Surg Endosc DOI 10.1007/s00464-015-4259-x
and Other Interventional Techniques
Robotic anatomic lung resections: the initial experience and description of learning in 102 cases ¨ zyurtkan1 • Erkan Kaba1 • Kemal Ayalp1 Alper Toker1 • Mehmet Og˘uzhan O 1 ¨ zkan Demirhan • Elena Uyumaz1 O
•
Received: 11 February 2015 / Accepted: 13 May 2015 Ó Springer Science+Business Media New York 2015
Abstract Background The aim of this study was to analyze our initial pulmonary resection experience with robotic surgery (Da Vinci, Intuitive Surgical, Inc., Mountain View, California, USA) and define the learning curve based on the duration of operations. Methods A retrospective review was conducted on patients undergoing robotic pulmonary resections from October 2011 to December 2014. The operating time, including the docking and console times, postoperative hospitalization, and peri- and postoperative complications were studied. Results Hundred patients underwent 102 robotic anatomic pulmonary resections due to various pathologies. Fifty-three percent of the patients underwent lobectomy procedure, whereas 45 % underwent segmentectomy. The mean operating time was 104 ± 34 min. The learning curve was calculated to be 14 patients (R2 = 0.57). The complication rate in our series was 24 % (n = 24) and higher in elderly patients (p = 0.03) and in patients with longer operating times (p = 0.03). Prolonged air leaks were observed in 10, and arrhythmia developed in nine patients. Two patients died, due to a concurrent lymphoblastic leukemia diagnosed at the postoperative period and exacerbation of interstitial fibrosis, respectively. Conclusions Robotic pulmonary resections prove to be safe and effective even at the initial learning experience. The duration of operations is considered to be acceptable. The learning curve could be established after 14 cases.
Keywords Learning curve Pulmonary surgical procedure Robotic surgical procedures Treatment outcome
¨ zyurtkan & Mehmet Og˘uzhan O [email protected]
Materials and methods
1
Department of Thoracic Surgery, Istanbul Bilim University and Group Florence Nightingale Hospital, Abide-i Hu¨rriyet Cad. No: 166, S¸ is¸ li, Istanbul, Turkey
The introduction of minimally invasive approaches in the early 1990s proved to be a major advance in thoracic surgery. Video-assisted thoracoscopic surgery (VATS) has gained wide acceptance as both diagnostic and therapeutic procedures despite its limited two-dimensional vision, unsteady camera platform, and limited maneuverability of the instruments [1]. More recently, the arrival of robotic surgery has led to the further refinement of the concept and practice of minimally invasive surgery. The use of the surgical robot has widely spread worldwide as it provides a high-resolution binocular view, three-dimensional imaging, no fulcrum effect, filtration of physiological tremor, ‘‘wrist-like’’ action of the instruments, and capability of fine dissections in confined spaces [2]. Today, several authors have performed robotic thoracic surgery with encouraging results [3–13]. As an academic thoracic surgery center, we have been performing minimally invasive anatomic lung resections with VATS for 8 years. We adopted the da Vinci Robotic System (Intuitive Surgical, Inc., Mountain View, California, USA) on October 2011. Our preliminary results on robotic thoracic surgery have been published [14, 15]. Here we report our experience with robotic thoracic pulmonary resections so far, define the learning period, and compare our results with the literature.
A retrospective review was conducted on patients undergoing robotic pulmonary resections from October 2011 to December 2014. All patients had chest computed tomography with
123
Surg Endosc
contrast enhancement; positron emission tomography was obtained in patients with a suspected malignancy. Additional studies included standard hematology and chemistry panel studies, cardiac evaluation, arterial blood gases, and pulmonary function testing. Mediastinoscopy was performed in cases in which the visual assessment of the positron emission tomography scan suggested mediastinal adenopathy. Patients with diagnosed lung cancer over stage IB had brain magnetic resonance imaging. In the beginning of the program, only patients with clinical T1a–b lesions or benign lung lesions were selected for anatomic lung resections. This period consisted of the first 30 patients. After gaining experience, we revised our indications and also included patients with cT2N1 lung cancer. In the second period, we also included patients with neoadjuvant treatment. However, we still perform VATS operations. The only clear indication in patient selection for robotic surgery is the patient’s will and the economical coverage of the expenditures by the special insurance system. All robotic resections were performed by a single thoracic surgeon (A.T.) at a single academic center with a previous experience of 330 anatomic lung resections, 370 thymectomies (resection of thymus and anterior mediastinal adipose tissue around thymus), and 70 thymomectomies (resection of thymoma and thymus) using videoassisted thoracic surgery. A dedicated team of surgeons (including table surgeon and first surgical assistant), nurses, and anesthetists involved during robotic lung resection procedures. The da Vinci Surgical System was used to perform individual dissection and isolation of the arterial, venous, and bronchial structures in anatomic lung resections. The ligation and division of these structures were accomplished by appropriately selected endoscopic staplers (Ethicon Endo-Surgery, Inc., Cincinnati, Ohio, and Covidien, Inc., CO, USA) and Hem-o-Lok (Teleflex Medical, Research Triangle Park, NC). In case of malignant diseases, systematic mediastinal lymph node dissection was performed. The operating time, including the docking and console times, postoperative hospitalization, and peri- and postoperative complications were recorded. The docking time was defined as time from the first skin incision to the start of driving the robotic arm while seated at the console. The console time was defined as the time that principal surgeon drove the robotic arm while seated at the console and performed the intrathoracic procedures (dissection, isolation of structures, ligation, division, and taking the specimen out). The closure time included the period between the undocking of the robotic arms and the closure of the last skin incision. The operating time was defined as the sum of the docking, console, and closure times. Chest tubes were
123
removed during the hospital stay as soon as the drainage was less than 300 mL/day and no air leaks were present, if the length of stay was shorter than 5 days. If the drainage or air leaks lasted longer and patients did not have any other problems, then the patients were discharged with chest tubes attached to the Heimlich valve. The morbidity was defined using the Society of Thoracic Surgery database’s definitions [16]. Operative mortality was defined as death within 30 days of the operation from any cause or before discharge. Outcomes reported included intraoperative time, whether the patient was converted from robotic to an open approach and the reason why, postoperative morbidity, pathological analysis, and length of hospital stay. The study received institutional review board approval, with individual patient consent being waived. Surgical technique The patient was positioned on lateral decubitus position after single-lung ventilation was established and confirmed with the fiberoptic bronchoscope. The table was tilted either anteriorly or posteriorly depending on the type of the operation to be performed before the robotic arms were docked. Three ports were opened while trying to keep 10 cm between each port and 10–15 cm from the target, which was the hilum of the targeted lobe. The camera was placed in the middle port. The robot was docked from the posterior of the patient with 30°–45° between the vertebral column of the patient and transverse axis of the cart. With the robotic camera in the up position, the ports and instruments were placed and the pleural symphyses were divided. The service port was performed at the 10th–11th intercostal space at the posterior part of the thoracic wall to be used for suctioning, retracting, and taking the specimens out. The rest of the operation was performed with the camera in the down position. Maryland or curved bipolar forceps for the right arm and prograsper for the left arm were used, and the positions were changed as needed. CO2 insufflation was used when there were tight pleural adhesions and in cases with emphysematous lungs. For the diagnosis of an indeterminate nodule, the resection was performed by using the camera of the robotic system only, without using the arms, for both economical reasons, and a better digital palpation. By this way, without using the arms, the cost of the procedure may be decreased in case the nodule turned out to be benign. We used robotic wedge, only if the lesion was obvious with inspection through the monitor, and we have high suspects for a lung malignancy with clinical findings. The specimen was sent for frozen examination for patients without preoperative diagnosis. During the resection and lymph node dissection, all the
Surg Endosc
resected materials were extracted through the service port which was covered with ALEXISÒ soft tissue skin retractor (Applied Medical, Rancho Santa Margarita, CA), and the main tissue containing the tumor was extracted using a plastic endobag. Individual dissection and division of the hilar structures were performed with endoscopic staplers introduced through the service port unless a specific introduction was needed, from either the right or left robotic arm ports. The incomplete fissures were divided either with a stapler introduced by the assistant surgeon through one of the ports, or with bipolar cauterization. Segmentectomies have been similarly performed [14, 15]. Glues or sealants were not used. In the case of malignant disease, systematic mediastinal lymph node dissections were performed using the same technique as in open surgery [17]. The chest was closed by placing a single 28 F chest tube from the camera port. Pain management Routine pain management was with intercostal blocks to two intercostal spaces upper and two intercostal spaces lower from inside the chest cavity with the aid of robotic arms and a male–female line attached to the needle using not more than 20 mL Marcaine (Astra Zeneca, Istanbul) applied perioperatively. Then 1 g Perfalgan (Bristol-Myers Squibb, New York City) intravenous infusion every 6 h and Voltaren SR 75 mg (Novartis, Basel) were given through the intramuscular route twice a day until the chest tube was removed. After the removal of the chest tube or discharge of the patient, oral medication with paracetamol and nonsteroid anti-inflammatory drugs were given.
