Impact of minimally invasive trans-cervical thymectomy on outcome in patients with myasthenia gravis

Abstract

Objective: To study the impact of minimally invasive trans-cervical thymectomy on the incidence of remission of patients with myasthenia gravis (MG) in a single institution over a 10-year period. Methods: A total of 120 consecutive patients referred for video-assisted trans-cervical thymectomy between 1991 and 2000 were included in the analysis. Complete remission was defined as no symptoms and no treatment for 6 months, and remission as minimal ocular symptoms (slight ptosis) or treatment with pyridostigmine only for 6 months. Results: There were 86 females and 34 males with a median age of 33 (range 14–79) and 36 years (range 12–68), respectively. Symptoms of MG lasted between 2 months and 17 years before thymectomy (median 10 months). Surgery was converted to a partial upper sternotomy in 23 cases (19%). The median hospital stay decreased from 2 days (range 1–8) before 1994 to 1 day (range 1–8) thereafter (p<0.0001). Postoperative complications occurred in four patients (3.3%). After a median follow-up of 48 months (range 6–117 months), 50% of the patients were in complete remission (41%) or in remission (9%). Kaplan–Meier estimates rate of complete remission were 30% after 5 years of follow-up and 91% after 10 years. Conclusions: Minimally invasive trans-cervical thymectomy can be performed with short hospital stay and low morbidity, and achieve excellent durable results at 10 years.

1 Introduction

Myasthenia gravis (MG) is an autoimmune disorder diagnosed clinically for the first time at the end of the 17th century and currently affecting approximately 25,000 people in the US. Intense research into the pathogenesis of the disease has led to the discovery of an antibody directed against the muscle nicotinic acetylcholine receptor (AchR) in 1973 and, more recently, to the finding of a new antibody directed against the muscle-specific receptor tyrosine kinase (MuSK) in a subgroup of patients that remain seronegative for AchR antibodies [1].

Blalock's initial experience in the early 1940s and subsequently published series of surgically treated patients led to the widespread acceptance of thymectomy in the treatment of MG despite the absence of a prospective randomized trial comparing surgery with medical treatment alone [2]. Controversies remain, however, with regard to the timing and extent of surgery to be performed. Some authors recommend thymectomy early in the course of the disease, whereas others reserve surgery for when medical therapy fails or if a thymoma is suspected. Different surgical approaches have been recommended including trans-cervical, trans-sternal, and more recently a trans-thoracic thoracoscopic thymectomy [3–5]. All procedures allow extracapsular resection of the thymus and vary somewhat in the extent of mediastinal fat removal, which may contain foci of thymic tissue [6,7]. The most extensive resection combines the trans-cervical and trans-sternal thymectomy procedures and includes removal of all mediastinal fatty tissue, both sheets of mediastinal pleura along with a sharp dissection of the pericardium [8].

The trans-cervical approach was first described at the turn of the 20th century for thymic enlargement in children and consisted of an enucleation of the thymus from within its capsule [9]. Although initially reported in an adult patient with MG by Sauerbruch in 1912 [10], the trans-cervical approach was modified to completely remove the thymus with its capsule and reintroduced in the 1960s for patients with MG [11]. Through this approach, Kark and Kirschner reported fewer postoperative complications when compared to the trans-sternal approach [11]. Consequently, their patients were operated on earlier in the course of the disease and were shown to have more rapid rate of improvement [12].

In our institution, we have used a similar approach associated with the routine use of a videothoracoscope introduced through the cervicotomy to improve visualization of the mediastinum. This approach combined with early surgical referral, optimization of preoperative medical status when necessary by plasmapheresis, and careful perioperative management has led to greatly improved care of patients with MG. Herein, we report the short and long-term outcome in a series of 120 consecutive patients operated upon in our center.

