Is thymectomy in non-thymomatous myasthenia gravis of any benefit?

Abstract

A best evidence topic in thoracic surgery was written according to a structured protocol. The question addressed was if thymectomy in non-thymomatous myasthenia gravis was of any benefit? Overall, 137 papers were found using the reported search, of which 16 represented the best evidence to answer the clinical question. The authors, journal, date and country of publication, patient group studied, study type, relevant outcomes and results of these papers are tabulated. The outcome variables were similar in all of the papers, including complete stable remission (CSR), pharmacological remission, age at presentation, gender, duration of symptoms, preoperative classification (Oosterhius, Osserman or myasthenia gravis Foundation of America (MGFA)), thymic pathology, preoperative medications (steroids, immunosuppressants), mortality and morbidity. We conclude that evidence-based reviews have shown that relative rates of thymectomy patients compared with non-thymectomy patients attaining outcome indicate that the former group of patients is more likely to achieve medication-free remission, become asymptomatic and clinically improve (54%, P < 0.01), particularly patients with severe and generalized symptoms (P = 0.007). Patients with generalized myasthenia gravis showed 11% stronger association with favourable outcomes after thymectomy. Some studies show early remission rates (RRs), as early as 6 months post-thymectomy, of 44%. Overall, the reported remission rate for non-thymomatous myasthenia gravis is between 38 and 72% up to 10 years of follow-up. Among these patients, those with thymic hyperplasia show the best complete stable remission rates (42%, P < 0.04) in the majority of studies. Age showed variability across the studies and the cut-off was also different among them. Overall age < 45 years showed a higher probability of achieving complete stable remission during follow-up (81% benefit rate (BR), P < 0.02). Pharmacological improvement is reported between 6 and 42%. However, the certainty of these benefits has not been established due to factors such as the confounding differences between myasthenia gravis patients receiving and not receiving thymectomy, the non-randomized nature of class II studies and the lack of Class I evidence to support its use. There is currently a randomized trial ongoing looking at thymectomy by sternotomy vs controls and the results are eagerly awaited.

INTRODUCTION

A best evidence topic was constructed according to a structured protocol. This is fully described in ICVTS [1].

THREE-PART QUESTION

In [non thymomatous Myasthenia Gravis] is [Thymectomy] of [any benefit].

CLINICAL SCENARIO

You see a 28-year old female in the outpatient clinic with a diagnosis of MGFA class III myasthenia gravis (MG), who has been referred to you by a neurologist. He wants to know your opinion on whether thymectomy will help control her symptoms and reduce the need for long-term immunosuppression. You feel that thymectomy could offer her a good outcome. You carry out a review of the literature.

SEARCH STRATEGY

Medline 1950 to May 2013 using OVID interface [thymectomy] AND [myasthenia OR myasthenia gravis] AND [improved outcome OR beneficial OR postoperative].

SEARCH OUTCOME

One hundred and thirty-seven papers were found using the reported search. From these, 16 papers were identified. That provided the best evidence to answer the question. These are presented in Table 1.

Table 1:

