Robotic thymectomy for thymoma in patients with myasthenia gravis: neurological and oncological outcomes

Abstract

 

Open in new tabDownload slide

OBJECTIVES

The goal of this study was to analyse the outcomes in 53 patients with thymoma, 34 of whom had myasthenia gravis (MG), who were treated with robotic surgery. The oncological outcomes of the whole series of patients were analysed. Furthermore, because consistent data are not yet available in the literature, the main focus was the analysis of the neurological results of the patients affected by MG and thymoma.

METHODS

The clinical outcomes of 53 patients with a diagnosis of thymoma who underwent robotic thymectomy between January 2014 and December 2019 in our institution were collected and evaluated; 34 of these patients had a concomitant diagnosis of MG. The neurological status of the patients was determined from a clinical evaluation according to the Osserman classification and on pre- and post-surgery Myasthenia Gravis Composite scores, whereas neurological clinical outcomes were assessed using the Myasthenia Gravis Foundation of America Post-Intervention Score. Reduction of steroid therapy was also considered. The recurrence rate, adjuvant radiotherapy and overall survival of the patients with a thymoma were evaluated.

RESULTS

Neurological outcomes: improvement of the clinical conditions was obtained in 26 patients (76.5%) following the operation: complete stable remission was observed in 5 patients (14.7%), pharmacological remission in 10 (29.4%) and minimal manifestation in 11 (32.3%). Four patients (11.8%) exhibited no substantial change from the pretreatment clinical manifestations or reduction in MG medication and 4 (11.8%) patients experienced worsening of clinical conditions. In 21 patients (61.7%) a reduction of the dosage of steroid therapy was obtained. Oncological outcomes: at an average follow-up of 36 months, the overall survival was 96%, 4 patients (7.5%) had pleural relapses and 12 patients (22.6%) underwent postoperative radiotherapy, according to their stage. In accordance with Masaoka staging, 34% were in stage I, 56.6% in stage II and 9.4% in stage III

CONCLUSIONS

Our results suggest that robotic surgical treatment of patients with thymoma and concomitant MG is effective in improving the neurological outcomes. Moreover, the oncological results obtained in this series confirm the efficacy of robotic surgery for the treatment of thymic malignancies, with results in line with those of open surgery. However, due to the indolent growth of thymomas, further observations with longer follow-up are necessary.

INTRODUCTION

Thymoma is a rare epithelial tumour of the thymus gland, generally characterized by slow growth with local invasion and rare distant metastases. Thymoma is frequently associated with several types of diseases, of which myasthenia gravis (MG) is the most common; it is diagnosed in 10–15% of these patients.

The first experiences of surgical treatment for thymoma in patients with MG were reported in 1939 by Blalock, who reported a series of cases treated by the sternotomy approach [1]. Over the years thymectomy has become the treatment of choice and is currently recognized as a standard effective treatment for thymoma; it has emerged from one of the major studies on the efficacy of thymectomy in patients with MG in respect to clinical outcomes [2].

Initially, the surgical treatment comprised primarily a subtotal thymectomy [3], but over time, thanks to the evolution of the surgical technique and the acquisition of neurological postoperative data, the importance of the complete removal of thymic tissue and its surrounding mediastinal fat became undeniably evident, it being a key element to improve neurological outcomes [4, 5].

Moreover, resection of the thymoma also remains the main therapeutic strategy from an oncological point of view: tumour stage, histological analysis and radicality of the surgery constitute the fundamental prognostic factors [6].

In recent years, considering the young age profile of patients with MG, the challenge for the surgeon treating patients with thymoma has been focused on the minimally invasive approach.

After the initial experiences with video-assisted thoracoscopic surgery (VATS) for thymectomy, despite the undeniable advantages, numerous surgeons were unwilling to use a minimally invasive approach in patients with MG due to the risk of incomplete tissue resection [7].