2 Material and methods

2.1 Patient characteristics

All patients undergoing trans-cervical thymectomy at the Toronto General Hospital between January 1, 1991 and December 31, 2000 were reviewed after approval by the Institutional Review Board. Patient's age, gender, duration of symptoms before surgery, and postoperative complications were retrieved from hospital charts. The referring neurologists graded MG according to the modified Osserman classification as 0= asymptomatic, 1 = ocular signs and symptoms, 2= mild generalized weakness, 3 = moderate generalized weakness, bulbar dysfunction, or both, and 4 = severe generalized weakness, respiratory dysfunction or both. Conversion to an upper sternotomy was determined from the operative report, and pathological findings were obtained from the pathology report.

2.2 Preoperative care and timing of operation

The diagnosis of MG was confirmed by physical examination, a positive response to anticholinesterase test, and single fiber electromyography (EMG). Detection of AchR antibodies by radioimmunoassay was not performed routinely. After the diagnosis was confirmed, MG symptoms were medically stabilized and patients were referred to the surgeon for thymectomy. In our institution, all patients with MG are referred to one surgeon (SK) for evaluation of thymectomy, and all patients undergo video-assisted trans-cervical thymectomy unless there is a contraindication, e.g. the presence of a thymoma on the computed tomography (CT) scan. Patients were usually referred to surgery unless they were over 50 years of age or with ocular symptoms only, in which case the decision to proceed with surgery was made on a case by case basis. Pyridostigmine was used as the first line of therapy. Prednisone and other immunosuppressive agents were started in the presence of persistent symptoms. A course of five plasmapheresis treatment was administered before surgery in patients with severe generalized weakness and/or with bulbar symptoms. Since 1998, a few patients with severe bulbar symptoms have been treated with high dose intravenous immunoglobulin (IVIg) instead of plasmapheresis prior to surgery. Over the years, we have progressively limited the administration of immunosuppressive therapy before surgery and preoperative stabilization is most often accomplished with pyridostigmine and plasmapheresis only.

All patients had a CT scan of the thorax before surgery in order to exclude a thymoma. If a thymoma was detected a trans-sternal approach was always chosen. Relative contraindications to a trans-cervical approach included prior cervico-mediastinal surgery and/or radiation, and cervical spine pathology limiting extension of the neck. Age, gender, obesity, and exposure to steroids were not considered contraindications to the trans-cervical approach.

2.3 Perioperative management

Anesthetic assessment was performed at an ambulatory preadmission clinic visit. Patients were admitted on the day of surgery. They either took their morning dose of pyridostigmine as usually scheduled or took it immediately prior to surgery. If surgery was delayed or scheduled for the afternoon, another dose of pyridostigmine was given before surgery. No other premedication was administered. Anesthesia was induced with propofol and fentanyl, and was maintained with isoflurane and nitrous oxide. Propofol was often used in addition to inhaled anesthetics for maintenance of anesthesia. Muscle relaxation was rarely required. Patients were routinely extubated at the end of surgery. Analgesia was given orally using acetaminophen with or without codeine. Morphine was rarely required. Oral pyridostigmine at the patient's usual dose was reintroduced 4–6 h after surgery. If patients were on steroid therapy before surgery, an intravenous dose was given preoperatively and oral steroids were continued the next morning. Patients were ready to be discharged the next day if their symptoms were stable and pain was controlled. Postoperative complications included all complications occurring within 30 days of surgery.

