If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. Address for correspondence: Sukhmani K. Padda, MD, Department of Medicine, Division of Oncology, Stanford Cancer Institute/Stanford University School of Medicine, 875 Blake Wilbur Drive, Stanford, CA 94305.
Department of Thoracic Surgery, Icahn School of Medicine at Mount Sinai Health System, New York, New YorkValley/Mount Sinai Comprehensive Cancer Care, Paramus, New Jersey
Thymic epithelial tumors (TETs) are associated with paraneoplastic/autoimmune (PN/AI) syndromes. Myasthenia gravis is the most common PN/AI syndrome associated with TETs.
Methods
The International Thymic Malignancy Interest Group retrospective database was examined to determine (1) baseline and treatment characteristics associated with PN/AI syndromes and (2) the prognostic role of PN/AI syndromes for patients with TETs. The competing risks model was used to estimate cumulative incidence of recurrence (CIR) and the Kaplan-Meier method was used to calculate overall survival (OS). A Cox proportional hazards model was used for multivariate analysis.
Results
A total of 6670 patients with known PN/AI syndrome status from 1951 to 2012 were identified. PN/AI syndromes were associated with younger age, female sex, thymoma histologic type, earlier stage, and an increased rate of total thymectomy and complete resection status. There was a statistically significant lower CIR in the group with a PN/AI syndrome than in the group without a PN/AI syndrome (10-year CIR 17.3% versus 21.2%, respectively [p = 0.0003]). The OS was improved in the group with a PN/AI syndrome compared to the group without a PN/AI syndrome (median OS 21.6 years versus 17.0 years, respectively [hazard ratio = 0.63, 95% confidence interval: 0.54–0.74, p < 0.0001]). However, in the multivariate model for recurrence-free survival and OS, PN/AI syndrome was not an independent prognostic factor.
Discussion
Previously, there have been mixed data regarding the prognostic role of PN/AI syndromes for patients with TETs. Here, using the largest data set in the world for TETs, PN/AI syndromes were associated with favorable features (i.e., earlier stage and complete resection status) but were not an independent prognostic factor for patients with TETs.
Paraneoplastic syndromes can precede the diagnosis of a tumor and tend to improve with treatment of the tumor. Although myasthenia gravis (MG) is the most common PN/AI syndrome associated with TETs,
Long-term response and outcome following immunosuppressive therapy in thymoma-associated pure red cell aplasia: a nationwide cohort study in Japan by the PRCA collaborative study group.
With regard to MG, 10% to 20% of patients with MG have thymoma and 30% of patients with thymoma either present with or are eventually diagnosed with MG.
These coexisting antibodies increase the likelihood of a thymoma diagnosis and may herald more severe disease. Although autoantibodies play a critical role in the development of MG in patients with thymoma, there are seronegative cases reported, implicating alternative mechanisms of immune dysregulation.
Although the exact pathogenesis remains unknown, it is not surprising that TETs are associated with autoimmune disorders, given that the thymus is critical in building the T-cell repertoire (1) through positive selection in thymic cortical epithelial cells and (2) in maintaining immune homeostasis and central self-tolerance through negative selection in the thymic medulla.
Thymomas have a deranged tumor microenvironment, carry out abnormal intratumoral thymopoiesis, and thus disseminate an abnormal circulating T-cell repertoire that has been insufficiently tolerized to self-antigens.
For example, one study showed that patients with thymoma-associated MG had increased circulating mature CD4-positive/CD45RA-positive T-cells, and another study showed that they had decreased regulatory T-cells compared to patients with thymoma without MG.
Finally, thymoma epithelial cells have decreased major histocompatibility class II expression, which in conjunction with decreased self-antigen expression results in a biased T-cell repertoire.
In the current study, we examined patients with TETs, including thymomas, thymic carcinomas, and neuroendocrine tumors of the thymus (NETTs), with PN/AI syndrome status recorded in the International Thymic Malignancy Interest Group (ITMIG) retrospective database.
Development of the international thymic malignancy interest group international database: an unprecedented resource for the study of a rare group of tumors.
Our objectives were twofold: (1) to determine patient, tumor, and treatment characteristics associated with PN/AI syndromes and (2) to determine whether the presence of a PN/AI syndrome is an independent prognostic factor for patients with TETs.
Materials and Methods
The ITMIG retrospective database of thymic malignancies is the result of an international collaboration of 56 institutions, which has been described in detail elsewhere.
