Abstract
Background
Selection between open thyroidectomy (OT) and minimally invasive (endoscopic/robotic) thyroidectomy for patients with thyroid cancer has been a subject of considerable debate. Comprehensive analysis of the short-term outcomes of endoscopic thyroidectomy (ET), robotic-assisted thyroidectomy (RT) and open thyroidectomy (OT) for thyroid cancer using a large-scale dataset is important.
Methods
This cohort study evaluated the outcomes of patients receiving ET, RT or OT for thyroid cancer from January 1, 2003, to December 31, 2022. Propensity score matching (PSM) was performed among patients treated with ET, RT or OT to balance covariates distribution. This study involved single-institution patients (aged 18–70) who had undergone ET, RT or OT for thyroid cancer.
Results
The study included 11,066 thyroid cancer patients (OT group– mean (SD) age: 42.45 (10.84) years; ET group– mean (SD) age: 36.75 (9.32) years and RT group– mean (SD) age: 40.27 (10.42) years). After PSM for demographic and clinical characteristics, 908 matched pairs of patients (ET vs OT) and 1480 matched pairs (RT vs OT) were included for further analysis. Complication analysis revealed that RT was associated with a lower rate of transient hypoparathyroidism (339 (22.9%) vs 687 (46.4%); P < 0.001), a lower rate of permanent hypoparathyroidism (4 (0.3%) vs 16 (1.1%); P = 0.012) and a lower rate of transient recurrent laryngeal nerve injury (63 (4.3%) vs 89 (6.0%); P = 0.037).
Conclusion
This cohort study analyzed the short-term outcomes between ET, RT and OT in a large sample of patients with thyroid cancer over a period of two decades. PSM provided a comparable cohort, and the results suggested the advantage of RT, which reduced Clavien–Dindo grade Ⅰ complications in the surgical treatment of thyroid cancer.
Introduction
Minimally invasive thyroidectomy (MT) has been performed for more than 2 decades since the first endoscopic thyroidectomy (ET) in 1997 (1). Conventional open thyroidectomy (OT) leaves scarring caused by an incision in the neck, whereas MT, including ET and robotic-assisted thyroidectomy (RT), avoids a neck scar and achieves a cosmetic result (2, 3). Neck scars can negatively affect the quality of life of these patients, especially Asian patients, who are far more likely to experience complications of hypertrophic and keloid scarring than Caucasian counterparts (2, 4). Due to the narrow working space, performing MT involves limitations regarding manipulation of the mechanical maneuvers and restricted vision, and it thus has a longer learning curve than OT for proficiency and safety (5, 6). The introduction of surgical robots has overcome the drawbacks of ET and provided technical improvements, including magnified three-dimensional (3D) vision, tremor-filtering systems and additional degrees of freedom for fine dissection of nerves and vessels (7, 8). Despite these technical advantages of MT, outcomes are controversial. Several studies and meta-analyses have been performed to evaluate the safety and advantages of endoscopic or transoral endoscopic and robotic surgeries vs OT, and MT is favorably received for producing invisible and small scars through hidden incisions (9, 10, 11, 12, 13). However, these studies also included patients with multiple benign thyroid diseases, resulting in confounding factors.
Comparative studies focusing on endoscopic, robotic and open surgeries in large samples of thyroid cancer patients are limited. Therefore, to study the merits of MT in thyroid cancer, we analyzed the short-term surgical outcomes of ET, RT and OT for thyroid cancer.
Methods
Study groups
This was a retrospective cohort study of patients undergoing OT or MT (ET and RT) for thyroid cancer at the First Affiliated Hospital of Army Medical University, Southwest Hospital, from January 1, 2003, to December 31, 2022. The study was approved by the Ethics Committee of the Southwest Hospital, First Affiliated Hospital of Army Medical University.
All patients (aged 18–70 years) who had undergone OT, ET or RT for thyroid cancer were included. Patients with a history of thyroid surgery, second primary malignancies, distant metastasis, histologic undifferentiated carcinoma, serious cardiovascular, blood and immune system diseases or incomplete information were excluded. Patients undergoing endoscopic or robotic conversion to open surgery were also excluded.
Data collection
Patients voluntarily chose OT, ET or RT depending on tumor size and their need for aesthetics. Thyroid ultrasonography, neck computed tomography, fine-needle aspiration testing, thyroid function tests and electronic fibrolaryngoscopy were performed for preoperative evaluation of patients. The bilateral areola approach was used in RT or ET, and axillary non-inflatable was used in ET. All surgeries were performed by highly experienced surgeons. The recurrent laryngeal nerve (RLN) monitoring device was used for neural monitoring, and carbon nanoparticle injection and intraoperative frozen sections were used for the identification of parathyroid glands. Tumor staging was performed according to the American Joint Committee on Cancer/Union for International Cancer Control (AJCC-UICC) 8th edition for staging of differentiated thyroid cancer (14). Perioperative and postoperative outcomes were collected.
Outcome
Basic patient information (including sex, age, height, weight and previous medical and surgical history), preoperative examination results (thyroid ultrasound and neck CT), test results (thyroid hormone, antibodies against thyroid antigens, parathyroid hormone (PTH) and calcium), surgical procedure, operating time, intraoperative bleeding, major postoperative complications, postoperative pathological examination results (including pathological type, lymph node status and staging) and length of stay were collected from the medical records. Parathyroid autotransplantation was defined as selective autologous transplantation of parathyroid tissue that appeared suspicious of parathyroid survival or that was inadvertently removed with the thyroid gland (15). Finely minced parathyroid tissue was transplanted into the sternocleidomastoid or deltoid muscle. Hypoparathyroidism was defined as a PTH level below the laboratory standard (15 pg/mL) at 24 h postoperatively. Permanent hypoparathyroidism was defined as a PTH level below laboratory standards lasting longer than 12 months postoperatively. RLN injury was defined as postoperative impaired movement of one or both vocal cords confirmed by clinical symptoms and laryngoscopy. Permanent RLN injury was defined as vocal cord palsy that did not recover within 6 months after the surgery. Bleeding was defined as obvious postoperative bleeding that needed compression or reoperation to stop. Lymphatic leakage was defined as postoperative mass clear secretion or milky drainage, with a triglyceride concentration in the drainage fluid higher than 100 mg/dL. Infection was defined as postoperative localized swelling with evidence of bacterial spread that needed antibiotic treatment. Reoperation was defined as postoperative need for reoperation due to complications such as bleeding, lymphatic leakage or dyspnea. Postoperative complications were classified according to the Clavien–Dindo classification (16, 17).
