Clinical Significance of WHO Classification and Cell Kinetic Study using PCNA and Ki-67 on Thymic Epithelial Neoplasms

Table 1. The definition of WHO classification of thymic epithelial tumors

Condensed abstract

Cell kinetic study to evaluate if there is a correlation between proliferation indices(PI) of the neoplastic thymic cells and histological type with stage and OS.

Table 2. Clinical stage and the incidence of invasive tumors according to WHO classification

Abstract Purpose

Thymic epithelial neoplasms(TEN) are rare tumors in which the importance of histology is controversial. We performed cell kinetic study to evaluate if there is a correlation between proliferation indices(PI) of the neoplastic cells and histological type with stage and OS.

Table 3. Correlation between PCNA and Ki-67 labeling indices and histological subtypes

Patients and methods

Retrospective study of patients diagnosed with TEN(1988–2000). Demographic and clinical variables and complete follow-up data were obtained. PI using PCNA and Ki-67 were analyzed.

Results. We studied 22 patients. Median age was 47 years(range 24-75). Histological classification: type A=6(27%), AB=4(18%), B1=1(4.5%), B2=6(27%), B3-C(thymic carcinoma)=5(23%). Invasive tumors were seen in 72.72% of cases. Although the mean values of PCNA and Ki-67 were higher in types B3-C(12.5% and 14.25%) than in types A-B1-B2(5.8% and 6.1%), there was no association with histological type. An association among histologic type (A-AB-B2 vs B3-C) and stage (I-II-III vs IV; p=0.001) was found. A complete resection was achieved in 74% of cases. Complete remission was achieved in 18/21 patients who received definitive treatment. There were 6(28.6%) perioperative deaths and 3 patients relapsed. Five-year DFS for patients with B3-C was 40% compared with 100% for other subtypes. The actuarial 5-year OS of patients with B3-C was 80%, whereas that of patients with other types was 88%. Stage(p=0.001) and histology B3-C(p=0.008) were the only significant variables in predicting recurrence. Stage IV (p=0.005) and age> 60years (p=0.03) increased the risk of death.

Discussion. The clinical behavior of TEN depends on stage and histology. There was no correlation between PI and stage or histology. A larger study is warranted.

Table 4. Univariate analysis of survival


Several classification systems for thymic epithelial neoplasm (TEN) are currently in use.1-5 The low incidence of these neoplasms and the lack of uniformity in the reported series made necessary the unification of criteria thus resulting in the World Health Organization (WHO) classification.6 A critical issue is the confuse nomenclature of these classifications and an absence of consensus regarding their prognostic value. Nowadays the most important predictive factors for outcome still are complete resection7 and stage according to Masaoka classification8 (I: totally encapsulated, II: invasion through capsule, III: invasion into organs -pericardium, lung, great vessels-, IVa: pleural or pericardial implants and IVb: hematogenous metastasis).

Efforts have been made to discover immunohistochemical markers that identify factors of poor prognosis9-11. Few studies have attempted to correlate the proliferation index (PI) with the histological type12. Cellular cycle is complex and some proteins are expressed exclusively in the phases in which they intervene and their presence can reflect the PI of the cell. Monoclonal antibodies directed against them have been developed allowing to study two groups of proteins: those acting on DNA synthesis during S-phase like PCNA (proliferating cell nuclear antigen)13 and those expressed in proliferating but not resting cells like Ki-6714,15. These methods have shown that in some neoplasms high proliferative rates correlate with poor prognosis. These data correlate with those obtained by flow cytometry16.

Figure 1. PCNA labeling indices vs histological types. There was no difference between the other histoloical types

Patients and Methods

Patients. Patients diagnosed with TEN in our institution between 1988 and 2000 were included in the analysis. Age, gender, clinical presentation, association with MG, stage, treatment, and complete follow-up date were obtained from clinical records.