Then again, the introduction of robotic surgery has provided several advantages such as the 10 times magnified 3-dimensional vision, the highly precise movements and the effortless visualization of the operating field. Furthermore, thanks to its advantages, robotic surgery allows the application of the ‘no-touch technique’, according to ITMIG recommendations, to maximize the radicality of surgical procedure.

Several papers have confirmed the safety and effectiveness of robotic surgery both for benign and malignant tumours of the thymus, but studies on the neurological postoperative results after thymectomy in patients with MG with thymoma are still extremely limited [8, 9].

This study analyses the effects of robotic surgery for the treatment of patients with MG and thymoma, focusing on the evaluation of long-term postoperative oncological and neurological outcomes.

METHODS

Ethical statement

The study design, the patient enrolment method and data collection methods were reviewed and approved by the institutional review board.

Each patient was adequately informed and had to sign the informed consent in order to be enrolled in the study. This retrospective study evaluated data of patients with a diagnosis of thymoma who underwent robotic thymectomy from January 2014 to December 2019. Patients with thymoma with or without a neurological and serological diagnosis of MG who were suitable for surgery were also included in the study. Patients with a benign thymic disease or with thymoma not eligible for surgery were excluded.

Preoperatively, all subjects with symptoms attributable to a myasthenic syndrome underwent neurological examinations, including blood tests to reveal the presence of acetylcholine receptor and muscle-specific serum kinase (anti-MuSK) antibodies and computed tomography (CT) scans of the thorax to detect unspecified lesions in the anterior mediastinum. Similarly, subjects with a CT scan showing an anterior mediastinal mass, suggestive of thymoma, underwent acetylcholine receptor and anti-MuSK antibody blood tests and a neurological evaluation, to evaluate for the possibility of associated MG. No patient included in the study tested positive for anti-MuSK antibodies. After all the routine clinical examinations were completed, all the patients had a robotic-assisted thymectomy.

Robotic thymectomies were performed with the third and fourth generations of the robotic system. The fourth generation was introduced in our daily surgical practice in 2015. The robotic approach for thymectomy requires 3 surgical incisions on the left or the right hemithorax. In our institution the left side is preferred, while we use the right approach only in case of thymomas extended to the right hemithorax. The side of surgical access is based on the surgeon’s experience and on the extent of the unilateral tumour. The patient is positioned supine with a roll placed under the spine to expose the mediastinal area. The first surgical access, used for the 30° 3-dimensional camera, is placed on the V intercostal space at the anterior axillary line; the second access is on the third intercostal space on the midaxillary line; and the third is on the V intercostal space on the mid-clavicular line. The most cranial port is used to remove the lesion. In case of a large thymoma (e.g. 10 cm), this access was widened by a few centimetres along with the intercostal space. This specific port is generally chosen due to the more distensible intercostal space at this level.

In all the cases analysed, an extended thymectomy was performed according to the recommendations of the International Thymic Malignancy Interest Group. After identification of the phrenic nerves sideways and the innominate vein in the rostral position, the extended thymectomy comprised the complete dissection of the thymic gland en bloc with the mediastinal adipose tissue. In all cases, an exploration of the perithymic lymph node stations was performed as recommended by the nodal mapping protocol of the International Association for the Study of Lung Cancer/International Thymic Malignancy Interest Group [10]. In our series, no N2 lymphadenectomies were performed because the chest CT scans of all patients were negative.

From an anaesthesia point of view, the management of the patient with MG represents a real challenge due to the neuromuscular and respiratory involvement of the disease and the elevated risk of a sudden myasthenic crisis.

During the robotic thymectomy procedure, the anaesthesia monitoring protocol includes checking blood pressure non-invasively, electrocardiography, oximetry, temperature, bi-spectral index and train of four intraoperative monitoring of the neuromuscular blockade. Induction of anaesthesia is conducted with propofol 2 mg/kg, fentanyl 1 μg/kg and rocuronium 0.6 mg/kg. Anaesthesia is maintained by continuous intravenous infusion of propofol (4–6 mg/kg/h) and remifentanil (0.3–0.5 µg/kg/min), whereas the postoperative analgesia consists of morphine 0.5 mg/kg, ketorolac 30 mg and acetaminophen 1 g at the end of the operation and wound infiltration with ropivacaine (10 mg/ml) and lidocaine (2%). After the operation, analgesia is performed with ketorolac infusion 90 mg/day and acetaminophen in case of pain.