Development of the international thymic malignancy interest group international database: an unprecedented resource for the study of a rare group of tumors.
This study was conducted with use of a limited data set with de-identified information and therefore without requirement of authorization or documentation of waiver of the institutional review board. After execution of a data use agreement, a limited data set from each institution was provided to the ITMIG for the sole purpose of research. Only the ITMIG statistical core and selected members of the database committee had full access to the data (X. Y., Y. S., A. A., F. D., and J. H.).
Of 7795 patients in the ITMIG retrospective database from 1951 to 2012, we identified 6670 with known PN/AI syndrome status at initial diagnosis and known pathologic diagnosis of thymoma, thymic carcinoma, or NETT. The final sample size was 6297 patients (Fig. 1). The PN/AI syndrome data field terms included MG, pure red cell aplasia, hypogammaglobulinemia, and other. Only other PN/AI syndromes with further description were included in the analysis. The vast majority (97%) of patients in this cohort were from the time period 1991 onward. Patient, tumor, and treatment characteristics were collected. Each site contributing to the database reported histologic findings per the 2004 WHO classification,
and there was no central pathology review. Patients with histologic types of metaplastic, micronodular, or other without further specification were excluded. There were only 90 cases of micronodular thymoma and 16 cases of metaplastic thymoma in the database. Of the cases with micronodular and metaplastic thymoma, three and one had a PN/AI syndrome, respectively, with missing information on PN/AI syndrome status in 58% and 50%, respectively.
Figure 1Flow diagram. *Numbers do not add up to 373 because there is overlap among patients with missing information for each variable. ITMIG, International Thymic Malignancy Interest Group; PN/AI, paraneoplastic/autoimmune syndrome.
The ITMIG/International Association for the Study of Lung Cancer staging project demonstrated no difference in outcomes between patients with stage IIA versus stage IIB disease or between the Masaoka and Masaoka-Koga staging systems.
Detterbeck F. (2013) MS16.2—Towards a TNM-based prognostic classification for thymic tumours. Abstract presented at: 15th World Conference on Lung Cancer. October 27–30, 2013; Sydney, Australia.
Therefore, stage IIA, IIB, and II were categorized as stage II for the purpose of analyses. Given the differences in outcomes observed between patients with stage IVA and IVB disease,
Detterbeck F. (2013) MS16.2—Towards a TNM-based prognostic classification for thymic tumours. Abstract presented at: 15th World Conference on Lung Cancer. October 27–30, 2013; Sydney, Australia.
cases with stage IV disease without further subset clarification were excluded.
Statistical Analyses
All analyses were performed with SAS 9.3 software (SAS Institute Inc., Cary, NC) and R version 3.1.2 (R Foundation for Statistical Computing, Vienna, Austria). Descriptive statistics were used to summarize the baseline and treatment characteristics, including continuous variables as medians and ranges and categorical variables as frequencies and relative percentages. The chi-square test and t test were used to compare categorical variables and continuous variables, respectively, between patients with and without PN/AI syndromes. The competing risks model was used to estimate cumulative incidence of recurrence (CIR) of thymic malignancy. In the competing risks model, recurrence is the event of interest and death is included as the competing event, though curves of death are not shown in the cumulative incidence plots. The difference in CIR between patients with and without PN/AI syndromes was assessed by using the Gray test.
The Kaplan-Meier method was used to calculate overall survival (OS), which was measured from the date of intervention to the date of death and censored at the date on which the patient was last known to be alive. The log rank test was used to compare OS of patients with and without a PN/AI syndrome. Outcomes were also examined as stratified by PN/AI syndrome status in subsets of thymoma and thymic carcinoma, stage subsets, and time periods.
A Cox proportional hazards model for multivariate analysis of recurrence-free survival (RFS) and OS was used. Of note, RFS was measured from the date of intervention to the date of death or date of malignancy recurrence, whichever occurred first, and censored at the date patient last known to be alive without recurrence. The model included all the factors detailed in Table 1 including age, sex, PN/AI syndrome status (present or absent), continent (North/South America, Asia, or Europe), pathologic diagnosis (thymoma, thymic carcinoma, or NETT), thymoma WHO subtype (A, AB, B1, B2, or B3), pathologic stage (I, II, III, IVA, or IVB), tumor size, extent of thymectomy (none, partial, total, or extended), chemotherapy (curative or palliative/none), radiation (curative or palliative/none), and resection status (R0, R1, or R2). Patients with missing data were excluded from this analysis; no imputations were performed. A two-sided p value less than 0.05 was considered statistically significant for any statistical test used.
