ICVTS Click here to goto Smart Canula website
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Interact CardioVasc Thorac Surg 2005;4:267-271. doi:10.1510/icvts.2004.097246
© 2005 European Association of Cardio-Thoracic Surgery
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Ad J.J.C. Bogers
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Kraker, M.
Right arrow Articles by Bogers, A. J.J.C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Kraker, M.
Right arrow Articles by Bogers, A. J.J.C.
Related Collections
Right arrow Mediastinum
Right arrow Minimally invasive surgery

Institutional report - Thoracic general

CT and myasthenia gravis: correlation between mediastinal imaging and histopathological findings{star}

Marjolein de Kraker1, Jolanda Kluin1,*, Nomdo Renken2, Alex P.W.M. Maat1 and Ad J.J.C. Bogers1

1 Department of Cardiothoracic Surgery, Erasmus MC, Dr. Molewaterplein 40, 3015 GD, Rotterdam, The Netherlands
2 Department of Radiology, Erasmus MC, Rotterdam, The Netherlands

Received 30 August 2004; received in revised form 22 February 2005; accepted 24 February 2005

{star} Presented at the joint 18th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 12th Annual Meeting of the European Society of Thoracic Surgeons, Leipzig, Germany, September 12-15, 2004.

*Corresponding author. Tel.: + 31- 647262040; fax: +31-104633993.

E-mail address: j.kluin{at}erasmusmc.n (J. Kluin).


   Abstract
Top
 Abstract
1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
 Appendix. Conference discussion
 Acknowledgments
 References
 
The surgical strategy in patients with myasthenia gravis (MG) is influenced by the suspicion of thymoma based on mediastinal imaging. Aim of this retrospective study was to analyse the accuracy of CT of the mediastinum in predicting the histological findings in patients with MG referred for thymectomy. Thirty-four CT-scans of MG patients referred for thymectomy between October 1989 and October 2003 were retrospectively evaluated by three cardio-thoracic surgeons and three radiologists. Data were analysed by Kappa statistics to judge inter-observer variance and were compared to the histopathological findings to determine predictive value. Observer agreement among the radiologists was fair (Kappa=0.28) and among the cardio-thoracic surgeons slight (Kappa=0.08). The average negative predictive value of no thymoma on CT was 91% (range 78–100%). The average positive predictive of thymoma on CT value was only 39% (range 29–58%). The average sensitivity of CT imaging in the study population was 75% (range 25–100%) and the average specificity was 62% (range 42–81%). In patients with MG undergoing thymectomy, CT is helpful in detecting thymoma, but the high inter-observer variation indicates that it remains difficult to distinguish lymphoid follicular hyperplasia from thymoma. This will influence the surgical strategy in patients with myasthenia gravis.

Key Words: Myasthenia gravis; Computed tomography; Multiple observer agreement; Thymoma


   1. Introduction
Top
 Abstract
 1. Introduction
2. Materials and methods
 3. Results
 4. Discussion
 Appendix. Conference discussion
 Acknowledgments
 References
 
Myasthenia gravis (MG) is an autoimmune disease that often is associated with thymic abnormalities. In most patients, thymic histology shows follicular hyperplasia, but 10–20% of patients with MG have thymoma [1]. Currently, the detection of ‘thymoma-specific’ antistratitional antibodies in the peripheral blood and mediastinal imaging by computed tomography (CT) are the most specific diagnostic procedures for the detection of thymoma. However, experience and previous publications have shown that it is difficult to differentiate thymic lymphoid follicular hyperplasia from thymoma on CT [1–4] Fig. 1.



View larger version (99K):
[in this window]
[in a new window]
  		Fig. 1. Mediastinal CT scan showing an invasive thymoma in a 55-year-old woman with myasthenia gravis.

 
Thymectomy has gained increasing acceptance as the most effective treatment for achieving sustained improvement in patients with generalised MG. In case of thymoma, radical thymectomy is indicated independent of the severity of MG, because of the risk of malignant degeneration. In MG it is generally accepted that complete removal of all thymic tissue is required to obtain the maximal therapeutic effect [5–9]. According to the recent review by Jaretzki, there is evidence that demonstrates that the extent of the various thymic resectional techniques is very variable and often incomplete. It shows again that the more complete the thymic resection, the better the results [8].