The demographic data and the intraoperative (side of the approach, operative time, conversion to open surgery) and postoperative results (dimension of thymoma, histopathological diagnosis, adjuvant therapy, oncological and neurological follow-up) were collected and analysed.

The neoplasm was histologically classified according to the World Health Organization classification of thymic epithelial tumours and the clinical staging system of Masaoka.

Neurological status was based on a clinical evaluation according to the Osserman–Jankins classification and the preoperative and postoperative Myasthenia Gravis Composite (MGC) score, whereas neurological clinical outcomes were assessed by the Myasthenia Gravis Foundation of America Post-Intervention Score (MGFA-PIS), in which complete stable remission (CSR) was defined as no symptoms and without medication for at least 1 year [11]. All the patients were examined by a dedicated neurological team preoperatively and at 6-month intervals to evaluate any neurological improvement and possible patient-reported signs and symptoms (12 h after pyridostigmine ingestion). Moreover, any eventual postoperative reduction of steroid dosage was recorded.

The main hypothesis of the study is the analysis of the oncological and neurological results of a robotic approach in patients suffering from MG and thymoma, to confirm, as seen in the literature, the advantages of using the robotic approach compared to other more traditional approaches.

Statistical analyses

Any categorical data collected were described by frequency and the continuous data, by the mean and standard deviation (SD).

To evaluate the normality of the continuous data distributions, the Shapiro–Wilk test was performed to show the normality of the distributions. To analyse the continuous and categorical variables, we performed the t-test for independent samples (two-tailed) and the χ² test, respectively.

To evaluate differences between the preoperative and postoperative MGC scores, we used the t-test for paired data (two-tailed). Statistical testing was performed at an α level of 0.05. All analyses were performed using SPSS v.27 (IBM-SPSS Inc., Armonk, NY, USA).

RESULTS

A total of 233 robotic thymectomies were performed in our institution from 2014 to 2019. These included surgery for benign tumours in MG patients and thymoma in MG and non-MG patients. Fifty-three patients with thymoma had a robotic-assisted thymectomy. The total cohort comprised 25 men and 28 women, with an average age of 56 years (SD 16). Among these patients, 34 had MG [18 women and 16 men with a mean age of 52.6 years (SD 18.1)]. Of the 34 patients with MG, 13 were given intravenous immunoglobulin (10 preoperatively, 3 postoperatively); no patient was treated with plasmapheresis. Five patients received immunosuppressant therapy (the mean pyridostigmine bromide dose was 38.12 mg). The average operation lasted was 126 min (SD 59). Conversion to open surgery was necessary in 1 case (1.9%) due to infiltration of the innominate vein and pericardium. All patients were extubated at the end of the surgical procedure. The average hospital stay was 5.1 days (SD 3.3).

Postoperative complications occurred in 4 patients (7.5%): 1 case of respiratory failure, requiring non-invasive ventilation; 1 case of pneumothorax, treated with a pleural drain; and 2 cases of chylothorax, 1 of whom underwent robotic surgical revision on the second postoperative day.

The mean diameter of the neoplastic lesion was 3.8 cm (SD 2.5), range 0.1–10 cm. In 16 (30.2%) cases lymphadenectomy was also performed, with negative histological results. The histopathological diagnosis was type A thymoma in 16 (30.2%) patients, AB in 8 (15.1%), B1 in 8 (15.1%), B2 in 14 (26.4%) and B3 in 7 (13.2%). According to the Masaoka staging, 18 (34%) patients were in stage I, 30 (56.6%) in stage II and 5 (9.4%) in stage III. All patients were found to be R0 on the definitive histological examination. Based on pathological staging and a multispecialist board evaluation, 12 patients (22.6%) underwent adjuvant radiotherapy.