Statistical analysis
Continuous data on demographic characteristics are provided as the mean (SD) and continuous data on outcomes as the median (interquartile range (IQR)). Continuous variables were analyzed using the independent paired t test (normal distribution) or the Mann–Whitney U test (abnormal distribution). Categorical data are reported as counts and percentages. Categorical data were analyzed using Pearson’s χ 2 test or Fisher’s exact test. In the groups showing significant differences of hypoparathyroidism after propensity score matching (PSM), univariate and multivariate logistic regression were performed to analyze the factors of hypoparathyroidism.
Due to the large amount of data, standardized mean difference (SMD) was used to assess covariate balance between the open and minimally invasive groups, whereby SMD < 0.1 usually indicates good balance. To achieve reliability and comparability between the open and minimally invasive (endoscopic/robotic) groups, PSM was used to balance the distribution of covariates; the propensity score was estimated using a 1:1 nearest neighbor matching algorithm. The type of surgery was the dependent variable, and age, sex, year of resection, body mass index (BMI), hypertension, diabetes, smoking, drinking, hyperthyroidism, Hashimoto’s thyroiditis, surgical extent, lateral lymph node dissection, unilateral/bilateral lymph node dissection, pathological stage, pathology type, tumor size and lymph node metastasis were covariates. Sensitivity analysis was performed using an optimal matching algorithm.
After PSM, the SMD between the open and minimally invasive groups was less than 0.1. In all statistical analyses, two-sided P < 0.05 indicated statistical significance. All statistical analyses were performed using R version 4.1.3 (RStudio; https://cran.r-project.org).
Results
Comparison of characteristics between OT and MT (ET/RT) groups before and after PSM
A total of 12,029 patients with thyroid cancer were identified. After the exclusion of 963 patients, 11,066 patients were included in the analysis. Of these, 8631 patients underwent open surgery, 943 patients underwent endoscopic surgery, and 1492 patients underwent robotic surgery (Fig. 1).
Flowchart of patient inclusion and exclusion.
Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0134
In unmatched cohorts (before PSM, Tables 1 and 2), patients undergoing ET or RT were more likely to be younger, especially in the ET group (mean (SD): 36.75 (9.32) vs 42.45 (10.84); P < 0.001) than patients undergoing OT. The same as age, other demographic and clinical characteristics of the study population between OT and ET or RT were largely unbalanced, with many potential confounding factors that could affect the reliability of results.
Demographic and clinical characteristics of the study population before and after PSM in OT and ET. Data are presented as mean ± SD or as n (%).
| Variable | Unmatched cohort (n = 9574) | 1:1 PSM (n = 1816) | ||||||
|---|---|---|---|---|---|---|---|---|
| OT (n = 8631) | ET (n = 943) | P value | SMD | OT (n = 908) | ET (n = 908) | P value | SMD | |
| Age, years | 42.45 ± 10.84 | 36.75 ± 9.32 | <0.001 | 0.564 | 37.28 ± 8.84 | 37.21 ± 9.17 | 0.872 | 0.008 |
| Sex | ||||||||
| Male | 2331 (27.0) | 153 (16.2) | <0.001 | 0.264 | 147 (16.2) | 153 (16.9) | 0.752 | 0.018 |
| Female | 6300 (73.0) | 790 (83.8) | 761 (83.8) | 755 (83.1) | ||||
| Year of resection | ||||||||
| 2003–2015 | 1205 (14.0) | 108 (11.5) | 0.038 | 0.075 | 109 (12.0) | 102 (11.2) | 0.660 | 0.024 |
| 2016–2022 | 7426 (86.0) | 835 (88.5) | 799 (88.0) | 806 (88.8) | ||||
| BMI | 23.70 ± 3.20 | 22.76 ± 2.99 | <0.001 | 0.302 | 22.74 ± 2.98 | 22.85 ± 2.98 | 0.398 | 0.040 |
| Hypertension | 774 (9.0) | 31 (3.3) | <0.001 | 0.239 | 28 (3.1) | 31 (3.4) | 0.791 | 0.019 |
| Diabetes | 221 (2.6) | 7 (0.7) | 0.001 | 0.143 | 8 (0.9) | 7 (0.8) | 1.000 | 0.012 |
| Smoking | 1102 (12.8) | 66 (7.0) | <0.001 | 0.194 | 66 (7.3) | 66 (7.3) | 1.000 | 0.012 |
| Drinking | 1477 (17.1) | 132 (14) | 0.017 | 0.086 | 131 (14.4) | 130 (14.3) | 1.000 | 0.003 |
| Hyperthyroidism | 202 (2.3) | 8 (0.8) | 0.004 | 0.119 | 5 (0.6) | 8 (0.9) | 0.578 | 0.039 |
| Hashimoto’s thyroiditis | 2202 (25.5) | 209 (22.2) | 0.027 | 0.079 | 204 (22.5) | 205 (22.6) | 1.000 | 0.003 |
| Extent of surgery | ||||||||
| Hemithyroidectomy | 1896 (22.0) | 597 (63.3) | <0.001 | 0.920 | 560 (61.7) | 562 (61.9) | 0.961 | 0.005 |
| Bilateral thyroidectomy | 6735 (78.0) | 346 (36.7) | 348 (38.3) | 346 (38.1) | ||||
| R1/R2 resection | 35 (0.4) | 1 (0.1) | 0.252 | 0.059 | 1 (0.1) | 1 (0.1) | 1.000 | <0.001 |
| Lateral LN dissection | 2018 (23.4) | 5 (0.5) | <0.001 | 0.753 | 6 (0.7) | 5 (0.6) | 1.000 | 0.014 |
| Unilateral central LN dissection | 4278 (49.6) | 891 (94.5) | <0.001 | 1.156 | 852 (93.8) | 856 (94.3) | 0.766 | 0.019 |
| Bilateral central LN dissection | 4353 (50.4) | 52 (5.5) | <0.001 | 1.156 | 56 (6.2) | 52 (5.7) | 0.766 | 0.019 |
| Pathological stage | ||||||||
| I–II | 8476 (98.2) | 942 (99.9) | <0.001 | 0.175 | 906 (99.8) | 907 (99.9) | 1.000 | 0.027 |
| III–IV | 155 (1.8) | 1 (0.1) | 2 (0.2) | 1 (0.1) | ||||
| Pathology type | ||||||||
| PTC | 8522 (98.7) | 941 (99.8) | 0.007 | 0.123 | 906 (99.8) | 907 (99.9) | 1.000 | 0.027 |
| Other | 109 (1.3) | 2 (0.2) | 2 (0.2) | 1 (0.1) | ||||
| Tumor size, cm | ||||||||
| ≤2 | 6671 (77.3) | 868 (92) | <0.001 | 0.418 | 2147 (89.9) | 2138 (89.5) | 0.703 | 0.012 |
| >2 | 1960 (22.7) | 75 (8.0) | 241 (10.1) | 250 (10.5) | ||||
| LN metastasis | ||||||||
| N1a | 3270 (37.9) | 341 (36.2) | 0.316 | 0.036 | 319 (35.1) | 332 (36.6) | 0.557 | 0.030 |
| N1b | 1629 (18.9) | 4 (0.4) | <0.001 | 0.658 | 4 (0.4) | 4 (0.4) | 1.000 | |
| Metastatic lymph nodes | 3.23 ± 5.27 | 0.97 ± 1.77 | <0.001 | 0.575 | 1 ± 1.95 | 0.99 ± 1.79 | 0.950 | 0.003 |
BMI, body mass index; R1, microscopic infiltration of the resection and dissection margins; R2, evidence of gross residual disease; LN, lymph node; PTC, papillary thyroid cancer; OT, open thyroidectomy; ET, endoscopic thyroidectomy; SMD, standardized mean difference; PSM, propensity score matching.
