Advanced Masaoka-Koga stage as an independent predictor of thymoma recurrence after complete resection: myasthenia gravis was not an independent factor

Introduction

Thymoma is a typically slow-growing anterior mediastinal tumor that often remains clinically silent and is frequently identified incidentally during radiologic evaluations or in the context of neuromuscular symptoms attributable to myasthenia gravis (MG). The association between thymoma and MG is well-documented, with thymoma present in approximately 10–15% of patients with MG, and conversely, MG observed in 30–50% of patients diagnosed with thymoma (1,2). Despite this association, the impact of MG on the prognosis of thymoma—particularly regarding recurrence—remains controversial.

Previous studies have proposed various predictors of thymoma recurrence, including World Health Organization (WHO) histologic subtype, Masaoka-Koga stage, completeness of resection, and tumor size (3). However, the prognostic value of MG status remains unclear, and its independent association with recurrence after surgery has yet to be established.

In this retrospective cohort study, we aimed to evaluate clinicopathologic differences between thymoma patients with and without MG, and to identify factors independently associated with thymoma recurrence after surgical resection by comprehensively analyzing established prognostic variables, including MG status, tumor stage, tumor size, histologic type, and extent of resection. We present this article in accordance with the STROBE reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2025-401/rc).

Methods

Subjects

We retrospectively reviewed the clinical records of adult patients (age >19 years old) who underwent thymectomy between January 2000 and December 2023 at Gil Medical Center, Gachon University College of Medicine. We included only thymoma (thymic epithelial tumor) after excluding thymic carcinoma, thymic hyperplasia, thymic cyst, and other mediastinal mass (such as teratoma, neuroendocrine tumor, myxoma, etc.) by pathologic confirmation. Micronodular thymomas with lymphoid stroma (n=3) were excluded due to their distinct histologic behavior and lack of association with MG, making them unsuitable for comparative analysis (4).

We classified patients into two groups: thymoma with MG and those without MG. Patients with MG were required to have typical clinical symptoms presenting fluctuating muscle weakness with fatigability and result of serum anti-AChR antibody by radioimmunoassay (considered positive when ≥0.5 nmol/L) with or without decremental response in Jolly test. Patients were classified into the MG group if they had a diagnosis of MG at any point during follow-up, regardless of the timing of onset relative to thymectomy. All patients should be followed at least 12 months.

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the Gil Medical Center (IRB No. GDIRB2024–154). Informed consent was waived in this retrospective study.

Data collection

Clinical data included age at MG symptom onset, age at thymectomy, sex, treatment, MG Foundation of America clinical classification (MGFA class) at initial visit, and MGFA postinterventional status at the latest visit. Surgical characteristics included surgical approach (transsternal vs. video-assisted thoracoscopy), type of resection [thymectomy (extended or total) vs. thymomectomy], completeness of tumor resection categorized as R0 (microscopically complete resection with negative margins), R1 (microscopic residual tumor), or R2 (macroscopic residual tumor) as determined by pathological assessment (5), tumor size, WHO classification of thymoma pathology (types A, AB, B1, B2, B3), and Masaoka-Koga stage (6). Additionally, information on MG perioperative management (e.g., steroid or immunosuppressive therapy, use of plasmapheresis or intravenous immunoglobulin), postoperative adjuvant radiotherapy or chemotherapy, thymoma recurrence, and mortality rate was collected. Radiotherapy was typically administered in patients with Masaoka-Koga stage II or higher, or based on the surgeon’s judgment.

Statistical analysis

Continuous variables were presented as mean ± standard deviation (SD) if normally distributed or as median with interquartile range (IQR) if non-normally distributed. Between-group comparisons were performed using the Mann-Whitney U test for continuous or ordinal variables and the Chi-squared test for categorical variables.

Univariable logistic regression analysis was first performed to evaluate potential predictors of thymoma recurrence. Subsequently, multivariable logistic regression analysis was conducted using the enter method, including variables considered clinically relevant or previously reported as potential predictors (MG status, Masaoka-Koga stage, WHO classification, completeness of resection, surgical approach, adjuvant treatment, tumor size, age at onset, and sex). Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. Adjuvant treatment was coded as a binary variable (administered vs. not administered) without distinguishing between radiotherapy, chemotherapy, or combined modalities. Completeness of resection was categorized as R0 versus R1–2 for statistical analyses. Complete case analysis was applied for missing data.

