Adjuvant radiotherapy for completely resected stage 2 thymoma
Abstract
BACKGROUND:
The clinical benefit of postoperative mediastinal radiation for completely resected Masaoka stage 2 thymoma remains controversial. Due to its indolent nature and infrequent recurrences, no study has definitively determined the optimal approach.
METHODS:
We retrospectively reviewed 175 consecutive patients who underwent thymic resection from January 1990 to July 2008 at the University of Pennsylvania. The primary endpoint was local recurrence, defined as recurrence within the surgical bed, treated by resection alone versus resection plus radiation. Patients with high recurrence risk were referred for adjuvant radiotherapy.
RESULTS:
Seventy-four Masaoka stage 2 patients were resected; 62 underwent complete resections with adequate postsurgical follow-up. Thirty-seven patients received adjuvant radiotherapy and 25 patients were observed. The median radiation dose was 5040 cGy. The median follow-up for all patients was 52 months. The local recurrence rate was 3.2%. The proportion of recurrences in patients observed after surgery was 8% versus 0% in those who received adjuvant radiotherapy (P = .15). Size was not an independent predictor of recurrence (P = .81). The tumor-related death rate was 0%, and overall death rate was 3.2%. One death occurred in each group, observation, and radiation. There were no grade 3 or 4 complications with radiation.
CONCLUSIONS:
Recurrence rates were low following resection of stage 2 thymoma either with or without adjuvant radiotherapy. Adjuvant radiotherapy, although well-tolerated, did not significantly decrease the local relapse rate. Differences may be observed in future studies of patients who are at higher risk for local recurrence, based on completeness of resection, World Health Organization histology, and tumor size. Cancer 2011. © 2011 American Cancer Society.
Thymomas are the most common anterior mediastinal neoplasm in adults and the most common tumor of the thymus. There are 5 types of WHO histology possible (Table 1). Clinical staging is performed using 2 systems: an anatomical Masaoka staging system with a 1994 modification (Table 2) and a surgical resection-based Groupe d'Etudes des Tumeurs Thymiques staging system.1 The best predictors of recurrence of thymoma are Masaoka stage, World Health Organization (WHO) histological classification, and extent of surgical resection. Many studies have shown that both Masaoka stage and WHO histology are independent predictors of recurrence.2-10 Other studies have also shown vessel invasion and larger size to be predictive of recurrence.11, 12 Studies of thymoma are challenging: the disease course is prolonged, with recurrences occurring as late as 30 years after initial tumor resection, making extended follow-up important.
| Type | Histological Description |
|---|---|
| A | Medullary thymoma |
| AB | Mixed thymoma |
| B1 | Predominantly cortical thymoma |
| B2 | Cortical thymoma |
| B3 | Well-differentiated thymic carcinoma |
| Masaoka Stage | Diagnostic Criteria |
|---|---|
| Stage 1 | Macroscopically completely encapsulated and microscopically no capsular invasion |
| Stage 2 | Macroscopic invasion into surrounding fatty tissue or mediastinal pleura, or microscopic invasion into capsule |
| Stage 3 | Macroscopic invasion into neighboring organs (pericardium, great vessels, or lung) |
| Stage 4A | Pleural or pericardial dissemination |
| Stage 4B | Lymphogenous or hematogenous metastasis |
Surgery remains the cornerstone of treatment, and its completeness is related to the patient's outcome.13 A complete resection is the most important prognostic marker for disease-free survival. Chemotherapy is only used in unresectable or metastatic thymomas.14, 15
Radiation therapy (RT) is also indicated in many situations, including locally advanced or metastatic disease, incompletely resected tumor, or positive resection margins. In stage 2-4a thymomas, adjuvant radiation can decrease the recurrence rate significantly and potentially increase the 5-year overall survival (62% vs 18%).16, 17 The correct dose of radiation has been debated, but generally 45-50 Gy is given in 1.8-Gy fractions, with the exception of stage 3 disease (50-54 Gy) or gross residual tumor (54-60 Gy). Studies investigating the role of RT have only been performed retrospectively and have a wide range of results.
