Minimally invasive vs. trans-sternal thymectomy for non-thymomatous myasthenia gravis: a multi-institutional longitudinal study examining neurologic outcomes

Introduction

Myasthenia gravis (MG) is an autoimmune disorder with a prevalence of 60,000 in the United States (1). Its primary manifestation is muscle weakness, which can be specific or generalized (2), and results from production of antibodies which bind receptors in the neuromuscular junction at the post synaptic membrane (3). Medical therapies such as acetylcholinesterase inhibitors, corticosteroids, and other immunosuppressive agents have long been cornerstones of treatment, but these are not without adverse effects and potential risks (4).

While the mechanism is incompletely understood, it is believed that the thymus plays a key role in the pathogenesis of MG for patients with acetylcholine receptor autoantibodies (5). Radical thymectomy was first performed as a treatment for MG in the early part of the 20^th century (6), and more recently gaining traction to achieve symptom control in certain patients. Multiple systematic reviews and smaller trials have demonstrated the efficacy of thymectomy for MG patients, especially those who are younger, have confirmed acetylcholine receptor antibodies, and generalized symptoms (7-9). The most important of these trials, the MGTX (Randomized Trial of Thymectomy in Myasthenia Gravis), demonstrated a clear benefit of trans-sternal thymectomy for inducing disease remission, improving symptoms, and reducing the need for immunomodulatory therapies (9). An important and unanswered question, however, is whether similar results can be achieved with less invasive surgical approaches.

Minimally invasive thymectomy (MIT), via video-assisted thoracoscopic (VATS) or robotic approaches, is an attractive alternative to trans-sternal thymectomy. Aside from the benefits of smaller incisions, improved cosmesis, and avoidance of a sternotomy, MIT is associated with other benefits such as reduced length of hospital stay and decreased pain, particularly important in this population, as pain can precipitate myasthenic crisis (10). On the other hand, there has been theoretical concern that MIT may leave behind residual thymic tissue when compared to a trans-sternal approach, reducing the efficacy of MIT for MG patients (11). MIT has been studied in mixed patient populations, including those with thymic malignancies, such as thymoma, and has largely shown equivalent efficacy compared to open approaches. Indeed, for thymic tumors, MIT has become the most common surgical approach (12). In addition, for mixed indications, MIT has favorable short-term results (13-16). The few studies that have been done in patients with MG have been limited to small single-institution studies with surgical outcome follow-up only (17,18), limiting interpretation of results for neurologic outcomes. In a recent database study of over 1,700 patients, perioperative surgical outcomes were superior with MIT compared to trans-sternal approaches (12). But for these patients, the most important question remains: can MIT achieve similar long-term neurologic outcomes to trans-sternal thymectomy for patients with non-thymomatous MG?

The purpose of this study is to determine just that, by collecting disease-specific neurologic short- and long-term data for MIT, and comparing it to a trans-sternal approach for patients with non-thymomatous MG. We hypothesize that MIT provides quicker recovery and equivalent long-term disease control compared to trans-sternal thymectomy. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-423/rc).

Methods

Patient data

Patient data were obtained through retrospective review of patients undergoing surgery for non-thymomatous MG at multiple institutions. Data were extracted from a combination of institutional Society of Thoracic Surgeons databases and review of the clinical record. Surgical data were abstracted by surgical staff, and neurologic data were abstracted by a neurologist or dedicated research staff. De-identified data were faxed securely to the University of Virginia and then uploaded to a secure REDCap database. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board (IRB) at the University of Virginia on 4/24/2019 (#14542), and IRB approval was obtained at each participating site. Individual consent for this retrospective analysis was waived.

Patients who underwent thymectomy for non-thymomatous MG between 2012 and 2020 were included. Review of pathology records confirmed the absence of thymoma, so that only patients undergoing surgery for non-thymomatous MG were included. Patients were stratified by operative approach (trans-sternal vs. MIT). MIT approaches included both VATS and robotic approaches. Neurological clinical follow-up was recorded at 3-month intervals for up to 2 years postoperatively.