Demographic and clinical characteristics of the study population before and after PSM in OT and RT >data are presented as mean ± SD or as n (%).
| Variable | Unmatched cohort (n = 10,123) | 1:1 PSM (n = 1816) | ||||||
|---|---|---|---|---|---|---|---|---|
| OT (n = 8631) | RT (n = 1492) | P value | SMD | OT (n = 1480) | RT (n = 1480) | P value | SMD | |
| Age, years | 42.45 ± 10.84 | 40.27 ± 10.42 | <0.001 | 0.205 | 40.45 ± 10.12 | 40.41 ± 10.34 | 0.930 | 0.003 |
| Sex | ||||||||
| Male | 2331 (27.0) | 334 (22.4) | <0.001 | 0.107 | 336 (22.7) | 330 (22.3) | 0.826 | 0.010 |
| Female | 6300 (73.0) | 1158 (77.6) | 1903 (79.7) | 1903 (79.7) | ||||
| Year of resection | ||||||||
| 2003–2015 | 1205 (14.0) | 30 (2.0) | <0.001 | 0.452 | 32 (2.2) | 30 (2.0) | 0.898 | 0.009 |
| 2016–2022 | 7426 (86.0) | 1462 (98.0) | 1448 (97.8) | 1450 (98) | ||||
| BMI | 23.70 ± 3.20 | 23.35 ± 3.20 | <0.001 | 0.108 | 23.34 ± 3.12 | 23.38 ± 3.20 | 0.779 | 0.010 |
| Hypertension | 774 (9.0) | 88 (5.9) | <0.001 | 0.117 | 82 (5.5) | 88 (5.9) | 0.693 | 0.017 |
| Diabetes | 221 (2.6) | 36 (2.4) | 0.806 | 0.009 | 29 (2.0) | 36 (2.4) | 0.452 | 0.032 |
| Smoking | 1102 (12.8) | 151 (10.1) | 0.005 | 0.083 | 160 (10.8) | 150 (10.1) | 0.589 | 0.022 |
| Drinking | 1477 (17.1) | 168 (11.3) | <0.001 | 0.168 | 181 (12.2) | 168 (11.4) | 0.494 | 0.027 |
| Hyperthyroidism | 202 (2.3) | 37 (2.5) | 0.814 | 0.009 | 42 (2.8) | 37 (2.5) | 0.648 | 0.021 |
| Hashimoto’s thyroiditis | 2202 (25.5) | 382 (25.6) | 0.967 | 0.002 | 388 (26.2) | 380 (25.7) | 0.769 | 0.012 |
| Extent of surgery | ||||||||
| Hemithyroidectomy | 1896 (22.0) | 244 (16.4) | <0.001 | 0.143 | 263 (17.8) | 244 (16.5) | 0.380 | 0.034 |
| Bilateral thyroidectomy | 1248 (83.6) | 1217 (82.2) | 1236 (83.5) | |||||
| R1/R2 resection | 35 (0.4) | 4 (0.3) | 0.672 | 0.024 | 5 (0.3) | 4 (0.3) | 1.000 | 0.012 |
| Lateral LN dissection | 2018 (23.4) | 242 (16.2) | <0.001 | 0.180 | 215 (14.5) | 236 (15.9) | 0.306 | 0.039 |
| Unilateral central LN dissection | 4278 (49.6) | 751 (50.3) | 0.602 | 0.015 | 750 (50.7) | 745 (50.3) | 0.883 | 0.007 |
| Bilateral central LN dissection | 4353 (50.4) | 741 (49.7) | 0.602 | 0.015 | 730 (49.3) | 735 (49.7) | 0.883 | 0.007 |
| Pathological stage | ||||||||
| I–II | 8476 (98.2) | 1487 (99.7) | <0.001 | 0.143 | 1477 (99.8) | 1475 (99.7) | 0.723 | 0.026 |
| III–IV | 155 (1.8) | 5 (0.3) | 3 (0.2) | 5 (0.3) | ||||
| Pathology type | ||||||||
| PTC | 8522 (98.7) | 1484 (99.5) | 0.022 | 0.077 | 1473 (99.5) | 1472 (99.5) | 1.000 | 0.010 |
| Other | 109 (1.3) | 8 (0.5) | 7 (0.5) | 8 (0.5) | ||||
| Tumor size, cm | ||||||||
| ≤2 | 6671 (77.3) | 1317 (88.3) | <0.001 | 0.294 | 1300 (87.8) | 1305 (88.2) | 0.821 | 0.010 |
| >2 | 1960 (22.7) | 175 (11.7) | 180 (12.2) | 175 (11.8) | ||||
| LN metastasis | ||||||||
| N1a | 3270 (37.9) | 606 (40.6) | 0.048 | 0.056 | 594 (40.1) | 600 (40.5) | 0.851 | 0.008 |
| N1b | 1629 (18.9) | 148 (9.9) | <0.001 | 0.257 | 123 (8.3) | 148 (10.0) | 0.126 | 0.059 |
| Metastatic lymph nodes | 3.23 ± 5.27 | 2.03 ± 3.60 | <0.001 | 0.266 | 1.80 ± 3.04 | 2.04 ± 3.62 | 0.056 | 0.070 |
BMI, body mass index; R1, microscopic infiltration of the resection and dissection margins; R2, evidence of gross residual disease; LN, lymph node; PTC, papillary thyroid cancer; OT, open thyroidectomy; RT, robotic thyroidectomy; SMD, standardized mean difference; PSM, propensity score matching.
