Thymomas: A Proposed Staging System
Thymomas: A Proposed Staging System
Our patient population included 120 males and 130 females between the ages of 13 and 92 years, with the great majority of cases occurring in the fifth, sixth, and seventh decades of life. The patients had diverse symptoms including chest pain, cough, and dyspnea. None of the patients had a history of myasthenia gravis. Of the patients, 6 had a history of malignancy: breast carcinoma, 2; prostatic carcinoma, 1; colonic adenocarcinoma, 2; and papillary thyroid carcinoma, 1. All patients underwent surgical resection of the mediastinal mass. Some of the most relevant clinical features and their statistical significance are given in Table 1. The median age was estimated at 57 years, and, when compared with the total number of cases, a significant P value was obtained (P < .0001). However, sex was not found to show statistical significance (P < .4954).
Grossly, tumor sizes varied from 3 to 20 cm in greatest diameter, with about 80% of the tumors 5 cm or more, with a median size of 7 cm. Tumor size, however, did not correlate with any particular stage. A nonsignificant statistical P value of .0702 was obtained (Table 1). Histologically, the tumors were divided according to their morphologic features following the grouping proposed by the World Health Organization. Type A thymoma represented 21.6% of the cases (54 cases), while thymomas B1, B2, and B3 represented 13.2%, 3.2%, and 9.2%, respectively. Approximately 53% of the tumors showed different morphologic growth patterns with combinations of types A, B1, B2, and B3. In addition, no statistically significant value was obtained by correlating histologic type with invasive or encapsulated tumors (P = .4074; Table 1). Table 2 depicts all histologic growth patterns that may be seen in thymomas in relation to the proposed staging system. These results support the concept that histologic type alone is not a predictor of clinical behavior.
At the time of diagnosis, according to our proposed staging system, 31 thymomas were stage 0 (encapsulated tumors), and 219 thymomas were invasive. Of the invasive thymomas, 128 tumors were stage I, 70 stage II (54, IIA; 14, IIB; 2, IIC), and 21 stage III (19, IIIA; 2, IIIB). By stratifying cases in these staging categories, we were able to obtain statistically significant overall survival curves with a P value of .044 Figure 1, while the recurrence-free survival P value was .015 Figure 2. Although we further stratified cases in stage II into A, B, and C, we did not obtain any statistically significant values among these categories. Similarly, we were not able to obtain a statistically significant P value between stage IIIA and IIIB. This latter issue may have been due to the lower number of cases among those categories.
(Enlarge Image)
Figure 1.
Prognosis analyzed by proposed stages 0-I vs II-III. Overall survival curve showing a statistically significant difference (P = .044) Kaplan-Meier analysis; cumulative proportion surviving, n = 231.
(Enlarge Image)
Figure 2.
Recurrence-free survival curve showing a statistically significant difference (P = .016). Kaplan-Meier analysis; cumulative proportion surviving, n = 231.
It is important to highlight that the overall survival and recurrence, based on our statistical analysis, are determined by the extent of infiltration. Patients with tumors with limited disease to the mediastinum (stages 0 and I) definitely fare much better than patients with tumors invading neighboring organs (stages II and III). Table 3 depicts the recurrence rate of these tumors with our proposed staging system and highlights that once the tumor invades adjacent structures, the chances of recurrence are higher than with tumors limited to the thymus. Also important to note is that once the tumor invades adjacent mediastinal structures, regardless of the structure affected, it is more likely that overall survival will be affected, as will recurrence-free survival, which in our analysis corresponded to P values of.044 and .015, respectively.
Follow-up information was obtained for 231 patients, ranging from 1 to 16 years (mean, 3.49 years). Of the patients, 185 were alive, of whom 22 had recurrent or metastatic disease, and 46 had a fatal outcome. We did not observe any recurrence in stage 0 cases, while there were recurrences in 7 of 115 in stage I, 15 of 66 in stage II, and 4 of 20 in stage III cases. It is interesting that 9 of 20 patients with stage IIIA tumors in whom follow-up was obtained died between 1 and 12 years after (Table 3). The only 2 patients with stage IIIB tumors, along with 19 other patients, were lost to follow-up. It is important to note that in not one of our cases did we identify "dissemination" of the tumor in pleura or pericardium. On the contrary, in all of our cases in which the tumor appeared to be invading those structures, it was by direct extension.