Other exclusion criteria included transcervical thymectomy, as this procedure is performed in very few centers and was not performed in any of our participating sites. Most importantly, we excluded sites that could not provide details on long-term neurologic follow-up. This reduced our original cohort of 23 academic medical centers to four participating sites.

Myasthenia outcomes

MG outcomes included glucocorticoid use and dosage, adjunctive immunomodulatory agents, myasthenia status, and two validated myasthenia instruments, the MG Composite and MG quality of life (QOL) measure. Myasthenia status was categorized on a nine-point scale with higher levels indicating worse disease control and scores of six or less indicating minimal manifestation status. This scoring system was put forth by the Task Force of the Medical Scientific Advisory Board of the Myasthenia Gravis Foundation of America in an effort to adopt a universally accepted classification system (19). The MG Composite is recommended by the MG Foundation to determine clinical response in a clinical trial setting and is scored on a range from 0 to 50 with higher scores indicating more severe symptoms. The MG-QOL15 measure is a survey which assesses patient QOL as it relates to myasthenia symptoms. It correlates with other MG-specific instruments and is scored on a scale of 0–60 with higher scores indicating more severe disease (20,21).

Statistical analysis

Univariate analysis was performed for continuous variables with median [interquartile range (IQR)] for skewed distributions and mean ± standard deviation for normally distributed variables, and for categorical variables with Chi-squared tests, respectively. Mixed effects modeling evaluated the impact of surgical approach on long-term myasthenia outcomes and prednisone dosing. Time-weighted means were calculated for key outcome variables stratified by operative approach. Statistical analysis was performed using SAS 9.4 (SAS Institute, Cary, NC, USA) using P less than 0.05 as the threshold for statistical significance.

Results

Baseline characteristic

A total of four United States institutions contributed data (Dartmouth University, Tufts University, University of California San Diego, and the University of Virginia). Fifty-four patients underwent thymectomy for non-thymomatous MG during the study period, with a median of 12 (IQR, 8.75–16.75) cases per center. A total of 29 thymectomies (53.7%) were performed minimally invasively with the remainder performed via sternotomy. Operative approach was at the discretion of the surgeon. Of the minimally invasive operations, 69% (20/29) were performed robotically, the remainder via a VATS approach. There were no major differences in baseline characteristics or comorbidities between groups except for some differences in racial distribution between the two groups (Table 1). Myasthenia severity was also largely similar between groups with equivalent median [MIT: 10 (IQR, 0–20) vs. sternotomy: 10 (IQR, 0–20) mg] and mean (11.4±15.9 vs. 9.7±9.5 mg, P=0.84) daily mg of prednisone used; similar functional myasthenia specific scores (7 and 6.16) and similar adjunctive medication usage prior to surgery (Table 2). For the group who underwent sternotomy, 15 of the 25 patients were on prednisone preoperatively, while four patients were on steroid-sparing immunosuppression regimens. In the minimally invasive group, 16 of the 29 patients were on prednisone preoperatively, while seven patients were on a steroid-free immunosuppression regimen. Of note, there were five patients in the minimally invasive group with ocular predominant symptoms. These patients were discussed in a multidisciplinary format, and the decision to proceed with surgery was unanimously agreed upon by the surgeon and neurologist. Overall, most patients at the time of surgery were classified as symptomatic on oral medications, with the next most common group receiving outpatient intravenous immunoglobulin (IVIG) or plasma exchange prior to surgery.

Operative outcomes

There was one conversion from a minimally invasive approach to sternotomy due to innominate vein bleeding. Operative time [MIT: 141 (IQR, 113–194) vs. sternotomy: 128 (IQR, 112–144) minutes, P=0.11] was similar, as was the rate of complications (Table 3) between groups. With minimally invasive surgery, patients were less likely to go to the intensive care unit (18.5% vs. 60%, P=0.002) and had a shorter hospital length of stay [2.5 (IQR, 1.5–3) vs. 5 (IQR, 4–5) days, P<0.001]. Specimen weight was similar between groups [45 (IQR, 30–81) vs. 40.5 (IQR, 31–77) g, P=0.96].