After matching by 1:1 PSM, 908 pairs of patients (908 OT patients and 908 ET patients) and 1480 pairs of patients (1480 OT patients and 1480 RT patients) were included for further analysis. Based on the nearest neighbor matching, the two groups were well balanced (SMD less than 0.1) across all baseline demographic and clinical characteristics (Tables 1 and 2; eFigs. 1, 2, 3, 4 in Supplement 1 (see section on Supplementary materials given at the end of the article)).
Comparison of surgical outcomes between OT and MT (ET/RT) groups before and after PSM
With optimal matching, the surgical outcomes of the OT and ET/RT groups were comparable. The outcomes of the OT and ET groups before and after PSM are shown in Table 3. The ET group had a longer operation time than the OT group (median (IQR), 140 (119–173) minutes vs 90 (72–115) minutes; P < 0.001) after PSM. The proportion of parathyroid autotransplantation was significantly lower in the ET group (295 (32.5%) vs 422 (46.5%); P <0.001). However, there was no significant difference in the occurrence of complications between the OT and ET groups.
Outcome variables of OT and ET before and after PSM. Data are presented as median (IQR) or as n (%).
| Variable | Before PSM | After PSM | ||||
|---|---|---|---|---|---|---|
| OT (n = 8631) | ET (n = 943) | P value | OT (n = 908) | ET (n = 908) | P value | |
| Operation time, min | 115 (85–159) | 140 (118–172) | <0.001 | 90 (72–115) | 140 (119–173) | <0.001 |
| Blood loss, mL | 50 (20–50) | 30 (20–50) | <0.001 | 30 (20–50) | 30 (20–50) | 0.565 |
| Postoperative stay, days | 4 (3–5) | 4 (3–5) | <0.001 | 4 (3–5) | 4 (3–5) | 0.007 |
| Parathyroid autotransplantation | 4970 (57.6) | 309 (32.8) | <0.001 | 422 (46.5) | 295 (32.5) | <0.001 |
| Bleeding | 30 (0.3) | 0 | 0.068 | 0 | 0 | >0.99 |
| Tracheotomy | 10 (0.1) | 0 | 0.131 | 0 | 0 | >0.99 |
| RLN injury | ||||||
| Transient | 533 (6.2) | 40 (4.2) | 0.021 | 43 (4.7) | 35 (3.9) | 0.418 |
| Permanent | 105 (1.2) | 5 (0.5) | 0.074 | 7 (0.8) | 4 (0.4) | 0.548 |
| Hypoparathyroidism | ||||||
| Transient | 3817 (44.2) | 181 (19.2) | <0.001 | 185 (20.1) | 177 (19.7) | 0.681 |
| Permanent | 97 (1.1) | 4 (0.4) | 0.043 | 3 (0.3) | 4 (0.4) | >0.99 |
| Infection | 44 (0.5) | 3 (0.3) | 0.622 | 4 (0.4) | 3 (0.3) | >0.99 |
| Lymphatic leakage | 257 (3.0) | 4 (0.4) | <0.001 | 4 (0.4) | 3 (0.3) | >0.99 |
| Reoperation | 41 (0.5) | 0 | 0.030 | 0 | 0 | >0.99 |
| Other complications | 28 (0.3) | 3 (0.3) | >0.99 | 3 (0.3) | 2 (0.2) | >0.99 |
RLN, recurrent laryngeal nerve; PSM, propensity score matching; OT, open thyroidectomy; ET, endoscopic thyroidectomy.
The outcomes of the OT and RT groups before and after PSM are shown in Table 4. The RT group had a longer operation time than the OT group (median (IQR): 144 (116–186) minutes vs 106 (83–140) minutes; P < 0.001). Intraoperative blood loss was lower in the RT group (median (IQR): 30 (20–50) mL vs 50 (20–50) mL; P < 0.001). The proportion of transient hypoparathyroidism was also lower in the RT group (339 (22.9%) vs 687 (46.4%); P < 0.001), similar to permanent hypoparathyroidism (8 (0.5%) vs 14 (0.9%); P = 0.012). Although the proportion of transient RLN injury was lower in the RT group (63 (4.3%) vs 89 (6.0%); P = 0.037), the proportion of permanent RLN injury was not significantly different between the two groups. The incidence of lymphatic leakage was lower in the RT group (18 (1.2%) vs 34 (2.3%); P = 0.036). Nevertheless, there was no significant difference in the incidence of reoperation, postoperative infection and other complications. The types of other complications are shown in eTable 1 in Supplement 1. The postoperative hospital stay was significantly longer in the RT group (median (IQR): 5 (4–6) days vs 4 (3–5) days; P < 0.001).
Outcome variables of OT and RT before and after PSM. Data are presented as median (IQR) or as n (%).
