Abstract
PURPOSE:
The aim of this study was to evaluate the efficacy, safety and costs of ultrasound guided percutaneous radiofrequency ablation (RFA) versus open thyroidectomy for treating low-risk papillary thyroid microcarcinoma (PTMC) by using propensity score matching (PSM).
PATIENTS AND METHODS:
157 patients who underwent RFA and 206 patients who underwent surgery for low-risk PTMC were included in the study. The patients were followed up at 1, 3, 6, 12 months after treatment, and every half year thereafter. A 1:1 PSM method was applied to balance the pretreatment data of the two groups. In the matched group (133 patients for each), the operative time, length of hospital stay, hospitalization expenses, cosmetic results, complications were assessed and compared between two groups.
RESULTS:
At last follow-up, 39 tumors (29.3%) in the RFA group completely disappeared. Between the well-matched groups, no local recurrence, lymph node metastasis or distant metastases were detected in either group during the follow-up period. After matching, the operation time and hospitalization time in RFA group were shorter than those in surgery group (both P < 0.05). The average hospitalization expense of the patients in RFA group was cheaper than that in surgery group (P < 0.05). Moreover, the cosmetic score was found to be higher in RFA group than that observed in surgery group (P < 0.05).
CONCLUSIONS:
RFA may be an effective and safe method for treating low-risk PTMC with a superior advantage of being low-cost and having a shorter operation time and hospital stay versus surgery.
Introduction
The incidence of papillary thyroid carcinoma is increasing due to great diagnostic performance of ultrasound [1], leading to detection of large reservoir of asymptomatic papillary thyroid microcarcinomas (PTMC) [2]. PTMC is defined as a tumor of which diameter is 1 cm or less in size. The indolent nat-ure of PTMC was proven with < 1%disease-specific mortality rate, 2%∼6%local-regional recurrence rates and 1%∼2%distant recurrence rates [3, 4]. Therefore, the optimal treatment for PTMC is still cont-roversial as the guidelines of 2015 American Thyroid Association recommended active follow-up rather than surgery unless there is the presence of evidence demonstrating clinically lymph node metastasis, local invasion or other high-risk factors. However, currently, there is no reliable method to distinguish relatively small number of PTMC patients inclining to develop clinically significant progression from low-risk ones [5]. More importantly, a “cancer” diagnosis might induce unavoidable patient anxiety during follow-up, especially when considering the fact that younger patients (< 40 years old) may have a relatively higher rate (8.9%) of clinical progression during active surveillance [6]. When it comes to open thyroidectomy, it carries certain potential complications including neck scars, hypothyroidism, hypoparathyroidism, and recurrent laryngeal nerve injury, which may severely affect the patient’s quality of life.
Therefore, thermal ablation with safe, effective and minimally invasive non-surgical treatment has been used to treat PTMC. Previous studies mainly focused on the role of radiofrequency ablation (RFA) procedure in treating benign thyroid nodules and locally recurrent papillary thyroid carcinoma after surgery [7, 8], all of which have shown absence of serious complications and desirable treatment efficacy, which showed significant reduction of tumor volume or complete tumor elimination. Some studies investigated the application and availability of RFA for primary low-risk PTMC with promising therapeutic outcomes [9]. Prior studies had also compared the outcomes between open thyroidectomy and ultrasound guided thermal ablation in patients with PTMC and concluded that these two approaches were comparable regarding patient outcome [10, 11]. However, with respect to potential confounders caused by retrospective nature, these studies were susceptible to selection bias. Thus, the aim of this study was to evaluate the efficacy, safety and costs of RFA versus open thyroidectomy for treating PTMC by using propensity score matching (PSM) analysis.
Materials and methods
Patients
This single-center retrospective study was approved by the ethics committee and institutional review board of the Affiliated Hospital of University (Approval number: J16013), and the written informed consent was waived. However, written informed consent for RFA and surgery were obtained from all patients before treatment.
The flowchart of patient enrollment is shown in Fig. 1. In our hospital, patient inclusion criteria for RFA were as follows: (a) nodule size <10 mm; (b) cytology-proven PTMC by fine needle aspiration (FNA) other than high-risk categories such as tall cell variant or sclerosis variant; (c) without extrathyroidal invasion; (d) no lymph node metastasis in imaging studies and (e) no history of neck irradiation. All patients were informed the advantages and disadvantages of RFA and thyroid surgery and were enrolled voluntarily for RFA or open surgery. The exclusion criteria concerning RFA procedure were defined as (a) pregnancy; (b) patients with severe heart, respiratory, liver diseases or renal failure; (c) patients with coagulation disorder or anticoagulant therapy. Between November 2017 and March 2020, 157 patients who underwent RFA and 206 patients who underwent surgery for low-risk PTMC were included in the study.