Results During the study period, 100 patients underwent 102 robotic anatomic pulmonary resections due to various pathologies. Preoperative data of the patients are given in Table 1. There were 74 men (73 %) and 28 women (27 %), with a mean age of 60 ± 13 years. Twenty-nine patients (28 %) were older than 70 years, including seven octogenarian patients (7 %). Fifty-eight patients (54 %) had comorbidities, mostly previous cancer history (n = 17, 17 %), and coronary artery disease (n = 16, 16 %). The mean forced expiratory volume in one second (FEV1) was 2444 mL (range 1300–4500 mL) and 80 % (range 46– 128 %). Most of the resections were right-sided (n = 63, 62 %). Thirty-seven patients (36 %) who had no preoperative tissue diagnosis had a final pathological diagnosis intraoperatively. Ten patients (10 %) had neoadjuvant treatment before the operation. Short-term outcomes, including operative details and postoperative course, are shown in Table 2. The mean operating time was 104 ± 34 min (range 50–230 min). The mean docking time was 17 ± 10 min (range 5–45 min). The mean consol time was 82 ± 26 min (range 45–200 min). Four conversions to open surgery were required (4 %). The reasons for open conversions were pulmonary arterial
Table 1 Preoperative clinical data Age (years)a
60 ± 13
C70 (%)
29 (28)
C80 (%)
7 (7)
Gender (M/F) (%)
74:28 (73/27)
Comorbidity, n (%)
55 (54)
Previous cancer
17 (17)
Coronary artery disease
16 (16)
Statistical analysis The data were collected and analyzed using Excel software (Microsoft Corp, Seattle, WA). Descriptive statistics were used to report the means and standard deviations of the continuous variables and number and percent of categoric variables. Statistical comparisons were conducted using Chi-square, and t tests, as appropriate. A p value of less than or equal to 0.05 was considered a statistically significant difference. A scatter plot was constructed to evaluate the relationship of docking, consol, and operating times to the extent of experience, defined as the number of consecutive cases. In this plot, a regression trendline was drawn, and the change in the slope of the curve corresponding to the beginning of the plateau defined the learning curve. The overall learning curve was defined as the mean ± SD of the sum of the individual learning curves [11].
Hypertension
12 (12)
COPD
11 (11)
Diabetes mellitus FEV1 (mL, range) (%, range)
8 (8) 2444 (1300–4500) 80 (46–128)
Location of lesion (n, %) Right upper lobe
33 (32)
Right lower lobe
25 (24)
Left upper lobe
23 (23)
Left lower lobe
16 (17)
Middle lobe
5 (4)
Preoperative neoadjuvant treatment (n, %)b
10 (10)
M male, F female, COPD chronic obstructive pulmonary disease, FEV1 forced expiratory volume in 1 s a
Mean ± standard deviation
b
Chemotherapy in 4, chemoradiotherapy in 6
123
Surg Endosc Table 2 Perioperative and postoperative outcomes Operation time (mean in min ± SD) Docking
17 ± 10
Consol
82 ± 26
Operating
104 ± 34
Conversion to thoracotomy (n, %)a
4 (4)
Resection types (n, %) Lobectomy
54 (53)
Segmentectomy
46 (45)
Pneumonectomy
2 (2)
Transfusion (n, %)
4 (4)
Largest diameter of the lesion (mean in mm ± SD)
32 ± 25 mm
Complication (n, %)
24 (24)
Prolonged air leak
10
Arrhythmia
9
Pneumonia Atelectasis
3 1
Chylothorax
1
Chest tube duration, days (mean ± SD)
4±3
Hospital stay, days (mean ± SD)
5±4
60-day mortality (n, %)b
2 (2)
Follow-up, month, range
13 (2–35)
Long-term mortality (n, %)c
3 (3)
a
Pulmonary artery laceration in 3, and requirement of sleeve procedure in 1 b Postoperatively developing acute leukemia c
Metastatic disease in 2, and chronic disease in 1
hemorrhage in two patients and the need for a bronchial sleeve resection due to positive bronchial margin in one. After the main pulmonary artery was mistakenly stapled instead of truncus anterior artery, the fourth patient required an open conversion because of a need for a pulmonary arterial sleeve resection and anastomosis. This patient had had neoadjuvant chemo- and radiotreatment (60 Gy). In a later more careful examination of the chest computed tomography, it was found that the truncus anterior had been retracted into the right upper lobe which caused the misinterpretation of the right main pulmonary artery as the truncus anterior artery. Lobectomies and segmentectomies were the most commonly performed surgical procedures (53 and 45 %, respectively). Seventytwo patients (71 %) required a stay in the intensive care unit (range 6 h–2 days). Ten patients (10 %) were discharged with chest tubes attached to the Heimlich valve. The learning curves for operative times (docking, console, and operating) are given in Fig. 1. Figure 1A depicts the docking time versus number of consecutive cases. A sharp change in the slope of the regression trendline correlated with case 14 (R2 = 0.72), demonstrating the learning curve at docking. Figure 1B demonstrates the console time versus the number of consecutive cases.
123
Based on the change in the slope of the regression trendline, the learning curve is 13 patients (R2 = 0.41) at console. Finally, Fig. 1C shows the total operating time versus number of consecutive cases. Again, based on the change in the slope, the learning curve is 14 patients (R2 = 0.57) at robotic lung resection surgery. The complication rate in our series is 24 % (n = 24). Prolonged air leak developed in 10 patients (10 %), and arrhythmia occured in nine patients (9 %). Except for two patients who required a VATS revision to control air leaks on the seventh and fifteenth postoperative days, the remaining patients with air leaks were treated conservatively. During VATS revision, the parenchymal tissues causing air leak were found and sutured primarily. For patients developing arrhythmia, appropriate cardiologic medical treatment was given. Chest physiotherapy was sufficient for the patient developing atelectasis, without the need for a bronchoscopy. Antibiotic treatment was administered in patients developing pneumonia and chylothorax resolved with observation under medium-chain-fat diet without further surgical intervention. There were no statistically significant differences in the gender, the operation types, or the presence of comorbidities in patients with or without complications. The morbidity rate was insignificantly higher in patients with lower pulmonary function test (77 vs 82 %, p = 0.09). Besides, the risk of morbidity was significantly higher in elderly patients (64 vs 59 years, p = 0.03) and in patients with longer operating times (112 vs 101 min, p = 0.03). Patients developing morbidities had a significantly longer chest tube duration (7 vs 3 days, p = 0.0006) and longer hospital stay (8 vs 5 days, p = 0.02). There were only two deaths in our series (2 %). One patient who had undergone a right upper lobectomy due to adenocarcinoma was readmitted 1 week later with an infiltration of the contralateral lung and a leucocytosis of 88.000/mL. He was diagnosed with a concurrent lymphoblastic lymphoma through the bone marrow aspiration biopsy and died of chemotherapy side effects. The second patient had undergone a left upper lobectomy due to squamous cell carcinoma. He developed pneumonia, and exacerbation of interstitial fibrosis, and died on the 42th day despite treatment. Pathologic findings are listed in Table 3. Malignant diseases were more common (n = 81, 79 %), especially in elderly patients (62 vs 56 years, p = 0.05). Among malignant diseases, most patients (n = 75, 93 %) had primary and six (7 %) had metastatic lung cancer. All patients with metastatic disease underwent segmentectomy. Of the patients with primary lung cancer, 38 had pathologic stage IA, 17 had stage IB, 7 had stage IIA, 2 had stage IIB, 6 had stage IIIA, and 5 had stage IV diseases. Five patients had single-level N2 disease. All patients with stage IV disease
Surg Endosc
Fig. 1 A Docking time versus consecutive cases. The regression trendline is shown with the beginning of the plateau at case 14. B Consol time versus consecutive cases. The regression trendline is
shown with a sharp decline after 13 cases. C Operating time versus consecutive cases. The regression line is shown to be decreasing after 14 cases
had synchronous brain metastasis and underwent metastasectomy via open surgery (n = 2) or gamma knife treatment (n = 3) prior to the pulmonary resections. For primary lung cancer, 50 patients had lobectomies, 23 had segmentectomies, and 2 had pneumonectomies. Tuberculosis (n = 7) was the most common benign pathology, followed by bronchiectasis (n = 4). Seventeen patients underwent segmentectomies, and four underwent lobectomies for benign disease. The mean console time of pulmonary resections in patients with malignant pathologies was insignificantly longer than that for nonmalignant diseases (82 vs 78 min, p = 0.3). The mean console time of lobectomy operations was insignificantly longer than that for segmentectomy operations (83 vs 79 min, p = 0.3). The mean follow-up time was 13 months (range 2–36 months). During the follow-up, three patients died (3 %). A patient who underwent right upper lobectomy due
to adenocarcinoma of the lung (stage IIA) developed multimetastatic disease and died 15 months after the operation. The second patient who underwent segmentectomy for squamous cell carcinoma died 21 months after the operation due to malignant pleural and pericardial effusions. Both patients had a previous history of breast cancer and treatment, consecutively 3 and 16 years ago, and received chemo- and ipsilateral thoracic radiotherapy. The third patient developed contralateral lung metastasis at 7 months and died of chemotherapy side effects. One patient had a metastatic (brain) recurrence at postoperative 6th month. Among the four patients who had metastasectomy operations, only the one with osteosarcoma developed recurrence within a follow-up of 9 months and scheduled for resurgery after the completion of the chemotreatment. The remaining patients with malignant diseases and all patients with benign pathologies are disease free.