| Variable | Before PSM | After PSM | ||||
|---|---|---|---|---|---|---|
| OT (n = 8631) | RT (n = 1492) | P value | OT (n = 1480) | RT (n = 1480) | P value | |
| Operation time, min | 115 (85–159) | 144 (116–187) | <0.001 | 106 (83–140) | 144 (116–186) | <0.001 |
| Blood loss, mL | 50 (20–50) | 30 (20–50) | <0.001 | 50 (20–50) | 30 (20–50) | <0.001 |
| Postoperative stay, days | 4 (3–5) | 5 (4–6) | <0.001 | 4 (3–5) | 5 (4–6) | <0.001 |
| Parathyroid autotransplantation | 4970 (57.6) | 60 (4.0) | <0.001 | 869 (58.7) | 59 (4.0) | <0.001 |
| Bleeding | 30 (0.3) | 6 (0.2 | 0.567 | 5 (0.3) | 6 (0.4) | >0.99 |
| Tracheotomy | 10 (0.1) | 0 | 0.131 | 0 | 0 | >0.99 |
| RLN injury | ||||||
| Transient | 533 (6.2) | 68 (4.6) | 0.017 | 89 (6.0) | 63 (4.3) | 0.037 |
| Permanent | 105 (1.2) | 10 (0.7) | 0.088 | 14 (0.9) | 8 (0.5) | 0.285 |
| Hypoparathyroidism | ||||||
| Transient | 3817 (44.2) | 343 (30.0) | <0.001 | 687 (46.4) | 339 (22.9) | <0.001 |
| Permanent | 97 (1.1) | 4 (0.3) | 0.001 | 16 (1.1) | 4 (0.3) | 0.012 |
| Infection | 44 (0.5) | 11 (0.7) | 0.361 | 5 (0.3) | 11 (0.7) | 0.209 |
| Lymphatic leakage | 257 (3.0) | 18 (1.2) | <0.001 | 34 (2.3) | 18 (1.2) | 0.036 |
| Reoperation | 41 (0.5) | 3 (0.2) | 0.198 | 6 (0.4) | 3 (0.2) | 0.507 |
| Other complications | 28 (0.3) | 2 (0.1) | 0.303 | 2 (0.1) | 2 (0.1) | >0.99 |
RLN, recurrent laryngeal nerve; PSM, propensity score matching; OT, open thyroidectomy; RT, robotic thyroidectomy.
As there were significant differences in hypoparathyroidism between OT and RT groups after PSM, univariate and multivariate logistic regression were conducted to evaluate the factors of hypoparathyroidism (Table 5). The multivariate analysis showed that sex (OR = 1.72; 95% CI: 1.39–2.13; P < 0.001), Hashimoto’s thyroiditis (OR = 1.36; 95% CI: 1.13–1.64; P = 0.001), type of surgery (OR = 0.39; 95% CI: 0.32–0.48; P < 0.001), extent of surgery (OR = 10.42; 95% CI: 6.93–15.66; P < 0.001), bilateral lymph node dissection (OR = 1.42; 95% CI: 1.18–1.70; P < 0.001) and parathyroid autotransplantation (OR = 1.68; 95% CI: 1.36–2.07; P < 0.001) may be associated with transient hypoparathyroidism. In the analysis of permanent hypoparathyroidism, type of surgery (OR = 0.20; 95% CI: 0.06–0.66; P = 0.008) and bilateral lymph node dissection (OR = 3.15; 95% CI: 1.01–9.84; P = 0.048) showed the difference.
Univariate and multivariate analysis with hypoparathyroidism.
| Variable | Transient hypoparathyroidism | Permanent hypoparathyroidism | ||||||
|---|---|---|---|---|---|---|---|---|
| UV analysis, OR (95%CI) | P value | MV analysis, OR (95%CI) | P value | UV analysis, OR (95%CI) | P value | MV analysis, OR (95%CI) | P value | |
| Age | 1.00 (0.99–1.01) | 0.752 | 0.97 (0.92–1.01) | 0.129 | ||||
| Sex | 1.67 (1.38–2.03) | <0.001 | 1.72 (1.39–2.13) | <0.001 | 0.87 (0.32–2.40) | 0.788 | ||
| BMI | 0.96 (0.94–1.01) | 0.135 | 0.96 (0.83–1.11) | 0.580 | ||||
| Hyperthyroidism | 1.10 (0.69–1.74) | 0.70 | 1.93 (0.26–14.61) | 0.524 | ||||
| Hashimoto’s thyroiditis | 1.56 (1.32–1.85) | <0.001 | 1.36 (1.13–1.64) | 0.001 | 0.32 (0.07–1.36) | 0.122 | ||
| Type of surgery | 0.34 (0.29–0.40) | <0.001 | 0.39 (0.32–0.48) | <0.001 | 0.25 (0.08–0.74) | 0.013 | 0.20 (0.06–0.66) | 0.008 |
| Extent of surgery | 11.73 (7.95–17.32) | <0.001 | 10.42 (6.93–15.66) | <0.001 | 7 × 106 (0–Inf) NA | 0.984 | ||
| R1/R2 resection | 0.53 (0.11–2.58) | 0.434 | 8 × 10−6 (0–Inf) NA | 0.988 | ||||
| Lateral LN dissection | 1.71 (1.40–2.10) | <0.001 | 1.16 (0.90–1.48) | 0.245 | 0.98 (0.29–3.36) | 0.976 | ||
| Bilateral central LN dissection | 2.33 (1.99–2.72) | <0.001 | 1.42 (1.18–1.70) | <0.001 | 4.12 (1.37–12.34) | 0.012 | 3.15 (1.01–9.84) | 0.048 |
| Parathyroid autotransplantation | 2.90 (2.46–3.40) | <0.001 | 1.68 (1.36–2.07) | <0.001 | 1.46 (0.60–3.59) | 0.406 | ||
| Pathological stage | 1.59 (0.32–7.91) | 0.569 | 1 × 105 (0–Inf) NA | 0.989 | ||||
| Pathology type | 1.65 (0.60–4.57) | 0.332 | 3 × 10−6 (0–Inf) NA | 0.990 | ||||
| Tumor size | 1.21 (0.97–1.52) | 0.098 | 1.84 (0.61–5.55) | 0.276 | ||||
| LN metastasis | ||||||||
| N1a | 1.01 (0.87–1.18) | 0.867 | 1.21 (0.50–2.93) | 0.670 | ||||
| N1b | 1.67 (1.30–2.15) | <0.001 | 1.14 (0.82–1.60) | 0.424 | 1.10 (0.25–4.78) | 0.895 | ||
| Metastatic lymph nodes | 1.04 (1.02–1.07) | <0.001 | 1.02 (0.98–1.05) | 0.268 | 1.08 (0.99–1.18) | 0.060 | ||
BMI, body mass index; R1, microscopic infiltration of the resection and dissection margins; R2, evidence of gross residual disease; LN, lymph node; OR, odd ratio; CI, confidence interval; UV, univariate; MV, multivariate.
We further assessed surgical complications according to the Clavien–Dindo grading system (Table 6). The incidence of Clavien–Dindo grade I complications was significantly lower in the RT group than in the OT group (380 (25.7%) vs 728 (49.2%); P < 0.001) after PSM. The incidence of Clavien–Dindo grade I complications showed no difference between the ET and OT groups. Moreover, there was also no significant difference in Clavien–Dindo grade II–IV complications between the groups.