Specifically, the clinical benefit of postoperative mediastinal radiation for patients with completely resected Masaoka stage 2 thymoma remains controversial. Due to the indolent nature and infrequent recurrences of this tumor, no study has definitively determined the optimal approach. There are a lack of prospective data for surgery, chemotherapy, and radiation; and there are limited retrospective analyses that are available to help guide treatment.18 The purpose of this analysis was to determine the role of adjuvant radiation therapy in Masaoka stage 2 thymomas. This study is a retrospective chart review of patients who underwent surgical resection for thymoma at the Hospital of the University of Pennsylvania. We examined the patterns of failure in thymoma patients based on initial tumor characteristics, patient characteristics, and treatment given. Our objective was to determine whether postoperative radiotherapy improves local control in stage 2 thymoma and to identify prognostic factors for tumor recurrence.
MATERIALS AND METHODS
Approval for this study was granted by the Institutional Review Board of the University of Pennsylvania.
Patient Population and Initial Evaluation
We retrospectively reviewed 175 consecutive patients who underwent complete thymic resection from January 1990 to July 2008 at the University of Pennsylvania Medical Center. There were 62 patients with stage 2 thymomas with a minimum 6 months of follow-up according to the Masaoka staging system.1 We reviewed the records of these patients to extract the following information: medical record number; date of birth; telephone number; age; sex; Karnofsky performance status; presentation; presence of paraneoplastic syndrome (myasthenia gravis, pure red blood cell aplasia, hypogammaglobulinemia, limbic encephalitis, rheumatoid arthritis [RA], anti-DNS antibodies); “B” symptoms; histology; Masaoka stage; tumor size; invasion through capsule; intrathoracic or extrathoracic metastases; surgical approach; extent of surgical resection; surgical margins; any mutations known; radiation (technique, total dose, dose/fraction, number of fractions); chemotherapy (regimen, number of cycles, date); any further surgery, radiation, or chemotherapy; second primary cancers; recurrence and disease-free survival; metastases; and death.
Pathological Review
Operative notes were reviewed to determine intraoperative suspicion of invasion, gross tumor extension into adjacent structures, and completeness of resection. Pathology reports were obtained for all patients. All pathology specimens underwent re-review in the Department of Pathology at the University of Pennsylvania by V. Livolsi or L. Litzky and were assigned a WHO histological grade. The tumors were classified retrospectively into the 5 WHO histopathological classification subtypes (Table 1).
Pathological Staging
Staging was based on the surgical and pathological criteria described by Masaoka and colleagues in 1981 and modified in 1994 (Table 2).1, 19 Staging was retrospectively performed based on the review of the pathology report and operative notes of the surgeon. The staging was confirmed by unblinded centralized dual observer review of the pathology for the purpose of WHO histology subtype classification.
Adjuvant Radiation
Patients who were felt to be at high risk for recurrence by the operating surgeon were referred for adjuvant radiotherapy. The total doses, dose fractionation, and complications were reviewed for patients who underwent radiation therapy and are shown in Table 3.
| Overall | Observation | Radiotherapy | P | |
|---|---|---|---|---|
| No. of patients | 62 | 25 | 37 | |
| Mean age | 59.2 | 62.5 | 57.7 | .18b |
| Men/Women | 26/36 | |||
| Associated disorders,a No. (%) | ||||
| Myasthenia gravis | 20 (32) | 6 (24) | 14 (38) | |
| Pure red blood cell aplasia | 2 (3) | 0 (0) | 2 (5) | |
| Systemic lupus erythematosus | 2 (3) | 0 (0) | 2 (5) | |
| WHO histology, No. (%) | ||||
| A | 8 (13) | 6 (24) | 2 (5) | .13c |
| AB | 16 (26) | 5 (20) | 11 (20) | .58c |
| B | 30 (48) | 11 (44) | 19 (51) | .82c |
| B1 | 13 (21) | 6 (24) | 7 (19) | .76c |
| B2 | 14 (23) | 5 (20) | 9 (24) | 1.0c |
| B3 | 3 (5) | 0 (0) | 3 (8) | .28c |
| Unknown | 8 (13) | 3 (12) | 5 (14) | |
| Mean tumor size, cm | 5.8 | 5.3 | 6.1 | .49b |
| Margin status, No. (%) | ||||
| Positive | 15 (24) | 3 (12) | 12 (32) | .23c |
| Negative | 33 (53) | 17 (68) | 16 (43) | .39c |
| Close | 12 (19) | 3 (12) | 8 (22) | .51c |
| Unknown | 2 (3) | 2 (8) | 0 (0) | .17c |
| Mean follow-up, mo | 52.0 | 42.7 | 58.2 | .2b |
- WHO indicates World Health Organization.
- a The following disorders were not observed in the cohort: hypogammaglobulinemia, limbic encephalitis, rheumatoid arthritis, Sjogren syndrome.