123
Surg Endosc Table 3 Pathological diagnosis Benign (n, %)
7
Bronchiectasis
4
Hamartoma
2
Sequestration
2
Interstitial lung disease
2
Othersa
4
Malignant (n, %) Adenocarcinoma
a
21 (21)
Tuberculosis
81 (79) 35
Squamous cell carcinoma
23
Metastaticb
6
Large cell carcinoma
5
Adenosquamous carcinoma
5
Typical carcinoid tumor
5
Othersc
2
Aspergilloma, hydatid cyst, pneumonia, sarcoidosis
b
Colon cancer in 2, breast cancer in 1, osteosarcoma in 1, liver carcinoma in 1, soft tissue sarcoma in 1
c
Small cell carcinoma, indifferentiated sarcoma
Discussion In this study, we report our experience with robotic thoracic pulmonary resections performed for various etiologies and analyze and compare our results with the literature. Robotic thoracic surgery reports by several authors within the past decade are presented [3–13]. Our study differs from these previous reports in several ways: (1) First of all, our study differs from the previous reports on the primary pathologies and the type of the operations. Nearly all previous series are based on patients with primary lung cancer, and a great majority of the operations are lobectomies. Contrary to this, 21 % of our patients had benign diseases, and the rate of segmentectomy operations was high (45 %). (2) Second, we calculate that 14 cases are enough to complete the learning curve for robotic pulmonary resections, being lesser than proposals of previous studies [4, 9, 11, 13]. Since we performed nearly equal numbers of lobectomy and segmentectomy (53 vs 45 %), our proposal of 14 cases could be considered more descriptive as the learning curve of robotic pulmonary resections. (3) Our total operating time is the shortest reported so far. The possible explanations for this are given below. An improved comfort and quality of life of the patient are more frequently observed as the technology evolves. Minimally invasive surgical techniques are claimed to reduce pain, surgical inflammation, and postoperative dysfunction [18, 19]. Thoracoscopic surgery has gained wide adoption with these goals. Still, factors restrict a wider
123
adoption of VATS. These include the nonergonomy and difficulty in the learning curve, the limitation of two-dimensional visual information, and the restricted maneuverability of the instruments (mostly because of the rigid axes fixed to the thoracic wall by the entry trocar) [20]. The advanced engineering embodied in the da Vinci System makes it possible to overcome many of the disadvantages of VATS. Cadieres et al. [21] first described the robotic surgery in humans in 1997. Since then, robotic surgery in general has evolved and become widespread in various fields of surgery. In 2002, Melfi et al. [22] reported the robotic lung resections. Several important and large series followed the initial publication experience, mostly on robotic lobectomies for lung cancer [3–13]. These studies demonstrate that robotic thoracic surgery is feasible and safe with encouraging results (Table 4). Previous series reported a mortality rate between 0 and 3 %, a morbidity rate between 10.5 and 43 %, and a conversion rate between 0 and 15.7 %. Our results are within the abovedescribed reported range (2, 24, and 4 %, respectively). In terms of the length of stay, our results were similar to the ones reported before. In this study, we find that the risk of morbidity is significantly higher in elderly patients, as in open and VATS, and in patients with longer operating times. Advanced age is considered as a risk factor for lung resections [23]. Arrhythmia and prolonged air leaks are the most commonly reported complications after robotic pulmonary resections with a rate between 3 and 19 %, and 2.8 % and 13, respectively [3–13]. In our series, prolonged air leaks occured in 10 %, and arrhythmia developed in 9 % of the patients in parallel with the literature. Several series had an operative time in robotic pulmonary resection ranging between 132 and 226 min [3– 13]. This study had the shortest operative time reported so far (104 min, range 50–230 min). The reasons of such a short duration in our series can be explained in several ways. First, as we mentioned above, we started our thoracic robotic program after an established experience of VATS program. Second, in almost all previously reported large series about robotic pulmonary resections, lobectomy procedures consisted of the great majority of the resections. But in our series, the rate of lobectomy was nearly equal to that of segmentectomy (53 vs 45 %), and as we mentioned above, the mean console time of segmentectomy operations was shorter than that of lobectomy operations. Finally, a great majority of the patients reported in previous studies had primary lung cancers, and their operative times included mediastinal lymph node dissections. Twenty-one percent of the patients in our series had benign pulmonary pathologies; thus, they did not require mediastinal staging, so their operative times were insignificantly shorter.