Complications of surgery stratified according to Clavien–Dindo classification after PSM. Data are presented as n (%).
| Variable | OT (n = 908) | ET (n = 908) | P value | OT (n = 1480) | RT (n = 1480) | P value |
|---|---|---|---|---|---|---|
| Clavien–Dindo grade a | ||||||
| I | 201 (22.1) | 187 (20.6) | 0.457 | 728 (49.2) | 380 (25.7) | <0.001 |
| II | 16 (1.8) | 11 (1.2) | 0.438 | 36 (2.4) | 24 (1.6) | 0.151 |
| III | ||||||
| IIIa | 0 | 1 (0.1) | >0.99 | 6 (0.4) | 3 (0.2) | 0.507 |
| IIIb | 0 | 0 | >0.99 | 0 | 0 | >0.99 |
| IV | ||||||
| IVa | 1 (0.1) | 0 | >0.99 | 1 (0.1) | 0 | >0.99 |
| IVb | 0 | 0 | >0.99 | 0 | 0 | > 0.99 |
| III–IV | 1 (0.1) | 1 (0.1) | >0.99 | 7 (0.5) | 3 (0.2) | 0.343 |
PSM, propensity score matching; OT, open thyroidectomy; MT, minimally invasive thyroidectomy.
Clavien–Dindo grades: I, any deviation from the normal postoperative course allowed therapeutic regimens are drugs as antipyretics, analgetics, electrolytes and physiotherapy; II, requiring pharmacological treatment with other drugs; IIIa, requiring surgical or endoscopic intervention not under general anesthesia; IIIb, requiring surgical or endoscopic intervention under general anesthesia; IV, life-threatening complication requiring intermediate care/intensive care unit management; IVa, single organ dysfunction; and IVb, multiorgan dysfunction (16).
Discussion
In this large cohort study, we compared open surgery with minimally invasive surgery for thyroid cancer at our center over the past two decades. This study demonstrated that the incidence of surgical complications from RT was lower than from OT for differentiated thyroid cancer after a 1:1 PSM analysis based on demographic and clinical characteristics. Prior to PSM, the demographic and clinical characteristics of the two groups were significantly different. The choice of surgical approach (OT, ET or RT) may be related to patient preference, pathology type and tumor size. After PSM, the baseline clinical characteristics and surgical type were essentially balanced, reducing the influence of confounding factors on surgical outcomes, which was an important basis for this study. The results showed that despite the longer operation time in the RT group than in the OT group, there were less intraoperative blood loss and a lower incidence of postoperative Clavien–Dindo grade I complications in the former. The operation time, of course, depends on the surgeon’s learning curve for endoscopic and robotic surgery, especially robotic surgery (18).
Although OT is the gold standard for thyroid cancer treatment, the development of ET and RT has occurred over the last two decades, altering the aesthetically and psychologically disruptive outcome of open surgery that produces neck scarring (7, 19, 20 and 21). Visible scars are known to negatively impact a patient’s quality of life, especially in certain Asian populations who are wary of neck scars (2, 22). Many studies have compared OT to ET or RT, but many of these studies have included surgery for benign thyroid tumors or had small sample sizes (10, 23, 24, 25 and 26). Whether ET, RT or OT is a better procedure for thyroidectomy in thyroid cancer patients remains unknown. In this study, we summarized a large sample size for thyroid cancer and found that endoscopic conversion to open surgery was 22 (20.4%) from 2003 to 2015, while 5 (0.7%) from 2016 to 2022, and robotic conversion to open surgery was 2 (6.7%) from 2003 to 2015, while 1 (0.1%) from 2016 to 2022, which also indicated that minimally invasive techniques were becoming more proficient after overcoming the learning curve.
Postoperative hypoparathyroidism is a common complication of thyroid surgery and remains a major challenge. Hypocalcemia due to hypoparathyroidism requires treatment with calcium and active vitamin D3 supplements, which affects patients' quality of life and increases their financial burden (15). In recent years, protection of parathyroid blood supply technology, supermeticulous capsular dissection, parathyroid autotransplantation technology and identification of parathyroid glands using near-infrared autofluorescence, carbon nanoparticles or indocyanine green fluorescent dye have been used to reduce the risk of postoperative hypoparathyroidism (27, 28, 29, 30). Regardless, hypoparathyroidism occurs in approximately 3–49% of patients after total thyroidectomy (31). Studies have shown that the incidence of postoperative hypoparathyroidism is lower in endoscopic or robotic thyroid surgery than in traditional open surgery (32, 33), but the amount of data is small. In this study, we performed analyses of parathyroid protection and found that parathyroid autotransplantation after ischemic or accidental removal of parathyroid glands was significantly lower in the ET group and the RT group than in the OT group. Moreover, the incidence of hypoparathyroidism, either transient or permanent, was significantly lower in the RT group than in the OT group. In multivariate analysis for transient hypoparathyroidism, lateral lymph node dissection, N1b and metastatic lymph nodes were non-significant factors. We think this may be due to confounding factors in the univariate analysis and multicollinearity among multiple independent variables. Moreover, the univariate and multivariate analyses showed that type of surgery and bilateral lymph node dissection were associated with transient and permanent hypoparathyroidism. The results suggested that robotic thyroid surgery provides significantly better intraoperative protection of the parathyroid glands than open surgery. This might be attributed to the high-resolution 3D magnifying camera and freedom of motion around various anatomical areas of the robotic-assisted system, which facilitates meticulous dissection of the thyroid true capsule and preservation of the parathyroid glands and their vascular components during surgical procedures. Although there was no significant difference in hypoparathyroidism after ET compared with OT, the parathyroid autotransplantation rate of the ET group was significantly lower than the OT group, indicating that endoscopic thyroid surgery also had a certain protection effect on the parathyroids. This might be attributed to the magnifying effect on endoscopic surgery.
RLN injury, which results in swallowing impairment or voice dysfunction, is one of the most common complications of thyroid surgery (34, 35 and 36). One study reported no significant difference in RLN injury between OT and transoral ET (23). Another study showed that the incidence of transient RLN palsy was lower in the conventional OT group than in the total ET group (32). In our large-sample study, we found that after PSM, the probability of temporary RLN injury in the RT group was slightly lower than that in the OT group. Considering the rarity of asymptomatic vocal cord palsy, as well as the high costs of frequent examinations to patients, postoperative laryngoscopy was performed just for patients with any postoperative dysphonia, respiratory symptoms, swallowing difficulties or abnormal signals during neuromonitoring of RLN, not for all thyroid surgery patients. This could also lead to potential bias in the analysis of the RLN injury. However, it cannot be denied that there are still margins for improvement in RLN protection with endoscopic or robotic thyroid surgery. In addition, intraoperative bleeding was less common in the RT group than in the OT group, and the probability of postoperative lymphatic leakage was lower in the RT group than in the OT group. This may be attributed to better hemostasis, improved identification and protection of lymphatic and blood vessels under the magnification and flexibility of the robotic system.
