Introduction

Thyroid nodules are common. Although most are benign and will remain relatively unchanged over time, some do cause symptoms necessitating surgical resection [1,2,3]. However, surgery is not without risks and requires a general anaesthesia and hospitalisation. As a result, there has been an increasing interest in developing less invasive, non-surgical techniques in treating benign thyroid nodules [4, 5]. For predominantly solid or wholly solid nodules, several image-guided thermal ablation techniques like laser ablation (LA), microwave and radiofrequency ablation (RFA) have been shown to be effective [4, 5].

High-intensity focused ultrasound (HIFU) is one of the newer ablation techniques that has been shown to be effective in not only causing significant nodule shrinkage but also alleviating obstructive symptoms shortly after single treatment [6,7,8,9]. Interestingly, apart its thermal effect, less intense HIFU energy is capable of causing mechanical or non-thermal ablation that could lead to various important pathological, immunological and therapeutic consequences both locally and systemically [10, 11].

However, despite being a promising form of ablation, the medium to long-term efficacy following a single HIFU treatment remains unclear. Although studies have shown that nodules generally continue to shrink while obstructive symptoms continue to improve over time [8, 9, 12,13,14], it is uncertain whether these results can continue beyond 12 months. Cases of nodule regrowth and recurrence have been reported following other forms of ablation technique when treated nodules were followed beyond 12 months [12,13,14,15]. Also, to ensure treatment success in the long-term, it would be important to identify patient or disease factors that may affect long-term treatment efficacy. With these issues, therefore, the present study aimed to evaluate the amount of physical shrinkage and symptom improvement beyond 12 months and also to identify what factors might influence treatment success at 2 years from single HIFU treatment.

Methods

This retrospective analysis was approved by the local institutional review board (UW 17-234). All relevant clinical and treatment data were recorded prospectively after obtaining informed consent. Consecutive patients who underwent HIFU ablation for a symptomatic, solid or predominantly solid (<30% cystic areas) benign thyroid nodule from 2015 to 2016 were analysed. At our institution, only patients who were indicated but unwilling to undergo thyroidectomy were considered for ablation. Details on the eligibility for ablation were previously described [9]. In brief, the nodule had to be proven benign on fine-needle aspiration cytology (Bethesda category II) and to have a low or very low suspicion sonographic pattern together with its centre located within the treatable depth of 7–30 mm from the skin surface. Also, the swelling (which could either be a solitary nodule or a dominant nodule in a multinodular gland) had to be causing obstructive symptoms and the longest diameter of the nodule had to be ≥20 mm but ≤60 mm on ultrasonography (US). For the present study, any patients with incomplete or less than 24 months of follow-up or had received two or more treatments to the same nodule within 24 months were excluded.

At each visit (baseline and 3, 6, 12, 18 and 24 months), the three orthogonal dimensions of the index nodule were measured on US by an independent experienced sonographer. Nodule dimensions were measured using the LOGIQ e (GE Healthcare, Chicago, IL, USA) scanner equipped with a 10- to 14-MHz linear matrix transducer. Three orthogonal diameters of the index nodule (its longest diameter and two other perpendicular diameters) were recorded. In general, the longest diameter was the cranio-caudal dimension (length) of the nodule, while the other two perpendicular diameters were the medio-lateral (width) and antero-posterior (depth) dimensions of the nodule. All measurements were to the nearest 0.1 mm. To estimate nodule volume, we used the formula: volume (mL) = [width (in cm) × depth (in cm) × length (in cm)] × (π/6) where π was taken as 3.1416. The volume reduction ratio (VRR) (%) was calculated based on the formula: [(baseline volume – volume at visit) / (baseline volume)] × 100. Since our intra-observer coefficient of variation for sonographic volume was around 4.5% [16], only a change in nodule volume >4.5% between time-points was taken as an actual change. Treatment success was defined as VRR >50%. Nodule regrowth was defined as a >20% increase in nodule volume relative to the 12-month volume given that the 12-month volume is normally the lowest measured volume on US in the first 24 months.

Clinical assessments

In addition to US assessments, the thyroid swelling was clinically graded by the WHO goitre grade system [17] before treatment. Also, at each clinic visit (baseline, 6-, 12- and 24-month) patients were asked to rate their obstructive symptoms on a visual analogue scale (VAS) (0–10) (0 = no obstructive symptoms; 10 = most significant obstructive symptoms).

Application of HIFU ablation

All treatments were carried out in a similar matter by one person (B.H.L.) using the same commercially available US-guided HIFU device (see later). All patients were instructed to be fasted overnight and to admit to the hospital in the early morning where baseline blood tests including serum thyroid function tests [free T4 and thyroid stimulating hormone (TSH) levels], thyroglobulin (Tg) and anti-thyroid autoantibodies were taken.