Time-to-event outcomes were analyzed using Kaplan-Meier methods. Recurrence-free survival (RFS) was assessed in the R0 resection subgroup, while progression-free survival (PFS) and overall survival (OS) in the total cohort were also estimated. Survival curves were compared using the log-rank test, and the numbers of patients at risk were displayed below each curve.

Independent predictors of recurrence were evaluated using Cox proportional hazards regression restricted to patients with R0 resection. Prespecified clinically relevant covariates included age, sex, MG status, tumor size, Masaoka-Koga stage, surgical approach, and adjuvant therapy. To avoid model overfitting given the limited number of events, covariates were restricted relative to the number of events, and proportional hazards assumptions were tested.

All statistical analyses were performed using SPSS Statistics, version 21.0 (IBM Corp., Armonk, NY, USA).

Results

Among 213 patients with thymic mass, forty-five patients were in the thymoma with MG group and 168 patients were in those without MG group. Total 79 patients met all the inclusion criteria (Figure S1). Of the 79 patients, 32 were classified as having thymoma with MG, including 4 who developed MG after thymectomy. The remaining 47 patients had no clinical evidence of MG throughout the follow-up period.

In the total cohort (n=79), the median age at thymectomy was 55 years (IQR, 47–62 years), and 44 patients (55.7%) were female. Histologically, B1–B3 thymomas accounted for 70.9% of cases. Advanced Masaoka-Koga stage (stage III–IV) was observed in 19 patients (24.1%). Complete resection (R0) was achieved in 64 patients (81.0%), while 8 (10.1%) and 7 (8.9%) underwent R1 and R2 resection, respectively. Among patients with advanced stage disease, R0 resection was achieved in 14 patients (73.7%), whereas 5 patients (26.3%) underwent R1/2 resection. Among the 64 patients who R0 resection, 12 (18.8%) developed recurrence during follow-up, of whom 11 had regional recurrence and 1 had distant recurrence.

When comparing thymoma patients with or without MG (Table 1), patients with thymoma with MG showed younger age at onset (median: 50 years), longer follow-up duration, and higher rate of transsternal extended thymectomy (96.9%), advance Masaoka-Koga stage (37.5%), thymoma recurrence (37.5%) and death (25.0%). Type B histology was more frequent in patients with thymoma with MG (81.2%) compared to those without MG (63.8%), with a trend toward statistical significance (P=0.07). The pattern of thymoma recurrence did not differ significantly between thymoma patients with and without MG.

Among patients with thymoma with MG, the majority presented with ocular symptoms (MGFA class I, 67.9%), while 32.1% had generalized symptoms (class II–III) before thymectomy. At the last follow-up, 65.6% remained in class I and 90.6% showed improved or stable status, although 28.1% experienced MG crisis requiring intubation after thymectomy during follow-up. Preoperative immunosuppressive treatment was rarely administered (3.1%), whereas 90.3% received postoperative steroid or immunosuppressive therapy. Anti-AChR antibody titers remained detectable during follow-up (median 7.06 nmol/L) (Table S1).

Univariate logistic regression analysis identified several factors associated with thymoma recurrence (Table 2). The presence of MG was significantly associated with recurrence (OR =4.1, 95% CI: 1.34–12.52, P=0.01), as were longer follow-up duration (OR =1.01, 95% CI: 1.00–1.01, P=0.007), transsternal surgical approach (OR =9.59, 95% CI: 1.19–77.02, P=0.03), adjuvant treatment (OR =3.97, 95% CI: 1.04–15.14, P=0.04), tumor size (OR = 1.22, 95% CI: 1.01–1.47, P=0.04), B2/B3 histology (OR =2.99, 95% CI: 1.01–8.86, P=0.048), and advanced Masaoka-Koga stage (OR =10.41, 95% CI: 3.12–34.72, P<0.001). Gender, age at thymoma operation, and extent of thymectomy were not significantly associated with recurrence, while incomplete resection showed borderline association.

In the multivariable logistic regression analysis, advanced Masaoka stage (stage III–IV) was identified as an independent predictor of thymoma recurrence, with an OR of 6.55 (95% CI: 1.12–38.20, P=0.04). Other variables, including MG status, completeness of resection, surgical approach, extent of thymectomy, tumor size, age, sex, and WHO classification, were not significantly associated with recurrence.

