all included were actively followed at out clinic.The Ethics Committee of Greater Copenhagen assessed that an ethics approval was not needed.The National Data Protection Agency approved the use of data.

Subjects
All patients with an MG diagnosis and an active association with CNMC were invited to participate.
Participants had to fulfill 2 of 3 diagnostic criteria (positive AChR, MuSK or LRP4 antibodies, evidence of neuromuscular transmission defect on neurophysiological examination, positive effect of pyridostigmine treatment).Participants received oral and written information and gave written consent.The study conforms with World Medical Association Declaration of Helsinki.Information from patients' first MG symptom (from 1982) until inclusion into the study was evaluated.

Data collection
Data were retrieved from the electronic patient records, physical record archives and an antibody database and were entered in a RedCap database.Due to the retrospective design, not all records contained sufficient information about every parameter evaluated in this study, thus variation in number of patients (n) regarding some parameters occurred.

Outcome measures
The parameters evaluated spanned the entire course of MG disease from first symptom until inclusion in this study and included: prevalence of oMG vs. gMG, thymus pathology, antibody profile (MuSK and LRP4 only tested if AChR negative) and titer (AChR and MuSK antibody tests are analyzed with radioimmune assays.LRP4 antibody tests are analyzed with cell-based assays), age at onset, sex, first symptom, other autoimmune diseases, diagnostic delay, examinations before MG diagnosis, MG medication and delay before well-controlled disease.
Sex in this research project was defined as the biological gender (male/female) based on chromosomal genotype and internal/external anatomy.

Statistical analysis:
Statistical analysis was performed using R, version 4.1.2and R studio, version 1.4.1717.
Descriptive analysis of characteristics of patients with oMG and gMG, were conducted with calculation of means for continuous variables, and proportions for categorical variables, unless indicated otherwise.Comparison of the two subgroups was assessed using chi 2 -test, fisher-test when the sample size was less than 5 and t-test when comparing means.
Except age at onset, the continuous variables did not follow a normal distribution, thus data were log transformed before comparative analysis, and transformed back for evaluation.
A p-value < 0.05 was considered statistically significant, with adjustment for multiple measures.

Results
350 of 407 patients accepted participation and fulfilled the diagnostic criteria.Of the 57 nonparticipators, 17 patients did not have sufficient information about diagnostic criteria, 8 did not fulfill the diagnostic criteria and were not further evaluated, 24 did not respond and 8 declined participation.

Subgroups
We defined oMG patients as patients only having pure ocular symptoms (diplopia or ptosis) at the date of inclusion.Of the included main cohort, 84.3% of the patients had gMG (n=295), leaving 15.7% with oMG (n=55).The included MG patients were diagnosed between 1954 and 2022 and 10 of the oMG patients were diagnosed within the recent two years.
The average follow-up period among 310 patients with sufficient information about dates (gMG=261, oMG=49), was 119.4 months (range: 2.3-537 months) from diagnosis to inclusion (oMG: 92 months, gMG: 124.5 months).15.2% of patients were followed for less than two years (12% gMG, 3.2% oMG).Among the oMG patients followed for less than two years (n=10), the first symptom appeared on average 30.7 months before inclusion, with 50% showing symptoms for more than two years.Of the five oMG patients having symptoms less than two years, the average initial positive AChR-ab titer was 11.5 nmol/L (range: 1.4-35.7 nmol/L).
Of the 17 patients excluded due to insufficient information regarding diagnosis, 47% were categorized as gMG (n=8) and 53% as oMG (n=9), which was a different distribution between the subtypes compared to the main cohort of included patients (p=0.0067).The excluded patients with a phenotype of oMG and gMG showed the same distribution between ab-positive and positive pyridostigmine responders however.