There is an ongoing debate about the most suitable surgical approach for thymectomy for the various indications [8]. In the Erasmus MC, since 2003, in patients with MG, thymectomy is performed by a video-assisted approach when no thymoma is suspected on CT. In case of thymoma thymectomy is not performed by a video assisted approach, because of the risk of incomplete resection that is attributable to technical difficulty and a limited field of vision. In patients with MG and thymoma suspected on CT, thymectomy is performed through a median sternotomy. Thus, CT diagnosis influences planning of the surgical strategy.

CT, as any clinical tool, be it qualitative or quantitative, is subject to some variability. Little is known about inter-observer variability of CT interpretation in MG patients. We therefore performed a retrospective study to evaluate the inter-observer agreement in CT-scanning in patients with MG referred for thymectomy among a panel of radiologists and thoracic surgeons. Additionally, CT-findings were correlated with the histopathological findings and the primary reviewer's findings.


   2. Materials and methods
Top
 Abstract
 1. Introduction
 2. Materials and methods
3. Results
 4. Discussion
 Appendix. Conference discussion
 Acknowledgments
 References
 
In our institution, between October 1989 and October 2003 preoperative CT scans of 34 MG patients who underwent thymectomy (study population) could be obtained (Fig. 2).



View larger version (88K):
[in this window]
[in a new window]
  		Fig. 2. Mediastinal CT scan of a 68-year-old male showing two lymph nodes (1–1.5 cm) in 1 mediastinal mass. Four out of 6 observers judged this CT scan positive for thymoma. Also antistratitional antibodies were positive in the peripheral blood. However, histology showed normal thymic tissue.

 
Thirty-one of these 34 patients underwent thymectomy by a median sternotomy and 3 patients by a thoracoscopic left sided three-trocar approach. Included were 14 males and 20 females, with a median age of 35 years (range 15–67 years). Thirty-three patients had generalised MG and 1 patient had only ocular MG. Six patients were positive for ‘thymoma-specific’ antistratitional antibodies.

2.1. Imaging protocol

Contiguous 0.5–1.0 cm collimation slides were obtained from Th1 to the diaphragm. Imaging parameters were 120–150 kV, 125–280 mA and slice thickness was 5–10 mm. Intravenously administered contrast medium was used in 27 patients.

CT scans were obtained with GE PACE (GE Healthcare, Fairlands, United States), Somaton Plus (Siemens, Erlangen, Germany) and Philips easy vision 4.3 scanners (Philips, Best, The Netherlands).

Scans were organised by patient and were then sorted into random order. All distinguishing patient and hospital information on the films was covered. All scans were interpreted independently.

The radiologist that evaluated the preoperative CT scans, upon which the surgical strategy was based, is referred as the ‘primary reviewer’. From 1989 until 2003, there have been several radiologists that have performed this pre-operative evaluation.

For the current study, three radiologist and three cardio-thoracic surgeons were selected to interpret the mediastinal CT-scans. The reviewers were only informed that patients were diagnosed with MG. They were blinded to the primary reviewer's interpretation, the assigned disease stage, the presence of antistratitional-antibodies, the patients’ age and outcome.

The scans were presented to all the reviewers in the same order. Reviewers were asked to interpret the CT-scans in terms of negative or positive for thymoma.

2.2. Microscopic evaluation

The microscopic diagnosis of lymphoid follicular hyperplasia was based on the presence of lymphoid follicles with germinal centres. Thymomas associated with MG are usually epithelial, lymphocytic or lymphoepithelial. Thymomas were histologically classified by the predominant cell type present in the tissue. The tumour extent was evaluated according to the Masaoka staging system [10].

2.3. Statistics

The diagnostic value of CT compared to histology was expressed in terms of sensitivity, specificity, positive predictive value and negative predictive value. Statistical analysis was performed with the {chi}2 test. Data were analysed by Kappa statistics to measure observer agreement for multiple reviewers, as outlined by Fleiss.

In summary, kappa is defined as the agreement possible beyond chance:

K=(O–C)/(1–C), where O is the observed agreement and C is the chance agreement. When K is zero, agreement is only at the level expected by chance. When K is 1, agreement has no element of chance. Kappa values of 0.00–0.20 represent slight agreement between observers, 0.21–0.40 represent fair agreement between observers, 0.41–0.60 represent moderate agreement between observers, 0.61–0.80 represent substantial agreement between observers and 0.81–1.00 represent almost perfect agreement between observers.


   3. Results
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
4. Discussion
 Appendix. Conference discussion
 Acknowledgments
 References
 
The primary reviewer suspected a thymoma on 12 out of 34 CT-scans (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1 Thymic histology, antistratitional antibodies and CT diagnosis of all individual myasthenia gravis patients referred for thymectomy

 
Histologic examination of these 12 patients showed thymoma in 7, hyperplasia in 2, normal thymic tissue in 1, sclerosed tissue in 1 and a cyst in 1, resulting in a positive predictive value of 58%.