Histological classification. Analysis of tissue samples and classification according to the WHO classification was carried out by a pathologist (AG) in a blinded fashion. TEN were divided in four groups depending on their nuclei characteristics (type A: spindle/oval shape, type B: dendritic or epitheloid appearance). Tumors combining both morphologies were designated as type AB, and type C (atypia and necrosis). Type B tumors were further subdivided into three categories (B1, B2 and B3) on the basis of the proportional increase in the epithelial component and cytologic atypia. (Table 1)

Immunohistochemistry. Studies were performed on paraffin embedded tissue sections in 14 cases using monoclonal antibodies directed against PCNA and Ki-67 (Dako, Carpinterra, CA) in an automated immunostaining equipment (Ventana, Nexes, Tucson, AZ) as suggested by the manufacturer. Labeling index was determined by light microscopy with an oil-immersion objective randomly counting 500 tumor epithelial cells and expressing the results as a percentage of positive cells.

Statistical analysis. Cross-tabulation between present status and clinical variables (age, gender, presence of MG, tumor weight, invasion, stage, histological type and PI using PCNA and Ki-67) were used to screen for significant associations using the Chi-square test for independence. The correlation between PCNA and Ki-67 labeling indices was carried out using the Spearman correlation test. Overall survival (OS) and disease-free survival (DFS) curves were plotted according to the method described by Kaplan and Meier and analyzed using the log-rank test. A p value was considered significant if it was below 0.05. The analysis was performed using the SPSS-10 commercial software package.

Clinical presentation and diagnostic methods. Twenty-six patients were diagnosed with TEN between 1988 and 2000. Four were excluded because of incomplete data. Sixteen patients (72%) male and 6(28%) female were included. Male-to-female ratio was 2.6:1. Mean age was 50.5 years (range 24-75). Sixteen patients (72%) had symptoms of MG at onset, 2(9%) presented with symptoms attributable to anterior mediastinal mass (i.e. shortness of breath and chest pain), 1 patient (4.5%) had disautonomia at onset and in 1(4.5%) the diagnosis was preceded 12 months by pure red blood cell aplasia. The tumor was discovered on routine chest radiograph in 2(9%) patients. No other paraneoplastic syndromes were found. According to Masaoka classification there were 6 patients in stage I, 5 in stage II, 5 in stage III and 6 in stage IV. Histological diagnosis was made in material obtained using fine-needle aspiration (4.5%), core biopsies (9%) or surgical resection (86.5%). According to WHO classification 27% were type A, 18% AB, 4.5% B1, 27% B2, and 23% B3-C types (Table 2).

The computed tomography (CT) was abnormal in all cases, suggesting an invasion that was proved by histological studies in 16 of them (72% sensitivity and 100% specificity). When invasive TEN were analyzed we found necrosis in 9/16 patients (56%; p<0.05) and mediastinal lymphadenopathy in 11/16 patients (69%; p<0.05).

Mean tumor weight was 70 grams (range 30–100) and mean diameter was 8.5 cm (range 3.5–15.1). There were no differences between invasive and no invasive TEN in weight (< 70 vs. > 70 gr.; p=0.09) or diameter (< 8.5 vs. > 8.5 cm.; p=0.6).

Determination of PCNA and Ki-67 labeling indices. Although mean values of both PI were higher in types B3-C (stage IV) as compared to types A-AB-B2 (Table 3 and figures 1 and 2), there was no association between levels of PCNA and Ki-67 (<7% vs. > 7%) and histological type (A-AB-B2 vs B3-C; p=0.099; p=0.48 respectively). The patients who relapsed showed higher PI (mean PCNA=12.5%; mean Ki-67=14.25%) than those who did not relapsed (PCNA=5.83% and Ki-67=6.31%), although the difference was not statistically significant. The Spearman’s correlation coefficient between PCNA and Ki-67 was 0.389. An association among histological type (A- AB-B2 vs B3-C) and stage (I-II-III vs IV) (p=0.001) was found.

Figure 2. Ki-67 labeling indices vs histological types. Thymic carcinoma showed higher labeling indices by ki-67 antibody compared with the A-AB-B2 but there were no significant differences.

Treatment and results

Complete resection was achieved in 16(74%) patients, partial resection in 5(23%) and 1(4.5%) was solely biopsed. The approach was medial sternotomy in 19 cases (86.71%) and in 2 cases (9.52%) it was posterolateral thoracotomy. Ten (45%) patients required resection of pericardium or lung. In this series 16(72%) patients presented with symptoms of MG. After definitive resection 8 (50%) patients remained asymptomatic, 7(44%) had partial improvement and 1(6%) did not respond. A patient develop gastric non-Hodgkin’s lymphoma 4 years after the thymoma diagnosis.