The mean oncological and neurological follow-up was 36 months (SD 17) [33.9 in the patients with MG (SD 13.2)]. During the follow-up period, 2 patients died: 1 of causes unrelated to thymoma (ictus) and 1 of disease progression. The 1 patient out of 34 who showed signs of a myasthenic crisis was treated with intravenous immunoglobulin. One patient was transferred to the intensive care unit postoperatively to be extubated in a protected area because of severe obesity.

Patients with thymoma A and AB had no recurrence, whereas in those with type B, recurrence was diagnosed in 4 cases (P = 0.069): 1 patient had a B1 thymoma; 1 patient, a B2; 1 patient, a B3; and 1 patient, aB2/B3; 3 were myasthenic. One patient with a B2/B3 thymoma was not diagnosed with MG. The size of the primary tumour for these patients was 2.4, 5, 3.5 and 7 cm, respectively These 4 patients were diagnosed with pleural relapses. Twelve patients underwent radiotherapy treatment after the thymectomy: 2 (16.7%) patients had a thymoma A or AB, whereas 10 (83.3%) a type B (P = 0.034).

When we evaluated the histopathological patterns, we found a significant statistical correlation (P = 0.030) between the histological type of the thymoma and the age of the patient: the average age of the patients with a type B thymoma, i.e. with a worse prognosis, was 52 years (SD 15), whereas type A or AB was more common in older people, in fact, the mean age was 61 years (SD 15).

MG was diagnosed in 34 (64.2%) patients: 8 (23.5%) were Osserman class I, 19 (55.9%) were class II and 7 (20.6%) were class III.

The multivariate analysis documented a statistically significant correlation between the lower grades of Osserman classes (I and II) and a lower Masaoka stage (I), whereas the higher Osserman classes (IIb, III) appeared to be associated with the higher Masaoka stages (II, III) (P = 0.035) (Table 1).

A similar trend was detected in the analysis of histological type: thymoma B appears to be correlated with Osserman IIb or III (P = 0.022) (Table 2).

All the patients with MG were evaluated with MGFA-PIS. Clinical conditions improved in 26 patients (76.5%) following the operation: CSR was observed in 5 patients (14.7%); pharmacological remission, in 10 (29.4%) and minimal manifestation in 11 (32.3%). Four patients (11.8%) exhibited no substantial change from the pretreatment clinical manifestation and no reduction in MG medication as defined in the protocol (U); 4 (11.8%) patients experienced worsening of clinical conditions (Table 3).

Furthermore, the average preoperative MGC score was 11.9 (SD 6.1), whereas the average postoperative MGC score was 3.5 (SD 4.3), showing a statistically significant (P < 0.0001) improvement in the postoperative clinical status according to MGC score.

When we analysed the effects of postoperative steroid therapy in patients with MG, we found a reduction in the steroid dosage in 21 cases (61.7%), in particular in patients with Osserman I or II, more than in Osserman III (P = 0.588). The mean preoperative dose of corticosteroid for each patient was 36.7 mg (SD 14.5), whereas the mean postoperative dose was 15.5 mg (SD 13.4).

DISCUSSION

Extended thymectomy is considered the gold standard for the treatment of early-stage thymoma [12]. The procedure consists of the radical resection of the tumour en bloc with the thymus and perithymic fat tissue. As established in the ITMIG guidelines, it is essential to remove the anterior mediastinal fat tissue after the identification of both the right and left phrenic nerves to ensure oncological and neurological radicality, proceeding with the dissection from the jugular to the anterior pericardiophrenic angle [10].

Trans-sternal thymectomy has long represented the surgical approach of choice in the treatment of patients with thymoma. Open thymectomy allows excellent visualization of the anterior mediastinum, ensuring the appropriate control of the nerves and vessels. Moreover, in the case of invasion of the neighbouring organs, sternotomy allows a feasible en bloc tumour resection with the adherent structures.