Consistently, Kaplan-Meier survival analysis demonstrated that patients with advanced Masaoka-Koga stage had significantly shorter PFS compared with those with early stage disease in the total cohort including R0 and R1/2 resections (log-rank P=0.001; Figure 1A), as well as shorter RFS in the R0 resection subgroup (log-rank P<0.001; Figure 1B). In the Cox proportional hazards model restricted to patients with R0 resection, in order to minimize the potential confounding effect of incomplete resection, both advanced stage disease (HR 4.57, 95% CI: 1.05–19.80, P=0.04) while larger tumor size showed only a borderline association (per 1-cm increase: HR 1.37, 95% CI: 0.99–1.91, P=0.06) (Table 3).

Figure 1 Kaplan-Meier survival curves according to Masaoka-Koga stage. (A) Progression-free survival in the total cohort (log-rank P=0.001). (B) Recurrence-free survival in patients who underwent complete (R0) resection (log-rank P<0.001). (C) Overall survival in the total cohort (log-rank P=0.047). Patients with advanced stage (III/IV) showed significantly worse outcomes compared with those with early stage (I/II). Numbers below the x-axis indicate the number of patients at risk at each time point. Statistical significance was assessed using the log-rank test.

Apart from advanced Masaoka-Koga stage (log-rank P=0.047; Figure 1C), no other variable, including tumor size, was a significant predictor of OS.

Discussion

In our single-center experience, the age at thymectomy had a median of 55 years. Younger age at the thymectomy in the thymoma with MG group may be attributed to the fact that thymomas are frequently asymptomatic and are more likely to be detected earlier when accompanied by MG-related symptoms. In contrast, a higher proportion of advanced Masaoka-Koga stage (III–IV) was observed in MG-associated thymoma in our study, which seems contrary to the aforementioned explanation. This paradox likely reflects a referral bias inherent to a tertiary neurology-thoracic surgery center, where patients with more complex or invasive disease are preferentially referred for combined management. Accordingly, our MG cohort may represent a more advanced subset of thymoma cases rather than the full clinical spectrum typically seen in population-based studies. Type B thymoma was the most frequently observed histological subtype in both groups. Tumor size and sex distribution were comparable between the groups, consistent with previous studies (2,7,8). Type B thymomas, particularly subtypes B1 and B2, are characterized by an abundance of lymphocytes and a prominent immunological milieu, as defined by their histopathological morphology, and are highly associated with MG (7-9). The predominance of transsternal extended thymectomy in thymoma with MG group may reflect the surgical preference for more aggressive resection in patients with MG to remove completely ectopic thymic tissue in addition to the tumor itself and to optimize neurological outcomes.

Thymoma has malignant potential, with the ability to invade adjacent great vessels, lymphatics, and cardiopulmonary structures. Accordingly, there remains a possibility of recurrence even after complete resection, although such cases are rare (10). The average relapse-free interval for thymoma has been reported to be approximately five years, with an overall recurrence rate of 7–8%. Although complete resection of thymoma is achieved, local recurrence may still occur long after a prolonged disease-free interval, with cases reported even 30 years postoperatively (10-12). A study from another Korean medical center reported an overall recurrence rate of 13.4% and a median disease-free interval of 52 months (13). In our cohort, the median time to recurrence was 55.5 months. The recurrence rate was 37.5% in patients with thymoma with MG, compared to 14.9% in those without MG. These findings are generally consistent with previous studies.

Several factors have been widely accepted as potential predictors of thymoma recurrence, including WHO histological subtype, advanced Masaoka stage, incomplete initial resection, extent of thymectomy, postoperative adjuvant therapy, site of recurrence, and larger tumor size and some of which may be interrelated (10,13-16). However, the prognostic significance of MG coexistence in thymoma, including its association with recurrence, remains controversial.

In the past, it was considered that unfavorable effect on postoperative outcomes might be raised from perioperative myasthenia crisis and increased perioperative morbidity. Independent of perioperative myasthenic crisis, some studies have suggested that the presence of MG may be a negative prognostic factor for long-term outcomes (17-19). In contrast, other studies have indicated that MG may not adversely affect prognosis and might even be associated with improved outcomes (20-22).

