Therapeutic dose and long-term efficacy of high-intensity focused ultrasound ablation for different types of uterine fibroids based on signal intensity on T2-weighted MR images

Abstract Objective To investigate the therapeutic dose and long-term efficacy of high-intensity focused ultrasound (HIFU) ablation for different types of uterine fibroids based on signal intensity on T2-weighted MR images (T2WI). Materials and methods Four hundred and one patients with a solitary uterine fibroid treated with HIFU were classified into four groups consisting of extremely hypointense, hypointense, isointense and hyperintense fibroids. Each group was further classified into two subtypes: homogeneous and heterogeneous, based on signal homogeneity of fibroids. The therapeutic dose and long-term follow-up results were compared. Results There were significant differences in treatment time, sonication time, treatment intensity, total treatment dosage, treatment efficiency, energy-efficiency factor (EEF) and non-perfused volume (NPV) ratio among the four groups (p<.05). The average NPV ratio achieved in patients with extremely hypointense, hypointense, isointense and hyperintense fibroids was 75.2 ± 14.6%, 71.1 ± 15.6%, 68.2 ± 17.3% and 67.8 ± 16.6%, respectively; the re-intervention rates at 36 months after HIFU were 8.4%, 10.3%, 12.5% and 6.1%, respectively. Sonication time, treatment intensity and total energy for heterogeneous fibroids were greater than that for homogeneous fibroids in patients with extremely hypointense fibroids (p<.05). The treatment time for heterogeneous fibroids was significantly longer than that for homogeneous fibroids in patients with isointense fibroids (p<.05). Multivariate ordered logistic regression analysis showed that the ablation volume of fibroids and treatment time were related to NPV ratio (p<.05). Conclusion Every group of patients obtained satisfactory long-term results. Hyperintense fibroids are difficult to treat by HIFU. Heterogeneous fibroids are more difficult to treat with HIFU than homogeneity fibroids.


Introduction
Uterine fibroids are the most common benign tumors in reproductive age women.The prevalence of uterine fibroids varies across races and age groups and ranges from 5.4% to 77% [1][2][3][4].Symptoms caused by uterine fibroids are related to the number, size and location of the fibroids, including heavy or prolonged menstrual flow, pelvic pain, urinary frequency or urgency, and constipation.Uterine fibroids also may affect fertility.Therefore, patients with symptomatic uterine fibroids require therapeutic intervention.
Conventional treatments for symptomatic uterine fibroids include medication, myomectomy and hysterectomy.Recently, uterine artery embolization (UAE) has become a routine treatment for uterine fibroids in developed countries.
Medication can be used to effectively control fibroid-related symptoms, but symptoms can easily recur after medication withdrawal.In addition, serious side effects of some medications have limited their role in the management of uterine fibroids.Myomectomy is a standard treatment for patients with uterine fibroids who wish to retain their uterus, but the cumulative recurrence rates at 12 and 24 months after myomectomy were high [5].Hysterectomy is a definitive treatment for uterine fibroids, but this operation is not suitable for patients who wish to remain fertile.UAE is less invasive than surgery and can be used to effectively control the symptoms caused by uterine fibroids, but its adverse effects on ovarian function, have limited the clinical application of this technique.
As a non-invasive treatment, high-intensity focused ultrasound (HIFU) ablation has been widely used in the management of uterine fibroids.Many studies have shown the safety and efficacy of HIFU treatment for uterine fibroids [6,7].Earlier studies have demonstrated that the histological variants of uterine fibroids are presented as different features on magnetic resonance imaging (MRI).Recently, uterine fibroids were classified as hypointense, isointense and hyperintense fibroids by comparing the signal intensity of uterine fibroids with skeletal muscle and uterine smooth muscle on T2-weighted MR images (T2WI).Several studies have also shown that the signal intensity of uterine fibroids on T2WI is correlated with therapeutic dose and treatment efficacy of HIFU.Hypointense fibroids are easier to treat with HIFU than isointense and hyperintense fibroids.Hyperintense fibroids are difficult to treat with this non-invasive technique [8,9].Therefore, delivering adequate acoustic energy on the basis of signal intensity of fibroids on T2WI is the key to ensure treatment efficacy and to reduce recurrence rate.
Zhao et al. further classified hyperintense fibroids as homogeneous and heterogeneous hyperintense fibroids, and found that there was a significant difference in treatment results between the two subtypes [10].In clinical practice, we also found that both homogeneous and heterogeneous fibroids exist in hypointense and isointense fibroids, but no comparative study was performed on HIFU dosage and treatment outcomes.In addition, a study from Funaki et al. concluded that hyperintense fibroids should be exempted from the application of magnetic resonance-guided focused ultrasound surgery (MRgFUS) [11].In addition, it has been shown that classifying fibroids as hyperintense solely through comparing T2WI signal intensity to uterine smooth muscle is inadequate.Therefore, we proposed to classify fibroids as four types on the basis of signal intensity of T2WI as follows: extremely hypointense (fibroids with signal intensity equal to or lower than skeletal muscle), hypointense (fibroids with signal intensity higher than that of skeletal muscle but lower than that of myometrium), isointense (fibroids with signal intensity equal to that of myometrium) and hyperintense (fibroids with signal intensity higher than that of myometrium).Then, each type was further divided into homogeneous and heterogeneous fibroid groups to investigate the differences in therapeutic dose and long-term efficacy of HIFU treatment.