The first thoracoscopic thymectomy, performed by Sugarbaker in 1993 [13], radically changed the surgical approach to thymic malignancies. Minimally invasive surgery has several advantages, including less surgical trauma, less postoperative pain, reduced duration of pleural drainage, shorter hospitalization and better aesthetic results in association with the same oncological radicality [14]. Furthermore, several studies on minimally invasive surgery in the treatment of thymus neoplasms have shown an early improved pulmonary recovery after surgery, which can be considered crucial in the postoperative management of patients with MG [15].

To date, robotic surgery is considered to be the extreme evolution of minimally invasive surgery. Thanks to its features, the surgeon is able to replicate the same steps of the open thymectomy with enhanced accuracy.

Several studies [10] have confirmed the safety, feasibility and effectiveness of robotic surgery when applied to anterior mediastinal lesions. However, no consistent neurological data regarding long-term outcomes in the robotic treatment of thymic neoplasms with concurrent MG are available. In particular, few data exist on the neurological outcomes in patients with MG affected by a thymoma.

In a 2015 study, Keijzers et al. [16] analysed the neurological results of patients with MG affected by several thymic pathological entities (hyperplasia, cysts, lipomas and thymomas) undergoing robotic surgery. They focused on the MGFA classification to determine preoperative and postintervention status, obtaining in patients with thymoma a 5.6% 3-year CSR.

In 2017, Kumar et al. [17] analysed oncological and neurological outcomes in 71 patients with MG who underwent robotic thymectomy for different diseases, confirming the safety and feasibility of robotic surgery in the treatment of these patients. The overall CSR rate was 38% after 17.5 months; in the 21 thymoma patients specifically, it was 19%.

Despite the large numbers of patients assessed in these studies and the good results obtained, the specific long-term results in patients with MG affected exclusively by thymoma are not consistent.

This study is one of the first experiences concerning long-term neurological and oncological outcomes in a homogeneous group of patients affected by MG and thymoma. In fact, as it is well known, the course of myasthenia is different in patients diagnosed with thymoma or benign disease [18].

In this experience, 53 patients with thymoma who underwent robotic thymectomy with a totally endoscopic technique were analysed, 34 of whom were affected by MG.

Concerning neurological results, in this series, an improvement of the symptoms (MGFA-PIS) was obtained in 76.5% of cases, with a CSR rate of 14.7% after 36 months. These results appear to be consistent when compared with those of other series [16, 17].

Moreover, a correlation of Osserman classes I–II with Masaoka stage III–IV (P = 0.035) and histotype B (P = 0.022) was observed. These results confirm the data found in the literature concerning the association between myasthenia symptoms and more aggressive forms of thymoma [19].

In our series, after an average follow-up of 36 months, when we analysed the oncological outcomes, the overall survival was 96% and the recurrence was 7.5%. Furthermore, when we compared our neurological and oncological results with data from the international scientific literature, we observed that they were in line with the open surgery outcomes [20], confirming the radicality achieved by robotic surgery [21]. This agreement is also due to the fact that all patients are examined postoperatively by a multidisciplinary board to determine the indications for follow-up or for radiotherapy (in stages IIb and III) in accordance with the international guidelines [22].

Moreover, when compared with a trans-sternal approach, robotic surgery is associated with lower complication rates, reductions in hospital stays and better postoperative quality of life [23].

The satisfying postoperative results of robotic surgery could be explained by the features of the robotic system: better visibility, greater dexterity and better manoeuvrability of the instruments, which allow the surgeon to reach more easily the remote regions of the anterior mediastinum, enhancing the precision of the procedure and thereby ensuring the execution of the extended thymectomy and the application of the ‘no-touch’ technique, which reduces the risk of the local diffusion of neoplastic cells. The technique implies that the thymoma is never touched; during the procedure, the normal thymic tissue and perithymic fat are used for grasping and for traction. This technique avoids any direct manipulation of the tumour, minimizing the risk of tumour seeding as a result of capsule damage. The thymus and perithymic fat are therefore dissected with safe surgical margins, according to the International Thymic Malignancy Interest Group criteria [10]. The analysis of oncological variables shows a statistically significant correlation between the World Health Organization classification and the relapse rate (P = 0.069), confirming the possibility of applying the no-touch technique and of performing a radical surgical procedure using the robotic approach.