Surg Endosc Table 4 Results of the recent reports in the field of robotic pulmonary resection Authors
Year
n
Age (years)
Anderson et al. [3]
2007
21
Gharagozloo et al. [4]
2009
100
Giulianotti et al. [5]
2010
38
Dylewski et al. [6]
2011
200
68 (20–92)
175 (82–370)
1.5
1.5
26
3 (1–44)
Fortes et al. [7]
2011
23
70 (51–86)
218 (74–323)
17
0
39
3 (1–13)
Cerfolio et al. [8]
2011
106
Veronesi et al. [9]
2011
91
Op. time (min)
C (%)
Mrt (%)
Mrb (%)
Stay
67 (36–86)
216 (60–384)
0
0
27
4 (2–10)
65 ± 8
216 ± 27
1
3
31
4 (3–42)
66 (16–78)
209 ± 66
15.7
2.6
10.5
10 (3–24)
66 (31–85)
132 ± 60
11.9
0
27
2 (1–7)
NP
226 (146–513)
11
0
27
5 (3–38)
Park et al. [10]
2012
325
66 (30–87)
206 (110–383)
8.3
0.3
25
5 (2–28)
Meyer et al. [11]
2012
185
65 ± 9
211 ± 60
1.6
1.6
16.8
4 (2–21)
Louie et al. [12]
2012
46
65
213
5.7
0
43
4 (2–21)
Lee et al. [13] This study
2014 2015
35 102
71 (52–84) 60 ± 13
161 (104–272) 104 (50–230)
2.9 4
0 2
11 24
3 (2–6) 5 (2–42)
Values are presented as mean ± standard deviation or median (range) N number, Op operation, C conversion, Mrt mortality, Mrb morbidity, NP nonprecise
The term ‘‘learning curve’’ has been frequently used in publications about surgical procedures as a reference to the process of gaining experience and improving skills in performing such procedures. The da Vinci System has a rapid learning curve [24]. Obtaining shorter operative time is regarded as a reference in several reports, mostly dealing with robotic lobectomies. Although operative times vary with the complexity of individual cases, it appears that the abrupt decrease in operative times corresponds to the learning curve [11]. Melfi et al. [25] suggested a minimum of 20 operations to acquire new skills for the whole surgical team. Gharagozloo et al. [4] and Veronesi et al. [9] separately observed that the first 20 patients had longer operative times. Based on operative times, Meyer et al. [11] proposed a learning curve of 18 ± 3 cases and Lee et al. [13] of 15–17 cases. Based on the operative times (docking, console, and total) in our current analyses, we calculate that 14 cases could be considered as completion of the learning curve for pulmonary resections. However, during the first 14 pulmonary resections, we also performed other thoracic surgical procedures such as thymectomies. Thus, 20 cases may be an appropriate learning period in robotic thoracic surgery. Both video- and robotic-assisted pulmonary resections offer minimally invasive approaches to the patients. Very recently, two studies have been published comparing VATS and robotic lobectomies [13], and VATS and robotic segmentectomies [14]. Both studies showed that both approaches were similar in terms of mortality, morbidity, and the length of stay. In addition, both studies demonstrated that the duration of both VATS lobectomies and segmentectomies was significantly shorter than that of robotic ones. Another common conclusion of both studies was that as the maturity in robotic surgery increased, the
duration of both operations reached similar levels. Meanwhile, the rate of conversion to thoracotomy was insignificantly lower in robotic approach [14]. As previously mentioned, the principal surgeon performing robotic resection in this study has had enough experience in VATS resections, and the fact that all the operations had been performed by the same surgeon eliminated any potential effect on surgeon variability on the results. Based on the personal knowledge and background in VATS, we can say that the beginning period of robotic-assisted thoracic surgery was more complicated, troublesome, and time-consuming. Thus, during this period, a VATS surgeon could easily regret the robotic surgery without understanding the real benefits of robotic-assisted thoracic surgery. Certainly, a minimally invasive chest surgeon will notice the precise dissection capabilities around the vessels, better visualization with stable camera platform, and dissection of lymph nodes around segmentary arteries and lobar bronchi. The second most important advantage, which has not been very often mentioned in the literature, is the capability of extra-routine dissection techniques. In VATS resections, a surgeon generally and traditionally dissects from anterior to posterior or from posterior to anterior. This may cause troublesome bleeding when one of the major structures has adhesions and difficulty in dissection. In robotic surgery, both arms can be used for dissection purposes. Thus, dexterity is a superiority when there is a difficulty in the dissection of major vessels, and the surgeon may use anterior, posterior, superior, or fissural approaches, whichever is suitable. Even, the surgeon can dissect in an extraordinary order. That superiority of the robotic surgery provides surgical technical comfort and may have the potential to decrease the conversion rate. We believe, as in VATS lobectomies, the
123
Surg Endosc
resection of the left upper lobe or superior/inferior bilobectomies of the right lung need extreme care and experience. We believe segmentectomies are better operations to start with. In particular, superior segmentectomies and common basal segmentectomies of the lower lobes are the best operations during completion of the learning curve. Robotic approaches for pulmonary resections are feasible and safe and have encouraging results even in the initial learning experience of the establishment of a robotic thoracic surgery program. This serves as a platform for further improvement in minimally invasive technologies. Disclosures No conflicts of interest or financial ties to disclose have ¨ zyurtkan, E. Kaba, K. Ayalp, been declared by A. Toker, MO. O ¨ . Demirhan, and E. Uyumaz. O
References 1. McKenna RJ Jr, Houck W, Fuller CB (2006) Video-assisted thoracic surgery lobectomy: experience with 1100 cases. Ann Thorac Surg 81:421–425 2. Ambrogi MC, Fanucchi O, Melfi F, Mussi A (2014) Robotic surgery for lung cancer. Korean J Thorac Cardiovasc Surg 47:201–210 3. Anderson CA, Hellan M, Falebella A, Lau CS, Grannis FW Jr, Kernstine KH (2007) Robotic-assisted lung resection for malignant disease. Innovations 2:254–258 4. Gharagozloo F, Margolis M, Tempesta B, Strother E, Najam F (2009) Robot-assisted lobectomy for early-stage lung cancer: report of 100 consecutive cases. Ann Thorac Surg 88:380–384 5. Giulianotti PC, Buchs NC, Caravaglios G, Bianco FM (2010) Robot-assisted lung resection: outcomes and technical details. Interact CardioVasc Thorac Surg 11:388–392 6. Dylewski MR, Ohaeto AC, Pereira JF (2011) Pulmonary resection using a total endoscopic robotic video-assisted approach. Semin Thorac Surg 23:36–42 7. Fortes DL, Tomaszek SC, Wigle DA (2011) Early experience with robotic-assisted lung resection. Innovations 6:237–242 8. Cerfolio RJ, Bryant AS, Skylizard L, Minnich DJ (2011) Initial consecutive experience of completely portal robotic pulmonary resection with 4 arms (2011). J Thorac Cardiovasc Surg 142:740–746 9. Veronesi G, Agoglia BG, Melfi F, Maisonneuve P, Bertoletti R, Bianchi PP, Rocco B, Borri A, Gasparri R, Spaggiari L (2001) Experience with robotic lobectomy for lung cancer. Innovations 6:355–360 10. Park BJ, Melfi F, Mussi A, Maisonneuve P, Spaggiari L, Da Silva RK, Veronesi G (2012) Robotic lobectomy for non-small cell lung cancer (NSCLC): long-term oncologic results. J Thorac Cardiovasc Surg 143:383–389
123
11. Meyer M, Gharagozloo F, Tempesta B, Margolis M, Strother E, Christenson D (2012) The learning curve of robotic lobectomy. Int J Med Robotics Comput Assist Surg 8:448–452 12. Louie BE, Farivar AS, Aye RW, Vallieres E (2012) Early experience with robotic lung resection results in similar operative outcomes and morbidity when compared with matched video-assisted thoracoscopic surgery cases. Ann Thorac Surg 93:1598–1605 13. Lee BE, Korst RJ, Kletsman E, Rutledge JR (2014) Transitioning from video-assisted thoracic surgical lobectomy to robotics for lung cancer: are there outcomes advantages? J Thorac Cardiovasc Surg 147:724–729 ¨ zkan B, Kaba E, Toker A (2014) Robotic 14. Demir A, Ayalp K, O and video-assisted thoracic surgery lung segmentectomy for malignant and benign lesions. Interact Cardiovasc Thorac Surg ivu 39. [Epub ahead of print] ¨ , Erus S 15. Toker A, Ayalp K, Uyumaz E, Kaba E, Demirhan O (2014) Robotic lung segmentectomy for malignant and benign lesions. J Thorac Dis 6(7):937–942 16. The Society of Thoracic Surgeons. Database. http://www.sts.org/ sections/stsnationaldatabase/publications/executive/artcile.html 17. Ma K, Chang D, He B, Gong M, Tian F, Hu X, Ji Z, Wang T (2008) Radical systematic mediastinal lymphadenectomy vs. mediastinal lymph node sampling in patients with clinical stage IA and pathological stage T1 non-small cell lung cancer. Cancer Res Clin Oncol 134:1289–1295 18. Ng CS, Wan S, Hui CW, Lee TW, Underwood MJ, Yim AP (2006) Video-assisted thoracic surgery for early stage lung cancer- can short term immunological advantages improve long-term survival? Ann Thorac Cardiovasc Surg 12:308–312 19. Nakata M, Saeki H, Yokoyama N, Kurita A, Takiyama W, Takashima S (2000) Pulmonary function after lobectomy: videoassisted thoracic surgery vs. thoracotomy. Ann Thorac Surg 70:938–941 20. Gharagozloo F, Margolis M, Tempesta B (2008) Robot-assisted thoracoscopic lobectomy for early-stage lung cancer. Ann Thorac Surg 85:1880–1886 21. Cadiere GB, Himpens J, Vertruyen M, Favretti F (1999) The World’s first obesity surgery performed by a surgeon at a distance. Obes Surg 9:206–209 22. Melfi FMA, Menconi GF, Mariani AM, Angeletti CA (2002) Early experience with robotic technology for thoracoscopic surgery. Eur J Cardiothorac Surg 21:864–868 23. Ferguson MK (2009) Pulmonary physiologic assessment of operative risk. In: Shields TW, LoCicero J, Reed CE, Feins RH (eds) General thoracic surgery. Lippincott Williams & Wilkins, Philadelphia, pp 325–338 24. Hernandez J, Bann S, Munz K, Moorthy K, Datta V, Martin S, Dosis A, Bello F, Darzi A, Rockall T (2004) Qualitative and quantitative analysis of the learning curve of a simulated task on the da Vinci system. Surg Endosc 18:372–378 25. Melfi FMA, Mussi A (2008) Robotically assisted lobectomy: learning curve and complications. Thorac Surg Clin 18:289–295
and Other Interventional Techniques
Robotic anatomic lung resections: the initial experience and description of learning in 102 cases ¨ zyurtkan1 • Erkan Kaba1 • Kemal Ayalp1 Alper Toker1 • Mehmet Og˘uzhan O 1 ¨ zkan Demirhan • Elena Uyumaz1 O
•
Received: 11 February 2015 / Accepted: 13 May 2015 Ó Springer Science+Business Media New York 2015
Abstract Background The aim of this study was to analyze our initial pulmonary resection experience with robotic surgery (Da Vinci, Intuitive Surgical, Inc., Mountain View, California, USA) and define the learning curve based on the duration of operations. Methods A retrospective review was conducted on patients undergoing robotic pulmonary resections from October 2011 to December 2014. The operating time, including the docking and console times, postoperative hospitalization, and peri- and postoperative complications were studied. Results Hundred patients underwent 102 robotic anatomic pulmonary resections due to various pathologies. Fifty-three percent of the patients underwent lobectomy procedure, whereas 45 % underwent segmentectomy. The mean operating time was 104 ± 34 min. The learning curve was calculated to be 14 patients (R2 = 0.57). The complication rate in our series was 24 % (n = 24) and higher in elderly patients (p = 0.03) and in patients with longer operating times (p = 0.03). Prolonged air leaks were observed in 10, and arrhythmia developed in nine patients. Two patients died, due to a concurrent lymphoblastic leukemia diagnosed at the postoperative period and exacerbation of interstitial fibrosis, respectively. Conclusions Robotic pulmonary resections prove to be safe and effective even at the initial learning experience. The duration of operations is considered to be acceptable. The learning curve could be established after 14 cases.