A Masaoka stage I thymoma (capsulated) was found in 2 cases, stage II in 2 cases and stage III in 3 cases.

In 22 patients, the primary reviewer diagnosed the CT scan as negative for thymoma. Histopathological examination showed hyperplasia in 18, normal tissue in 2 and lipoma in 1. In only one patient, histology showed a thymoma (stage I).

The accuracy of the CT diagnoses (sensitivity, specificity, negative predictive value and positive predictive value) for each independent reviewer and the average values are shown in Table 2.


View this table:
[in this window]
[in a new window]
  		Table 2 Accuracy of CT diagnosis for each independent reviewer

 
The Kappa statistics for observer agreement among the radiologists was fair (kappa=0.28) and slight among the cardio-thoracic surgeons (kappa=0.08)

Overall agreement showed a Kappa value of 0.15.

Complete agreement was found in only 3/34 CT scans (9%, patient number 14, 27 and 29, Table 1). Five out of 6 reviewers agreed on the diagnosis in 11 scans (32%), 4 out of 6 reviewers agreed in 15/34 scans (44%) and 3 out of 6 reviewers agreed in 5/34 cases (15%).

As noted before, CT diagnosis and the presence of antistratitional-antibodies in the peripheral blood are currently the most specific diagnostic procedures in predicting the presence of thymoma in MG patients.

In 30/34 patients (88%), the presence or absence of antistratitional-antibodies was indicative for the presence or absence of thymoma.

In 3 of the remaining 4 patients (nr. 6, 20 and 24) the radiological diagnosis of most reviewers correlated well with the histopathological findings. In one patient (nr. 17) both antistratitional-antibodies and most reviewers predicted thymoma, while histology showed normal thymic tissue (Fig. 3).



View larger version (65K):
[in this window]
[in a new window]
  		Fig. 3. Preoperative CT scan showing thymic hyperplasia in a 31-year-old man with myasthenia gravis.

 

   4. Discussion
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
Appendix. Conference discussion
 Acknowledgments
 References
 
Nowadays, in many institutions the surgical strategy in patients with MG is influenced by the suspicion of thymoma based on mediastinal imaging. Currently, CT is thought to be the best imaging technique in detecting mediastinal abnormalities, but one must be familiar with potential pitfalls and controversies [3]. The current study was set up to evaluate the value of CT imaging in MG patients referred for thymectomy.

First we looked at the individual results. The primary reviewer detected all but one (7/8) thymoma. However, 5 patients with no thymoma on histology were mistaken for thymoma on CT by the primary reviewer.

Half of the review panel had more difficulty differentiating hyperplasia from thymoma, resulting in a less favourable specificity and positive predictive value.

The high negative predictive value of no thymoma on CT indicates a small risk of conversion from a video assisted to a transsternal approach due to the unexpected presence of thymoma. However, the rather low positive predictive value of thymoma on CT results in an unnecessary high number of more invasive procedures.

Besides the individual results, the high inter-observer variation indicates that it remains difficult to differentiate between hyperplasia and thymoma. This suggests that the evaluation is a highly subjective process and that CT has a limited role in the preoperative evaluation. Previous studies also have shown significant inter-observer variation in the detection of other (intrathoracic) malignancies by CT imaging [11–13].

We believe that better results probably could have been obtained when the presence of antistratitional-antibodies was known by the review panel, since these antibodies are predictive for the presence of a thymoma. However, this information was in most cases also not known by the primary reviewer.

The time range in which the scans were made, conducted over 14 years (1989–2003). This resulted in quality differences, differences in slice thickness and difference in the use of contrast. Though, most CT scans were conducted in the last 10 years and in 27 CT scans contrast was used. Thin-section CT enables the observation of lesion details; however, its effectiveness to enable the diagnosis of hyperplasia is unknown. Although contrast-enhanced CT reveals more abnormalities than unenhanced CT, the two independent observers in the study conducted by Nicolau et al. concluded that the use of intravenous contrast material was not particularly more helpful in enabling them to diagnose thymic abnormality [1,11,14,15]. The current review panel did report more difficulty with old scans and in scans without contrast.

The smaller agreement among the cardio-thoracic surgeons compared to the radiologists might have been explained by the difference in level of experience. Experience has been reported as an important influencing factor in observation errors. However, the study conducted by Nicolau et al. concluded differently [1]. They found experience to be of limited value. In the present study, years of experience did not correlate with predictive values.