At treatment, all patients were placed in a supine position with neck slightly extended and received a bolus of intravenous Diazepam (Actavis, Dublin, Ireland) (10–15 mg) and Pethidine (Martindale Pharma, Wooburn Green, Buckinghamshire, UK) (50–100 mg). Patients’ heart rate, blood pressure, respiration rate and peripheral oxygenation were monitored throughout the procedure. Patients were asked to show a hand sign without moving the neck if the pain became too severe during treatment. In that situation, either the energy setting was lowered or more medications were administrated.

The US-guided HIFU device comprised an energy generator, a treatment head, a skin-cooling device and a touch-screen interface for planning. The treatment head incorporated an image transducer (7.5 MHz, 128 elements, linear array) and HIFU transducer (3 MHz, single element, 60 mm in diameter) (Fig. 1). The treatment head was placed on the neck of the patient on the side of the index nodule and was finely adjusted until the nodule was within the treatable depth of 7-30mm from skin surface. Once marked on the treatment screen, the device computer (Beamotion version no. TUS 3.2.2; Theraclion, Malakoff, France) automatically divided the nodule into multiple ablation voxels. Each voxel measured approximately 7.3 mm in thickness and 5 mm in width and received a continuous 8-s pulse of HIFU energy followed by 20–30 s of cooling time before the beam was moved to the adjacent voxel (Fig. 2). To ensure safety, nearby structures like the carotid artery, trachea and skin were marked out on the treatment screen before the start of treatment by the operator. To avoid inadvertent heat injury to important surrounding structures, the device automatically selected the safety margins for the skin, the trachea and recurrent laryngeal nerve and from the ipsilateral carotid artery. A laser-based movement detector enabled immediate power interruption when the patient moved or swallowed during ablation. To avoid skin burn, the skin was cooled by a balloon (filled with 10oC liquids) at the tip of the treatment head. Both the total amount of energy delivered to the nodule (in kilojoules) and the “on-beam” (sonification) time taken (in minutes) were automatically recorded by the device’s computer. The “on-beam” treatment time was the duration between the first to the last pulse (in minutes). Oral diet was resumed immediately afterwards and patients were discharged home 2–3 h after treatment. Afterwards, a transcutaneous laryngeal US was done to assess the mobility of both vocal cords [18]. Vocal cord palsy (VCP) was defined as having an impaired or absent movement in one of the vocal cords corresponding to the ablated side. Any other possible complications including skin burn or Horner’s syndrome were recorded.

Fig. 1
figure 1

A picture of the ultrasound-guided high intensity focused ultrasound (HIFU) device placed at the cranial end of the treatment bed

Full size image
Fig. 2
figure 2

A picture captured immediately after a treatment pulse with the formation of microbubbles (white arrows) at the focus of the beam

Full size image

Statistical analysis

Continuous variables were expressed as mean ± SD and/or median and interquartile range (IQR) when appropriate. For comparison between groups, either t-test or the Mann-Whitney U test was used. Chi-squared tests were used to compare categorical variables. For correlation between continuous variables, the Pearson correlation test was performed. Both the univariate and multivariate analyses were done by logistic regression analysis. For multivariate analysis, only parameters considered to be associated with ablation success were entered together using the step-down procedure. Obstructive symptom was rated on a VAS of 0-10 at baseline and at 6, 12 and 24 months. All statistical analyses were performed using SPSS version 18.0 (SPSS, Chicago, IL, USA). All significance tests were two-tailed and those with a p value less than 0.05 were considered statistically significant.

Results

A total of 136 patients underwent HIFU ablation of a symptomatic, benign thyroid nodule. Of these, 4 (2.9%) patients were lost to follow-up within the first 24 months, while 24 (17.6%) patients either received two or more treatments in the same session or two or more treatments within 24 months. After excluding these 28 patients, 108 (79.4%) patients were eligible for analysis. This cohort included outcomes of 22 patients previously published [9]. At the time of analysis, the mean (± SD) follow-up was 28.7 ± 4.0 months. Baseline characteristics and treatment parameters are shown in Table 1. In terms of complications, three (2.8%) patients developed unilateral VCP afterwards but they all recovered (i.e. regained normal movement) within the first 3 months. One other patient (0.9%) suffered Horner’s syndrome on the side of the treatment. Her mild ptosis improved gradually over a period of 6 months.