Keywords Learning curve Pulmonary surgical procedure Robotic surgical procedures Treatment outcome
¨ zyurtkan & Mehmet Og˘uzhan O [email protected]
Materials and methods
1
Department of Thoracic Surgery, Istanbul Bilim University and Group Florence Nightingale Hospital, Abide-i Hu¨rriyet Cad. No: 166, S¸ is¸ li, Istanbul, Turkey
The introduction of minimally invasive approaches in the early 1990s proved to be a major advance in thoracic surgery. Video-assisted thoracoscopic surgery (VATS) has gained wide acceptance as both diagnostic and therapeutic procedures despite its limited two-dimensional vision, unsteady camera platform, and limited maneuverability of the instruments [1]. More recently, the arrival of robotic surgery has led to the further refinement of the concept and practice of minimally invasive surgery. The use of the surgical robot has widely spread worldwide as it provides a high-resolution binocular view, three-dimensional imaging, no fulcrum effect, filtration of physiological tremor, ‘‘wrist-like’’ action of the instruments, and capability of fine dissections in confined spaces [2]. Today, several authors have performed robotic thoracic surgery with encouraging results [3–13]. As an academic thoracic surgery center, we have been performing minimally invasive anatomic lung resections with VATS for 8 years. We adopted the da Vinci Robotic System (Intuitive Surgical, Inc., Mountain View, California, USA) on October 2011. Our preliminary results on robotic thoracic surgery have been published [14, 15]. Here we report our experience with robotic thoracic pulmonary resections so far, define the learning period, and compare our results with the literature.
A retrospective review was conducted on patients undergoing robotic pulmonary resections from October 2011 to December 2014. All patients had chest computed tomography with
123
Surg Endosc
contrast enhancement; positron emission tomography was obtained in patients with a suspected malignancy. Additional studies included standard hematology and chemistry panel studies, cardiac evaluation, arterial blood gases, and pulmonary function testing. Mediastinoscopy was performed in cases in which the visual assessment of the positron emission tomography scan suggested mediastinal adenopathy. Patients with diagnosed lung cancer over stage IB had brain magnetic resonance imaging. In the beginning of the program, only patients with clinical T1a–b lesions or benign lung lesions were selected for anatomic lung resections. This period consisted of the first 30 patients. After gaining experience, we revised our indications and also included patients with cT2N1 lung cancer. In the second period, we also included patients with neoadjuvant treatment. However, we still perform VATS operations. The only clear indication in patient selection for robotic surgery is the patient’s will and the economical coverage of the expenditures by the special insurance system. All robotic resections were performed by a single thoracic surgeon (A.T.) at a single academic center with a previous experience of 330 anatomic lung resections, 370 thymectomies (resection of thymus and anterior mediastinal adipose tissue around thymus), and 70 thymomectomies (resection of thymoma and thymus) using videoassisted thoracic surgery. A dedicated team of surgeons (including table surgeon and first surgical assistant), nurses, and anesthetists involved during robotic lung resection procedures. The da Vinci Surgical System was used to perform individual dissection and isolation of the arterial, venous, and bronchial structures in anatomic lung resections. The ligation and division of these structures were accomplished by appropriately selected endoscopic staplers (Ethicon Endo-Surgery, Inc., Cincinnati, Ohio, and Covidien, Inc., CO, USA) and Hem-o-Lok (Teleflex Medical, Research Triangle Park, NC). In case of malignant diseases, systematic mediastinal lymph node dissection was performed. The operating time, including the docking and console times, postoperative hospitalization, and peri- and postoperative complications were recorded. The docking time was defined as time from the first skin incision to the start of driving the robotic arm while seated at the console. The console time was defined as the time that principal surgeon drove the robotic arm while seated at the console and performed the intrathoracic procedures (dissection, isolation of structures, ligation, division, and taking the specimen out). The closure time included the period between the undocking of the robotic arms and the closure of the last skin incision. The operating time was defined as the sum of the docking, console, and closure times. Chest tubes were
123
removed during the hospital stay as soon as the drainage was less than 300 mL/day and no air leaks were present, if the length of stay was shorter than 5 days. If the drainage or air leaks lasted longer and patients did not have any other problems, then the patients were discharged with chest tubes attached to the Heimlich valve. The morbidity was defined using the Society of Thoracic Surgery database’s definitions [16]. Operative mortality was defined as death within 30 days of the operation from any cause or before discharge. Outcomes reported included intraoperative time, whether the patient was converted from robotic to an open approach and the reason why, postoperative morbidity, pathological analysis, and length of hospital stay. The study received institutional review board approval, with individual patient consent being waived. Surgical technique The patient was positioned on lateral decubitus position after single-lung ventilation was established and confirmed with the fiberoptic bronchoscope. The table was tilted either anteriorly or posteriorly depending on the type of the operation to be performed before the robotic arms were docked. Three ports were opened while trying to keep 10 cm between each port and 10–15 cm from the target, which was the hilum of the targeted lobe. The camera was placed in the middle port. The robot was docked from the posterior of the patient with 30°–45° between the vertebral column of the patient and transverse axis of the cart. With the robotic camera in the up position, the ports and instruments were placed and the pleural symphyses were divided. The service port was performed at the 10th–11th intercostal space at the posterior part of the thoracic wall to be used for suctioning, retracting, and taking the specimens out. The rest of the operation was performed with the camera in the down position. Maryland or curved bipolar forceps for the right arm and prograsper for the left arm were used, and the positions were changed as needed. CO2 insufflation was used when there were tight pleural adhesions and in cases with emphysematous lungs. For the diagnosis of an indeterminate nodule, the resection was performed by using the camera of the robotic system only, without using the arms, for both economical reasons, and a better digital palpation. By this way, without using the arms, the cost of the procedure may be decreased in case the nodule turned out to be benign. We used robotic wedge, only if the lesion was obvious with inspection through the monitor, and we have high suspects for a lung malignancy with clinical findings. The specimen was sent for frozen examination for patients without preoperative diagnosis. During the resection and lymph node dissection, all the
Surg Endosc
resected materials were extracted through the service port which was covered with ALEXISÒ soft tissue skin retractor (Applied Medical, Rancho Santa Margarita, CA), and the main tissue containing the tumor was extracted using a plastic endobag. Individual dissection and division of the hilar structures were performed with endoscopic staplers introduced through the service port unless a specific introduction was needed, from either the right or left robotic arm ports. The incomplete fissures were divided either with a stapler introduced by the assistant surgeon through one of the ports, or with bipolar cauterization. Segmentectomies have been similarly performed [14, 15]. Glues or sealants were not used. In the case of malignant disease, systematic mediastinal lymph node dissections were performed using the same technique as in open surgery [17]. The chest was closed by placing a single 28 F chest tube from the camera port. Pain management Routine pain management was with intercostal blocks to two intercostal spaces upper and two intercostal spaces lower from inside the chest cavity with the aid of robotic arms and a male–female line attached to the needle using not more than 20 mL Marcaine (Astra Zeneca, Istanbul) applied perioperatively. Then 1 g Perfalgan (Bristol-Myers Squibb, New York City) intravenous infusion every 6 h and Voltaren SR 75 mg (Novartis, Basel) were given through the intramuscular route twice a day until the chest tube was removed. After the removal of the chest tube or discharge of the patient, oral medication with paracetamol and nonsteroid anti-inflammatory drugs were given.