In the present study, advanced Masaoka-Koga stage emerged as the key independent predictor of thymoma recurrence in the final multivariate model, underscoring the central role of tumor invasiveness in recurrence risk. Although MG showed an association with recurrence in univariate analysis, this relationship did not remain significant in multivariate logistic regression or Kaplan-Meier analyses after adjustment for stage and other clinical variables. In our study, recurrence analyses were performed separately for the overall cohort and the R0 resection subgroup to minimize confounding by incomplete resections. Because recurrence after R1/R2 resection often reflects residual disease rather than new tumor biology, restricting Cox regression and Kaplan–Meier analyses to R0 cases enabled more precise assessment of true prognostic factors. While previous large-scale series have consistently identified Masaoka-Koga stage as a major determinant of recurrence, the prognostic significance of MG has remained uncertain. Our findings refine this understanding by demonstrating that, even within a smaller but clinically well-characterized single-center cohort, MG status does not independently influence recurrence once tumor invasiveness is accounted for. This clarification offers clinical value by reinforcing the dominance of tumor stage over MG in postoperative risk stratification.

Tumor size also appeared to affect recurrence risk. Larger tumors were significantly associated with recurrence in univariate logistic regression and Kaplan-Meier analyses, whereas the association was borderline in multivariate logistic regression (P=0.08) and Cox regression (P=0.06). These findings suggest that tumor size may contribute to recurrence risk, although its independent effect was attenuated after adjustment for stage and other covariates. Given the limited number of events, tumor size should be regarded as a potential prognostic factor requiring validation in larger cohorts.

The median time to recurrence was numerically longer in the MG group than in the non-MG group (72.5 vs. 28.0 months; P=0.14), and overall mortality was also higher in the MG group. However, these findings are unlikely to reflect a direct effect of MG. Rather, they are better explained by confounding factors, particularly the higher frequency of extended thymectomy among MG patients, which may delay early recurrence, and the greater need for long-term immunosuppressive or symptomatic management, which naturally prolongs surveillance. Taken together, these results suggest that the observed differences are largely attributable to tumor stage, surgical extent, and treatment-related factors, rather than to MG itself.

Although the sample size is limited, this single-center cohort offers several advantages that complement findings from large multicenter registries. The homogeneity of surgical approach, pathological evaluation, and follow-up protocols minimizes inter-institutional variability, allowing more consistent assessment of clinicopathologic correlations. In addition, the integration of detailed neurological data—such as MG severity, antibody titers, and postoperative immunosuppressive management—is rarely available in multicenter surgical databases. Collectively, this study provides clinically detailed evidence supporting the dominant role of tumor invasiveness rather than MG status in predicting recurrence after complete resection.

Nevertheless, several limitations should be acknowledged. First, the retrospective, single-center design and relatively small sample size may limit the generalizability of our findings. Despite the long observational period, selection and information biases inherent to retrospective data collection could not be entirely excluded.

Second, the follow-up duration and mortality differed between the MG and non-MG groups, potentially influencing recurrence estimates through underdetection of late recurrences in the non-MG group and informative censoring in the MG group. Postoperative immunosuppressive therapy was administered to almost all MG patients but not to any non-MG patients, precluding its inclusion as a covariate in multivariable analyses and leaving the possibility of residual confounding. Moreover, as most patients were treated before the introduction of recent molecular-targeted therapies for MG, the immunosuppressive regimens largely reflected conventional corticosteroid, azathioprine, or calcineurin inhibitor-based therapy, which may not fully represent current practice.

Third, the proportion of incomplete resections (R1/R2) in our cohort was relatively high compared to typical rates reported in thymoma surgery, reflecting the complexity of cases referred to a tertiary center.

Finally, due to the limited number of recurrence and death events—particularly in subgroup analysis—certain associations may be underpowered or unstable. Larger prospective studies are warranted to confirm the recurrence predictors identified in this study.

Conclusions

In conclusion, this observational study indicates that MG itself is not directly associated with thymoma recurrence or mortality. Advanced Masaoka–Koga stage consistently emerged as the strongest predictor, while tumor size showed only a borderline association. These findings highlight tumor invasiveness, rather than MG, as the key determinants of outcome underscoring the need for intensive surveillance in patients with advanced disease.

Acknowledgments

The authors thank Dr. Dong-Hyun Lee for his help with English editing and statistical advice. This contributor had no involvement in the study design or data interpretation.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-2025-401/rc

Data Sharing Statement: Available at https://gs.amegroups.com/article/view/10.21037/gs-2025-401/dss

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-2025-401/prf

Funding: This work was supported by the Gachon University research fund of 2023 (GCU-202309260001) and the research fund of 2024 (GCU-202410410001).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-2025-401/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. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the Gil Medical Center (IRB No. GDIRB2024–154). Informed consent was waived in this retrospective study.