Best evidence papers

Author, date, journal, country Study type (level of evidence)	Patient group	Outcomes	Key results	Comments/weakness
Mantegazza et al. (1990), J Neurol, Italy [2] Retrospective study (level 2b)	1152 patients with MG from 6 centres 829 (72%) underwent thymectomy, 531 transcervical, 249 trans-sternal and 49 unspecified Osserman classes I–IV were included Mean follow-up was 4.9 years	Complete remission (CR) with thymectomy Pharmacological remission	11% 6%	Thymectomy seemed to raise the RR mostly for patients operated shortly after the diagnosis, generalized mild–moderate MG and normally involuted thymus Study period not specified, not clear indications for thymectomy; different and unspecified surgical techniques used; patients with missing data were included
Frist et al. (1994), Ann Thorac Surg, USA [3] Retrospective study (level 2b)	A total of 46 MG patients from 1971 until 1992 operated at a single institution with combined surgical and medical therapy The surgical approach was trans-sternal for all patients NT n = 42 Thymoma n = 4 Mean follow-up 75 months Oosterhius classification used	Age Sex Preoperative stage Duration of symptoms Mortality and morbidity	Patients <45 years had a better outcome (P = 0.0044) Female patients showed a better outcome (P = 0.06) Response to thymectomy improved dramatically with a more advanced preoperative stage (P = 0.02) No significant influence on outcome No operative/hospital deaths No recurrent nerve injuries	Thymectomy for MG is an effective therapeutic modality. Of the total, 87% patients had improvement and 28% achieved permanent remission Close correlation between postoperative status at 1, 6 and 12 months and status at last follow-up. Once patients clinically improve, they maintain that clinical state over time. Status at 1 month post-thymectomy strongly correlates with long-term outcome Retrospective uncontrolled study; small number of patients. Thymoma patients included
Masaoka et al. (1996), Ann Thorac Surg, Japan [4] Retrospective study (level 2b)	384 MG patients (286 non-thymomatous, 98 thymomatous) operated at 2 institutions from 1973 until 1993 Surgical approach for all non-thymomatous patients was trans-sternal extended thymectomy (en bloc resection of anterior mediastinal fat tissue, around upper poles of thymus, both brachiocephalic veins and on pericardium) Severity of disease by their own MG classification system Mean age 35 years Female predominance Follow-up to 20 years	Remission rate (RR) % (3 months, 6 months, 1 year, 3 years, 5 years, 10 years, 15 years, 20 years) Palliation rate (PR) % (3 months, 6 months, 1 year, 3 years, 5 years, 10 years, 15 years, 20 years) Age at time of operation Duration of disease Preoperative steroids Mortality	15.2, 15.9, 22.4, 36.9, 45.8, 55.7, 67.2, 50 74.2, 79.5, 86.3, 91.6, 92.2, 98.2, 91.7 <34 years old showed better RR (P < 0.05) <23 months showed better RR No difference in outcome when comparing patients who received steroids preoperatively and patients who did not receive them No operative deaths	Extended thymectomy is an excellent operative procedure in both non-thymomatous and thymomatous MG. The RR in the non-thymomatous group of patients continued to rise even after 5 years. PRs were stable after 3 years. These data suggest that the longer the postoperative period, the better are the results Thymoma patients included use of own classification system for disease severity; different protocols used for steroid administration
Venuta et al. (1999), Eur J Cardiothorac Surg, Italy [5] Retrospective study (level 2b)	A total of 217 patients [155 with non thymomatous myasthenia gravis (NTMG)] underwent thymectomy over a 27-year period at a single institution MG was graded as per Osserman classification 3 surgical techniques were used (cervicotomy, partial upper sternal splitting and complete sternotomy) Female predominance Follow-up 6–234 months (mean 119)	Complete remission Reduction of medication and/or clinical improvement Stable disease Age at operation Duration of symptoms Presence of thymoma MG severity Mortality (operative)	25% 46% 18% <45 years old correlated with remission or clinical improvement <18 months’ duration correlated with increased RR (P = 0.049) and PR (P = 0.041) Negatively influenced the prognosis RR Class IIA 86.4%, Class IIB 55.8, 38.5% Class III (P = 0.028) 2 patients (0.92%)	Thymectomy is effective in the management of patients with MG at all stages with low morbidity An upper split incision was favoured by this group as it allows extensive removal of thymic tissue; offers reduced surgical trauma with shorter length of stay Patients with thymoma present a less favourable outcome Different surgical techniques were employed Both patients with thymoma and non-thymomatous MG were included with no clear differentiation of results between the two groups