Results During the study period, 100 patients underwent 102 robotic anatomic pulmonary resections due to various pathologies. Preoperative data of the patients are given in Table 1. There were 74 men (73 %) and 28 women (27 %), with a mean age of 60 ± 13 years. Twenty-nine patients (28 %) were older than 70 years, including seven octogenarian patients (7 %). Fifty-eight patients (54 %) had comorbidities, mostly previous cancer history (n = 17, 17 %), and coronary artery disease (n = 16, 16 %). The mean forced expiratory volume in one second (FEV1) was 2444 mL (range 1300–4500 mL) and 80 % (range 46– 128 %). Most of the resections were right-sided (n = 63, 62 %). Thirty-seven patients (36 %) who had no preoperative tissue diagnosis had a final pathological diagnosis intraoperatively. Ten patients (10 %) had neoadjuvant treatment before the operation. Short-term outcomes, including operative details and postoperative course, are shown in Table 2. The mean operating time was 104 ± 34 min (range 50–230 min). The mean docking time was 17 ± 10 min (range 5–45 min). The mean consol time was 82 ± 26 min (range 45–200 min). Four conversions to open surgery were required (4 %). The reasons for open conversions were pulmonary arterial
Table 1 Preoperative clinical data Age (years)a
60 ± 13
C70 (%)
29 (28)
C80 (%)
7 (7)
Gender (M/F) (%)
74:28 (73/27)
Comorbidity, n (%)
55 (54)
Previous cancer
17 (17)
Coronary artery disease
16 (16)
Statistical analysis The data were collected and analyzed using Excel software (Microsoft Corp, Seattle, WA). Descriptive statistics were used to report the means and standard deviations of the continuous variables and number and percent of categoric variables. Statistical comparisons were conducted using Chi-square, and t tests, as appropriate. A p value of less than or equal to 0.05 was considered a statistically significant difference. A scatter plot was constructed to evaluate the relationship of docking, consol, and operating times to the extent of experience, defined as the number of consecutive cases. In this plot, a regression trendline was drawn, and the change in the slope of the curve corresponding to the beginning of the plateau defined the learning curve. The overall learning curve was defined as the mean ± SD of the sum of the individual learning curves [11].
Hypertension
12 (12)
COPD
11 (11)
Diabetes mellitus FEV1 (mL, range) (%, range)
8 (8) 2444 (1300–4500) 80 (46–128)
Location of lesion (n, %) Right upper lobe
33 (32)
Right lower lobe
25 (24)
Left upper lobe
23 (23)
Left lower lobe
16 (17)
Middle lobe
5 (4)
Preoperative neoadjuvant treatment (n, %)b
10 (10)
M male, F female, COPD chronic obstructive pulmonary disease, FEV1 forced expiratory volume in 1 s a
Mean ± standard deviation
b
Chemotherapy in 4, chemoradiotherapy in 6
123
Surg Endosc Table 2 Perioperative and postoperative outcomes Operation time (mean in min ± SD) Docking
17 ± 10
Consol
82 ± 26
Operating
104 ± 34
Conversion to thoracotomy (n, %)a
4 (4)
Resection types (n, %) Lobectomy
54 (53)
Segmentectomy
46 (45)
Pneumonectomy
2 (2)
Transfusion (n, %)
4 (4)
Largest diameter of the lesion (mean in mm ± SD)
32 ± 25 mm
Complication (n, %)
24 (24)
Prolonged air leak
10
Arrhythmia
9
Pneumonia Atelectasis
3 1
Chylothorax
1
Chest tube duration, days (mean ± SD)
4±3
Hospital stay, days (mean ± SD)
5±4
60-day mortality (n, %)b
2 (2)
Follow-up, month, range
13 (2–35)
Long-term mortality (n, %)c
3 (3)
a
Pulmonary artery laceration in 3, and requirement of sleeve procedure in 1 b Postoperatively developing acute leukemia c
Metastatic disease in 2, and chronic disease in 1
hemorrhage in two patients and the need for a bronchial sleeve resection due to positive bronchial margin in one. After the main pulmonary artery was mistakenly stapled instead of truncus anterior artery, the fourth patient required an open conversion because of a need for a pulmonary arterial sleeve resection and anastomosis. This patient had had neoadjuvant chemo- and radiotreatment (60 Gy). In a later more careful examination of the chest computed tomography, it was found that the truncus anterior had been retracted into the right upper lobe which caused the misinterpretation of the right main pulmonary artery as the truncus anterior artery. Lobectomies and segmentectomies were the most commonly performed surgical procedures (53 and 45 %, respectively). Seventytwo patients (71 %) required a stay in the intensive care unit (range 6 h–2 days). Ten patients (10 %) were discharged with chest tubes attached to the Heimlich valve. The learning curves for operative times (docking, console, and operating) are given in Fig. 1. Figure 1A depicts the docking time versus number of consecutive cases. A sharp change in the slope of the regression trendline correlated with case 14 (R2 = 0.72), demonstrating the learning curve at docking. Figure 1B demonstrates the console time versus the number of consecutive cases.
123
Based on the change in the slope of the regression trendline, the learning curve is 13 patients (R2 = 0.41) at console. Finally, Fig. 1C shows the total operating time versus number of consecutive cases. Again, based on the change in the slope, the learning curve is 14 patients (R2 = 0.57) at robotic lung resection surgery. The complication rate in our series is 24 % (n = 24). Prolonged air leak developed in 10 patients (10 %), and arrhythmia occured in nine patients (9 %). Except for two patients who required a VATS revision to control air leaks on the seventh and fifteenth postoperative days, the remaining patients with air leaks were treated conservatively. During VATS revision, the parenchymal tissues causing air leak were found and sutured primarily. For patients developing arrhythmia, appropriate cardiologic medical treatment was given. Chest physiotherapy was sufficient for the patient developing atelectasis, without the need for a bronchoscopy. Antibiotic treatment was administered in patients developing pneumonia and chylothorax resolved with observation under medium-chain-fat diet without further surgical intervention. There were no statistically significant differences in the gender, the operation types, or the presence of comorbidities in patients with or without complications. The morbidity rate was insignificantly higher in patients with lower pulmonary function test (77 vs 82 %, p = 0.09). Besides, the risk of morbidity was significantly higher in elderly patients (64 vs 59 years, p = 0.03) and in patients with longer operating times (112 vs 101 min, p = 0.03). Patients developing morbidities had a significantly longer chest tube duration (7 vs 3 days, p = 0.0006) and longer hospital stay (8 vs 5 days, p = 0.02). There were only two deaths in our series (2 %). One patient who had undergone a right upper lobectomy due to adenocarcinoma was readmitted 1 week later with an infiltration of the contralateral lung and a leucocytosis of 88.000/mL. He was diagnosed with a concurrent lymphoblastic lymphoma through the bone marrow aspiration biopsy and died of chemotherapy side effects. The second patient had undergone a left upper lobectomy due to squamous cell carcinoma. He developed pneumonia, and exacerbation of interstitial fibrosis, and died on the 42th day despite treatment. Pathologic findings are listed in Table 3. Malignant diseases were more common (n = 81, 79 %), especially in elderly patients (62 vs 56 years, p = 0.05). Among malignant diseases, most patients (n = 75, 93 %) had primary and six (7 %) had metastatic lung cancer. All patients with metastatic disease underwent segmentectomy. Of the patients with primary lung cancer, 38 had pathologic stage IA, 17 had stage IB, 7 had stage IIA, 2 had stage IIB, 6 had stage IIIA, and 5 had stage IV diseases. Five patients had single-level N2 disease. All patients with stage IV disease
Surg Endosc
Fig. 1 A Docking time versus consecutive cases. The regression trendline is shown with the beginning of the plateau at case 14. B Consol time versus consecutive cases. The regression trendline is
shown with a sharp decline after 13 cases. C Operating time versus consecutive cases. The regression line is shown to be decreasing after 14 cases
had synchronous brain metastasis and underwent metastasectomy via open surgery (n = 2) or gamma knife treatment (n = 3) prior to the pulmonary resections. For primary lung cancer, 50 patients had lobectomies, 23 had segmentectomies, and 2 had pneumonectomies. Tuberculosis (n = 7) was the most common benign pathology, followed by bronchiectasis (n = 4). Seventeen patients underwent segmentectomies, and four underwent lobectomies for benign disease. The mean console time of pulmonary resections in patients with malignant pathologies was insignificantly longer than that for nonmalignant diseases (82 vs 78 min, p = 0.3). The mean console time of lobectomy operations was insignificantly longer than that for segmentectomy operations (83 vs 79 min, p = 0.3). The mean follow-up time was 13 months (range 2–36 months). During the follow-up, three patients died (3 %). A patient who underwent right upper lobectomy due
to adenocarcinoma of the lung (stage IIA) developed multimetastatic disease and died 15 months after the operation. The second patient who underwent segmentectomy for squamous cell carcinoma died 21 months after the operation due to malignant pleural and pericardial effusions. Both patients had a previous history of breast cancer and treatment, consecutively 3 and 16 years ago, and received chemo- and ipsilateral thoracic radiotherapy. The third patient developed contralateral lung metastasis at 7 months and died of chemotherapy side effects. One patient had a metastatic (brain) recurrence at postoperative 6th month. Among the four patients who had metastasectomy operations, only the one with osteosarcoma developed recurrence within a follow-up of 9 months and scheduled for resurgery after the completion of the chemotreatment. The remaining patients with malignant diseases and all patients with benign pathologies are disease free.