Table 1Cohort Characteristics of 6297 Patients from the ITMIG Retrospective Database
Other included two cases of Cushing syndrome; four of rheumatoid arthritis; and one each of anemia not otherwise specified, hemolytic anemia, erythrocytosis, thrombocytosis, neutropenia, polymyositis, Goodpasture syndrome, Grave’s disease, and polymyalgia rheumatica.
Of 42 patients with a thymic carcinoma and a PN/AI syndrome, 38 had myasthenia gravis and one each of pure red cell aplasia, hypogammaglobulinemia, Cushing syndrome, and rheumatoid arthritis. NETTs included 37 typical carcinoid, 55 atypical carcinoid, 28 large cell or small cell neuroendocrine carcinoma, nine carcinoid not otherwise specified, and 10 others not further classified. All patients with a NETT and a PN/AI syndrome had myasthenia gravis and included one each with typical and atypical carcinoid and three with poorly differentiated neuroendocrine carcinoma.
Of 42 patients with a thymic carcinoma and a PN/AI syndrome, 38 had myasthenia gravis and one each of pure red cell aplasia, hypogammaglobulinemia, Cushing syndrome, and rheumatoid arthritis. NETTs included 37 typical carcinoid, 55 atypical carcinoid, 28 large cell or small cell neuroendocrine carcinoma, nine carcinoid not otherwise specified, and 10 others not further classified. All patients with a NETT and a PN/AI syndrome had myasthenia gravis and included one each with typical and atypical carcinoid and three with poorly differentiated neuroendocrine carcinoma.
Of 42 patients with a thymic carcinoma and a PN/AI syndrome, 38 had myasthenia gravis and one each of pure red cell aplasia, hypogammaglobulinemia, Cushing syndrome, and rheumatoid arthritis. NETTs included 37 typical carcinoid, 55 atypical carcinoid, 28 large cell or small cell neuroendocrine carcinoma, nine carcinoid not otherwise specified, and 10 others not further classified. All patients with a NETT and a PN/AI syndrome had myasthenia gravis and included one each with typical and atypical carcinoid and three with poorly differentiated neuroendocrine carcinoma.
Curative includes neoadjuvant, adjuvant, and definitive (for chemotherapy, when given definitively with radiation).
2134 (41.9)
700 (42.9)
1434 (41.5)
p = 0.35
Palliative/none
2955 (58.1)
933 (57.1)
2022 (58.5)
Missing
1208
—
—
Resection status
R0
4768 (83.5)
1729 (87.2)
3039 (81.5)
p < 0.0001
R1
465 (8.1)
165 (8.3)
300 (8.1)
R2
480 (8.4)
90 (4.5)
390 (10.5)
Missing
584
—
—
Note: Percentages calculated with exclusion of missing values and column percentages reported.
ITMIG, International Thymic Malignancy Interest Group; PN/AI, paraneoplastic/autoimmune syndrome; NETT, neuroendocrine tumors of the thymus.
a Surgery field (yes/no) was derived from many separate surgical fields in the database for completeness and accuracy.
b Curative includes neoadjuvant, adjuvant, and definitive (for chemotherapy, when given definitively with radiation).
c Other included two cases of Cushing syndrome; four of rheumatoid arthritis; and one each of anemia not otherwise specified, hemolytic anemia, erythrocytosis, thrombocytosis, neutropenia, polymyositis, Goodpasture syndrome, Grave’s disease, and polymyalgia rheumatica.
d Of 42 patients with a thymic carcinoma and a PN/AI syndrome, 38 had myasthenia gravis and one each of pure red cell aplasia, hypogammaglobulinemia, Cushing syndrome, and rheumatoid arthritis. NETTs included 37 typical carcinoid, 55 atypical carcinoid, 28 large cell or small cell neuroendocrine carcinoma, nine carcinoid not otherwise specified, and 10 others not further classified. All patients with a NETT and a PN/AI syndrome had myasthenia gravis and included one each with typical and atypical carcinoid and three with poorly differentiated neuroendocrine carcinoma.