- b Calculated using t test.
- c Calculated using chi-square test.
Follow-up
Survival and recurrence information were collected through chart review, patient or relative contact, and contact with personal physicians. Patients were contacted via telephone; if the patient could not be reached, a physician of that patient was contacted. Patients with ≤6 months of follow-up were excluded from statistical analysis.
Statistical Analysis
We compared outcomes of patients with stage 2 thymomas treated via resection alone (n = 25) versus resection plus radiation (n = 37). Some of these patients form the basis of a previous report.20 The Fisher exact test was used to estimate overall death rate and local recurrence rate. The Cox proportional hazards regression model was used to determine prognostic variables. P ≤ .05 was considered statistically significant. Survival time and time to recurrence was calculated from the date of surgery.
RESULTS
Patient Characteristics
Between January 1990 and July 2008, 175 patients underwent surgical resection for thymoma. Of these, 74 Masaoka stage 2 patients were resected. Sixty-two patients were identified with a minimum of 6 months of follow-up after surgery; myasthenia gravis was present in 20 (32.3%) of these patients. Two patients had pure red blood cell aplasia, and 2 patients had systemic lupus erythematosus. The mean tumor size was 5.8 cm (Table 3). Thirty-seven (61%) patients received adjuvant radiotherapy, and 25 (40%) patients were observed without intervention. The median radiation dose was 5040 cGy delivered in either 180 or 200 cGy fractions. The median tumor size in patients undergoing observation was 5.3 cm and was 6.1 cm for those receiving adjuvant radiation (P = .49). The treatment volume at the Hospital of the University of Pennsylvania varied: the dose ranged from 4500 to 7460 cGy.
Clinical Outcomes
The 5-year and 10-year survival of all patients was 96.7%. Local recurrence was defined as relapse within the surgical bed or radiated field. Two patients treated with surgery alone recurred (8.3%), whereas no patients treated with combination surgery and adjuvant radiotherapy recurred. No patients experienced distant relapse (outside of surgical bed or irradiated field).
The median follow-up of all patients was 52 months (42.7 months in the observation group and 58.2 months in the adjuvant radiation group; P = .2). The overall crude local recurrence rate was 3.2% (2/62). The proportion of recurrences in patients observed after surgery was 8.3% (2/24) and 0% (0/38) in patients who received adjuvant radiotherapy (P = .15, Fisher exact test). The first recurrence was a metastasis to the lung parenchyma and pleura at 26 months and was then treated with surgery, chemotherapy, and radiation. The second recurrence was a local 15 × 24 mm anterior mediastinal recurrence at 27 months that was treated with surgery and radiation only. The mean time to recurrence was 26.5 months. Both patients are now disease-free at 56 and 96 months, respectively. Both recurrences occurred in WHO histology B2.
Disease-free survival was 100%, and the overall death rate was 4.8% (3/62). One death occurred in the observation group and one in the radiation group (P = 1.0). The patients died of unrelated causes, and no patients were identified to have thymoma recurrence at the time of death (Table 4).
| Overall | Observation | Radiotherapy | Pa | |
|---|---|---|---|---|
| Local recurrence | 2 (3.2%) | 2 | 0 | .15 |
| Overall death rate | 2 (3.2%) | 1 | 1 | 1.0 |
- a Calculated using Fisher exact test.
Toxicity
Patients who underwent surgery alone experienced 8% grade 1-2 toxicity and 16% grade 3-4 toxicity. There were no postoperative deaths and 1 postoperative wound complication. Patients who underwent radiotherapy experienced 45.9% grade 1-2 esophagitis; 5% grade 1-2 pneumonitis; and no grade 3 or higher toxicity.
All patients underwent a total thymectomy via either a median sternotomy, transcervical, combined, or thoracotomy approach (Table 5). All patients were considered at the time of surgery to have undergone complete resections. There were no perioperative mortalities. Morbidity included 3 patients who developed atrial fibrillation postoperatively, a deep venous thrombosis, gastrointestinal bleed, and a sternal wound infection (Table 6). There were no mortalities due to radiation, nor any grade 3-4 toxicity. Grade 1-2 toxicities included erythema, dysphagia, odynophagia, cough, and fatigue (Table 6).