Gronseth et al. (2000), Neurology, USA [6] Meta-analysis (level 2a)	A definitive study of the effectiveness of thymectomy has never been done A Medline search was carried out to find studies (controlled non-randomized and uncontrolled case series) describing outcomes in patients with or without thymectomy Patients with thymoma were excluded	Survival Improvement since diagnosis Asymptomatic on or off medication Asymptomatic off medication MG severity Age and gender Medical therapy	Positive associations in most studies between thymectomy and MG remission and improvement There are confounding differences in baseline characteristics of prognostic importance between thymectomy and non-thymectomy patient groups in all studies Persistent positive associations between thymectomy and improved MG outcomes after controlling for single confounding variables such as age, gender and severity of MG Conflicting associations between thymectomy and improved MG outcomes in studies controlling for multiple confounding variables simultaneously	For patients with NTMG, thymectomy is recommended as an option to increase the probability of remission or improvement (Class II) There is no conclusive evidence of the superiority of one surgical technique to another as RRs data comes from uncontrolled studies
Budde et al. (2001), Ann Thorac Surg, USA [7] Retrospective study (level 2b)	A total of 113 patients underwent thymectomy for MG between 1974 and 1999 at a single institution 2 surgical techniques were employed - T incision limited upper sternotomy in 84% (2 × 2 in. T-shaped skin incision with the horizontal limb over the second intercostal space and vertical limb from the midpoint of the T down to the fourth intercostal space. A superior skin flap is created and the sternum is divided from the notch to the fourth intercostal space) - Full sternotomy in 16% Both NTGM and thymoma patients were included Follow-up was obtained in 92 patients (81%). Mean follow-up 51 ± 59 months	Remission Improved Remission + improvement (‘benefit’) Unchanged Worse Symptom duration Osserman classification Age Sex Thymic pathology Preoperative treatment Mortality Morbidity	21% 54% 75% 14% 11% did not correlate with outcome (average length of symptoms for worse patients was the lowest) Class 2 of 3 tended to benefit more (79%) <50 years (81% benefit vs 55% if >50 years, P = 0.02) Improved outcome for female patients in univariate analysis No difference in non-thymomatous MG No significant correlation with the outcome 1 patient (2 weeks postoperatively of unknown causes) 14%	Thymectomy may be safely performed by a limited T upper sternal incision. The proportion of overall improvement (75%) compares favourably with other large studies. The RR of 21% is slightly lower. However, comparison among studies is often difficult because of different exclusion criteria Complete stable remission (CSR) was achieved in patients at an average of 66 ± 71 months postoperatively, which is longer than the average follow-up of 51 months Multivariate analysis of patient sex was not statistically significant (P = 0.34) Preoperative stabilization of the disease included different regimes given along the period of study Two surgical techniques were employed The number of patients within each group is not clearly defined
De Perrot et al. (2001), Respiration, Switzerland [8] Retrospective study (level 2b)	A total of 35 patients underwent thymectomy between 1979 and 1999 at a single institution (23 patients with non-thymomatous MG) Complete follow-up available for 33 patients, with a mean of 96 months Modified Osserman classification was employed 4 different surgical techniques employed: transcervical (preferred), median sternotomy (if thymoma was suspected or complete thymectomy could not be performed through a cervicotomy), right or left thoracotomy (for tumours extending into the pleural cavity)	Age Sex Duration of symptoms Osserman stage Histological findings Mortality	The cut-off was 40 years of age. Remission and improvement were not statistically significant (P = 0.8 and 0.7, respectively) There was no statistically significant difference between male and female rates for remission and improvement (P = 0.4 and 0.8, respectively) No statistically significant difference for remission or improvement (P = 0.4 and 0.9, respectively) Greater clinical improvement in IIB, III, IV Class patients (P = 0.04) RRs better in the presence of thymic hyperplasia (P = 0.04) None for the non-thymomatous group	The large majority of patients (81%) improved after surgery. Postoperative clinical improvement was greater when extensive myasthenic involvement was present preoperatively RR was greater in patients with stage I and IIA (not statistically significant but this finding has been observed by other authors) Shorter duration of symptoms did not affect the RR in contrast to several other reports Small series including thymomatous and non-thymomatous MG Different surgical approaches
