Ten-year experience of mediastinal robotic surgery in a single referral centre

Abstract

OBJECTIVE

The aim of this study was to report a single referral centre experience in the exeresis of mediastinal lesions with robotic surgical system. The outcomes of all patients were retrospectively investigated with special regard to myasthenic and thymomatous patients.

METHODS

From February 2001 to December 2010, 69 patients (30 males, 39 females), with a mean age of 55.4 years (range 20–81), underwent robotic surgery for treatment of mediastinal lesions. There were 39 thymectomies, 13 paravertebral neurogenic tumour removals, 9 pleuropericardial cyst removals, 3 teratoma excisions, 4 lymphonodal removals and 1 enterogenous cyst excision. Operative time, conversion rate, morbidity and mortality were analysed.

RESULTS

There were no intra-operative complications, and no mortality. The mean operative time was 124.3 min (range 45–240). Conversion to open surgery occurred in three cases (4.3%). All post-operative complications (7.2%) were conservatively treated. The mean post-operative stay was 4.3 days (range 3–10). For the myasthenic patients (mean follow-up of 18 months), the Myasthenia Gravis Foundation of America change in status resulted improved in 18 (90%) patients and unchanged in 2 (10%) patients. In cases of thymoma, the pathological analysis revealed Masaoka stage I (seven cases), IIA (three cases), IIB (two cases) and III (one case). No disease recurrence occurred at a mean follow-up of 16 months.

CONCLUSIONS

Our experience suggests that the surgical system was safe in performing mediastinal mass resection, with low morbidity and conversion rate. The robotic system proved useful, especially in a tiny space such as the mediastinum, and appeared to be an effective treatment for myasthenia gravis and for early-stage thymoma.

INTRODUCTION

Video-assisted thoracic surgery (VATS) has been accepted but is not routinely performed as a therapeutic approach for mediastinal mass resections [1–4]. It permits less operative trauma with respect to open surgery, but in some cases, especially for anterior and upper mediastinal lesions, there have been surgeons who hesitated to perform VATS [1–4]. This fact is probably due to the limitation in exposure of the interested region, and, above all, to the limited freedom of movement of conventional thoracoscopic instruments. At the beginning of this century, some case reports described preliminary experiences in mediastinal mass removal and in thymectomy with the robotic system [5–8]. Subsequently, some authors have reported their results on small series, performing robotic thymectomy, thymic or pleural–pericardial cyst excision and paravertebral tumour resection [9–11]. Therefore, we want to report the outcomes of our 10-year experience with the robotic system in surgical resections of mediastinal masses, analysing morbidity, mortality and survival, comparing them with those reported in the literature.

MATERIALS AND METHODS

From our surgical database (Ormawind 2000, Avelco srl, Genova, Italy) we retrospectively selected all patients who underwent surgical resection of mediastinal masses with the da Vinci System, between February 2001 and December 2010. The main aim of this study was to evaluate the feasibility and safety of the da Vinci System in performing heterogeneous surgical procedures for mediastinal lesion. We evaluated operative time (defined as the time from skin incision and skin closure), docking time (defined as the time necessary for surgical cart positioning and robotic arms placement in the surgical field), conversion rate, mortality, morbidity and hospitalization.

The second aim was the evaluation of the efficacy of robotic thymectomy performed for myasthenia gravis (MG) or for early-stage thymoma.

Selection criteria

The lesions were considered suitable for a complete robotic removal if the diameter was <5 cm, with no evidence of extracapsular extension or contiguous structure infiltration. Indication for thymectomy was the presence of thymic cyst, MG and early-stage thymoma observed at computed tomography scan. Diagnosis of MG was determined by neurologists based on clinical criteria, electromyography and the titre of acetylcholine receptor antibodies, and it was stratified according to the Myasthenia Gravis Foundation of America (MGFA) classification [12]. The correct timing of thymectomy for MG was established according to our neurologists. Additionally, patients with clinical evidence of pleural adhesions, with previous thoracic intervention, previous chemotherapy or previous radiotherapy of the chest were not referred for the robotic approach. Moreover, additional parameters, observed during the pre-operative evaluation course, had to be satisfied (Table 1).

Pre-operative evaluation

Patients were evaluated by a thoracic surgeon, based on chest radiograph, chest CT scan imaging with contrast media enhancement and positron emission tomography (PET) scan in some cases. In all cases pre-operative functional assessment consisted of medical history, physical examination, basic blood tests, blood gas analysis, spirometry and cardiologic evaluation with electrocardiogram and echocardiography. Stress testing and other assays of myocardial function were performed when indicated by the cardiologist in case of a history of coronary artery disease, suspicious symptoms or electrocardiographic abnormality. An informed consensus was obtained from all patients. The Institutional Review Board at the University of Pisa approved the study.