The mean follow-up was 58 months (range 12–228). Six patients (27%) received postoperative radiotherapy in doses varying from 40 to 60 Gy (complete resection=1 and partial resection=5). Of these 3(13.5%) had stage III (type A=1 and B2=2) TEN, and 3(13.5%) had stage IV thymic carcinoma. One patient did not accept treatment (stage IVa, B2) and is alive 24 months after diagnosis. Of the 21 patients who received definitive treatment, 18 achieved complete remission. Of these 12 (54.5%) were alive without relapse at last follow-up. There were 3(13.5%) perioperative deaths (2 stage II and 1 stage IV) due to myasthenic crisis, 3(13.6%) patients died of unrelated causes. Three patients (13.5%) relapsed during the follow-up period (B3=2, C=1) at 41, 61 and 24 months respectively. The relapse site was lung in 2 patients and retroperitoneum in 1. In all cases complete resection was possible. Two patients are disease-free at 12 and 18 months and the other developed a new recurrence 5 months after resection and is currently being treated with chemotherapy.

The recurrence probability of TEN was 51% at 24 months and 40% at 60 months. When A-AB-B2 and B3-C groups were compared, 5-year DFS was 100% and 40% respectively. On univariate analysis only stage (I, II, III vs IV; p=0.001) and histological classification (A, AB, B1, B2 vs B3 and C; p=0.008) were of value for the prediction of recurrence(Figure 3).

The actuarial 5-year survival rate for the entire group was 86%. For patients with type B3-C it was 80% at 5 year, whereas for patients with other types it was 88%. On univariate analysis both stage IV (p=0.005) and age > 60 years (p=0.03) were risk factors for death(Table 4). Histology (A-B1-B2 vs. B3-C) was correlated with stage (I-II-III vs IV, p=0.001). The actuarial 5-year survival rate according to Masaoka classification was: stage I=100%(6/6), II=60%(3/5), III=100%(5/5) and IV=83.3% (5/6)(Figure 4).

Higher absolute values of PCNA and Ki-67 expression was found in types B3-C than in types A, AB, B1, B2 TEN, thus reflecting a low growth proliferation in non-invasive or minimally invasive thymic epithelial neoplasias, although no statistically significant differences were observed.

In the reported series of TEN male and female patients16 are equally affected and 70% of cases occur during the sixth and seventh decades of life (mean age 53)17. In our series we found a slight male predominance(male-to-female ratio 2.6:1) and patients were younger(mean age 47). We also found a high prevalence of associated systemic syndromes and incidental finding was made only in 9% of our patients in contrast to other published data in which this occurs in 40 to 50% of patients18.
Imaging studies play an important role in the detection and staging of TEN. CT is the method of choice in the evaluation of a mediastinal mass with a 72% sensitivity and 100% specificity for invasive TEN. CT can differentiate cystic from solid lesions and the presence of fat or calcium within the lesion. The findings suggestive of invasion include necrosis and mediastinal lymphadenopathy as well as obliteration of mediastinal fat planes or poor demarcation of tumor from adjacent structures. Baron et al19 showed that in patients with proven thymomas, the CT showed an ovoid soft-tissue density mass in the anterior mediastinum and necrosis in 100% of patients. An important consideration to be made is whether the lesion determines the treatment plan.

Prognostic factors that have been associated with better survival include non-invasion, lower stage and complete surgical resection. Invasion is reported in 30% to 40% of the patients8. Bernatz et al.18 demonstrated an overall 15-year survival rate of 12.5% in the invasive group and 47% in the non invasive group. Several studies have examined the effect of the extent of surgical resection on OS and DFS. In 241 operative cases, Maggi and colleagues20 found an 82% OS rate in those patients whose tumors underwent complete resection and 26% survival rate at 7 years in those undergoing biopsy alone. Yagui et al21 reported excellent long-term survival with extended resection in patients with stages III–IV. Their overall 5- and 10-year survival rates were 77% and 59%, respectively. Ten out of 12 patients who underwent resection of the superior vena cava had a long term survival without evidence of recurrence. Nakahara and coworkers22 have shown that the survival rate in patients with stage III disease undergoing complete resection was comparable to those patients with stage I and II disease. Therefore, regardless of stage, tumor resection is one of the import predictors of treatment outcome. The Masaoka stage is also a prominent prognostic factor8. Stages I-II can be considered together as a group with a favorable prognosis; stages III-IV have a significantly worse prognosis. In general, complete resection rate is about 70%, in nearly all stage I-II and in 27%-44% of stage III patients14. In this series the complete resection rate was 71% (stage I-II: 100%, stage III-IV: 40%) which does not differ from published data.