Gronseth et al. (2002), Neurology, USA [9] Review	Patients with MG from non-randomized trials dating back to 1953	Measured outcomes included survival, improvement since diagnosis, becoming asymptomatic on medication and medication-free remission	No Class I studies of the effectiveness of thymectomy have been performed. Most Class II studies demonstrated higher MG. RRs in patients undergoing thymectomy. However, these studies were consistently confounded by differences between MG patients	MG patients undergoing thymectomy in most studies were more likely to achieve medication-free remission, become asymptomatic and improve Due to the confounding differences between MG patients receiving and not receiving thymectomy, a benefit of thymectomy in MG patients has not been established
Mantegazza et al. (2003), J Neurol Sci, Italy [10] Prospective cohort study (1b)	A total of 206 NTMG patients Video assisted thoracoscopic extended thymectomy (VATET) n = 159 (transcervical incision with removal of thymic tissue in the neck. Subsequently, the sternum is lifted and trocars are introduced in the left pleural cavity to proceed via video-assisted thoracoscopy) Extended trans-sternal n = 47 Osserman classification I–V patients were included 6-year follow-up	CSR at the 6-year follow-up Mortality	Thymic hyperplasia, treatment with anticholinesterase drugs and age of onset <40 years were associated with a significantly greater probability of achieving CSR (P = 0.0001, <0.0001 and 0.037, respectively) Sex, time of onset to surgery, presence of anti-Ach-R antibody and type of surgery did not affect CSR 0	The absence of perioperative mortality and very low morbidity during the postoperative period demonstrate that VATET is safe CSR was achieved in 53.9% of patients at 6 years
Tansel et al. (2003), Surg Today, Turkey [11] Retrospective study (level 2b)	A total of 204 NTMG patients underwent thymectomy between 1980 and 2001 Surgical technique via partial median sternotomy (94%) and median sternotomy (6%) Modified Osserman classification employed Medical treatment included anticholinesterase therapy, steroids, combination of both, steroids + immunosuppressant or no medication Mean follow-up 7.2 ± 1.2 years with last follow-up in 79% of patients	Preoperative classification Preoperative treatment Age Duration of symptoms Gender Histological findings. Mortality	No significant correlation was found (P = 0.43) although patients with Class I and IIc2 tended to benefit more Not significant influence on outcome (P = 0.35) Did not influence remission or improvement rates (P = 0.42) Did not influence remission or improvement rates (P = 0.67) No influence on remission (P = 0.97) Favourable trend towards remission in patients with thymic hyperplasia (P = <0.001) No perioperative deaths	The early RR (44%) achieved after thymectomy had increased significantly by the end of the first year (72%, P = <0.001) Early and late RRs of 6 months and 1 year seem arbitrary
El-Medany et al. (2003), Asian Cardiovasc Thorac Ann, Saudi Arabia [12] Retrospective study (level 2b)	A total of 100 MG patients at a single institution between 1986 and 2001 (93 non-thymomatous and 7 thymoma-associated MG) Surgical technique: maximal thymectomy (combined transcervical and trans-sternal approach. Through the transcervical incision, thymic and fatty tissue anterior and anterolateral to the trachea from below the thyroid gland to the superior mediastinum is removed. In the mediastinum, the removal of tissue extends fatty tissue extending down to the diaphragm, between phrenic nerves, cardiophrenic tissue, retroinnominate and AP window) Osserman classification employed Follow-up 8–180 months (mean 91 months)	CR BR Age Sex Duration of symptoms (<1 year) Preoperative steroids Histology Ectopic thymic tissue	38.7% at last point of follow-up (increased progressively to reach a peak of 75% at 15 years) 86% (increased to 100% at 15 years) <50 years had better outcome (P = 0.0044) No statistically significance between female and male (BR 85 vs 75%) No significant correlation found with CR (36 vs 35.5%) CR in 27% of patients on steroids compared with 40% not on steroids (not statistically significant) CR in 42% of patients with hyperplastic thymus Poor prognostic factor (P = 0.0001)	Maximal thymectomy is an effective and safe procedure for treatment of MG The CR and total BR are prone to increase over time Univariate analysis showed that age, histology and ectopic thymic tissue are significant prognostic factors for outcome
Kawaguchi et al. (2007), Clin Neurol Neurosurg, Japan [13] Retrospective study (level 2b)	A total of 34 late-onset (age of onset >50 years) NTMG patients were selected 20 patients underwent thymectomy (approach not specified) and 14 medical treatment Clinical grade evaluated according to MGFA Subgroup analysis of MGFA Class 2 was performed Clinical course and outcomes over 2 years Mean follow-up for thymectomy patients was 11.7 years and 7.8 years for non-thymectomy patients	Minimal symptoms Generalized symptoms (thymectomy group vs non-thymectomy) Clinical remission	Present in 50% of thymectomy group patients at the end of the follow-up period 30 vs 75% (P < 0.05) 50 vs 17% (P = 0.11)	Thymectomy is a potentially effective treatment for late-onset NTMG with mild generalized symptoms Small sample, only assessing subgroup analysis