Results
Clinical Features
Our patient population included 120 males and 130 females between the ages of 13 and 92 years, with the great majority of cases occurring in the fifth, sixth, and seventh decades of life. The patients had diverse symptoms including chest pain, cough, and dyspnea. None of the patients had a history of myasthenia gravis. Of the patients, 6 had a history of malignancy: breast carcinoma, 2; prostatic carcinoma, 1; colonic adenocarcinoma, 2; and papillary thyroid carcinoma, 1. All patients underwent surgical resection of the mediastinal mass. Some of the most relevant clinical features and their statistical significance are given in Table 1. The median age was estimated at 57 years, and, when compared with the total number of cases, a significant P value was obtained (P < .0001). However, sex was not found to show statistical significance (P < .4954).
Pathologic Features
Grossly, tumor sizes varied from 3 to 20 cm in greatest diameter, with about 80% of the tumors 5 cm or more, with a median size of 7 cm. Tumor size, however, did not correlate with any particular stage. A nonsignificant statistical P value of .0702 was obtained (Table 1). Histologically, the tumors were divided according to their morphologic features following the grouping proposed by the World Health Organization. Type A thymoma represented 21.6% of the cases (54 cases), while thymomas B1, B2, and B3 represented 13.2%, 3.2%, and 9.2%, respectively. Approximately 53% of the tumors showed different morphologic growth patterns with combinations of types A, B1, B2, and B3. In addition, no statistically significant value was obtained by correlating histologic type with invasive or encapsulated tumors (P = .4074; Table 1). Table 2 depicts all histologic growth patterns that may be seen in thymomas in relation to the proposed staging system. These results support the concept that histologic type alone is not a predictor of clinical behavior.
Staging
At the time of diagnosis, according to our proposed staging system, 31 thymomas were stage 0 (encapsulated tumors), and 219 thymomas were invasive. Of the invasive thymomas, 128 tumors were stage I, 70 stage II (54, IIA; 14, IIB; 2, IIC), and 21 stage III (19, IIIA; 2, IIIB). By stratifying cases in these staging categories, we were able to obtain statistically significant overall survival curves with a P value of .044 Figure 1, while the recurrence-free survival P value was .015 Figure 2. Although we further stratified cases in stage II into A, B, and C, we did not obtain any statistically significant values among these categories. Similarly, we were not able to obtain a statistically significant P value between stage IIIA and IIIB. This latter issue may have been due to the lower number of cases among those categories.
(Enlarge Image)
Figure 1.
Prognosis analyzed by proposed stages 0-I vs II-III. Overall survival curve showing a statistically significant difference (P = .044) Kaplan-Meier analysis; cumulative proportion surviving, n = 231.
(Enlarge Image)
Figure 2.
Recurrence-free survival curve showing a statistically significant difference (P = .016). Kaplan-Meier analysis; cumulative proportion surviving, n = 231.
It is important to highlight that the overall survival and recurrence, based on our statistical analysis, are determined by the extent of infiltration. Patients with tumors with limited disease to the mediastinum (stages 0 and I) definitely fare much better than patients with tumors invading neighboring organs (stages II and III). Table 3 depicts the recurrence rate of these tumors with our proposed staging system and highlights that once the tumor invades adjacent structures, the chances of recurrence are higher than with tumors limited to the thymus. Also important to note is that once the tumor invades adjacent mediastinal structures, regardless of the structure affected, it is more likely that overall survival will be affected, as will recurrence-free survival, which in our analysis corresponded to P values of.044 and .015, respectively.
Follow-up information was obtained for 231 patients, ranging from 1 to 16 years (mean, 3.49 years). Of the patients, 185 were alive, of whom 22 had recurrent or metastatic disease, and 46 had a fatal outcome. We did not observe any recurrence in stage 0 cases, while there were recurrences in 7 of 115 in stage I, 15 of 66 in stage II, and 4 of 20 in stage III cases. It is interesting that 9 of 20 patients with stage IIIA tumors in whom follow-up was obtained died between 1 and 12 years after (Table 3). The only 2 patients with stage IIIB tumors, along with 19 other patients, were lost to follow-up. It is important to note that in not one of our cases did we identify "dissemination" of the tumor in pleura or pericardium. On the contrary, in all of our cases in which the tumor appeared to be invading those structures, it was by direct extension.
Source...