Most TEN are slow-growing tumors and at diagnosis they present with a high tumor volume, the mean diameter in this series was of 8.5 cms. Caruso et al24 reported a mean diameter of 9.9 cm. Some authors suggest that recurrence rate seem to be related to the size and stage of TEN. No recurrence was observed if the tumor was less than 5 cm in diameter and a recurrence rate of 15% and 33% was observed if the tumor diameter was 5–15 cm and more than 15 cm. resepectively25. In our series this fact could not be confirmed but perhaps this is due to the number of patients studied.

When TEN was suspected and the preoperative evaluation suggested a resectable tumor, the surgical approach was medial sternotomy (86%), which provides a wide exposure of the anterior mediastinum26. Previous reports yielded an approximate 25% of invasive tumors.

In our series we found 22.3% non invasive and 72.7% invasive tumors. These percentages were higher than reported because carcinomas (22.72% in our series) were included in the analysis. Difference between these data and previously reported data is due to the fact that in previous reports carcinomas were not included. Using the WHO classification Okumura et al reported 53.6% of invasive tumors (12.14% of which were carcinomas)27. These data does not differ to much from ours.

The TEN are biologically heterogeneous tumors as it is demonstrated by their association with paraneoplastic syndromes28. We found a higher than expected prevalence of MG (72%), probably because our institution is a national referral center for complex diseases. Improvement in myasthenic symptoms is almost always recognized following thymectomy, and complete remission rates varies from 7% to 63%1. In our series it was 50%. The prevalence of association with MG in reference to Masaoka classification was higher among stages I-II (100%) than among stages III (60%) and IV (66%), but no significant difference in survival was noted between patients with invasive and noninvasive thymomas.

Currently, death after TEN resection in the perioperative period is rare and should be less than 6%29. In this series 3 deaths (13.6%) occurred, all of them in patients with poorly controlled MG and respiratory complications. Modern preoperative preparation, intensive care facilities and plasmapheresis have reduced this risk.

Verley and Hollman2, Maggi and colleagues20 and Lewis and co-workers3 did not find a significant difference in the prognosis of patients with and without MG. In the present study patients without MG had a higher survival rate (100%) than patients with MG (84%), although difference was not significant. The presence of MG may even confer a survival advantage, but this may be due to a predominance of incidentally discovered early-stage tumors in myasthenic patients.

Some authors consider the use of radiotherapy in all patients after either a complete or partial surgical resection and even as primary treatment29. In this series 27.3% received postoperative radiotherapy with adequate local control. In 10/11 cases in stages III-IV, complete remission was achieved without adjuvant therapy. These data demonstrate that the most important factor predicting outcome is the complete resection of the tumor. In patients who have residual macroscopic disease, radiation therapy achieves local control in 60% to 90% of cases regardless the stage. Three (13.6%) patients had distant relapses, similar to the reported 9-11%30. The actuarial 5-year survival rate was 86% in our series.

Figure 3. Disease-free survival for patients with types A-AB-B1-B2 and B3-C

Our results differ from those reported in regard to age, male-to-female ratio, association with MG, proportion of thymic carcinomas and OS rate. These differences cannot be explained only by a selection bias and they could show great variability in cell composition, histologic grow patterns and different biological behavior in our population. However, since this is the only study carried out in our country, we cannot fully support this notion.
We examined the clinical significance of histologic classification of TEN proposed by WHO. We analyzed separately the group of malignant thymoma and thymic carcinoma. Seventy-seven percent of cases were types A-AB-B1-B2. The higher proportion corresponded to types A and B2 (35.2%) and the less frequent was the B1 (4.55%). MG was associated to type A-AB-B1-B2 in 66.6%, 75%, 100% and 100% respectively. We found thymic carcinoma in 23% of cases, all in stage IV. The mean age in these patients was 48 years, similar to previously reported31. We did not found correlation between stage and invasion with PI.