A total of 6297 patients were included in the analysis (see Fig. 1). Cohort characteristics are reported in Table 1. Most patients were from Asia (42%) followed by Europe (33%) and North America (25%). As expected, thymoma predominated (86%), with the most common histotypes being AB and B2. Thymic carcinoma was relatively well represented (12%) in this cohort but NETT was rare (2%). Approximately one-third of patients had a PN/AI syndrome, with the vast majority having MG. However, there were also cases of pure red cell aplasia (n = 47) and hypogammaglobulinemia (n = 13). Pathologic stage was biased toward early-stage disease (stage I–II), representing two-thirds of the cohort. Almost all patients (98%) underwent an operation, with most patients undergoing a total thymectomy (81%) and achieving a complete R0 resection (84%).
Baseline and Treatment Characteristics Associated with PN/AI Syndromes
Patients with a PN/AI syndrome were younger (median age 50 years) and predominantly female and European (see Table 1). Patients without a PN/AI syndrome were older (median age 55 years) and predominantly male and Asian. In both groups with and without a PN/AI syndrome, the most common pathologic diagnosis was thymoma. However, the most common thymoma WHO histotype was B2 for the group with a PN/AI syndrome versus AB for the group without a PN/AI syndrome. In addition, the group without a PN/AI syndrome had a higher proportion of thymic carcinoma histologic type (17% versus 2%, respectively). Pathologic stage was mostly early (stage I–II) in both groups with and without a PN/AI syndrome; however, the group without a PN/AI syndrome had a higher proportion of advanced-stage disease (stage III–IVB). Both the groups with and without a PN/AI syndrome most commonly underwent a total thymectomy, received no (or palliative) chemotherapy, received no (or palliative) radiotherapy, and achieved a complete R0 resection. However, when the group with a PN/AI syndrome was compared to the group without a PN/AI syndrome, the rates of total thymectomy, no (or palliative) chemotherapy, and complete R0 resections were higher. There was no significant difference in radiation practices between the groups. The findings of these analyses were the same when only the PN/AI syndrome of MG was examined (data not shown).
CIR and OS
The median follow-up for the whole group was 3.7 years. The median follow-up for the group with a PN/AI syndrome was 4.4 years, and for the group without a PN/AI syndrome was 3.3 years. Recurrence information was available on 4375 patients. Overall, there was a statistically significant lower cumulative incidence of recurrence (CIR) in the group with a PN/AI syndrome than in the group without a PN/AI syndrome (p = 0.0003 [Fig. 2]): a 10-year CIR of 17.3% versus 21.2%, respectively; a 20-year CIR of 27.2% versus 28.1%, respectively; a 30-year CIR of 29.5% versus 39.4%, respectively; and a 40-year CIR of 29.5% versus not available, respectively. There was no difference in CIR between the groups with and without a PN/AI syndrome when examining the subgroups of thymoma (p = 0.93), thymic carcinoma (p = 0.76) (Supplementary Figs. 1 and 2), and stage (Supplementary Figs. 3–7).
Figure 2Cumulative incidence of recurrence in all patients with paraneoplastic/autoimmune syndrome (PN/AI(+)) versus without paraneoplastic/autoimmune syndrome (PN/AI(-)).CI, confidence interval.
Survival information was available on 4962 patients. In the overall population, OS was improved in the group with a PN/AI syndrome compared to the group without a PN/AI syndrome (hazard ratio [HR] = 0.63, 95% confidence interval [CI]: 0.54–0.74, p <0.0001, median OS 21.6 years versus 17.0 years, respectively [Fig. 3]). In the thymoma subgroup, there was a statistically significant improvement in OS for the group with a PN/AI syndrome compared to the group without a PN/AI syndrome (HR = 0.82, 95% CI: 0.70–0.97, p = 0.02) (Fig. 4A). In the thymic carcinoma subgroup, there was also a trend toward improved OS in the group with a PN/AI syndrome (HR = 0.56, 95% CI: 0.29–1.10, p = 0.09), with the survival curves separating early (Fig. 4B). When stratified by stage, the improved OS in the group with a PN/AI syndrome compared to the group without a PN/AI syndrome remained in the stage III and stage IVB subgroups (Supplementary Figs. 8–12). The aforementioned analyses of CIR and OS were also repeated for only those patients who had MG, and the results were the same (data not shown).