123
Surg Endosc Table 3 Pathological diagnosis Benign (n, %)
7
Bronchiectasis
4
Hamartoma
2
Sequestration
2
Interstitial lung disease
2
Othersa
4
Malignant (n, %) Adenocarcinoma
a
21 (21)
Tuberculosis
81 (79) 35
Squamous cell carcinoma
23
Metastaticb
6
Large cell carcinoma
5
Adenosquamous carcinoma
5
Typical carcinoid tumor
5
Othersc
2
Aspergilloma, hydatid cyst, pneumonia, sarcoidosis
b
Colon cancer in 2, breast cancer in 1, osteosarcoma in 1, liver carcinoma in 1, soft tissue sarcoma in 1
c
Small cell carcinoma, indifferentiated sarcoma
Discussion In this study, we report our experience with robotic thoracic pulmonary resections performed for various etiologies and analyze and compare our results with the literature. Robotic thoracic surgery reports by several authors within the past decade are presented [3–13]. Our study differs from these previous reports in several ways: (1) First of all, our study differs from the previous reports on the primary pathologies and the type of the operations. Nearly all previous series are based on patients with primary lung cancer, and a great majority of the operations are lobectomies. Contrary to this, 21 % of our patients had benign diseases, and the rate of segmentectomy operations was high (45 %). (2) Second, we calculate that 14 cases are enough to complete the learning curve for robotic pulmonary resections, being lesser than proposals of previous studies [4, 9, 11, 13]. Since we performed nearly equal numbers of lobectomy and segmentectomy (53 vs 45 %), our proposal of 14 cases could be considered more descriptive as the learning curve of robotic pulmonary resections. (3) Our total operating time is the shortest reported so far. The possible explanations for this are given below. An improved comfort and quality of life of the patient are more frequently observed as the technology evolves. Minimally invasive surgical techniques are claimed to reduce pain, surgical inflammation, and postoperative dysfunction [18, 19]. Thoracoscopic surgery has gained wide adoption with these goals. Still, factors restrict a wider
123
adoption of VATS. These include the nonergonomy and difficulty in the learning curve, the limitation of two-dimensional visual information, and the restricted maneuverability of the instruments (mostly because of the rigid axes fixed to the thoracic wall by the entry trocar) [20]. The advanced engineering embodied in the da Vinci System makes it possible to overcome many of the disadvantages of VATS. Cadieres et al. [21] first described the robotic surgery in humans in 1997. Since then, robotic surgery in general has evolved and become widespread in various fields of surgery. In 2002, Melfi et al. [22] reported the robotic lung resections. Several important and large series followed the initial publication experience, mostly on robotic lobectomies for lung cancer [3–13]. These studies demonstrate that robotic thoracic surgery is feasible and safe with encouraging results (Table 4). Previous series reported a mortality rate between 0 and 3 %, a morbidity rate between 10.5 and 43 %, and a conversion rate between 0 and 15.7 %. Our results are within the abovedescribed reported range (2, 24, and 4 %, respectively). In terms of the length of stay, our results were similar to the ones reported before. In this study, we find that the risk of morbidity is significantly higher in elderly patients, as in open and VATS, and in patients with longer operating times. Advanced age is considered as a risk factor for lung resections [23]. Arrhythmia and prolonged air leaks are the most commonly reported complications after robotic pulmonary resections with a rate between 3 and 19 %, and 2.8 % and 13, respectively [3–13]. In our series, prolonged air leaks occured in 10 %, and arrhythmia developed in 9 % of the patients in parallel with the literature. Several series had an operative time in robotic pulmonary resection ranging between 132 and 226 min [3– 13]. This study had the shortest operative time reported so far (104 min, range 50–230 min). The reasons of such a short duration in our series can be explained in several ways. First, as we mentioned above, we started our thoracic robotic program after an established experience of VATS program. Second, in almost all previously reported large series about robotic pulmonary resections, lobectomy procedures consisted of the great majority of the resections. But in our series, the rate of lobectomy was nearly equal to that of segmentectomy (53 vs 45 %), and as we mentioned above, the mean console time of segmentectomy operations was shorter than that of lobectomy operations. Finally, a great majority of the patients reported in previous studies had primary lung cancers, and their operative times included mediastinal lymph node dissections. Twenty-one percent of the patients in our series had benign pulmonary pathologies; thus, they did not require mediastinal staging, so their operative times were insignificantly shorter.