Although TEN have been traditionally separated into benign and malignant categories, a reliable distinction cannot be made on the basis of either histopathologic or electron microscopic findings. Malignancy can only be demonstrated by the finding of invasion of the tumor capsule or surrounding organs, or by the presence of metastasis31.

The traditional classification based on invasiveness, number of lymphocytes and epithelial cell architecture is descriptive but straightforward and based on numerous clinicopathologic studies. The most consistent prognostic factor is the presence of invasion through the capsule1-3. In our series we found correlation between DFS and stage with WHO classification only in cases of thymic carcinoma. This fact could be explained because of the aggressive clinical course of this disease.

Few studies have correlated the proliferative index with the histological type. Chilosi and colleagues32 examined the proportion of proliferation in 8 cases of thymoma. Using the Ki-67 antigen they found a large increase in the activity of all examined samples ranging from 35% to 80% and the expression of this antigen was much higher than that seen in age matched control thymuses.

These authors speculated whether if this phenomenon might explain the pathogenesis of autoimmune diseases in these patients. Woo-Ick et al13 found a correlation between PCNA and Ki-67 IP (Spearman coefficient = 0.72). We could not confirm this finding. In summary analysis of a larger group of patients will be required to determine whether proliferation fraction as determined by this method can predict outcome.

Figure 4. Actuarial overall survival and disease-free survival for entire group of 22 thymic epithelial neoplasm

Address. Dr. Hugo Raúl Castro Salguero. Grupo Ángeles, 2da calle 25-19 zona 15, Vista Hermosa I, Edificio Multimédica, oficina 10-15. Telefax: (502) 23857572. E-mail.

Running title. Cell kinetics in thymic epithelial neoplasms
Key words: Thymic epithelial neoplasms, proliferation indices, histologic type, WHO classification