Figure 3Overall survival (OS) in all patients with paraneoplastic/autoimmune syndrome (PN/AI(+)) versus without paraneoplastic/autoimmune syndrome (PN/AI(-)).HR, hazard ratio; CI, confidence interval.
Figure 4Overall survival (OS) with paraneoplastic/autoimmune syndrome (PN/AI(+)) versus without paraneoplastic/autoimmune syndrome (PN/AI(-)) in patients with thymoma (A) and patients with thymic carcinoma (B). HR, hazard ratio; CI, confidence interval.
Survival was also examined over three time periods (1951–1970, 1971–1990, and 1991–2012) that included 50, 228, and 5678 patients, respectively, with a prevalence of PN/AI syndrome of 70.0%, 52.2%, and 32.0%, respectively. For the group with a PN/AI syndrome, the OS was significantly improved in the most recent time period (p = 0.001 [Supplementary Fig. 13]). However, for the group without a PN/AI syndrome, there was no difference in OS among the different time periods (p = 0.87).
Multivariate Model for RFS and OS
There were complete data available on 2193 patients and 2352 patients for the multivariate analysis of RFS and OS, respectively. In a multivariate model (Table 2), the following characteristics were independently associated with increased recurrence: older age, thymic carcinoma and NETT histologic types, advanced-stage (stage III–IVB) disease, larger tumor size, and R2 resection status. Characteristics independently associated with decreased recurrence included residence on the Asian continent and receipt of curative radiation (i.e., neoadjuvant, adjuvant, and definitive setting). In a multivariate model, the following characteristics were independently associated with worse OS: older age, thymic carcinoma and NETT histologic types, advanced-stage (stage III–IVB) disease, receipt of curative chemotherapy (i.e., neoadjuvant, adjuvant, and in definitive setting with radiation), and R2 resection status. Characteristics independently associated with improved OS included receipt of curative radiation and B1 histotype. The presence of PN/AI syndrome was not independently associated with clinical outcomes of RFS and OS.
To our knowledge, this is the largest study and the first multicontinent study, examining PN/AI syndromes in patients with TETs. This was achieved using the ITMIG retrospective database, which serves as a paradigm for performing research on rare malignancies. Previously, there have been conflicting data about the prognostic role of PN/AI syndromes for patients with TETs.
Our main finding is that the presence of PN/AI syndromes is associated with favorable prognostic factors but is not an independent prognostic factor for patients with TETs. Our results were identical when examining only the PN/AI syndrome of MG (data not shown), which as expected comprised the majority (96%) of cases of PN/AI syndrome in this cohort. Given the small number of cases of pure red cell aplasia and hypogammaglobulinemia, we were not able to independently assess the clinical impact of these syndromes and thus, they were analyzed in aggregate with MG. In the ITMIG cohort, approximately one-third of patients with TETs had a PN/AI syndrome. This is similar to the prevalence of MG reported in national retrospective database studies,
There is overlap with these national databases and the ITMIG database cohort, including 31% of cases from the ChART database and 14% of cases from the ESTS database (Supplementary Fig. 14), whereas the JART database was not included.
PN/AI Syndrome and Patient, Tumor, and Treatment Factors
In our cohort, the group with a PN/AI syndrome was associated with known favorable prognostic factors. With regard to demographic factors, the group with a PN/AI syndrome was younger by a median of 5 years and had a higher proportion of females, confirming prior findings in the literature.
Does myasthenia gravis influence overall survival and cumulative incidence of recurrence in thymoma patients? A retrospective clinicopathological multicentre analysis on 797 patients.
With regard to pathologic diagnosis, the group with a PN/AI syndrome had a higher proportion of favorable thymoma histologic type and a lower proportion of thymic carcinoma. Type B2 was the most common thymoma histotype (37%) in the group with a PN/AI syndrome, whereas type AB (30%) was the most common in the group without a PN/AI syndrome. These thymoma histotypes have also been previously identified as the most common when stratified by PN/AI syndrome status
The consistent association of PN/AI syndromes with type B2 thymoma across large database studies may be explained by B2 thymomas containing heavy lymphocyte regions, including CD4-positive/CD8-positive T-cells that are vulnerable to altered positive selection in the setting of thymic epithelial cells with low expression of HLA-DR.
Of note, 6% of patients in our cohort with thymic carcinoma had a PN/AI syndrome. An increasing number of patients with thymic carcinoma and PN/AI syndrome are being reported,
Does myasthenia gravis influence overall survival and cumulative incidence of recurrence in thymoma patients? A retrospective clinicopathological multicentre analysis on 797 patients.