Follow-up myasthenia outcomes

Median duration of follow-up was 21 months (IQR, 9–24 months). At the most recent follow-up, 62% of patients were off prednisone with similar rates of freedom from glucocorticoids in both groups (64% vs. 60%, P=0.77). Average daily prednisone dose was similar between groups at each 3-month follow-up period. Time-weighted mean prednisone dose (8.2 vs. 7.3, P=0.57) and myasthenia status (3.6 vs. 3.8, P=0.83) were similar between groups. MG Composite and MG-QOL15 scores were similar at each time point with the exception of the MG Composite at 6 months, which was lower in the minimally invasive group [0 (IQR, 0–1) vs. 5.5 (IQR, 3–9), P=0.007]. At the most recent follow-up, there were no differences in the proportion of patients that had achieved minimal manifestation status or better (44% vs. 56%, P=0.72). During the 24-month study period, there was a trend towards decreasing prednisone dose and MG Composite scores over time (Figures 1,2).

Figure 1 Median prednisone dose over time in mg.

Figure 2 Median MG composite scores over time. MG, myasthenia gravis.

Discussion

This multi-institution longitudinal study of neurologic outcomes following trans-sternal vs. MIT for non-thymomatous MG demonstrated similar impact on MG regardless of operative approach. In addition, there were perioperative benefits, including fewer intensive care admissions and shorter hospital length of stay with the MIT approach. Both arms yielded significant improvements in outcomes during the 24-month follow-up period, substantiating the findings of the value of thymectomy in the MGTX trial. Importantly, there were no differences in prednisone dose or myasthenia-specific functional outcomes between surgical approaches, suggesting in our cohort of patients MIS thymectomy is an acceptable way to treat non-thymomatous MG.

The seminal MGTX study, demonstrating efficacy of thymectomy for non-thymomatous MG, evaluated time-weighted MG score and time-weighted prednisone dose as co-primary endpoints (9). For this reason, we evaluated similar end points in our study. Prednisone dosage in the MGTX trial was substantially higher than what we observed, both at baseline and in their time-weighted analyses. This is possibly due to different medication profiles as specified in the trial design compared to real-world practice. The treatment arm in the MGTX study saw a reduction in prednisone dose by approximately 45%. In our study, we saw a comparable decrease in both arms with a 35% reduction in prednisone dose overall and 62% of patients that were prednisone-free at last follow-up. These data suggest a preserved benefit from thymic resection outside of the trial setting and one that is agnostic to surgical approach.

One of the important strengths of this study was the longitudinal follow-up of myasthenia-specific outcomes as well as prednisone dosing over time. Few studies have evaluated the long-term impact of surgical approach on myasthenia outcomes and many of those that do combine outcomes for non-thymomatous patients and those with thymomas. In these studies, rates of complete stable remission (CSR) have generally been reported without specific myasthenia-related functional scores (11,15-26). Myasthenia-related quality outcomes were followed in our study both with MG Composite score and MG-QOL15 questionnaires at 3-month intervals. At each time point, except for the 6-month MG Composite, there were no differences in disease control between groups. The difference at the 6-month time point may reflect an earlier recovery from minimally invasive surgery, but given the observed difference in a single outcome measure at a single time point, it is also possible that this is due to multiple comparisons rather than a true difference in myasthenia-related functionality.

Previous reports have summarized neurologic outcomes with thymectomy by denoting patients with CSR at last follow-up. In contrast, this study examined prednisone dose and other myasthenia-related outcomes at multiple time intervals. CSR is generally defined in most studies as being symptom-free without medications for at least 1 year. The rates of CSR vary widely from study to study making, it challenging to compare results across studies. Prior studies have reported the CSR rates as high as 60% and 75% at 8 and 10 years, respectively, while others have reported CSR rates closer to 30% (22-27). Perhaps the differences across studies can be partially attributed to follow-up duration with studies that examined longer follow-up periods generally having higher rates of CSR. Another potential explanation is patient baseline disease severity is variable with lower disease severity in studies with higher rates of CSR. Rather than CSR, based on the 2016 Myasthenia Gravis Foundation of America 2016 treatment guidelines, the MGTX 5-year extension looked at the proportion of patients achieving minimal manifestation status. Eighty-eight percent of those treated with thymectomy achieving minimal manifestation status at 5 years (4,28). This had increased from 67% at 3 years, further supporting the notion that there are ongoing benefits of thymectomy over time. Our results were concordant with the MGTX study, with approximately half of all patients achieving minimal manifestation status at a median of 21 months by either surgical approach.