Flowchart of patient selection. PTMC, papillary thyroid microcarcinoma; US, ultrasound; RFA, radiofrequency ablation; LNM, lymph node metastasis; FNA, fine needle aspiration.
All eligible patients underwent cervical conventional ultrasound (US) and contrast-enhanced ultrasound (CEUS) examination 3 days before biopsy or treatment procedures. The US and CEUS examination were performed by one radiologist (YH) with 7 years of experience in thyroid US. A standard US device MyLab Twice (Esaote, Genoa, Italy) machine equipped with an L523 (4∼13 MHz) linear-array transducer was used for conventional US and CEUS. The size, location, echogenicity, shape, margin and calcification of thyroid nodules and cervical lymph node were evaluated. On conventional US, three diameters of the tumor were measured, and the tumor volume was calculated according to equation V = πabc/6 (a: longitudinal diameter, b: anteroposterior diameter, c: transverse diameter).
RFA procedures
A radiofrequency generator (VIVA; STARmed, Goyang, Korea) along with an 18-gauge monopolar internally cooled electrode (VIVA; STARmed) was used to generate RF energy. Continuous cooling was applied by distilled water circulation inside the antenna to keep the shaft from overheating. And the shaft antenna was designed to prevent tissue adhesion and to control easily.
The patient was placed in a supine position with the neck slightly extended. After the local anesthesia with 1%lidocaine administrated, fluid was injected into the area between the tumor and the trachea or adjacent significant tissues, to keep them far from the tumor for avoiding heat injury and then the electrode was positioned at the bottom of tumor under US guidance. An energy output of 20∼40 W was used. The ablation was not stopped until the vaporization covered the whole tumor. Immediately after the ablation, CEUS was performed to evaluate the ablation zone. In this study, all the lesions were less than 1 cm in diameter. If any residual tumor was suspected by CEUS, the RFA was repeated to achieve complete ablation. Compression of neck lasted for 30 min. Operation time and intraoperative complications were recorded. Each patient was observed for 1∼2 hours in the hospital after the ablation.
Surgical procedures
Open thyroidectomy was performed with the patient in the supine position and the neck hyperextended. A 5-6 cm transverse cervicotomy was performed, and the midline was explored. After the laryngeal nerve and parathyroids were localized, total thyroidectomy (TT) and thyroid lobectomy (TL) was performed according to the guidelines of the National Comprehensive Cancer Network [12]. The surgical time, intraoperative complications and the size of lesion were recorded.
Post-procedural observation and follow-up
Follow-up examinations were performed by means of neck ultrasonography at 1, 3, 6, 12 months after surgery, and intervals of half year thereafter, to evaluate tumor recurrence. Lymph node status was evaluated and the suspicious ones were taken for biopsy. Abnormal findings suggestive of cervical lymph node metastasis included a globular shape, loss of the normal echogenic hilum, presence of peripheral rather than hilar flow, heterogeneity with cystic components, and presence of microcalcifications [13]. The percentage reduction of the ablation area in volume was calculated as follows: Volume reduction ratio (VRR) = [(initial volume - final volume)×100]/initial volume (%). Ultrasound-guided fine needle aspiration biopsy (FNAB) of the ablated area was performed three months after RFA. On CEUS examination, the enhancement pattern of ablation zone was evaluated by adopting the criteria as follows: (a) complete ablation defined by whole ablation area showing non-enhancement; (b) recurrence or incomplete ablation was defined as enhancement in any portion of ablation zone. Complications during follow-up were assessed using the reporting criteria of image-guided tumor ablation [14]. Cosmetic result was evaluated by a numeric scale, ranging from 0 to 10 [15]. All patients were asked to grade the cosmetic appearance of their wound 2 months after ablation or surgery.
Additionally, the clinical outcomes including operative time, hospital stay, intraoperative (major vessel laceration, major nerve injury, and penetration into adjacent vital structures such as trachea or esophagus), postoperative complications (pain, hemorrhage, hematoma, seroma, chylous leakage, and neurologic complications) and costs were evaluated.