Surg Endosc Table 4 Results of the recent reports in the field of robotic pulmonary resection Authors
Year
n
Age (years)
Anderson et al. [3]
2007
21
Gharagozloo et al. [4]
2009
100
Giulianotti et al. [5]
2010
38
Dylewski et al. [6]
2011
200
68 (20–92)
175 (82–370)
1.5
1.5
26
3 (1–44)
Fortes et al. [7]
2011
23
70 (51–86)
218 (74–323)
17
0
39
3 (1–13)
Cerfolio et al. [8]
2011
106
Veronesi et al. [9]
2011
91
Op. time (min)
C (%)
Mrt (%)
Mrb (%)
Stay
67 (36–86)
216 (60–384)
0
0
27
4 (2–10)
65 ± 8
216 ± 27
1
3
31
4 (3–42)
66 (16–78)
209 ± 66
15.7
2.6
10.5
10 (3–24)
66 (31–85)
132 ± 60
11.9
0
27
2 (1–7)
NP
226 (146–513)
11
0
27
5 (3–38)
Park et al. [10]
2012
325
66 (30–87)
206 (110–383)
8.3
0.3
25
5 (2–28)
Meyer et al. [11]
2012
185
65 ± 9
211 ± 60
1.6
1.6
16.8
4 (2–21)
Louie et al. [12]
2012
46
65
213
5.7
0
43
4 (2–21)
Lee et al. [13] This study
2014 2015
35 102
71 (52–84) 60 ± 13
161 (104–272) 104 (50–230)
2.9 4
0 2
11 24
3 (2–6) 5 (2–42)
Values are presented as mean ± standard deviation or median (range) N number, Op operation, C conversion, Mrt mortality, Mrb morbidity, NP nonprecise
The term ‘‘learning curve’’ has been frequently used in publications about surgical procedures as a reference to the process of gaining experience and improving skills in performing such procedures. The da Vinci System has a rapid learning curve [24]. Obtaining shorter operative time is regarded as a reference in several reports, mostly dealing with robotic lobectomies. Although operative times vary with the complexity of individual cases, it appears that the abrupt decrease in operative times corresponds to the learning curve [11]. Melfi et al. [25] suggested a minimum of 20 operations to acquire new skills for the whole surgical team. Gharagozloo et al. [4] and Veronesi et al. [9] separately observed that the first 20 patients had longer operative times. Based on operative times, Meyer et al. [11] proposed a learning curve of 18 ± 3 cases and Lee et al. [13] of 15–17 cases. Based on the operative times (docking, console, and total) in our current analyses, we calculate that 14 cases could be considered as completion of the learning curve for pulmonary resections. However, during the first 14 pulmonary resections, we also performed other thoracic surgical procedures such as thymectomies. Thus, 20 cases may be an appropriate learning period in robotic thoracic surgery. Both video- and robotic-assisted pulmonary resections offer minimally invasive approaches to the patients. Very recently, two studies have been published comparing VATS and robotic lobectomies [13], and VATS and robotic segmentectomies [14]. Both studies showed that both approaches were similar in terms of mortality, morbidity, and the length of stay. In addition, both studies demonstrated that the duration of both VATS lobectomies and segmentectomies was significantly shorter than that of robotic ones. Another common conclusion of both studies was that as the maturity in robotic surgery increased, the
duration of both operations reached similar levels. Meanwhile, the rate of conversion to thoracotomy was insignificantly lower in robotic approach [14]. As previously mentioned, the principal surgeon performing robotic resection in this study has had enough experience in VATS resections, and the fact that all the operations had been performed by the same surgeon eliminated any potential effect on surgeon variability on the results. Based on the personal knowledge and background in VATS, we can say that the beginning period of robotic-assisted thoracic surgery was more complicated, troublesome, and time-consuming. Thus, during this period, a VATS surgeon could easily regret the robotic surgery without understanding the real benefits of robotic-assisted thoracic surgery. Certainly, a minimally invasive chest surgeon will notice the precise dissection capabilities around the vessels, better visualization with stable camera platform, and dissection of lymph nodes around segmentary arteries and lobar bronchi. The second most important advantage, which has not been very often mentioned in the literature, is the capability of extra-routine dissection techniques. In VATS resections, a surgeon generally and traditionally dissects from anterior to posterior or from posterior to anterior. This may cause troublesome bleeding when one of the major structures has adhesions and difficulty in dissection. In robotic surgery, both arms can be used for dissection purposes. Thus, dexterity is a superiority when there is a difficulty in the dissection of major vessels, and the surgeon may use anterior, posterior, superior, or fissural approaches, whichever is suitable. Even, the surgeon can dissect in an extraordinary order. That superiority of the robotic surgery provides surgical technical comfort and may have the potential to decrease the conversion rate. We believe, as in VATS lobectomies, the
123
Surg Endosc
resection of the left upper lobe or superior/inferior bilobectomies of the right lung need extreme care and experience. We believe segmentectomies are better operations to start with. In particular, superior segmentectomies and common basal segmentectomies of the lower lobes are the best operations during completion of the learning curve. Robotic approaches for pulmonary resections are feasible and safe and have encouraging results even in the initial learning experience of the establishment of a robotic thoracic surgery program. This serves as a platform for further improvement in minimally invasive technologies. Disclosures No conflicts of interest or financial ties to disclose have ¨ zyurtkan, E. Kaba, K. Ayalp, been declared by A. Toker, MO. O ¨ . Demirhan, and E. Uyumaz. O
References 1. McKenna RJ Jr, Houck W, Fuller CB (2006) Video-assisted thoracic surgery lobectomy: experience with 1100 cases. Ann Thorac Surg 81:421–425 2. Ambrogi MC, Fanucchi O, Melfi F, Mussi A (2014) Robotic surgery for lung cancer. Korean J Thorac Cardiovasc Surg 47:201–210 3. Anderson CA, Hellan M, Falebella A, Lau CS, Grannis FW Jr, Kernstine KH (2007) Robotic-assisted lung resection for malignant disease. Innovations 2:254–258 4. Gharagozloo F, Margolis M, Tempesta B, Strother E, Najam F (2009) Robot-assisted lobectomy for early-stage lung cancer: report of 100 consecutive cases. Ann Thorac Surg 88:380–384 5. Giulianotti PC, Buchs NC, Caravaglios G, Bianco FM (2010) Robot-assisted lung resection: outcomes and technical details. Interact CardioVasc Thorac Surg 11:388–392 6. Dylewski MR, Ohaeto AC, Pereira JF (2011) Pulmonary resection using a total endoscopic robotic video-assisted approach. Semin Thorac Surg 23:36–42 7. Fortes DL, Tomaszek SC, Wigle DA (2011) Early experience with robotic-assisted lung resection. Innovations 6:237–242 8. Cerfolio RJ, Bryant AS, Skylizard L, Minnich DJ (2011) Initial consecutive experience of completely portal robotic pulmonary resection with 4 arms (2011). J Thorac Cardiovasc Surg 142:740–746 9. Veronesi G, Agoglia BG, Melfi F, Maisonneuve P, Bertoletti R, Bianchi PP, Rocco B, Borri A, Gasparri R, Spaggiari L (2001) Experience with robotic lobectomy for lung cancer. Innovations 6:355–360 10. Park BJ, Melfi F, Mussi A, Maisonneuve P, Spaggiari L, Da Silva RK, Veronesi G (2012) Robotic lobectomy for non-small cell lung cancer (NSCLC): long-term oncologic results. J Thorac Cardiovasc Surg 143:383–389
123
11. Meyer M, Gharagozloo F, Tempesta B, Margolis M, Strother E, Christenson D (2012) The learning curve of robotic lobectomy. Int J Med Robotics Comput Assist Surg 8:448–452 12. Louie BE, Farivar AS, Aye RW, Vallieres E (2012) Early experience with robotic lung resection results in similar operative outcomes and morbidity when compared with matched video-assisted thoracoscopic surgery cases. Ann Thorac Surg 93:1598–1605 13. Lee BE, Korst RJ, Kletsman E, Rutledge JR (2014) Transitioning from video-assisted thoracic surgical lobectomy to robotics for lung cancer: are there outcomes advantages? J Thorac Cardiovasc Surg 147:724–729 ¨ zkan B, Kaba E, Toker A (2014) Robotic 14. Demir A, Ayalp K, O and video-assisted thoracic surgery lung segmentectomy for malignant and benign lesions. Interact Cardiovasc Thorac Surg ivu 39. [Epub ahead of print] ¨ , Erus S 15. Toker A, Ayalp K, Uyumaz E, Kaba E, Demirhan O (2014) Robotic lung segmentectomy for malignant and benign lesions. J Thorac Dis 6(7):937–942 16. The Society of Thoracic Surgeons. Database. http://www.sts.org/ sections/stsnationaldatabase/publications/executive/artcile.html 17. Ma K, Chang D, He B, Gong M, Tian F, Hu X, Ji Z, Wang T (2008) Radical systematic mediastinal lymphadenectomy vs. mediastinal lymph node sampling in patients with clinical stage IA and pathological stage T1 non-small cell lung cancer. Cancer Res Clin Oncol 134:1289–1295 18. Ng CS, Wan S, Hui CW, Lee TW, Underwood MJ, Yim AP (2006) Video-assisted thoracic surgery for early stage lung cancer- can short term immunological advantages improve long-term survival? Ann Thorac Cardiovasc Surg 12:308–312 19. Nakata M, Saeki H, Yokoyama N, Kurita A, Takiyama W, Takashima S (2000) Pulmonary function after lobectomy: videoassisted thoracic surgery vs. thoracotomy. Ann Thorac Surg 70:938–941 20. Gharagozloo F, Margolis M, Tempesta B (2008) Robot-assisted thoracoscopic lobectomy for early-stage lung cancer. Ann Thorac Surg 85:1880–1886 21. Cadiere GB, Himpens J, Vertruyen M, Favretti F (1999) The World’s first obesity surgery performed by a surgeon at a distance. Obes Surg 9:206–209 22. Melfi FMA, Menconi GF, Mariani AM, Angeletti CA (2002) Early experience with robotic technology for thoracoscopic surgery. Eur J Cardiothorac Surg 21:864–868 23. Ferguson MK (2009) Pulmonary physiologic assessment of operative risk. In: Shields TW, LoCicero J, Reed CE, Feins RH (eds) General thoracic surgery. Lippincott Williams & Wilkins, Philadelphia, pp 325–338 24. Hernandez J, Bann S, Munz K, Moorthy K, Datta V, Martin S, Dosis A, Bello F, Darzi A, Rockall T (2004) Qualitative and quantitative analysis of the learning curve of a simulated task on the da Vinci system. Surg Endosc 18:372–378 25. Melfi FMA, Mussi A (2008) Robotically assisted lobectomy: learning curve and complications. Thorac Surg Clin 18:289–295