Sonett et al. (2008), Ann N Y Acad Sci, USA [14] Review	Review attempting to clarify some of the controversial issues concerning the selection of a thymectomy technique in the treatment of NTMG and to make limited recommendations based on the best available evidence This analysis consists of uncontrolled retrospective studies with an extense number of confounding factors making the analysis speculative	Extent of thymic tissue removal Remission Disease severity Duration of symptoms Surgical technique	The more thymus removed, the higher the RR 51 and 50% at 5 years for VATET and combined transcervival–trans-sternal Less severe disease has better outcomes Better outcomes for shorter duration Combined transcervical and trans-sternal maximal thymectomy should remain the benchmark	At 5 years maximal thymectomy continues to produce the most effective overall response Controlled well-designed studies are required to begin to resolve the many conflicting statements and unanswered questions that exist concerning the selection of thymectomy in the treatment of MG
Pompeo et al. (2009), Eur J Cardiothorac Surg, Italy [15] Retrospective study (level 2b)	A total of 32 patients with NTMG underwent extended thoracoscopic thymectomy (4-trocars access with removal of the entire thymus, anterior mediastinal perithymic tissues, fatty tissue in the aortocaval groove, AP window, cardiophrenic sinuses and lower cervical area) MGFA clinical classification was used CR and symptomatic improvement were assessed Follow-up 60–156 months (median 119)	Sex Symptom duration (<12 months) MGFA class Oropharyngeal involvement Histology Ectopic thymic tissue Anti-AChRab Anti-MuSKab Mortality	No significant difference between male and female (P = 0.1) Shorter duration significantly correlated with CR and improvement (P = 0.006) Did not impact on CR or improvement (P = 1.0) No involvement significantly correlated with CR and improvement (P = 0.01) No statistically significant difference (P = 0.06) Presence was associated with negative impact on CR and improvement (P = 0.05) Does not influence outcome (P = 0.12) Significant correlation with CR and improvement (P = 0.0007) None	Extended thymectomy results in highly satisfactory long-term outcome in non-thymomatous MG with a 10-year remission of 50% and an overall response rate of 90% Patients who did not respond to thymectomy had a positive titre of anti-MuSKab (predictor of no response at univariate analysis) Results of thymectomy continue to improve over time and adequate length of follow-up is needed to assess the efficacy of any thymectomy technique Small cohort
Lin et al. (2010), Eur J Cardiothorac Surg, Taiwan [16] Retrospective study (level 2b)	A total of 60 NTMG patients underwent thymectomy from 1995 to 2004 at a single institution Preoperative status was classified according to MGFA classification Two surgical approaches were employed (trans-sternal thymectomy n = 22 and video assisted thoracoscopic surgery (VATS) thymectomy n = 38) Follow-up range 12–131 months (median 44 months)	Crude CSR rate Sex, disease duration, MGFA classification, anti-AchR antibody, preoperative plasma exchange, preoperative medication and operative method Age of onset (<40 years), presence of hypothyroidism and thymic hyperplasia Mortality Morbidity	32% at 38.5 month mean follow-up Did not influence CSR rate (P = 0.7, 0.21, 0.79, 0.32, 0.88, 0.3, 0.91, respectively) Higher probability of achieving CSR (P = 0.022, 0.003, 0.041, respectively) 0 5%	VATS thymectomy is an advantageous procedure for treating NTMG patients compared with trans-sternal approach offering equivalent CSR rates The acceptable CSR, high improvement rate, short hospital stay, low conversion rate, low morbidity rate and no mortality demonstrate that VATS thymectomy is a safe and effective method for NTGM patients The role of thymectomy for ocular MG remains controversial Small sample including 2 surgical techniques and comparing data between them MGFA classification was different in both groups
Spillane et al. (2013), J Neurol, UK [17] Retrospective study (level 2b)	A total of 89 MG patients underwent extended trans-sternal thymectomy over a 12-year period (1999–2011) Thymoma and non-thymoma patients included MGFA classification used for preoperative assessment of disease severity and post-intervention status Follow-up 0.5–11 years, last clinical review mean of 3.8 years	CSR (at last clinical review) PR Improved status Steroid requirement after thymectomy Duration of symptoms (<2 years) Thymic histology (hyperplasia vs other histology) Mortality Morbidity	34% 33% 13% Fell from 73% preoperatively to 47% (P < 0.01) CSR 40 vs 33% (P = 0.19) CSR 42 vs 26% (P < 0.05) 0 9%	Thymectomy is a safe and well-tolerated procedure generally followed by a long-term substantial improvement in myasthenic symptoms in the majority of patients. No correlation between duration of symptoms and response to thymectomy