Because there was no central pathology review in the ITMIG retrospective database, it is possible that a proportion of thymomas were misclassified as thymic carcinomas as poor interobserver reproducibility has been demonstrated between B3 thymoma and thymic carcinoma.
The enlightenments from ITMIG Consensus on WHO histological classification of thymoma and thymic carcinoma: refined definitions, histological criteria, and reporting.
Does myasthenia gravis influence overall survival and cumulative incidence of recurrence in thymoma patients? A retrospective clinicopathological multicentre analysis on 797 patients.
TETs that are associated with a PN/AI syndrome may be caught at an earlier stage because of the presence of symptoms, and it is also possible that they are associated with better biology.
With regard to treatment, a higher proportion of patients in the group with a PN/AI syndrome underwent a total thymectomy (90% with a PN/AI syndrome versus 77% without a PN/AI syndrome). It has been hypothesized that to increase the chances of remission of MG, a total thymectomy is required. This was recently supported by the results of an international randomized trial demonstrating that thymectomy, including en bloc resection of all mediastinal tissue containing either gross or microscopic thymic tissue, is more beneficial than prednisone therapy alone for the treatment of nonthymomatous MG.
Myasthenia gravis: recommendations for clinical research standards. Task Force of the Medical Scientific Advisory Board of the Myasthenia Gravis Foundation of America.
(i.e., the severity of MG symptoms) and MGFA postintervention status (i.e., clinical state after treatment such as thymectomy) in the ITMIG database, the data for these fields was not consistently collected. Therefore, we cannot comment on how surgical intervention affected the severity of MG symptoms in our cohort. Preoperative MGFA classification was examined in the JART database and showed no difference in recurrence or survival for patients with thymoma when stratified by MGFA classification.
Partial thymectomy can be considered in patients with early-stage tumors without MG because prior retrospective studies have shown similar recurrence rates for partial versus total thymectomy in this group.
This could account for the higher proportion of partial thymectomies performed in the group without a PN/AI syndrome in our cohort.
PN/AI Syndrome, Recurrence, and Survival
The group with a PN/AI syndrome had statistically significant improved outcomes compared to the group without a PN/AI syndrome for CIR and OS in the overall cohort. However, when the analysis was performed for thymoma, thymic carcinoma, and stage subgroups, there was no difference in CIR between the groups with and without a PN/AI syndrome, indicating that these may be confounding factors. When the same subgroup analyses were performed for OS, the improved OS among patients with a PN/AI syndrome remained significant in the thymoma subgroup, trended in the thymic carcinoma subgroup, and remained significant in advanced-stage (stages III and IVB) disease. The trend toward improved OS in the thymic carcinoma subgroup for patients with a PN/AI syndrome has been shown in other series, and is probably due to the association with favorable prognostic factors, including smaller tumors, earlier Masaoka stage, and higher rates of complete resection.
The improved OS in advanced-stage disease suggests that the presence of PN/AI syndromes may indicate better biology or a type of selection bias among advanced-stage tumors that is not readily identified in this retrospective cohort. With regard to time period, we found that the group with a PN/AI syndrome had significantly improved OS in the most recent time period (1991–2012) but there was no difference in OS in the group without a PN/AI syndrome over time. This observation is possibly due to improved treatment and supportive care of MG, as reports before the 1980s showed that MG negatively affected OS for patients with TETs.
The most important finding in our study is that PN/AI syndrome status was not an independent factor associated with RFS or OS for patients with TETs. Independent factors associated with RFS and OS in a multivariate model in this study included age (with older age being worse), histologic type (with nonthymoma histotype being worse), stage (with advanced stage [III–IVB] being worse), and resection status (with incomplete R2 resection status being worse). PN/AI syndrome status was likely not an independent prognostic factor because of its association with these strong prognostic factors. Additional unfavorable factors found in the multivariate analysis for RFS and OS were larger size
and receipt of curative intent chemotherapy, respectively. The latter could be because chemotherapy agents (i.e. anthracyclines) have long-term toxicities, including cardiomyopathy and secondary malignancies. Curative intent radiation was an independent favorable factor for both RFS and OS, possibly because radiation is capable of destroying microscopic islands of thymus not surgically excised. Independent favorable features for RFS included residence on the Asian continent, and for OS included type B1 thymoma.