The main drawback associated with endoscopic and robotic thyroid surgery compared to open surgery is the high cost of expensive equipment, especially for robotic thyroid surgery (37, 38). The results from this large-sample study found that the minimally invasive approach was associated with a longer mean operating time than the open approach. The main reason is that it takes time to establish sufficient workspace for ET and RT; RT also requires time for robot docking.
The major limitation is that this study is a single-center retrospective cohort study, which could lead to selection bias and limit the generalizability of the results. Although a large amount of data and PSM analysis were performed to reduce selection bias between the groups, there may still be selection bias from unknown information without strict randomization. In addition, the study focused on short-term postoperative outcomes, and the long-term outcomes on recurrence and survival still need further evaluation through long-term follow-up.
Conclusion
In conclusion, the results of this cohort study suggest that RT for thyroid cancer is associated with lower Clavien–Dindo grade I complications than OT after balancing the distribution of covariates. However, there were no significant differences in Clavien–Dindo grades II–IV between the ET/RT and OT groups after PSM. MT, especially RT, may be a good option for appropriately selected patients to optimize postoperative quality of life.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/ETJ-24-0134.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work reported.
Funding
This work did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Patient consent
Written consent has been obtained from each patient after full explanation of the purpose and nature of all procedures used.
Author contribution statement
TL, YG and LC contributed to the study concept and design. TL, XW and YG contributed to the acquisition, analysis or interpretation of data. TL contributed to the drafting of the manuscript. LC contributed to the critical revision of the manuscript for important intellectual content. TL, YG and SL contributed to the statistical analysis. LC, XC, XY and JL contributed to the administrative, technical or material support.
Ethical approval
The study was approved by the Ethics Committee of the Southwest Hospital, First Affiliated Hospital of Army Medical University (number: (B) KY2023001). The procedures used in this study adhere to the tenets of the Declaration of Helsinki.
References
- 1↑
Lirici MM , Hüscher CSG , Chiodini S , et al. Endoscopic right thyroid lobectomy. Surg Endosc 1997 11 877. (https://doi.org/10.1007/s004649900476)
- 2↑
Arora A , Swords C , Garas G , et al. The perception of scar cosmesis following thyroid and parathyroid surgery: a prospective cohort study. Int J Surg 2016 25 38–43. (https://doi.org/10.1016/j.ijsu.2015.11.021)
- 3↑
Duek I , Duek OS & Fliss DM . Minimally invasive approaches for thyroid surgery-pitfalls and promises. Curr Oncol Rep 2020 22 77. (https://doi.org/10.1007/s11912-020-00939-2)
- 4↑
Arora A , Garas G , Sharma S , et al. Comparing transaxillary robotic thyroidectomy with conventional surgery in a UK population: a case control study. Int J Surg 2016 27 110–117. (https://doi.org/10.1016/j.ijsu.2016.01.071)
- 5↑
Kandil EH , Noureldine SI , Yao L , et al. Robotic transaxillary thyroidectomy: an examination of the first one hundred cases. J Am Coll Surg 2012 214 558–564. (https://doi.org/10.1016/j.jamcollsurg.2012.01.002)
- 6↑
Kim SK , Kang SY , Youn HJ , et al. Comparison of conventional thyroidectomy and endoscopic thyroidectomy via axillo-bilateral breast approach in papillary thyroid carcinoma patients. Surg Endosc 2016 30 3419–3425. (https://doi.org/10.1007/s00464-015-4624-9)
- 7↑
Rossi L , Materazzi G , Bakkar S , et al. Recent trends in surgical approach to thyroid cancer. Front Endocrinol 2021 12 699805. (https://doi.org/10.3389/fendo.2021.699805)
- 8↑
Lira RB & Kowalski LP . Robotic head and neck surgery: beyond TORS. Curr Oncol Rep 2020 22 88. (https://doi.org/10.1007/s11912-020-00950-7)
- 9↑
Liu H , Wang Y , Wu C , et al. Robotic surgery versus open surgery for thyroid neoplasms: a systematic review and meta-analysis. J Cancer Res Clin Oncol 2020 146 3297–3312. (https://doi.org/10.1007/s00432-020-03418-0)
- 10↑
Zhang Y , Dong Z , Li J , et al. Comparison of endoscopic and conventional open thyroidectomy for graves' disease: a meta-analysis. Int J Surg 2017 40 52–59. (https://doi.org/10.1016/j.ijsu.2017.02.054)
- 11↑
de Vries LH , Aykan D , Lodewijk L , et al. Outcomes of minimally invasive thyroid surgery - a systematic review and meta-analysis. Front Endocrinol 2021 12 719397. (https://doi.org/10.3389/fendo.2021.719397)
- 12↑
Alramadhan M , Choe JH , Lee JH , et al. Propensity score-matched analysis of the endoscopic bilateral axillo-breast approach (BABA) versus conventional open thyroidectomy in patients with benign or intermediate fine-needle aspiration cytology results, a retrospective study. Int J Surg 2017 48 9–15. (https://doi.org/10.1016/j.ijsu.2017.09.077)
- 13↑
Kang YJ , Stybayeva G & Hwang SH . Surgical completeness and safety of minimally invasive thyroidectomy in patients with thyroid cancer: a network meta-analysis. Surgery 2023 173 1381–1390. (https://doi.org/10.1016/j.surg.2023.02.021)
- 14↑
Wang TS & Sosa JA . Thyroid surgery for differentiated thyroid cancer–recent advances and future directions. Nat Rev Endocrinol 2018 14 670–683. (https://doi.org/10.1038/s41574-018-0080-7)
- 15↑