Table 1 Baseline characteristics (n = 108)
Full size table

Extent of nodule shrinkage (VRR) and rate of treatment success in the first 24 months

Following a single HIFU ablation, the overall nodule shrinkage rate or VRR at 3, 6, 12, 18 and 24 months were 51.32 ± 20.71%, 62.99 ± 22.05%, 68.66 ± 18.48%, 69.76 ± 17.88% and 70.41 ± 17.39%, respectively (Fig. 3). The rate of treatment success at 3, 6, 12, 18 and 24 months was 42/108 (41.7%), 66/108 (61.1%), 73/108 (67.6%), 75/108 (69.4%) and 76/108 (70.4%), respectively. In the first 12 months, all of the nodules became progressively smaller in volume and the 12-month nodule volume was the lowest recorded volume relative to baseline. However, interestingly, despite the overall volume reduction over the 24 months, 63 (58.3%) of nodules had a further volume reduction (i.e. >4.5%) from 12 to 24 months. In this subgroup, the mean volume reduction (%) was 9.9 ± 9.5%. There were 23 (21.3%) nodules which remained static in volume from 12 to 24 months and there were 22 (20.4%) nodules which had a volume increase (i.e. >4.5%) during this period. In this latter subgroup, the mean volume increase (%) was 13.62 ± 9.60%. In terms of absolute volume increase, the mean was 0.45 ± 0.67 mL and the largest absolute increase was 3.10 mL (a nodule which measured 18.01 mL at 12 months became 21.11 mL at 24 months or a 17.2% increase in volume). Nevertheless, none of these treated nodules fulfilled the criteria for nodule regrowth (i.e. >20% from lowest recorded volume) in the first 24 months of follow-up.

Fig. 3
figure 3

Change in mean (± SD) volume reduction ratio (VRR) before and after single HIFU treatment

Full size image

Change in symptom score (by VAS) in the first 24 months

At baseline, the mean (± SD) symptom score by VAS was 4.12 ± 1.22 (median = 4.0; IQR = 2.0). Following treatment, the mean symptom scores at 6, 12 and 24 months were lowered relative to the baseline [2.57 ± 1.35 (median = 2.0; IQR = 1.0), 1.56 ± 1.05 (median = 2.0; IQR = 1.0), 1.27 ± 1.04 (median = 1.0; IQR = 2.0), respectively] (Fig. 4). At 6 months, 85 (78.7%) patients had a VAS score lower than that at baseline. Among those who did not have a lower VAS score, 20 (18.5%) patients had the same VAS score as that of baseline, while only 3 (2.8%) patients reported a VAS score higher than that of baseline. At 24 months, 103 (95.4%) patients had a lower VAS score than that of baseline and only 5 (4.6%) patients had the same VAS score as the baseline.

Fig. 4
figure 4

Change in mean (± SD) obstructive symptom score by visual analogue scale before and after single HIFU treatment

Full size image

Association between VRR and symptom score (by VAS) over time

At 6 months, patients reporting an improved symptom score (i.e. reporting a lower 6-month VAS score than baseline) (n = 85) had a significantly greater VRR than those who reported a similar (n = 20) or a higher VAS (n = 3) than baseline (67.45 ± 17.87% vs 51.71 ± 25.71%, p < 0.001 and 67.45 ± 17.87% vs 11.90 ± 17.13%, p = 0.001, respectively). Similarly, at 24 months, patients reporting an improved symptom score (i.e. a lower VAS than that of baseline) (n = 103) had a significantly greater VRR than those with the same VAS as baseline (n = 5) (71.50 ± 16.85% vs 48.07 ± 13.99%, p = 0.005). There was a significant correlation between VRR at 24 months and the absolute difference in VAS score at 24 months from baseline (r = -0.820, p = 0.040).

Factors determining treatment efficacy at 24 months

Table 2 compares the baseline characteristics and treatment parameters between those who had VRR >50% (i.e. treatment success) at 24 months (group I) and those who had VRR ≤50% at 24 months (group II). Age, sex ratio, body mass index, thyroid function, side and location of index nodule, as well as number of other nodules within the thyroid gland, were not significantly different between the two groups. However, group II had a significantly higher proportion of WHO grade 3 nodules than group I (46.2% vs 23.2%, p = 0.039). Also, in concordance with this, the nodule width, length and depth were all significantly larger in group II than group I (3.01 cm vs 2.55 cm, p = 0.037; 4.25 cm vs 3.29 cm, p = 0.001; 2.46 cm vs 2.00 cm, p = 0.003, respectively). As a result, the nodule volume at baseline in group II was almost twice of that in group I (20.34 mL vs 10.80 mL, p = 0.003) and also, the “on-beam” treatment time was significantly longer in group II than group I (55.20 min vs 38.51 min, p = 0.038). Despite having similar baseline symptom score (by VAS), the scores at 12 and 24 months were significantly lower in group I (p = 0.003 and p = 0.005, respectively). Relative to baseline, VAS scores at 6, 12 and 24 months were significantly less in group I (p < 0.001), while in group II only VAS scores at 12 and 24 months were significantly less (p < 0.001) (Table 3).