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. Cheney RT. The biologic spectrum of thymic epithelial neoplasms: current status and future prospects. J Natl Compr Canc Netw 2010;8:1322-8. [Crossref] [PubMed]
2. Romi F. Thymoma in myasthenia gravis: from diagnosis to treatment. Autoimmune Dis 2011;2011:474512. [Crossref] [PubMed]
3. Safieddine N, Liu G, Cuningham K, et al. Prognostic factors for cure, recurrence and long-term survival after surgical resection of thymoma. J Thorac Oncol 2014;9:1018-22. [Crossref] [PubMed]
4. von der Thüsen J. Thymic epithelial tumours: histopathological classification and differential diagnosis. Histopathology 2024;84:196-215. [Crossref] [PubMed]
5. Hermanek P, Wittekind C. The pathologist and the residual tumor (R) classification. Pathol Res Pract 1994;190:115-23. [Crossref] [PubMed]
6. Koga K, Matsuno Y, Noguchi M, et al. A review of 79 thymomas: modification of staging system and reappraisal of conventional division into invasive and non-invasive thymoma. Pathol Int 1994;44:359-67. [Crossref] [PubMed]
7. Marx A, Chan JK, Coindre JM, et al. The 2015 World Health Organization Classification of Tumors of the Thymus: Continuity and Changes. J Thorac Oncol 2015;10:1383-95. [Crossref] [PubMed]
8. Okumura M, Ohta M, Tateyama H, et al. The World Health Organization histologic classification system reflects the oncologic behavior of thymoma: a clinical study of 273 patients. Cancer 2002;94:624-32. [Crossref] [PubMed]
9. Marx A, Ströbel P, Weis CA. The pathology of the thymus in myasthenia gravis. Mediastinum 2018;2:66.
10. Luo T, Zhao H, Zhou X. The clinical features, diagnosis and management of recurrent thymoma. J Cardiothorac Surg 2016;11:140. [Crossref] [PubMed]
11. Awad WI, Symmans PJ, Dussek JE. Recurrence of stage I thymoma 32 years after total excision. Ann Thorac Surg 1998;66:2106-8. [Crossref] [PubMed]
12. Xu C, Feng QF, Fan CC, et al. Patterns and predictors of recurrence after radical resection of thymoma. Radiother Oncol 2015;115:30-4. [Crossref] [PubMed]
13. Bae MK, Lee CY, Lee JG, et al. Predictors of recurrence after thymoma resection. Yonsei Med J 2013;54:875-82. [Crossref] [PubMed]
14. Weis CA, Yao X, Deng Y, et al. The impact of thymoma histotype on prognosis in a worldwide database. J Thorac Oncol 2015;10:367-72. [Crossref] [PubMed]
15. Chiappetta M, Lococo F, Zanfrini E, et al. The International Thymic Malignancy Interest Group Classification of Thymoma Recurrence: Survival Analysis and Perspectives. J Thorac Oncol 2021;16:1936-45. [Crossref] [PubMed]
16. Aprile V, Korasidis S, Bacchin D, et al. Thymectomy in Myasthenic Patients With Thymoma: Killing Two Birds With One Stone. Ann Thorac Surg 2021;112:1782-9. [Crossref] [PubMed]
17. Zhang Z, Cui Y, Jia R, et al. Myasthenia gravis in patients with thymoma affects survival rate following extended thymectomy. Oncol Lett 2016;11:4177-82. [Crossref] [PubMed]
18. Miura K, Doi T, Tanaka Y, et al. Effect of myasthenia gravis on the surgical outcomes of patients with thymoma. Asian Cardiovasc Thorac Ann 2022;30:924-30. [Crossref] [PubMed]
19. Tian W, Li X, Sun Y, et al. Myasthenia gravis affects overall survival in patients with thymoma: an analysis of multicentre database using propensity score matching. Interact Cardiovasc Thorac Surg 2021;33:250-7. [Crossref] [PubMed]
20. Zhao K, Liu Y, Jing M, et al. Long-term prognosis in patients with thymoma combined with myasthenia gravis: a propensity score-matching analysis. Front Med (Lausanne) 2024;11:1407830. [Crossref] [PubMed]
21. Margaritora S, Cesario A, Cusumano G, et al. Thirty-five-year follow-up analysis of clinical and pathologic outcomes of thymoma surgery. Ann Thorac Surg 2010;89:245-52; discussion 252. [Crossref] [PubMed]
22. Kondo K, Monden Y. Thymoma and myasthenia gravis: a clinical study of 1,089 patients from Japan. Ann Thorac Surg 2005;79:219-24. [Crossref] [PubMed]