Rao SS , Rao H , Moinuddin Z , et al. Preservation of parathyroid glands during thyroid and neck surgery. Front Endocrinol 2023 14 1173950. (https://doi.org/10.3389/fendo.2023.1173950)
- 16↑
Dindo D , Demartines N & Clavien PA . Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004 240 205–213. (https://doi.org/10.1097/01.sla.0000133083.54934.ae)
- 17↑
Hsieh LB , Yen TWF , Dream S , et al. Perioperative management and outcomes of hyperthyroid patients unable to tolerate antithyroid drugs. World J Surg 2020 44 3770–3777. (https://doi.org/10.1007/s00268-020-05654-4)
- 18↑
Kassite I , Bejan-Angoulvant T , Lardy H , et al. A systematic review of the learning curve in robotic surgery: range and heterogeneity. Surg Endosc 2019 33 353–365. (https://doi.org/10.1007/s00464-018-6473-9)
- 19↑
Miccoli P , Berti P , Conte M , et al. Minimally invasive surgery for thyroid small nodules: preliminary report. J Endocrinol Invest 1999 22 849–851. (https://doi.org/10.1007/BF03343657)
- 20↑
Ohgami M , Ishii S , Arisawa Y , et al. Scarless endoscopic thyroidectomy: breast approach for better cosmesis. Surg Laparosc Endosc Percutan Tech 2000 10 1–4. (https://doi.org/10.1097/00129689-200002000-00001)
- 21↑
Lee J , Yun JH , Nam KH , et al. Perioperative clinical outcomes after robotic thyroidectomy for thyroid carcinoma: a multicenter study. Surg Endosc 2011 25 906–912. (https://doi.org/10.1007/s00464-010-1296-3)
- 22↑
Choi Y , Lee JH , Kim YH , et al. Impact of postthyroidectomy scar on the quality of life of thyroid cancer patients. Ann Dermatol 2014 26 693–699. (https://doi.org/10.5021/ad.2014.26.6.693)
- 23↑
Anuwong A , Ketwong K , Jitpratoom P , et al. Safety and outcomes of the transoral endoscopic thyroidectomy vestibular approach. JAMA Surg 2018 153 21–27. (https://doi.org/10.1001/jamasurg.2017.3366)
- 24↑
Dralle H , Machens A & Thanh PN . Minimally invasive compared with conventional thyroidectomy for nodular goitre. Best Pract Res Clin Endocrinol Metab 2014 28 589–599. (https://doi.org/10.1016/j.beem.2013.12.002)
- 25↑
Sun H , Wang X , Zheng G , et al. Comparison between transoral endoscopic thyroidectomy vestibular approach (toetva) and conventional open thyroidectomy for patients undergoing total thyroidectomy and central neck dissection: a propensity score-matching analysis. Front Oncol 2022 12 856021. (https://doi.org/10.3389/fonc.2022.856021)
- 26↑
Bo XW , Lu F , Yu SY , et al. Comparison of efficacy, safety, and patient satisfaction between thermal ablation, conventional/open thyroidectomy, and endoscopic thyroidectomy for symptomatic benign thyroid nodules. Int J Hyperthermia 2022 39 379–389. (https://doi.org/10.1080/02656736.2022.2040608)
- 27↑
Kong DD , Wang W & Wang MH . Superior parathyroid blood supply safety in thyroid cancer surgery: a randomized controlled trial. Int J Surg 2019 64 33–39. (https://doi.org/10.1016/j.ijsu.2019.02.019)
- 28↑
Wang B , Zhu CR , Liu H , et al. The accuracy of near infrared autofluorescence in identifying parathyroid gland during thyroid and parathyroid surgery: a meta-analysis. Front Endocrinol 2021 12 701253. (https://doi.org/10.3389/fendo.2021.701253)
- 29↑
Dip F , Alesina PF , Anuwong A , et al. Use of fluorescence imaging and indocyanine green during thyroid and parathyroid surgery: results of an intercontinental, multidisciplinary delphi survey. Surgery 2022 172 S6–S13. (https://doi.org/10.1016/j.surg.2022.09.004)
- 30↑
Qin X , Luo J , Ma J , et al. Prospective cohort study of parathyroid function and quality of life after total thyroidectomy for thyroid cancer: robotic surgery vs open surgery. Int J Surg 2023 109 3974–3982. (https://doi.org/10.1097/JS9.0000000000000725)
- 31↑
Privitera F , Centonze D , La Vignera S , et al. Risk factors for hypoparathyroidism after thyroid surgery: a single-center study. J Clin Med 2023 12 1956. (https://doi.org/10.3390/jcm12051956)
- 32↑
Jiang WJ , Yan PJ , Zhao CL , et al. Comparison of total endoscopic thyroidectomy with conventional open thyroidectomy for treatment of papillary thyroid cancer: a systematic review and meta-analysis. Surg Endosc 2020 34 1891–1903. (https://doi.org/10.1007/s00464-019-07283-y)
- 33↑
Lee IA , Kim K , Kim JK , et al. Comparison of surgical outcomes between robotic transaxillary and conventional open thyroidectomy in pediatric thyroid cancer. Cancers 2021 13 3293. (https://doi.org/10.3390/cancers13133293)
- 34↑
Liu N , Chen B , Li L , et al. Mechanisms of recurrent laryngeal nerve injury near the nerve entry point during thyroid surgery: a retrospective cohort study. Int J Surg 2020 83 125–130. (https://doi.org/10.1016/j.ijsu.2020.08.058)
- 35↑
Mahoney RC , Vossler JD , Murayama KM , et al. Predictors and consequences of recurrent laryngeal nerve injury during open thyroidectomy: an American college of surgeons national surgical quality improvement project database analysis. Am J Surg 2021 221 122–126. (https://doi.org/10.1016/j.amjsurg.2020.07.023)
- 36↑
Huang TY , Yu WHV , Chiang FY , et al. How the severity and mechanism of recurrent laryngeal nerve dysfunction during monitored thyroidectomy impact on postoperative voice. Cancers 2021 13 5379. (https://doi.org/10.3390/cancers13215379)
- 37↑
Cabot JC , Lee CR , Brunaud L , et al. Robotic and endoscopic transaxillary thyroidectomies may be cost prohibitive when compared to standard cervical thyroidectomy: a cost analysis. Surgery 2012 152 1016–1024. (https://doi.org/10.1016/j.surg.2012.08.029)
- 38↑
Choi YS , Shin WY & Yi JW . Single surgeon experience with 500 cases of the robotic bilateral axillary breast approach (BABA) for thyroid surgery using the da-vinci xi system. J Clin Med 2021 10 4048. (https://doi.org/10.3390/jcm10184048)