Robotic system

The da Vinci^TM Robotic System (Surgical Intuitive, Sunnyvale, CA, USA) consists of a master console, a computer controller and a three-arm surgical manipulator with fixed remote centre kinematics connected via electrical cables and optic fibres. The master console was connected to a surgical manipulator with instrument-arms and an endoscope arm. The endoscope was 12 mm in diameter and contained two separate 5 mm high-resolution cameras, and the two cameras showed two separated images. Specific robotic surgical instruments were introduced through special ports (8 mm) and attached to the arms of the robot. The surgeon's movements are transmitted from the handles to the tips of the robotic instruments.

Surgical technique

All surgical procedures were performed by the same surgical team. All patients underwent general anaesthesia with double-lumen intubation. All patients were extubated in the operating room immediately after reversal of anaesthesia. During all procedures, the hemithorax was inflated through the camera port with CO₂ ranging between 4 and 8 mmHg. All specimens were removed with Endoscopic Pouches (Inzii^TM, Applied Medical) through the trocar incision and underwent pathological examination. A drainage tube was inserted, generally 32Ch.

Thymectomy

The patient was positioned left side up at a 30-degree angle. The camera was introduced in the fifth intercostal space at the midaxillary line, the right arm was introduced in the third intercostal space at the midaxillary region, while the left arm was introduced in the fifth intercostal space at the midclavicular line. The right arm had a Spatula (EndoWrist; Intuitive Surgical) with electric cautery function in order to perform dissection, while the left arm had a Cadiere forcep (EndoWrist; Intuitive Surgical), an atraumatic instrument for grasping the thymus. The dissection started inferiorly, first from the left side, at the pericardiophrenic angle till the left inferior horn was dissected, and then on the right side, from the retrosternal area, finding the right mediastinal pleura and then the right inferior horn. Consequently, the dissection continued upward to the neck until the superior horns were identified. The thymic veins were identified and separately clipped. All thymus and perithymic fat were dissected, according to the Masaoka criteria [13]. The lesions were removed from the cavity through the port incision in the midaxillary region. If necessary, the incision was enlarged (3 cm maximum), but no additional utility incision was used.

Anterior mediastinal lesion removal

The patient's position was lateral decubitus position. The camera was generally introduced at the sixth intercostal space on the midaxillary or posterior-axillary line. The best position of the robotic arms was established according to the site of the lesion, generally at the fourth intercostal space, with a minimum distance of 6 cm from the camera. A Spatula (EndoWrist; Intuitive Surgical) with electric cautery function was positioned on the right arm, while the left arm had a Cadiere forcep (EndoWrist; Intuitive Surgical), in order to grasp the lesion and soft tissue of the mediastinum. The lesions were removed through the anterior port incision and, if necessary, the incision was enlarged (maximum 3 cm). No additional utility incision was performed.

Posterior mediastinal lesion removal

Also in these cases, the patient was at a lateral decubitus position. The camera was generally introduced at the sixth intercostal space on the midaxillary or anterior-axillary line. The position of the robotic arms was evaluated according to the site of the lesion, generally at the fourth intercostal space, with a minimum distance of 6 cm from the camera. A Spatula (EndoWrist; Intuitive Surgical) with electric cautery function was generally utilized on the right arm, while the left arm had a Cadiere forcep (EndoWrist; Intuitive Surgical), in order to grasp the lesion. The lesions were removed through the anterior port incision and if necessary, the incision was enlarged (maximum 3 cm). No additional utility incision was performed.

Post-operative evaluation and follow-up

The safety assessment included identification of treatment-related complications. These were those occurring within 30 days of treatment. Complications were stratified according to the Clavien–Dindo classification [14].

Benign lesion follow-up consisted of a chest radiograph 3 months after the procedure, a chest CT scan after 6 months and then 1 year. Additionally, all patients with MG were followed up by a neurologist, and the post-operative status of MG was stratified according to MGFA classification [12]. In cases of malignant lesion, follow-up consisted of a chest radiograph after 3 months, a chest CT scan after 6 months and then every year. When histological examination revealed the presence of unexpected microscopically invasive thymoma, adjuvant radiotherapy was proposed to the patients.