There are limitations of the recurrence and survival analyses in our study. The recurrence analysis was limited by the likely variable frequency of follow-up imaging at each institution and thus, when a recurrence was first detected. The ITMIG standards recommend follow-up after surgical resection with a computed tomography chest scan each year for 5 years and thereafter, alternating chest radiograph and computed tomography chest scan until year 11 at the minimum.
However, these follow-up standards were not likely applied, given that the cohort spanned multiple countries and was retrospective. Limitations of the OS analysis for our study included a significant proportion of missing data for cause of death (>90% missing), the relatively short median follow-up time of the group (<4 years), and lead time bias as patients with a PN/AI syndrome may be diagnosed earlier.
Here, we place our findings in the context of the existing literature. Similar to our study, the ChART database demonstrated a decreased rate of recurrence in patients with MG even in those with advanced Masaoka stage (stage III/IV) disease (15.7% versus 31.7%, respectively), although there were significantly lower rates of thymic carcinoma and smaller tumor size in the MG cohort.
However, the JART database showed similar recurrence-free intervals (RFI) in patients with and without MG (5-year RFI of 93% versus 92% and 10-year RFI of 89% versus 87%, respectively), although only patients with thymoma were included in this analysis.
Of these databases, the ESTS database was the only one that examined recurrence in a multivariate model and, unlike our study, showed that MG was an independent prognostic factor associated with lower CIR.
Similar to our findings, the ChART authors observed that OS was improved in patients with MG with advanced stage III/IV disease, but unlike our study, they also found that OS was worse in patients with MG with stage I disease.
Despite the strength in numbers of this study, it has the standard limitations of a retrospective database, including missing information and selection bias. We found PN/AI syndromes to be associated with favorable features such as younger age, thymoma histologic type, earlier stage, and increased rate of complete resection status. However, the presence of PN/AI syndrome was not an independent prognostic factor for patients with TETs for either RFS or OS. Importantly, our study confirms prior national database studies that PN/AI syndrome status (represented by MG in these studies) is not an independent factor associated with OS.
Long-term response and outcome following immunosuppressive therapy in thymoma-associated pure red cell aplasia: a nationwide cohort study in Japan by the PRCA collaborative study group.
Development of the international thymic malignancy interest group international database: an unprecedented resource for the study of a rare group of tumors.
Detterbeck F. (2013) MS16.2—Towards a TNM-based prognostic classification for thymic tumours. Abstract presented at: 15th World Conference on Lung Cancer. October 27–30, 2013; Sydney, Australia.
Does myasthenia gravis influence overall survival and cumulative incidence of recurrence in thymoma patients? A retrospective clinicopathological multicentre analysis on 797 patients.
The enlightenments from ITMIG Consensus on WHO histological classification of thymoma and thymic carcinoma: refined definitions, histological criteria, and reporting.
Myasthenia gravis: recommendations for clinical research standards. Task Force of the Medical Scientific Advisory Board of the Myasthenia Gravis Foundation of America.
Drs. Wakelee and Badve equally contributed to this article.
Disclosure: Dr. Wakelee has served as a consultant for/received honoraria from Peregrine, ACEA, Pfizer, Helsinn, and Genentech (uncompensated) and has received grants/research funding from Clovis, Exelixis, AstraZeneca/Medimmune, Genentech/Roche, BMS, Gilead, Novartis, Xcovery, Pfizer, Celgene, Pharmacyclics, and Lilly. Dr. Padda has served as a consultant for/received honoraria from G1 therapeutics, Janssen, and AstraZeneca and has received research funding from EpicentRx and Forty Seven Inc.
In this issue of the Journal of Thoracic Oncology, Padda et al. report on the features of paraneoplastic and autoimmune disorders in a retrospective cohort of 6670 patients registered in the International Thymic Malignancy Interest Group retrospective database.1 The results of this analysis confirm the previously reported significant association of autoimmune disorders with lymphocyte-rich, type B1 thymoma, early stage of the disease and, ultimately, complete resection of the tumor, with a tendency toward a better outcome in terms of recurrence and survival rates.
33064-2&id=gr1.jpg){kind=link}
33064-2&id=gr2.jpg){kind=link}
33064-2&id=gr3.jpg){kind=link}
33064-2&id=gr4.jpg){kind=link}