A concern that has been raised with minimally invasive approaches is that the radicality of the thymectomy may not be the same as with trans-sternal maximal thymectomy (9,11,19). In the MGTX study, trans-sternal thymectomy was chosen because “it provides reproducible resection of the maximal amount of thymic tissue” (9). While this is difficult to assess from a registry such as this, we evaluated specimen weights as a surrogate for the extent of thymectomy, as has been reported previously. The similar specimen weights between groups in our study support the equivalent extent of thymic resection regardless of incision.

Several prior studies have evaluated the perioperative outcomes of different approaches to thymectomy for various indications. The results of our study are largely consistent with the prior literature demonstrating slightly longer operative times with minimally invasive techniques but with fewer postoperative intensive care admissions and shorter duration of hospital stay. Interestingly, much of the data on hospital length of stay comes from outside of the United States, where the reported duration of stay is much longer than observed in our study, with a mean length of stay in a recent meta-analysis reported at 6.3 vs. 8.8 days for a minimally invasive and trans-sternal approach, respectively (29). This compares to 2.5 and 5 days in our study. We did not anticipate obtaining meaningful data from the estimated blood loss, so this was not recorded as in other studies, but the overall rate of blood transfusion was low, with just one case requiring a transfusion in the entire series. A recent Society of Thoracic Surgeons database study, including 1,725 patients, evaluated perioperative outcomes of different approaches to thymectomy (12). The main finding was a rapid adoption of robotic approach to thymectomy in the last decade. In addition, in a more limited subset of patients from that study that had adequate data (n=311), there was a 4.2% conversion rate. Despite this, there was a non-statistically significant reduction in odds of perioperative complication with a minimally invasive surgical approach. While this important study by Raja et al. (12) sheds light on some perioperative benefits of minimally invasive surgery, it does not address the effectiveness of the operative approach, which can only be confirmed with long-term follow-up of neurologic outcomes. Overall, our study and the previous body of literature suggest that perioperatively, both approaches are safe with shorter recovery after minimally invasive surgery. This is supported by a recent meta-analysis (30).

Our study is limited by its retrospective nature and the potential for selection bias. Moreover, given the retrospective nature, not all endpoints were available for every patient. In addition, our study was limited to a 2-year follow-up period (rather than 3 years in MGTX), and the benefits of the operation on disease control are likely to extend beyond this time frame. However, prior studies have demonstrated that the benefits begin prior to 2 years, and we anticipate this to be a sufficient duration of follow-up to observe changes. We also did not include any trans-cervical approaches to thymectomy, which is another minimally invasive approach which may also have similar rates of disease control, but is not performed in most centers. Finally, given the relatively small sample size, the study is at risk of type two error, and we may be underpowered to determine a small difference in approach. Unfortunately, the diagnosis of non-thymomatous MG in patients with refractory disease referred to surgery is relatively low. A prospective randomized control study would be ideal to truly compare MIT to trans-sternal thymectomy for non-thymomatous MG. The MGTX trial took 14 years to develop, enroll, and report its results in 126 subjects. The costs and time associated with conducting a similar trial to parse the role of surgical approach are likely prohibitive. There are no preexisting databases that collect both surgical and neurologic outcomes for this rare disease; thus, multi-institutional collaboration, as we pursued, was the most practical way to attempt to answer the question of whether the surgical approach to thymectomy makes a difference on the impact on MG treatment.