| Type of Resection | No. of Patients |
|---|---|
| Transsternal | 26 |
| Transcervical | 14 |
| Combined cervical and sternal incision | 12 |
| Thoracotomy | 2 |
| Unknown | 8 |
| Complication rates of surgery | Grade 1-2 Toxicity | Grade 3-4 Toxicity | |
|---|---|---|---|
| Atrial fibrillation | 1 | 2 | |
| DVT | 1 | 0 | |
| Gastrointestinal bleed | 0 | 1 | |
| Sternal wound infection | 0 | 1 | |
| Complication rates of adjuvant radiation | Grade 1 Toxicity | Grade 2 Toxicity | Grade 3-4 Toxicity |
| Erythema | 8 | 2 | 0 |
| Esophagitis | |||
| Dysphagia | 7 | 3 | 0 |
| Odynophagia | 1 | 0 | 0 |
| Other | 5 | 1 | 0 |
| Cough | 2 | 0 | 0 |
| Fatigue | 1 | 0 | 0 |
| Pericardial effusion | 1 | ||
Prognostic Factors for Recurrence
The following factors were examined for correlation with tumor recurrence: type and extent of surgical resection, margin status, adjuvant radiotherapy, WHO grade, tumor diameter, sex, and so forth. Given the small number of recurrences, there were no statistically significant factors identified that were prognostic for recurrence. Cox regression modeling demonstrated that size (P = .49) was not an independent predictor of recurrence. The distribution of WHO histologies between the observation and adjuvant radiation groups was not statistically significant. In the observation cohort, 12% of patients had positive margins and 32% in the adjuvant radiation cohort (P = .23) (Table 3). The majority of thymomas were found incidentally on imaging, 44% in the observation group, and 43% in the radiation group (Table 7).
| Presentation | Observation | Radiotherapy |
|---|---|---|
| Incidental finding on CXR, CT, MRI, or cardiac stress | 11 (44%) | 16 (43%) |
| Myasthenia gravis | 4 (16%) | 10 (27%) |
| Local symptoms (cough, dyspnea, chest pain) | 3 (5%) | 7 (19%) |
| Systemic lupus erythematosus | 0 (0%) | 1 (2.7%) |
| Unknown | 7 (28%) | 3 (8.1%) |
- CXR indicates chest x-ray; CT, computed tomography; MRI, magnetic resonance imaging.
DISCUSSION
Our report is amongst the largest retrospective studies of stage 2 thymoma. The results suggest that recurrence rates are low after complete resection of patients with stage 2 thymoma either with or without adjuvant radiotherapy; the rate of recurrence of reported here (3.8%) is lower than in other studies (up to 10%).10 We demonstrate no significant benefit to postoperative RT in decreasing local recurrence rates in stage 2 thymoma. There are likely subsets of patients who benefit from RT; these are patients who are generally considered by the operating surgeon to be at higher risk for recurrence based on completeness of resection, WHO histology, and tumor size.
Our patients had no radiation-induced grade 3-4 toxicity and only grade 1-2 toxicity, although the potential adverse effects of mediastinal radiation include radiation pneumonitis, pulmonary fibrosis, pericarditis with pericardial effusion, esophageal structure, and mediastinal fibrosis. Therefore, adjuvant radiotherapy in this patient population was well tolerated.
Classically, thymomas are considered to be at greatest risk for local rather than distant relapse after complete surgical resection; therefore, some have advocated for the use of radiation in stage 2 disease.21, 22 There are no prospective studies examining the role of adjuvant radiation in stage 2 thymoma and there are multiple retrospective studies that demonstrate conflicting results (Table 8). Curran et al17 reported 6 out of 18 mediastinal relapses in stage 2 thymoma patients without postoperative RT versus 0% for those with postoperative RT (n = 1). Monden et al22 at the Osaka University Medical school reported a reduction in recurrence rate from 29% to 8% with postoperative RT. Similarly, Kundel et al23 concluded that radiation therapy doses above 45 Gy in stage 2 disease reduce the risk of local recurrence (5-year relapse rate 63% in <45 Gy and 0% for >45 Gy). Notably the recurrences included 2 patients with thymic carcinoma. Forquer et al24 investigated the role of postoperative radiation in the Surveillance, Epidemiology, and End Results database. They found that, in stage 2-3 patients, the overall survival was improved with postoperative RT (76% vs 66%, P = .01), but there was no benefit if the patient had radical or total thymectomy.