Statistical analysis

Statistical analysis was performed using Statistica software version 7.0 for PC (Stat-Soft, Inc., Tulsa, OK, USA). The demographic data of patients' age, sex, comorbidities, operating time, docking time, conversion rate, mortality, morbidity and length of stay were collected. Statistical analysis was expressed in terms of frequency, mean and range. The Student's t-test was used to compare the distributions of continuous data. A probability value <0.05 was considered statistically significant. Disease-free survival was defined as the time from surgery to the first diagnosis of recurrence. Overall survival was defined as the period of time from treatment to death or last follow-up.

RESULTS

Demographics and clinical data

There were 69 patients: 30 males and 39 females. The mean age was 55.4 years (range 20–81). The most important comorbidities are summarized in Table 2. Twenty patients suffered from MG (11 patients were in class I, 9 patients in class IIA, according to the MGFA clinical classification) [12]. There were 39 thymectomies: in 19 cases for MG, in 13 cases for suspected thymoma and in 7 cases for thymic cyst. One patient with thymoma suffered from MG, nine patients had a pleuro-pericardial cyst and three patients had a lesion in the antero-superior mediastinum. Four patients had pathologic mediastinal lymphonodal enlargement, which was confirmed at PET scan (aorto-pulmonary window in two cases, cardio-phrenic angle in one case and pre-vascular in one case). Fourteen patients were presented with a lesion in the posterior mediastinum.

Surgical data and morbidity analysis

No intra-operative complications occurred. Conversion to open surgery was necessary in three (4.3%) cases, due to an unexpected degree of tumour invasion (two thymomas), and tenacious adhesions of contiguous structures in one case (one heterogenous cyst). The mean operating time was 124.3 min (range 45–240), with a mean docking time of 13 min (range 5–15), the mean duration of post-operative drainage was 1.3 days (range 1–3) and the mean hospital stay was 4.3 days (range 3–10).

No mortality occurred. Morbidity (grade II complications according to Clavien–Dindo classification) occurred in five (7.2%) patients. A woman affected by diabetes mellitus and hypertension, who had undergone thymectomy for a thymic cyst, experienced general convulsions immediately after reversal of anaesthesia. This event was promptly treated and rapidly resolved pharmacologically. The patient was evaluated by a neurologist, and underwent brain CT scan that revealed no abnormality. In this case hospitalization was prolonged to the 10th post-operative day. The other four cases were represented by pleural effusion (two cases), treated and resolved with medical therapy, anaemia requiring blood transfusion (one case) and biliary colic (one case), treated with medical therapy.

Learning curve

With regard to thymectomies alone, operative time and docking time were separately investigated. The mean operative time and docking time were, respectively, 134.1 min (range 90–240 min) and 9.5 min (range 5–15 min). If we separately analysed the first and the last 10 thymectomies, significant reductions in mean operative time and docking time were observed (Table 3).

Post-operative histology and staging

The mean diameter of the lesions was 3.1 cm, ranging from 1.6 to 4.2 cm. Histopathologic analysis revealed 13 neurinomas of paravertebral gutter, 9 pleuropericardial cysts, 3 teratomas, 1 enterogenous cyst, 2 lymph nodal metastasis of thyroid tumour, 1 sarcoidosis and 1 Castelman's disease. For the 39 extended thymectomies the histological examination revealed 10 thymic hyperplasia, 7 normal thymus, 2 thymolipoma, 7 thymic cyst and 13 thymomas. The WHO classification and Masaoka stage are, respectively, reported in Tables 4 and 5.

Follow-up and survival analysis

The median follow-up for all patients was 17 months (range 3–92), with no disease recurrence. For the patients affected by MG, the median follow-up was 20 months (range 3–68) and the MGFA change in status resulted improved in 18 (90%) patients and unchanged in 2 (10%) patients. All patients affected by Masaoka stage II–III thymoma underwent adjuvant radiotherapy, except for two cases that refused it. At a median follow-up of 16 months (range 5–96), no clinical or radiological sign of disease recurrence was observed.

COMMENT

Over the last two decades, the robotic system has been introduced in the thoracic field and has added several improvements over standard thoracoscopy. VATS technique is characterized by a planned vision and a limited manoeuvrability of thoracoscopic instruments that makes it difficult to operate around corners. This may be the reason that many surgeons hesitate to perform the VATS approach in cases of malignant tumour [1–4]. Robotic system can be considered a natural progression of VATS, offering a high-definition three-dimension vision, a tremor filtration and a 7-degree articulation of the instruments inside the chest cavity. These features permit a precise isolation of the anatomic structures and a safe manipulation of the tissues. In fact, some case reports regarding the application of the robotic system for thymectomy [6, 7] and for removal of a posterior-mediastinal mass [5, 8] were reported by several authors at the beginning of this century, emphasizing its safety and feasibility. As a result, several authors reported their outcomes on small series of patients affected by mediastinal lesion. In 2004, Bodner et al. utilized the robotic system for the exeresis of 14 mediastinal masses, including 9 thymectomies, reporting a short operating time and a low conversion rate (7%) [9]. Also Rea et al. [10] reported interesting results on 33 cases of thymectomy for MG with a short operating time and no conversion to sternotomy. In a recent study, Fleck et al. [11] described their experience on 18 thymectomies treated with the robotic system, with a mean operating time of 175 min and an acceptable conversion rate (5%). In our larger, but heterogeneous series, the mean operating time appeared to be comparable with those reported in the literature, that ranged between 120 and 175 min [9–11]. No intra-operative complication occurred, and the conversion rate was also adequately low.