1. Loehrer P. Thymomas: current experience and future direction in therapy. Drugs 45(4):477-487, 1993
2. Verley JM, Hallman KH. Thymoma: a comparative study of clinical features, histologic features and survival in 200 cases. Cancer 55:1074-1086, 1985
3. Lewis RJ, Wick MR, Scheithaur BW, Taylor WF. Thymoma. A clinicopathologic review. Cancer 60:2727-2743, 1987
4. Marino M, Müller-Hemerlink HK. Thymoma and thymic carcinoma: relation of thymoma epithelial cells to the cortical and medullary differentiation of thymus. Virchows Arch 407:119-149, 1985
5. Suster S, Moran CA. Thymoma, atypical thymoma, and thymic carcinoma: a novel conceptual approach to the classification of thymic epithelial neoplasms. Am J Clin Pathol 111:826-833, 1999
6. Rosai J, Sobin LH. Histological typing of tumors of the thymus, International Histologic Classification of tumors, 2nd ed. New York, Berlin: Springer, 1999, pp 1 – 25
7. Loehrer P. Current approaches to the treatment of thymoma. Ann Med 31:73-9, 1999 (suppl 2)
8. Masaoka A, Monden Y, Nakahara K, Tanioka T. Follow-up study fo thymomas with special reference to their clinic stage. Cancer 48:2485-2492, 1981
9. Shimizu J, Hayashi Y, et al. Primary thymic carcinoma: a clinicopathological and immunohistochemical study. J Surg Oncol 56(3):159-64, 1994
10. Ring NP, Addis BJ. Thymoma: an integrated clinicopathological and immunohitochemical study. J Pathol 149(4):327-37, 1986
11. Pich A, Chiarle R, Chiusa L, et al. Argyrophilic nucleolar organizer region counts predict survival in thymoma. Cancer 74:1568-1574, 1994
12. Pich A, Chiarle R, Chiusa L, et al. Long-term survival of thymoma patients by histologic patients by histologic pattern and proliferative activity. Am Surg Pathol 19:918-926, 1995
13. Yang W, Efird J, Quintanilla L, Harris N. Cell kinetic study of thymic epithelial tumors using PCNA (PC10) and Ki-67 (MIB-1) antibodies. Human Pathol 27(1):70-76, 1996
14. Pan CC; Ho DM; Chen WY; Huang CW; Chiang H. Ki-67 labelling index correlates with stage and histology but no significantly with prognosis in thymoma. Histopatology 33(5):453-8, 1998
15. Gilhus NE; Jones M; Turley H; Gatter KC; Nagvekar N; et al. Oncogene protein and proliferation antigens in thymomas: increased expression of epidermal growth factor receptor and Ki-67 antigen. J Clin Pathol 48(5):447-55, 1995
16. Hainsworth J, Greco A. Mediastinal tumors in Haskell M (edd) Cancer treatment. Philadelphia, W.B. Saunders Company, 2001 pp 652 – 657
17. Lanfenfeld J, Graeber GM. Current management of thymoma. Surg Oncol Clin N Am 8(2):327-39, 1999
18. Bernatz PE, Khonsari S, Harrison EG, et al. Thymoma: factors incluencing prognosis. Surg Clin North Am 53:885-893, 1973
19. Baron R, Joseph K, Sagel S, Levitt R. Computed tomography of the abnormal thymus. Radiology 142:127-134, 1982
20. Maggi G, Casadio C, Caballo A, Cianci R, Molinatti M, Ruffini E. Thymoma: results of 241 operated cases. Ann Thorac Surg 51:152-9, 1991
21. Yagi K, Hirata T, Gukuse T, et al. Surgical treatment for invasive thymoma, especially when the superior vena cava is invaded. Ann Thorac Surg 61:521-524, 1996
22. Nakahara K, Ohno K, Hashimoto J, et al. Thymoma: results with complete resection and adjuvant postoperative irradiation in 141 consecutive patients. J Thorac Cardiovasc Surg 95:1041-1047, 1988
23. Caruso ES, Vasallo BC, Beveraggi EJ, Dalurzo L. Quistes y tumores del mediastino. Análisis de 100 observaciones. Rev Argen Ciruj. 1996 pp: 22 – 30
24. Gawrychowski J, Rokicki M, Gabriel A, et al. Thymoma: the usefulness of some prognostic factors for diagnosis and surgical treatment. Eur J of Surg Oncol 26:203-208, 2000
25. Adkins RB, Maples MD, Hainsworth JD. Primary malignant mediastinal tumors. Ann Thorac Surg 38(6):648-59, 1984
26. Okumura M, Miyoshi S, Yoshitaka F, et al. Clinical and functional significance of WHO classification on human thymic epithelial neoplasms: a study of 146 consecutive tumors. Am J Surg Pathol 25(1)103-110, 2001
27. Davis RD, Oldham HN, Sabiston DC. Primary cist and neoplasms of the mediastinum: recent chages in clinical presentation, methods of diagnosis, management and results. Ann Thorac Surgery 44:229, 1987
28. Wilkins EW Jr. Thymoma: surgical management, in Wood DE, Thomas CR (eds): Mediatinal tumors: Update 1995; Medical Radiology-diagnostic imaging and radiation oncology volume. Heidelberg, Germany, Springer-Verlag, 1995, pp 19-25
29. Hejna M, Habert I, Raderer M. Nonsurgical management of malignant thymoma. Cancer 88(9):1871-1884, 1999
30. Suster S, Rosai J. Thymic carcinoma: a clinicopathologic study of 60 cases. Cancer 67:1025-1032, 1991
31. Lanfenfeld J, Graeber GM. Current management of thymoma. Surg Oncol Clin N Am 8(2):327-39, 1999
32. Chilosi M, Iannucci A, Menestrina F, Lestani M, Scarpa A, et al. Immunohistochemidcal evidence of active thymocyte proliferation in thymoma. Its possible role in the patho 27:70 – 76, 1996
33. Woo-IcK Y, Efird J, Quintanilla-Martinez L, Choi N, Harris N. cell kinetic study of thymic epithelial tumors using PCNA (PC10) and Ki-67 (MIB-1) antibodies. Human Pathol 27:70 – 76, 1996

We are at your service

Visit Us

Learn more about our research, treatments and equipment

Make an appointment

We can help you on diagnostics and treatments

The right diagnostics

We offer the best equipment for early diagnostics