| Study | No. of Stage 2 Patients | No. of Patients Receiving Adjuvant RT | No. of Recurrences in Adjuvant RT Group | No. of Recurrences in Observation Group | P of Recurrence in Adjuvant RT Group vs Observation Group |
|---|---|---|---|---|---|
| Current study | 62 | 37 | 0 | 2 | .15 |
| Rena et al27 | 58 | 26 | 3 | 2 | .432 |
| Wright et al26 | 59 | Unknown | 1 recurrence in all stage 2 | NA | |
| Singhal et al20 | 40 | 20 | 1 | 0 | .72 |
| Ruffini et al28 | 58 | 13 | 4 | 2 | .02 |
| Blumberg et al25 | 30 (with CR) | 17 | “Similar” recurrence rates | .21 | |
| Curran et al17 | 21 | 1 | 0 | 6 | .12 |
| Monden et al22 | 34 | Unknown | 8% | 29% | NA |
- RT indicates radiotherapy; NA, not available; CR, complete resection.
The retrospective results are mixed, with some series not showing a local control benefit to adjuvant radiation. Memorial Sloan Kettering Cancer Center reported on 30 stage 2 thymoma patients and found a nonsignificant difference between the radiation (n = 17) and no radiation groups (n = 9) in terms of local relapse.25 Massachusetts General Hospital has investigated stage 2 thymoma in great detail. They first suggested that patients with mixed or medullary histology could be spared adjuvant radiation.3 In a follow-up study by Wright et al,26 they found that adjuvant radiation was not an independent predictor of local recurrence on multivariable analysis in stage 2 and 3 patients. Another more recent study from Italy27 similarly did not show a significant difference in disease-free survival between radiated and nonradiated stage 2 thymomas. Finally, we have previously reported that adjuvant radiation did not improve survival in patients who undergo complete surgical resection for stage 2 thymoma.20 This report confirms and augments these findings in a larger patient cohort.
WHO histology has been shown to be useful in estimating prognosis, as WHO histology strongly correlates with tumor invasiveness and thus Masaoka stage. Wright et al26 proposed that the 6 histological classifications could be condensed for clinical prognostication into 4 groups: A (A and AB), early B (B1 and B2), B3, and C. The latter 2 are the more aggressive. Chen et al29 proposed that WHO subtypes A, AB, and B1 should not have postoperative RT. In our study, both recurrences occurred in B2 histology. Therefore, although we agree that Chen's conclusions are reasonable that A, AB, and B1 may avoid radiation. It seems as though B2 may fall into a more aggressive category, although given the small numbers of this study, this is not statistically substantiated.
Although increasing tumor size has been shown to be associated with greater recurrence, our study was not able to evaluate this factor due to the low number of recurrences. Studies that have shown a significant relationship between tumor size and recurrence have looked at all stages or stage 3 alone; our study only investigated stage 2 thymoma. One study found a threshold effect at 8 cm, with a large increase in recurrence compared with tumors <8 cm.26 In addition, our small sample size restricted the ability to find a statistically significant effect of tumor size. Thus, in addition to WHO histology, tumor size may be useful in prognostication and thus decision whether or not to irradiate stage 2 thymomas.
There are several limitations to our study. As a retrospective study, there may be selection bias as to which patients received radiation. If this were the case, the cases selected for radiation would be the ones perceived to be more aggressive, and if adjusted, this would likely not change our conclusion. Our study was also limited to stage 2 patients; this causes limitations in the ability to achieve statistical significance due to the small sample size. Also, thymomas are indolent tumors and can recur as late as 10 years after surgery. Finally, this was a single-institution experience, and as such these results may not be applicable to all patients with thymoma.
In contrast to other studies, the primary endpoint of this study was local control. We chose this endpoint because thymoma has a relatively indolent clinical course, with many patients dying of unrelated causes. We therefore examined the possibility of potential harmful adverse effects of radiation or surgery. Our analysis revealed no serious adverse events with radiation, and the mild adverse events resolved without intervention.
Conclusion
Postoperative radiation is well tolerated in patients with stage 2 thymoma. Our results show no benefit to postoperative RT in stage 2 thymoma as a group; however, a high-risk subset of patients may demonstrate a clear benefit from radiation in future studies. A prospective study of this disease with stratification by size and WHO histology may be useful in deciding which subset of stage 2 patients would benefit most. At our institution, the current policy is for all thymoma patients to be discussed in our multidisciplinary conference, ideally in advance of surgical resection. A comprehensive approach is then devised that is tailored to the individual patient based on completeness of resection, WHO histology, and tumor size.
CONFLICT OF INTEREST DISCLOSURES
The authors report no conflict of interests.