2.4 Surgical procedure

Surgery is performed in the supine position. The neck and full anterior chest were prepped in case a sternotomy is required. A curvilinear incision is made in the skin at the base of the neck, one finger breadth above the sternal notch, and extended on each side to the medial border of the sternocleidomastoid muscle. The superior poles of the thymus gland are dissected, and the thymus gland is followed inferiorly to the thoracic inlet. A retrosternal space is cleared to accommodate the placement of the Cooper retractor [13] and the upper hand retractor (Poly-Tract, Pilling Company, Fort Washington, PA) is then set up with the Cooper thymectomy retractor blade (Pilling Company), which is then placed beneath the manubrium to elevate it and open the thoracic inlet. Care is taken to make sure that the patient's head is not elevated off the operating table by the sternal retraction. A 30° videothoracoscope is then placed at the right lateral aspect of the neck incision to provide light for direct operating and a video magnified view of the operating field on a monitor for the surgeon and assistants. The thymic veins draining into the innominate vein are identified posteriorly and divided between stainless steel clips. The arterial vessels entering the gland laterally from the internal thoracic artery branches are also clipped with stainless steel clips. The dissection is carried down along the pleura to the inferior poles of the gland on both sides and along the pericardium. The assistance of the videothoracoscope provides good visualization of the lower mediastinum, down to the diaphragm if necessary. Once the gland is excised, if there is any further mediastinal fatty tissue present that is suspicious for being thymic tissue, this is excised or biopsied for frozen section analysis to ensure that no residual thymic tissue is left behind. If a complete thymectomy cannot be performed by removing the thymus and its capsule, the operation is usually converted to a partial upper sternotomy. This is carried out by the addition of a vertical skin incision extending down from the sternal notch to the lower end of the manubrium. The sternal incision is then extended laterally in the third or fourth intercostal space with the oscillating saw, to create a partial upper sternotomy, which provides sufficient exposure to easily complete the operation.

2.5 Follow-up

Fifty percent of the patients were followed by one neurologist with a special interest in the care of patients with MG (VB). These patients were reviewed at various intervals according to their status, but were seen at least every 12 months and as frequently as every 1–3 months. The remaining patients were followed by the attending surgeon (SK) on a yearly basis (30%) or were contacted by telephone (20%). Follow-up was complete if patients were reviewed or contacted between January 1 and December 31, 2000 with a minimum of 6 months follow-up since their surgery.

In order to compare our results with those of previous publications, we used definitions similar to those in other series. Complete remission is defined as asymptomatic without weakness and without any MG medications for at least 6 months; remission is defined as minimal ocular symptoms (slight ptosis) or treatment with pyridostigmine only for 6 months. The palliation rate included all patients on immunosuppression or with persistent mild to moderate symptoms despite MG treatment.

2.6 Statistical analysis

To document the evolution in our management and our results, we divided our study into three periods: the periods ranged from 1991 to 1994, 1995 to 1998, and 1999 to 2000. Data were analyzed by Fisher's exact test and Student's t-test where appropriate. Results are expressed as mean±SD, or as median and range. Life table analysis was tabulated by the Kaplan–Meier method with complete remission as the event of interest. Life table analysis with the log-rank test was used to assess the effect of the variables on the distribution of complete remission over time. Probability values <0.05 were considered to be statistically significant.

3 Results

A total of 120 consecutive patients underwent video-assisted trans-cervical thymectomy during the study period (Table 1 ). There were 86 females with a median age of 33 years (range 14–79) and 34 males with a median age of 36 years (range 12–68). Time elapsed between onset of symptoms and thymectomy varied from 2 months to 17 years (median, 10 months). The mean preoperative Osserman grade was 2.7±0.9 (range 1–4). A total of 78 patients were treated with pyridostigmine only. Plasmapheresis was performed in 30 patients and IVIg was administered to three patients before surgery.

Surgery was converted to an upper sternotomy in 23 cases (19%). Poor exposure was cited in 17 cases, a mass suggesting a thymoma in four cases, and previous neck surgery with adherent tissue in two cases. The risk of conversion was significantly higher in males than in females, and in patients with thymic involution rather than hyperplasia (Table 2 ). The rate of conversion decreased from 22% in 1991–1994 to 9% in 1999–2000 (p=0.4). During the same period, the proportion of patients treated without immunosuppression prior to surgery increased from 69% in 1991–1994 to 95% in 1999–2000 (p=0.7), and the time elapsed between the onset of MG symptoms and thymectomy decreased from 30±48 months in 1991–1994 to 14±14 months in 1999–2000 (p=0.3).