Statistical analysis
A 1:1 PSM analysis was performed to balance the pretreatment data of the two groups. Propensity scores for all the patients were estimated by multiple logistic-regression models using the following baseline characteristics as covariates: sex, age, nodule echogenicity, microcalcification, shape, margin, initial nodule diameter, initial nodule volume, TT3 level, TT4 level, follow-up period. Before and after the 1:1 matching,the statistical analysis was performed by using the following tests: t test for age, the Mann-Whitney U test for tumor volume and largest diameter, operative time, hospitalization stay, hospitalization expenses, thyroid related hormones and cosmetic result (both for numeric scale and verbal response scale), and the χ2 test or Fisher exact test for gender, US characteristics of the lesion (location, echogenicity, calcification, homogeneity, margin, shape) and complications. The software used was SPSS (SPSS Inc, Chicago, ¢ó). A P-value less than 0.05 was considered statistically significant.
Results
Baseline characteristics before and after matching
Baseline clinical and demographic parameters are shown in Table 1. Before matching, due to the inevitable selection bias, compared with patients in the RFA group, those in the surgery group had a higher prevalence of microcalcification, elder age, larger tumor diameter and tumor volume. After performing PSM for the entire population, a total of 133 matched patient pairs were enrolled. Baseline characteristics were not statistically significant between two groups P > 0.05). Demographic characteristics of the study participants after matching are also summarized in Table 1.
Demographic and US characteristics of patients before and after propensity score matching
Demographic and US characteristics of patients before and after propensity score matching
RFA: radiofrequency ablation; TT3: Total Triiodothyronine; TT4: Total Thyroxine. *Continuous variables are expressed as the mean±standard deviation. †Variables are expressed as n, data in parentheses are percentages (%).
The mean RFA power was 29.06±3.85 W (range, 20∼40 W). The changes in trend of the mean vol-umes of thyroid tumor are displayed in Table 2. The size of the tumor before RFA and the CEUS fea-tures before ablation and after ablation are illustrated in Fig. 2. After three months follow-up with RFA, the size of ablated lesion had decreased, and FNA was performed. The cytological results sho-wed that no residual tumor cells were found in the ablation zone. At last follow-up, 39 tumors (29.3%) completely disappeared, including 10 lesion disappearing in 3 months, 18 in 6 months, 9 in 12 months, 2 in 18 months. No local recurrence or distant metastases were detected in either group during the follow-up period. Before matching, there was one patient with LNM in the surgery group. But for all patients in well-matched groups, no local recurrence, LNM or distant metastases were detected during the follow-up period (Table 3). Regarding complications, no dysphagia, permanent hoarseness, hematoma, seroma, chylous leakage, and neurologic complications were observed. In the RFA group, one patient had transient hypoparathyroidism and 2 patients experienced transient voice change. In contrast, in the surgery group, one patient had transient hypoparathyroidism and 4 patients experienced transient voice change. All these minor complications disappeared spontaneously within 2 weeks. There was no significant difference of complication rate between two groups (P > 0.05) (Table 3).
Changes in volume and volume reduction ratio (VRR) of the thyroid lesion before and 1, 3, 6, 12, 18, and 30 months after RFA
Changes in volume and volume reduction ratio (VRR) of the thyroid lesion before and 1, 3, 6, 12, 18, and 30 months after RFA
RFA: radiofrequency ablation; NA: not applicable. *P lt; 0.05 compared with last follow-up time.
A 49-year-old male with papillary thyroid microcarcinoma (PTMC) in the right lobe of the thyroid gland. (A) Before the radiofrequency ablation, US scan showed a markedly hypoechoic nodule with taller-than-wide shape. (B) The nodule was revealed as hypo-enhancement on contrast-enhanced ultrasound (CEUS) examination (red arrow). (C) Electrode (red arrow) and vaporization (white arrows) were observed during radiofrequency ablation. (D) A well-defined ablation zone was observed on B-mode one month after radiofrequency ablation. (E) Ablation zone was presented with non-enhancement on CEUS one month after ablation (red arrows). (F) Six months after ablation, ablation zone was presented with a scar-like area (red arrow).
Complications, recurrence, lymph node metastasis, distant metastasis condition before and after propensity score matching analysis
RFA: radiofrequency ablation; LNM: lymph node metastasis; NA: not applicable.
The operative time was 26.7 minutes for RFA group and 65.9 minutes for open thyroidectomy group after matching (P < 0.05). The hospital stay in RFA group was statistically shorter than that in surgery group (3.8 days vs. 6.8 days, ∼P < 0.05). The hospitalization expenses in group RFA was cheaper than that in surgery group (RMB 17086.3254 vs. RMB 26012.29; ∼P < 0.05) (Table 3). The cosmetic result in RFA group was significantly higher than that in surgery group (P < 0.05) (Table 4).