Differences between oMG and gMG
Mean age at MG onset was significantly lower for patients with gMG (51 years) vs. oMG (57 years) (table 1, figure 1b).
Mean value of the first as well as the average of all measured AChR-antibody titers was higher in the generalized vs. ocular group (table 1).None of the patients with oMG had an AChR-antibody titer at diagnosis of more than 100nmol/L compared to gMG patients who had titers of more than 700nmol/L.Sixty percent of the entire cohort had ocular manifestations at MG onset (55 patients with oMG and 154 with gMG).Among the 209 MG patients presenting with an ocular onset, 154 patients subsequently developed generalized symptoms.Therefore, the conversion rate from initial purely oMG to gMG was 73.7%.Approximately half of the generalized patients had ocular manifestations first.Diplopia was the most frequent symptom at onset (63.6%) in oMG patients (table 1).
Among the gMG patients who initially presented with oMG symptoms (n=154), the first generalized symptom was recorded by a specialist, on average, 259 days after diagnosis (n=114, range: -876-4857 days) and 757 days after initial symptom onset (n=115, range: 3-7372 days).
Medication was divided into six categories (figure 1a) illustrating the approach to treatment with increasing intensity on the X-axis [7].There was a significant difference between the treatment of patients with gMG and oMG.Patients with oMG constituted most patients receiving pyridostigmine as monotherapy and one steroid sparing agent.Thirty-four percent of patients with gMG were treated with IVIG/PLEX compared to 7% of patients with oMG.
Patients with gMG had a longer symptom-to-first-specialist delay (p=0.00037), while patients with oMG had a longer first-specialist-to-diagnosis delay (p=0.0012).
Furthermore oMG had a 100-day shorter delay vs. gMG before the specialist and patient considered the disease as well-controlled (table 1).Specialist referral pattern before diagnosis differed between gMG and oMG (table 1).Not surprisingly, almost double the percentage of patients with oMG visited an ophthalmologist compared to patients with gMG.
Of the ab-positive oMG patients, 16.4% had a first-measured-positive-titer 0.3-1 nmol/l compared to 4.4% of the gMG patients.We included gray zone patients as positive, with 5.6% of the antibody-positive oMG patients and 1.8% of the antibody-positive gMG patients falling within the grey zone.
A larger percentage of patients with gMG tended to have thymoma and thymus hyperplasia (p=0.053)than patients with oMG (table 1).Enlarged thymus on imaging was found in 28.5% (n=84) of gMG patients and 12.7% (n=7) of oMG patients.Thymectomy due to an enlarged thymus was performed in 95% of the gMG patients and 100% of the oMG patients.Histology showed that 13.6% of all patients with gMG had thymoma compared to 5.5% of all patients with oMG.
A direct comparison of thymoma incidence between the two subgroups indicated that there was no significant difference between oMG and gMG (p=0.145).Among the oMG patients, thymectomy was conducted on average 12 months after date of diagnosis (n=5, range; 2.3-23.2months).No significant differences were observed between the two subgroups in terms of the examinations conducted prior to diagnosis, the presence of co-existing autoimmune diseases, prednisolone treatment (including both accumulated dosage and treatment duration), prednisolone side effects, azathioprine treatment (including both accumulated dosage and treatment duration).The analysis of sex distribution (male and female) also revealed no difference between the subgroups (table 1).