Moreover, in our series the morbidity rate was 7.2%. These data turned out to be comparable with the 6–7% reported by other authors who experienced chylothorax and haemothorax, conservatively treated [10], or transient palsy of the left laryngeal recurrent nerve [9]. Fleck et al. [11] reported a high incidence of complications (27%), principally due to pain at the port incision sites, which, however, was observed neither in other papers [9, 10] nor in this study. The low morbidity, in addition to the absence of intra-operative complications, and the low conversion rate suggested that many different types of mediastinal lesion, located in different positions, can be safely resected with the robotic system. Additionally, in our opinion, resections of neurinoma or pleuropericardial cyst represented an ideal training model for the ‘robotic team’ (the surgeon, the surgical assistant, the nurses and all personnel in the operating room), because they provided the means for learning basic procedures combined with a relatively simple technique. After an initial phase of the learning curve, the utilization of precise operative schemes for each singular type of procedure permitted to standardize and perform robotic surgery in a straightforward manner, as demonstrated by the reduction of operative and docking times observed in the last 10 thymectomies of our series.

Particular attention must be reserved for myasthenic and thymomatous patients. Even if on small series with short follow-up, in our study, the MGFA change in status resulted in improvement in 90% of cases. Similar results, with similar cohort and comparable follow-up period, were reported by other authors, who obtained an improvement of MGFA change in status, ranging between 91 and 93% [10, 11]. A comparative study by Cakar et al. [15] on 29 patients demonstrated that robotic thymectomy for MG provides at least the same positive effect as open transsternal thymectomy with regard to neurologic outcomes. Additionally, in a recent paper, Ruckert et al. [16] compared robotic and thoracoscopic thymectomy for MG on large cohorts, reporting an improved outcome for the robotic group, with a complete remission rate of 39.23 versus 20.3%, respectively. These papers suggest that the robotic system permits an excellent individuation of all thymic tissue, and thus the entire thymus en-bloc with perithymic tissue can be safely and radically removed.

Additionally, another important criticism concerned the surgical approach for patients with thymoma. In the past, many surgeons hesitated to perform VATS resection of thymomas, and median sternotomy was Generally accepted as a gold standard [1–3]. However, in 2011 a comparative study conducted by Pennathur et al. [17] on 40 patients revealed that VATS thymectomy for early-stage thymoma provides similar oncologic outcomes to those of open surgery. They reported a 5-year overall survival of 100%, with a mean follow-up period of 36 months. In our study, despite the small number of patients and the short follow-up, no disease recurrence was observed. We believe that the robotic system can overcome the limitations of traditional VATS, facilitating the individuation of thymoma capsule and normal thymic tissue, and so permitting safe and accurate manipulation and radical resection.

However, one of the most important limitations of the robotic system is represented by its cost: the initial capital cost, the annual maintenance cost and the cost of the disposable materials. At our department, this limitation was partly overcome, thanks to a multidisciplinary management of the robotic system that was fully utilized during the week by different surgical services.

Our study is subject to certain limitations. First, it included heterogeneous surgical procedures. Moreover, the main intent of this study was to report an all-inclusive experience of robotic surgery for mediastinal lesion. Thus, we did not separately analyse the results of each singular type of surgical procedure. Another limitation is represented by the retrospective nature of the study that may have resulted in a selection bias.

Nevertheless, these preliminary outcomes confirm the safety and feasibility of the robotic approach for different mediastinal lesions. Thymectomy for MG and early-stage thymoma appeared to be efficacious. The minimally invasive approach does not mean minimal accessibility, nor a lower degree of accuracy. However, further studies on larger cohorts with longer follow-ups are necessary in order to confirm the efficacy of robotic thymectomy, especially for an indolent disease such as thymoma. In this sense, prospective randomized trials should be designed in order to compare traditional open transsternal and robotic approaches for thymectomy, with regard to MG and thymoma.

Conflict of interest: none declared.

REFERENCES