Comparison of the operative time, hospitalization stay, hospitalization expenses and cosmetic results between RFA group and surgery group before and after propensity score matching analysis
Comparison of the operative time, hospitalization stay, hospitalization expenses and cosmetic results between RFA group and surgery group before and after propensity score matching analysis
RFA: radiofrequency ablation. *Continuous variables are expressed as the mean±standard deviation.
The increasing detection of low-risk PTMC constantly poses a challenge in clinical field regarding how to appropriately handle patient’s psychological burden and cosmetic concern. For high-risk surgical individuals or those refusing surgery, previous literature concluded that ultrasound-guided thermal ablation might be a possible resolution. For example, laser ablation was applied for single low-risk PTMC with the results showing that 10 ablation zones had totally disappeared, and 20 ablation zones remained as scar-like lesions during the last follow-up [16]. Another thermal ablation technique, RFA, has also been introduced as a promising tool for clinical management of PTMCs per Korean guidelines [17]. Several studies suggested that RFA could be a cost-effective treatment for thyroid nodules and replacement for surgery with fewer complications [7]. In the study of Zhang et al. [18], low-risk PTMCs were treated by RFA with significant reduction of volume ratio and completely diminished carcinomas in a follow-up of post-RFA 6 and 12 months, respectively. This result is consistent with our present study, in which 39 tumors (29.3%) completely disappeared and other nodules showed no microvascular signal on CEUS.
However, prior studies selected paired patients in both groups only based on chronological order or a retrospective consecutive enrollment of all patients who met criteria [19], although they pointed out that no significant baseline features were observed between RFA and surgery group. Therefore, in the present paper, a specific method, PSM analysis, was applied. The results showed that before matching, the patients in the surgery group were much more elder with larger nodules than those in the RFA group. After matching, the confounding baseline bias were eliminated by the logic regression model, which highlighted the importance of PSM in reducing confounding bias. As for safety concern, no neck haematoma, skin burns, dysphonia, local infection or injury to the vital organs of neck was observed. At the final follow-up, no patients had cervical lymph nodes or distant metastasis. It should be noted that the RFA procedure in our study centered on low-risk PTMCs rather than ones with suspicious extrathyroidal extension. The inclusion criteria adopted by this paper accords with the American Thyroid Association guidelines [5] and the study of Oda et al. [20] due to our consideration that extrathyroidal extension of PTMC should be selected as immediate surgery candidates.
There were earlier studies of thermal ablation versus surgery for benign thyroid nodules, recurrent thyroid cancers and PTMCs [7, 8], which revealed the main advantages of thermal ablation as being less invasive, more cost-effective than surgery. In RFA Group, there were only 2 cases with voice change, which disappeared spontaneously within 2 weeks. A previous study had compared the outcomes between open thyroidectomy and ultrasound guided RFA in patients with PTMC [10]. However, it was a retrospective non-randomized study. Furthermore, this study demonstrated the operative time and the hospitalization time of RFA were shorter than those of surgery. The total hospitalization expense of each patient in RFA group was cheaper than that in surgery group, indicating high cost-effective performance for PTMC ablation.
There were some limitations in the current study. First, the study was performed in a single center with a relatively shorter follow-up time. Therefore, a multi-center randomized clinical trial with more eligible patients is needed. In addition, only imaging approaches were utilized to identify suspicious lymph node or distant metastasis during follow-up, which may cause false-negative results. Several studies have reported that the rate of central compartment lymph nodes micrometastasis in patients with PTMC ranged from 16.9∼42.4%, and none of the presurgical features analyzed, including BRAF mutation, was able to predict LNM presence [21–24].
In conclusion, RFA may be an effective and safe method for treating low-risk papillary thyroid microcarcinoma with low-cost and a shorter operation time and hospitalization time. A multicenter randomized study is needed to evaluate its advantages and potential setbacks.
Abbreviations
RFA: radiofrequency ablation; PTMC: papillary thyroid microcarcinoma; US: ultrasound; TT: total thyroidectomy; TL: thyroid lobectomy; FT3: free triiodothyronine; FT4: free tetraiodothyronine; TSH: thyroid stimulating hormone; RFA: radiofrequency ablation; FNA: fine needle aspiration; CEUS: contrast-enhanced ultrasound; VRR: volume reduction ratio; VAS: visual analog scale; FNAB: fine needle aspiration biopsy.