Postoperative complications occurred in four patients. Two patients who required conversion to a sternotomy experienced myasthenic crisis and were re-intubated during the postoperative course. One patient had a hemothorax and one had a pneumothorax after the trans-cervical approach, both of which were sufficiently minor to be treated conservatively. No postoperative deaths occurred. Among patients with an uncomplicated course (i.e. excluding the four patients above), the median postoperative hospital stay was 2 days (range 1–8 days) after trans-cervical thymectomy, and 3.5 days (range 2–8 days) after conversion to an upper sternotomy (p<0.0001). Following the trans-cervical approach, the median postoperative hospital stay decreased from 2 days (range 1–8 days) before 1994 to 1 day (range 1–8 days) thereafter (p=0.0001). In contrast, the postoperative hospital stay after conversion to an upper sternotomy did not significantly change over time.

On pathologic examination, the thymus was involuted in 60 cases, hyperplastic in 55 cases, and contained an incidental thymoma in five cases. All five thymomas were undetected on the preoperative CT scan and were discovered at the time of surgery only. Their size ranged from 1 to 3 cm (median 2 cm). Three of them required conversion to a sternotomy, whereas two small (1 cm), encapsulated thymomas located in the upper part of the thymus were safely removed through the trans-cervical approach. All five patients are alive without recurrence after 39–111 months (median 97 months). The proportion of thymic involution was significantly higher among patients treated with steroids before surgery than among those not taking steroids (38 versus 16%, respectively; p=0.01).

One hundred patients were available for complete follow-up. Ten patients were lost to follow-up and 10 have been followed for less than 6 months and were not included in the outcome analysis. After a median follow-up of 48 months (range 6–117 months), 41% of the patients were in complete remission and 9% were in remission. The improvement in Osserman grade was not different between patients whose thymectomy was completed via the trans-cervical approach and those who required conversion to an upper sternotomy (Table 3 ). Two patients presented with persistent myasthenic symptoms refractory to medical treatment and underwent re-exploration through a sternotomy 25 and 42 months after the trans-cervical approach despite the absence of residual thymic tissue on chest CT scan. One had thymic tissue in the anterior mediastinum discovered during the second surgery and subsequently became asymptomatic, whereas the other had no residual thymic tissue.

Life table analysis showed that 91% of the patients were in complete remission at 10 years (Fig. 1 ). Stratification according to gender, age at surgery (cut-off value of 34 years), latency of surgery (cut-off value of 10 months), preoperative Osserman grade (stages I and II versus III and IV), and pathological findings (involution versus hyperplasia) showed no significant differences at 10 years (Fig. 2 ). Further stratification of age (≤40, 41–60, and >60 years), latency of surgery (≤8, 9–12, >12 months), and preoperative Osserman grade (stages I–IV) also showed no significant differences (p=0.5, p=0.2, and p=0.9, respectively).

Out of 14 patients with ocular symptoms (Osserman grade I), six (43%) achieved complete remission after a median follow-up of 63 months (range 20–104 months). Among the remaining eight patients, two were improved after thymectomy, four were unchanged (including two patients who had surgery within less than 6 months), and two presented with worsening of symptoms.

Stratification according to the need for preoperative immunosuppressive therapy showed no significant difference at 10 years when compared to patients receiving only pyridostigmine before surgery (p=0.1). However, complete remission was achieved more rapidly in the group of patients treated without preoperative immunosuppression, reaching statistically significant differences after 5 years of follow-up (p=0.04) (Fig. 2).

4 Discussion

Thymectomy is well accepted as a therapeutic option for patients with MG despite the absence of a randomized control trial of outcome. Minimally invasive approaches allow better preservation of lung function perioperatively than median sternotomy [4]. Early referral has been shown to allow rapid and durable clinical benefit, and may avoid the introduction of immunosuppressive therapy [12,14,15].