Table 2 A comparison of baseline characteristics, treatment parameters and volume reduction radio between those who had VRR >50% (i.e. treatment success) at 24-month (group I) and those who did not at 24-month (group II)
Full size table
Table 3 Logistic regression analysis of factors predictive of treatment success at 24 months
Full size table

In the univariate analysis by logistic regression, only larger pre-ablation nodule (OR = 0.945, 95% CI = 0.911–0.980, p = 0.003) and longer total “on-beam” time (OR = 0.972, 95% CI = 0.946–0.998, p = 0.035) were significant factors for treatment success at 24 months. In the multivariate analysis, none of the factors turned out to be significant after the step-down procedure (Table 4).

Table 4 Multivariate analysis with step-down procedure of factors predictive of treatment success at 24 months
Full size table

Discussion

To our knowledge, this was the first study to evaluate the clinical efficacy of HIFU ablation beyond 12 months. In terms of major findings, first, our data showed that there was a small but significant overall reduction in nodule volume (i.e. physical shrinkage) from 12 to 24 months (from 68.66 ± 18.48% to 70.41 ± 17.39% p < 0.001) and the rate of treatment success increased incrementally from 67.6 to 70.4% over this period. These findings were consistent with the experience from other forms of ablation technique [12,13,14,15]. One previous study reported that with a single application of RFA treatment, there was still an incremental reduction in nodule volume from 78.6% at 12 months to 79.4% at 24 months relative to baseline volume [14]. Another study reported a similar amount of reduction from 59% at 12 months to 60% at 24 months with a single LA application [14]. Regarding why the actual shrinkage rate varied between studies, one explanation might be due to the differences in case selection, operator experience and delivered energy [19].

However, it is interesting to note that when the volume of each nodule was looked at individually, only less than two-thirds (or 58.3%) of nodules had an actual volume reduction from 12 to 24 months, with another one-fifth (or 21.3%) remaining static in volume (± 4.5%). Therefore, the actual amount of shrinkage was relatively small when compared to that in the first 12 months. Also, worth noting was the fact that another one-fifth (or 20.4%) of nodules actually had a volume increase of >4.5%, with the largest absolute volume increase of 3.10 mL or 17.2% from 12 to 24 months. Therefore, even though none of the 108 nodules observed fulfilled the criteria for nodule regrowth of 20% relative to volume at 12 months, our data implied that perhaps a longer and closer follow-up is indicated and might prove to be worthwhile. This is why some authors have advocated the use of multiple treatments to optimise the long-term treatment efficacy and success [9, 20].

In addition to physical nodule shrinkage, our data showed that over the 24-month period, the severity of obstructive symptom became less or milder as reported by patients. This was despite the fact that the physical nodule volume appeared to plateau after 12 months. There was an overall decrease in mean symptom score (by VAS) from 12 to 24 months [from 1.56 ± 1.05 (median = 2.0; IQR = 1.0) to 1.27 ± 1.04 (median = 1.0; IQR = 2.0)]. In terms of association with nodule shrinkage, our data found that those with an improved symptom score had significantly greater nodule shrinkage (i.e. VRR) than those with similar symptom score at 6 and 24 months (p < 0.001 and p = 0.005, respectively). Also, there was a significant correlation between those with larger VRR and greater improvement in symptom score relative to baseline (p = 0.040). These findings had previously been reported [8,9,10,11,12].

In terms of factors determining treatment success at 24 months, the initial or pre-ablation nodule volume was a significant factor in the univariate analysis (OR = 0.945, 95% CI = 0.911–0.980, p = 0.003). Relative to those who did not achieve treatment success (or VRR >50%) at 24 months, those who did achieve treatment success had significantly smaller-volume nodules (10.80 ± 9.97 mL vs 20.34 ± 16.73 mL, p = 0.003). This finding was consistent with the experience of others and ours [21]. We believe this might be due to the fact that a more complete ablation could be achieved with a single ablation in smaller nodules, while perhaps more than a single ablation might be required for larger nodules. As a result, our group have begun to treat larger nodules (volume >15-20 mL) with two sequential treatments in the same session using a similar technique as the one for multinodular goitre [22]. Although ablation time was also found to be significant in the univariate analysis, this was simply because larger nodules needed longer ablation time than smaller nodules. As a result, it did not turn out to be an independent factor for treatment success.

Despite these findings, we would like to acknowledge several shortcomings. First, our study was only a moderately sized study and so some of the findings were prone to type II errors. Second, despite being the first study to evaluate longer-term efficacy after single HIFU treatment, an even longer period of follow-up would be necessary for the proper evaluation of events like nodule growth and recurrence.

Conclusions

The majority of patients had a further nodule volume reduction and symptom improvement beyond 12 months following a single HIFU ablation, but since about one-fifth of them had a notable nodule volume increase beyond 12 months, a longer period of follow-up would seem to be necessary. Pre-ablation nodule volume was a significant factor for treatment success at 24 months.