Interpretation criteria like calcifications, thymic size and irregularity were not applied as standardised interpretation criteria in the independent analysis. The reason for this was that these criteria were not used as standardised criteria in our institution, by our radiologists, before we conducted this study. We believed that then false accuracy would have been applied in our study. Our review panel analysed the CT scans according to their technical knowledge and experience.

In conclusion, CT is helpful in detecting thymoma in patients with MG referred for thymectomy, but as indicated by a high inter-observer variation, it is difficult to distinguish lymphoid follicular hyperplasia from thymoma. To obtain the most accurate preoperative diagnosis, a CT scan with contrast should be made and all patient characteristics, especially, age, stage of MG and the presence of antistratitional-antibodies should be assessed. Overall, our results indicate the need of a randomised prospective study with standardised stratification and analysis criteria.


   Appendix. Conference discussion
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
 Appendix. Conference discussion
Acknowledgments
 References
 
Dr W. Walker (Edinburgh, UK): This is a study over 14 years. There must have been a change in your scanning apparatus over time. Did that influence that outcome of the comparisons?

Dr de Kraker: Yes. We believe that over a period of 14 years there are large quality differences in the scans because of the use of contrast, smaller slice thickness, and we believe that that is one of the limitations of our study, and that may have resulted in the less favorable results that we have found.

Dr A. Toker (Istanbul, Trukey): I would like to ask about the surgeons who reevaluated the computerized tomographies. How many thymic operations did the surgeons perform annually?

Dr de Kraker: I'm not quite sure how many operations they have performed, but we selected the three cardiothoracic surgeons who performed the most thymectomies in our center, and so they are also the surgeons who evaluate the CT scans the most in our center.

Dr Toker: Because I think surgeons who perform thymic operations frequently, could have better radiological interpretations.

Dr M. Zielinski (Zakapane, Poland): I agree with you that in cases of thymoma the transsternal approach should be the standard. Nevertheless, I would like to ask, what was the total number of patients operated on? The accuracy of CT is much lower than in my experience. In my experience, the accuracy is very high. I don't know why it was so low in your material. The other remark is, you said that in 15% of cases of myasthenia there is thymoma and 85% is hyperplasia. There are also some other causes.[fjIn nonthymomatous myasthenia, not only hyperplasia is the factor. There is also thymic involution and there is also the normal thymus.

Dr de Kraker: I would liek to say that the percentages of 15% and 85% was an overall amount. Indeed there is fatty involution, involution of the thymus gland, and even nomral thymic tissue in patients with MG. However, in this study we asked the observers to judge the scans in terms of positive or negative for thymoma, and the most problems occur when differentiating the hyperplasia from thymoma, and we know that by our own experience and by previous studies. Therefore we noted the relation between how rare a thymoma is and how common hyperplasia is. But you are right that there are still cases of normal tissue or involuted thymic tissue. To answer your question about the thoracic surgeons, we perform about 5 thymectomies a year in patients with MG, so it's a quite rare operation. It's not one of the most common operations performed in our institute.

Dr Van Raemdonck (Leuven, Belgium): Did you have a chance in your institution to look at PET scan to see if this could discriminate between a nodular lesion and follicular hyperplasia?

Dr de Kraker: In our instituation, at the moment we are investigating the use of MRI to see if that is better in differtiating the invasion of the tissue in the surrounding structures. So that is what we are investigating at the moment and looking to see if that can improve the preoperative evalution next to CT. That is the only other mediastinal imaging procedure that we are assessing right now, and for the future we would like to do CT scans always with contrast, and looking at the age of the patient, the presence of thymoma-specific antibodies, looking at all those features, we believe that that is when we can be the most accurate.


   Acknowledgments
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
 Appendix. Conference discussion
 Acknowledgments
References
 
Special thanks go to J.A. Bekkers, MD, A.P. Kappetein, MD, PhD, S. Frerichs, MD, R. Dwarskarsing, MD, for reviewing the CT scans and P.G.M. Mulder, PhD, for performing the statistical analysis.