Discussion of interesting findings
The included patients in this study constituted about one third of all MG patients in Denmark.The fraction of included patients with oMG and gMG agrees with the literature [1][2][3].
The main findings of the study are the equal occurrence of AChR-antibodies in oMG and gMG and that AChR-titers were significantly lower in oMG patients.The findings suggest that an AChR titer above 100 nmol/L at diagnosis predicts generalization of the myasthenia, which is consistent with previous findings of a positive correlation between AChR-antibody-titers and clinical severity in a general MG population [8,9].These findings suggest that the first ab-titer (before intervention) could aid the diagnostic process and prediction of subtype at MG onset.This study found that more than 90% of both subgroups were antibody positive for AChR, MuSK or LRP4.For oMG, this is considerably higher than previously reported [2,4,5,[10][11][12].However, the discrepancy might partly be explained by better testing methods or higher testing rate than in the past.Another consideration is whether the MG patients, and especially oMG patients, classified as antibody-positive had titers close to the cut-off values.We found that a higher proportion of oMG patients (16.4%) had a first-measured positive AChR-ab titer of 0.3-1 nmol/L, which is near the cutoff value of 0.3-0.5 nmol/L, compared to 4.4% of gMG patients.Additionally, 5.6% of the abpositive oMG patients and 1.8% of the gMG patients fell within the grey zone (0.3-0.5 nmol/L).This finding supports the observation that oMG patients typically exhibit lower antibody titers, yet still display a higher-than-expected rate of antibody positivity.Given the small percentage of patients in the grey zone, it can be assumed that their influence on the overall results is minimal.It is also notable that a few patients initially tested negative for AChR antibodies before later converting to positive titers.While LRP4 antibodies may not be as specific as other antibodies, only two oMG patients and one gMG patient in our study were LRP4 positive.This finding aligns with the observation that more double seronegative MG patients present with ocular symptoms [1].Furthermore, the three patients with only LRP4 positivity all exhibited a positive response to pyridostigmine and had an abnormal EMG examination.Therefore, the detection of LRP4 antibodies was not critical for these patients to fulfill two out of three diagnostic criteria and be included in this study.The findings further suggest that a more favorable treatment response is associated with a speedier route to satisfaction according to when the patient and the doctor considered the disease as wellcontrolled (table 1).
The finding of no difference in thymus pathology between the two subgroups disagrees with other reports, who have suggested that thymoma is less frequent in oMG, than found in this study [4,11,13].A relatively small sample size of 55 oMG patients could explain this difference.
Other results demonstrated that the sex distribution did not differ between the two subgroups, which disagrees with the data from our study.A review from 2014 [6] and a chart review from 2020 [11] reported a higher frequency of men with oMG compared to gMG, while we found an equal distribution, indicating that sex did not influence the examined parameters.
Patients with gMG had a longer latency to visit first specialist.A possible explanation might be, that generalized weakness can be interpreted as the result of age or other diseases.The latency from specialist to diagnosis, on the other hand was longer for oMG, which seemed to be caused by a delayed referral from the ophthalmologist to the neurologist.
Limitations of a retrospective study like ours are inevitable.Ten patients were diagnosed within two years, and their MG disease and management may not have fully stabilized.Other studies have argued that only patients who have been ocular for at least two years are considered 'true' ocular (82% of the patients with oMG in this study) [1][2][3][4].Among the oMG patients who were followed for less than two years (n=10), 50% showed symptoms for more than two years.Of the five oMG patients who had symptoms for less than two years, the average initial positive AChR-ab titer was 11.5 nmol/L.This suggests a lower risk of conversion to gMG for those with symptoms for more than two years prior to inclusion, thus minimizing the influence of the number of oMG patients and the rate of ab-positive oMG patients.The last five patients (9% of all oMG cases) remained in a grey area of uncertainty.However, the range of the initial positive AChR-ab titer of these five oMG patients (1.4-35.7 nmol/L) was far from 100 nmol/L, thereby not predicting generalization (as suggested by findings earlier in this study).Of all MG patients in this study (n=350), 44% (n=154) of those who were initially oMG progressed to generalized MG, which agrees with other reports [1,4]

Conclusion
In conclusion, both patient groups with oMG and gMG are more than 90% antibody-positive and additionally, showed no difference in thymus pathology.oMG is likely a milder form of MG characterized by lower AChR ab-titers, a milder phenotype and a quicker response to a less aggressive treatment.Furthermore, oMG patients were 6 years older at disease onset and had a longer delay from first visit at a specialist to time of diagnosis

Declaration of interests
☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
☒ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Nanna Witting reports financial grant was provided by UCB Biopharma SRL John Vissing reports a relationship with UCB Biopharma SRL that includes: consulting or advisory, speaking fees, and travel reimbursement.
If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The proposition that oMG represents a less severe variant of gMG, aligns with the observations indicating that oMG patients exhibit a more favorable response to less aggressive therapeutic interventions.More patients with oMG only needed first-and second-line drugs (figure1.a),thus obviating the necessity for intensify therapeutic interventions (using intravenous immunoglobulin therapy or plasma exchange) in comparison to patients with gMG.

Figure 1 .
Figure 1.MG medication and age at disease onset

Table 1 .
Differences and similarities between the subgroupsTable showing parameters evaluated in this study.All values are given in percentages (count) or means (95%CI, count) unless other indicated.† =p values are given in chi 2 , fisher-test or t-test.‡ =not calculated.¶ =days from first symptom to diagnosis.§ =Satisfaction was defined as days from first visit at a neuromuscular expert to 1) the patient obtained drug stability (no change in medication done or planned within three months), 2) well-controlled disease according to the doctor and 3) well-controlled disease according to the patient.Well-controlled according to the doctor was defined as the first visit, where no MG symptoms or signs were noted.Well-controlled according to the patient was defined as the first visit, where the patient appeared satisfied with