RESULTS

Mantegazza et al. [2] in their large multicentre follow-up reported a complete stable remission (CSR) of 11% at the 5-year follow-up. They concluded that thymectomy raised the remission rate (RR) in patients operated shortly after diagnosis, generalized mild-to-moderate MG and involuted thymus.

Frist et al. [3] reported that age <45 years (P = 0.004), female sex (P = 0.03) and preoperative stage (P = 0.021) had a positive impact on outcome.

Masaoka et al. [4] found in their study an RR of 67.2% at 15 years and 50% at 20 years after extended thymectomy. Age <34 years (P < 0.05) and shorter duration of symptoms (<2 years, P < 0.05) were favourable prognostic factors.

Venuta et al. [5] in their 27-year experience review reported a 25% CSR at the 10-year follow-up. CSR in patients with <18 months’ duration of symptoms preoperatively was 26.8 vs 17.1% (P = 0.049). More importantly, palliation (improvement in activities and/or receiving less medication) in this group of patients was also significant (51.8 vs 19.5%, P = 0.041).

Gronseth et al. [6] reviewed the literature available before the year 2000 and developed evidence-based recommendations. After systematically reviewing the non-randomized studies describing outcomes, they recommended thymectomy as an option to increase the probability of remission or improvement. In patients who underwent thymectomy for severe MG (symptoms and Osserman >2b) relative rates of better outcomes were higher (P 0.06 and 0.007, respectively).

Budde et al. [7] conducted a retrospective review on 113 patients after radical thymectomy and found a 21% CSR, 54% improvement and total benefit of 75%. Patients <50 years (81 vs 55%, P = 0.02) and female sex correlated with better outcome (P = 0.045).

De Perrot et al. [8] found that 81% of their patients improved after thymectomy. Postoperative improvement was greater in patients with advance MG and thymic hyperplasia (P = 0.04).

Gronseth et al. [9] reported in 2002 yet with another review. They concluded that because of confounding differences between surgical and non-surgical therapies for MG the role of thymectomy remains controversial. However, patients with severe MG who undergo thymectomy show better outcomes.

Mantegazza et al. [10] reported that video assisted thoracoscopic extended thymectomy (VATET) is a safe procedure for the treatment of non thymomatous myasthenia gravis (NTMG) with CSR rate of 53.9% at 6 years. It offers similar efficacy to extended trans-sternal approach. They found that thymic hyperplasia, age ≤40 years and anticholinesterase medications were associated with greater probability of achieving CSR (P = 0.0001, P < 0.0001, P = 0.037, respectively).

Tansel et al. [11] found early and late postoperative RRs of 44.6 and 73%, respectively. They found no correlation between sex, age, duration of symptoms and preoperative classification. However, symptomatic improvement increased in late follow-up.

El-Medany et al. [12] reported that patients <50 years had better outcome (P = 0.0044). CSR and benefit rate (BRs) achieved a peak of 75 and 100% at 15 years, respectively.

Kawaguchi et al. [13] reported on a small series of patients with late-onset MG (>50 years old). They performed subgroup analysis on MGFA Class II patients showing a better clinical RR although this was not statistically significant (P = 0.11).

Sonett et al. [14] made recommendations on the best surgical approach based on the best evidence available. The data favour the use of combined trans-sternal–transcervical maximal thymectomy. However, they acknowledge that due to the lack of stronger evidence it is not possible to state with certainty which is the best technique.

Pompeo et al. [15] reported a 44% RR after 119 months of follow-up. They found thoracoscopic extended thymectomy a reliable alternative to standard open procedures.

Lin et al. [16] analysed the prognostic factors of VATS thymectomy vs trans-sternal thymectomy. Owing to the small numbers of patients in each cohort they found no statistically significant difference in achieving CSR (P = 0.91).

Spillane et al. [17] reported CSR in 34% of their patients at a mean follow-up of 3.8 years and 33% pharmacological remission. Hyperplastic thymus was associated with better outcome (P < 0.05).

CLINICAL BOTTOM LINE

A definitive study on the effectiveness of thymectomy in NTMG patients has not been done. Therefore, the role of this procedure remains uncertain after its introduction by Blalock in 1936. The evidence we have currently is from Class II studies that are affected by different classification systems and reporting crude rates of remission and improvement. Moreover, patients selected for thymectomy usually have aggressive preoperative medication regimes. Also, retrospective studies do not assess the heterogeneity of MG. Terminology employed such as remission and improvement was widely used with different meanings and only recently these terms have been standardized. A comparative analysis based on existing data is therefore difficult to perform. The existing data seem to support an extended resection that removes as much thymic tissue as possible. The ideal method to overcome these problems is a prospective randomized trial, Class I evidence. As highlighted by Gronseth [9], thymectomy in this group of patients should only be considered as an option to increase the chances of remission or improvement. There is an ongoing trial of thymectomy vs no thymectomy in patients receiving steroids. Results are eagerly awaited.

Conflict of interest: none declared.