   References
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
 Appendix. Conference discussion
 Acknowledgments
 References
 

  1. Nicolau S, Muller SL, Li DKB, Oger JJF. Thymus in myasthenia gravis: comparison of CT and pathological findings and clinical outcome after thymectomy. Radiology 1996;201:471–474.[Abstract/Free Full Text]
  2. Pirronti T, Rinaldi P, Batocchi AP, Evoli A, Di Schino C, Marano P. Thymic lesions and myasthenia gravis. Diagnosis based on mediastinal imaging and pathological findings. Acta Radiol 2002;43:380–384.[CrossRef][Medline]
  3. Brown LR, Muhm JR, Sheedy PF, Unni KK, Bernatz PE, Hermann RC jr. The value of computed tomography. AJR 1983;140:31–35.[Free Full Text]
  4. Camera L, Brunetti A, Romano M, Larobina M, Marano I, Salvatore M. Morphological imaging of thymic disorders. Ann Med 1999;31:57–62.
  5. Mack MJ. Video-assisted thoracoscopic thymectomy for myasthenia gravis. Chest Surg Clin North Am 2001;11:389–405.[Medline]
  6. Roth T, Ackermann R, Stein R, Inderbitzi R, Rösler K, Schmid RA. Thirteen years follow-up after radical transsternal thymectomy for myasthenia gravis. Do short-term results predict long-term outcome? Eur J Cardiothorac Surg 2002;21:664–670.[Abstract/Free Full Text]
  7. Mantegazza R, Baggi F, Bernasconi P, Antozzi C, Confalonieri P, Novellino L, Spinelli L, Ferro MT, Beghi E, Cornelio F. Video-assisted thoracoscopic extended thymectomy and extended transsternal thymectomy (T-3b) in non-thymomatous myasthenia gravis patients: remission after 6 years of follow-up. J Neurol Sci 2003;212:31–36.[CrossRef][Medline]
  8. Jaretzki A III. Thymectomy for Myasthenia Gravis: Analysis of controversies- patient management. The Neur 2003;9:77–92.
  9. Zielinski M, Kuzdal J, Szlubowski A, Soja J. Transcervical-subxiphoid-videothoracoscopic ‘maximal’ thymectomy-operative technique and early results. Ann Thorac Surg 2004;78:404–410.[Abstract/Free Full Text]
  10. Masaoka A, Yamakawa Y, Niwa H, Fukai I, Kondo S, Kobayashi M, Fujii Y, Monden Y. Extended thymectomy for myasthenia gravis patients: a 20-year review. Ann Thorac Surg 1996;62:853–859.[Abstract/Free Full Text]
  11. Fletcher BD, Glicksman AS, Gieser P. Interobserver variability in the detection of cervical-thoracic Hodgkin's disease by computed tomography. J Clin Oncol 1999;17:2153–2159.[Abstract/Free Full Text]
  12. Guyatt GH, Lefcoe M, Walter S, Cook D, Troyan S, Griffith L, King D, Zylak C, Hickey N, Carrier G. Inter-observer variation in computed tomographic evaluation of mediastinal lymph node size in patients with potentially resectable lung cancer. Chest 1995;107:166–119.
  13. Willimas JA, Kaste SC, Kaufmann WM, Winer-Muram H, Morris R, Luo X, Boyett JM. Use of chest computed tomography in the staging of pediatric Wilms’ tumor: inter-observer variability and prognostic significance. J Clin Oncol 1997;15:2631–2635.[Abstract/Free Full Text]
  14. Thorvinger B, Lyttkens K, Samuelsson L. Computed tomography of the thymusgland in myasthenia gravis. Acta Radiol 1987;28:399–401.[Medline]
  15. Cascade PN, Gross BH, Kazerooni EA, Quint LE, Francis IR, Strawderman M, Korobkin M. Variability in the detection of enlarged mediastinal lymph nodes in staging lung cancer: a comparison of contrast-enhanced and unenhanced CT. AJR 1998;170:927–931.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Eur. J. Cardiothorac. Surg.Home page
H. El-Bawab, A. A. Al-Sugair, M. Rafay, W. Hajjar, M. Mahdy, and K. Al-Kattan
Role of flourine-18 fluorodeoxyglucose positron emission tomography in thymic pathology
Eur. J. Cardiothorac. Surg., April 1, 2007; 31(4): 731 - 736.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Ad J.J.C. Bogers
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Kraker, M.
Right arrow Articles by Bogers, A. J.J.C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Kraker, M.
Right arrow Articles by Bogers, A. J.J.C.
Related Collections
Right arrow Mediastinum
Right arrow Minimally invasive surgery

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
ANN THORAC SURG ASIAN CARDIOVASC THORAC ANN EUR J CARDIOTHORAC SURG
J THORAC CARDIOVASC SURG ICVTS ALL CTSNet JOURNALS