Furthermore, the radicality of the procedure could be improved by the latest upgrades of the robotic system such as the intraoperative use of fluorescence, which in the hands of a skilled surgeon could play an important role in obtaining neurological and oncological radicality, giving the ability to perform a more precise and complete resection of the neoplasm, thymus and perithymic fat.

In our series, patients affected by a Masaoka stage III thymoma also underwent robotic surgery. This result shows that the evolution of robotic technology has allowed the treatment of locally advanced forms of the disease, avoiding the execution of a trans-sternal approach. Furthermore, our experience suggests that the volume of the neoplasm is not a contraindication to the robotic approach. We have in fact been able to treat lesions with a large diameter (up to 10 cm) without complications, thanks to the wide range of movements of the robotic instruments and the magnified 3-dimensional view.

However, the VATS approach cannot be overlooked. In fact, according to the data in the literature, VATS was safe and feasible for the treatment of early-stage thymomas and was associated with good oncological prognosis [24]. In 2014, a British group demonstrated good oncological short- and mid-term outcomes achieved in a cohort of patients with thymoma compared with the open approach [7].

Our oncological results are comparable with those obtained with VATS and open techniques [23]; however, due to the slow growth of the thymoma, further studies are necessary to evaluate long-term comparability with other techniques.

Limitations

This monocentric retrospective study has some limitations: the number of patients and the duration of the follow-up period, given the indolent growth of thymomas, should be increased to confirm the oncological results. A prospective comparative multicentric study with other surgical techniques (sternotomy and VATS) would be appropriate to better analyse oncological and neurological outcomes.

CONCLUSION

This study presents the long-term follow-up of 53 patients with thymoma, 34 of whom have MG. The clinical improvement of neurological symptoms in patients with MG and the oncological results demonstrate that the robotic technique is a technically feasible and safe procedure with low complication rates and short hospital stays.

From an oncological point of view, the reportedresults confirm that minimally invasive surgery should always be considered, where applicable, in accordance with the international guidelines. Furthermore, the neurological results confirmed the effectiveness of robotics from the long-term clinical point of view.

Conflict of interest: Prof. Franca Melfi and Dr Carmelina Cristina Zirafa are Official Intuitive proctors.

Author contributions

Gaetano Romano: Conceptualization; Data curation; Investigation; Visualization; Writing—original draft; Writing—review & editing. Carmelina Cristina Zirafa: Conceptualization; Investigation; Writing—original draft; Writing—review & editing. Ilaria Ceccarelli: Conceptualization; Data curation; Investigation. Melania Guida: Conceptualization; Data curation; Investigation. Federico Davini: Visualization. Michelangelo Maestri: Conceptualization; Data curation; Investigation. Riccardo Morganti: Statistics. Roberta Ricciardi: Conceptualization; Data curation; Formal analysis; Investigation; Methodology. Teresa Hung Key: Formal analysis; Methodology; Writing—original draft; Writing—review & editing. Franca Melfi: Conceptualization; Data curation; Investigation; Methodology; Supervision; Writing—original draft; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Bernhard Moser, Bernard Park, Alper Toker and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

Presented at the 34th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Barcelona, Spain, 8–10 October 2020.

REFERENCES

ABBREVIATIONS

 
• CSR

  Complete stable remission

 
• CT

  Computed tomography

 
• MG

  Myasthenia gravis

 
• MGC

  Myasthenia Gravis Composite

 
• MGFA-PIS

  Myasthenia Gravis Foundation of America Post-Intervention Score

 
• SD

  Standard deviation

 
• VATS

  Video-assisted thoracoscopic surgery