Conclusions

In a real-world, multicenter study evaluating disease control of MG, there is a decrease in glucocorticoid use following thymectomy for patients with non-thymomatous MG with either a trans-sternal or a minimally invasive surgical approach. These data support thymectomy for MG regardless of the surgical approach. While a randomized controlled trial would help definitively determine whether MIT is superior to the trans-sternal approach, given the time and resources required to execute the MGTX trial, it is likely not feasible to recreate a prospective randomized trial of surgical approaches for MG. Instead, based on these data, we suggest surgeons choose the approach with which they are most comfortable, tracking perioperative and long-term neurologic results to ensure benefit. Minimally invasive approaches do, however, provide additional benefits, including decreased length of stay, reduction of need for intensive care unit admission, avoidance of sternotomy, and better cosmesis.

Acknowledgments

This abstract was presented at the 102^nd Annual Meeting of the American Association for Thoracic Surgery, Boston, MA, USA, May 14, 2022. We dedicate this work to the late Ted Burns, MD (1966–2022, https://www.legacy.com/us/obituaries/dailyprogress/name/ted-burns-obituary?id=35487986), who led the MG program at UVA for much of his career. He was instrumental in establishing the MG Composite score system. He was an inspiration and provided comfort to many.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-423/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-423/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-423/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-423/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board (IRB) of the University of Virginia (#14542), and IRB approval was obtained at each participating site. Individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Phillips LH. The epidemiology of myasthenia gravis. Semin Neurol 2004;24:17-20. [Crossref] [PubMed]
2. Gilhus NE. Myasthenia Gravis. N Engl J Med 2016;375:2570-81. [Crossref] [PubMed]
3. Gilhus NE, Skeie GO, Romi F, et al. Myasthenia gravis - autoantibody characteristics and their implications for therapy. Nat Rev Neurol 2016;12:259-68. [Crossref] [PubMed]
4. Sanders DB, Wolfe GI, Benatar M, et al. International consensus guidance for management of myasthenia gravis: Executive summary. Neurology 2016;87:419-25. [Crossref] [PubMed]
5. Marx A, Pfister F, Schalke B, et al. The different roles of the thymus in the pathogenesis of the various myasthenia gravis subtypes. Autoimmun Rev 2013;12:875-84. [Crossref] [PubMed]
6. Blalock A, Harbey AM, Ford FR, et al. The treatment of myasthenia gravis by removal of the thymus gland preliminary report. JAMA 1941;117:1529-33.
7. Cataneo AJM, Felisberto G Jr, Cataneo DC. Thymectomy in nonthymomatous myasthenia gravis - systematic review and meta-analysis. Orphanet J Rare Dis 2018;13:99. [Crossref] [PubMed]
8. Cea G, Benatar M, Verdugo RJ, et al. Thymectomy for non-thymomatous myasthenia gravis. Cochrane Database Syst Rev 2013;2013:CD008111. [Crossref] [PubMed]
9. Wolfe GI, Kaminski HJ, Aban IB, et al. Randomized Trial of Thymectomy in Myasthenia Gravis. N Engl J Med 2016;375:511-22. [Crossref] [PubMed]
10. Wendell LC, Levine JM. Myasthenic crisis. Neurohospitalist 2011;1:16-22. [Crossref] [PubMed]
11. Jaretzki A 3rd. Thymectomy for myasthenia gravis: analysis of controversies--patient management. Neurologist 2003;9:77-92. [Crossref] [PubMed]
12. Raja SM, Guptill JT, McConnell A, et al. Perioperative Outcomes of Thymectomy in Myasthenia Gravis: A Thoracic Surgery Database Analysis. Ann Thorac Surg 2022;113:904-10. [Crossref] [PubMed]