Video-assisted trans-cervical thymectomy combines video technology with the minimally invasive trans-cervical technique. This approach provides excellent visualization of the mediastinum along both pleura down to the diaphragm if necessary, and permits extended resection of the thymus gland and perithymic tissue both in the neck and in the mediastinum. There is no need for placement of a double lumen endotracheal tube, the pleural space need not be breached, and there is minimal requirement for more than oral analgesia in the postoperative period. The morbidity is extremely low with minimal complications (one hemothorax and one pneumothorax treated conservatively in our series). In addition, 63% of our patients were hospitalized for 24 h or less and 83% for 48 h or less. These results were significantly better than for the group of patients that required conversion to an upper sternotomy. Indeed, the majority of these patients were hospitalized for 3–4 days because they required patient-controlled analgesia with intravenous morphine sulfate, and two patients in this group experienced a myasthenic crisis during their postoperative course.

Although the rate of complete remission in the long-term is certainly the most important endpoint in determining the efficacy of any surgical approach for patients with MG, direct comparison between series has been difficult. Most authors report crude remission rates, which correspond to the overall number of remissions per number of thymectomies performed, after a mean length of follow-up [16]. The time from thymectomy to complete remission is, however, of prime importance as remission is a time-dependent event. In order to correct for variable length of follow-up and for patients lost to follow-up, Jaretzki has recommended the use of life table analysis to determine the rate of remission [16]. This mode of analysis has, however, been adopted by only a few authors [5,17–19].

We agree that life table analysis, of which the Kaplan–Meier method is a refinement, is a more relevant statistical technique for the evaluation of remission when there is no competing risk, because it provides a complete description of the rate of remission over time. Crude data and life table analysis of complete remission comparing our experience with that reported in the recent literature are presented in Table 4 . In our study, the crude rate of complete remission was 41% after a mean follow-up of 4.3 years, which is similar to most other studies reporting results after trans-cervical and/or trans-sternal thymectomy. The life table analysis, however, showed a complete remission rate of 30% at 5 years and 91% at 10 years of follow-up. These results demonstrate that most of our patients achieved complete remission between 5 and 10 years after thymectomy. A complete remission rate of 91% at 10 years compares favorably with other series reporting life table analyses [5,17–19]. Durelli et al. [17] reported a complete remission rate of 30% at 5 years after trans-sternal thymectomy, and Jaretzki et al. [5] observed a complete remission rate of 81% at 8 years after extended (radical) cervico-mediastinal thymectomy.

When performing a thymectomy through a trans-cervical approach, one should not forget that conversion to a sternotomy should not be considered as treatment failure. In our series, a total of 23 patients required conversion to an upper sternotomy. This relatively low threshold for conversion may explain why only two patients (1.7%) required re-exploration for persistent refractory symptoms, whereas the re-exploration rate has been as high as 27% in some other series [25]. In our experience, the cause of conversion was mainly due to poor exposure in men with a relatively small thoracic inlet and an atrophic thymus. Video-assisted technology also provides much better visualization of the mediastinum and, thus, may have accounted for part of the conversion rate because some of the mediastinal fatty tissues that were visualized by videothoracoscopy might have been missed by direct vision alone.

In conclusion, considerable improvement has been made since the time when only selected patients with severe MG unresponsive to medical therapy underwent thymectomy. Surgery was associated with prolonged postoperative mechanical ventilation leading to a high morbidity and mortality rates. Currently, trans-cervical thymectomy requires hospitalization for less than 24 h in the majority of cases and is associated with very little morbidity. Patients therefore tend to be referred earlier in the course of their disease when they are in a more stable clinical condition. In the long-term, this approach achieved complete remission in 91% of the patients.

The authors would like to acknowledge David Liang for his help in retrieving the data and Peter Lewycky for his assistance with the statistical analyses.

References