13. Weksler B, Tavares J, Newhook TE, et al. Robot-assisted thymectomy is superior to transsternal thymectomy. Surg Endosc 2012;26:261-6. [Crossref] [PubMed]
14. Buentzel J, Straube C, Heinz J, et al. Thymectomy via open surgery or robotic video assisted thoracic surgery: Can a recommendation already be made?. Medicine (Baltimore) 2017;96:e7161. [Crossref] [PubMed]
15. Xie A, Tjahjono R, Phan K, et al. Video-assisted thoracoscopic surgery versus open thymectomy for thymoma: a systematic review. Ann Cardiothorac Surg 2015;4:495-508. [Crossref] [PubMed]
16. Gu Z, Chen C, Wang Y, et al. Video-assisted thoracoscopic surgery versus open surgery for Stage I thymic epithelial tumors: a propensity score-matched study. Eur J Cardiothoracic Surg 2018;54:1037-44. [Crossref] [PubMed]
17. Cakar F, Werner P, Augustin F, et al. A comparison of outcomes after robotic open extended thymectomy for myasthenia gravis. Eur J Cardiothorac Surg 2007;31:501-4; discussion 504-5. [Crossref] [PubMed]
18. Renaud S, Santelmo N, Renaud M, et al. Robotic-assisted thymectomy with Da Vinci II versus sternotomy in the surgical treatment of non-thymomatous myasthenia gravis: early results. Rev Neurol (Paris) 2013;169:30-6. [Crossref] [PubMed]
19. Jaretzki A 3rd, Barohn RJ, Ernstoff RM, et al. Myasthenia gravis: recommendations for clinical research standards. Task Force of the Medical Scientific Advisory Board of the Myasthenia Gravis Foundation of America. Ann Thorac Surg 2000;70:327-34. [Crossref] [PubMed]
20. Benatar M, Sanders DB, Burns TM, et al. Recommendations for myasthenia gravis clinical trials. Muscle Nerve 2012;45:909-17. [Crossref] [PubMed]
21. Burns TM, Sadjadi R, Utsugisawa K, et al. International clinimetric evaluation of the MG-QOL15, resulting in slight revision and subsequent validation of the MG-QOL15r. Muscle Nerve 2016;54:1015-22. [Crossref] [PubMed]
22. Siwachat S, Tantraworasin A, Lapisatepun W, et al. Comparative clinical outcomes after thymectomy for myasthenia gravis: Thoracoscopic versus trans-sternal approach. Asian J Surg 2018;41:77-85. [Crossref] [PubMed]
23. Rückert JC, Sobel HK, Göhring S, et al. Matched-pair comparison of three different approaches for thymectomy in myasthenia gravis. Surg Endosc 2003;17:711-5. [Crossref] [PubMed]
24. Lin MW, Chang YL, Huang PM, et al. Thymectomy for non-thymomatous myasthenia gravis: a comparison of surgical methods and analysis of prognostic factors. Eur J Cardiothorac Surg 2010;37:7-12. [Crossref] [PubMed]
25. Chen Z, Zuo J, Zou J, et al. Cellular immunity following video-assisted thoracoscopic and open resection for non-thymomatous myasthenia gravis. Eur J Cardiothorac Surg 2014;45:646-51. [Crossref] [PubMed]
26. Bagheri R, Boonstani R, Sadrizadeh A, et al. Thymectomy for Nonthymomatous Myasthenia Gravis: Comparison of Video-Assisted Thoracoscopic and Transsternal Thymectomy. Innovations (Phila) 2018;13:77-80. [Crossref] [PubMed]
27. Manlulu A, Lee TW, Wan I, et al. Video-assisted thoracic surgery thymectomy for nonthymomatous myasthenia gravis. Chest 2005;128:3454-60. [Crossref] [PubMed]
28. Wolfe GI, Kaminski HJ, Aban IB, et al. Long-term effect of thymectomy plus prednisone versus prednisone alone in patients with non-thymomatous myasthenia gravis: 2-year extension of the MGTX randomised trial. Lancet Neurol 2019;18:259-68. [Crossref] [PubMed]
29. Solis-Pazmino P, Baiu I, Lincango-Naranjo E, et al. Impact of the Surgical Approach to Thymectomy Upon Complete Stable Remission Rates in Myasthenia Gravis: A Meta-analysis. Neurology 2021;97:e357-68. [Crossref] [PubMed]
30. Lee Y, Samarasinghe Y, Patel J, et al. The short and long-term effects of open vs minimally invasive thymectomy in myasthenia gravis patients: a systematic review and meta-analysis. Surg Endosc 2023;37:3321-39. [Crossref] [PubMed]