Materials and methods
The protocol for this retrospective study was approved by the ethics committee at our institute (CQHF-2021-009, date of approval: 1 August 2021) and the requirement for informed consent was waived.

Patients
Patients with a solitary uterine fibroid who were treated with HIFU in Chongqing Haifu Hospital from October 2015 to October 2019 were included in this study.
Inclusion criteria: (1) patients with a solitary uterine fibroid whose diagnosis was confirmed by ultrasound and MRI; (2) patients who completed HIFU treatment and underwent preand post-MRI in our hospital.
Exclusion criteria: (1) patients with MRI contraindications or did not complete MRI examination; (2) patients with a solitary uterine fibroid and adenomyosis.
A total of 426 patients with a solitary uterine fibroid were treated with HIFU at our institution, 25 patients were excluded from this study and 186 patients were lost during the follow-up period (Figure 1).

MRI examination
MRI examinations were performed using a 1.5 T MRI system (Shanghai United Imaging Medical Technology Co., Ltd., Shanghai, China).Patients underwent MRI scans before and one day after HIFU treatment, using a standardized protocol.T2-weighted imaging and T1-weighted imaging before and after administration of gadolinium were obtained in three planes.Typical parameters used for T2WI were as follows: TR 4692 ms, TE 75 ms, layer thickness 5 mm and layer spacing 1 mm.Typical parameters used for T1WI were as follows: TR 196 ms, TE 8.18 ms, layer thickness 6 mm and layer spacing 1.2 mm.Typical parameters used for enhanced T1WI were as follows: TR 4 ms, TE 2 ms, layer thickness 5 mm and layer spacing 1.2 mm.

Pre-HIFU MRI evaluation
MR images from every patient were evaluated by three experienced radiologists independently.The following data included position of the uterus, type of uterine fibroids, location of uterine fibroids (anterior wall of the uterus, posterior wall of the uterus, lateral wall of the uterus, fundus of the uterus), size of uterine fibroids, thickness of the abdominal wall, thickness of the fat in abdominal wall, distance from the anterior surface of the fibroid to the skin, and distance from posterior side of the fibroid to the sacrococcyx were recorded.The fibroids were classified as four types on the basis of signal intensity of T2WI.Each of the four types of uterine fibroids was further classified into homogeneous and heterogeneous subtypes according to the signal intensity on T2WI as follows: (1) homogeneous: uniform signal intensity within the fibroid was observed, no area of linear high-signal intensity or heterogeneous signal intensity area was less than 25% of the fibroids; (2) heterogeneous: obvious area of linear high-signal intensity or heterogeneous signal intensity was more than 25% of the lesion (Figure 2).

High-intensity focused ultrasound ablation
The device used for HIFU treatment was a Focused Ultrasound Tumor Therapeutic System (model JC200 or JC, Chongqing Haifu Medical Technology Co., Ltd., Chongqing, China).In this study, therapeutic ultrasound beams were generated by a transducer with a frequency of 0.8-1.0MHz, a focal length of 15 cm and a diameter of 20 cm.A Mylab 70 ultrasound imaging device (Esaote, Genova, Italy) was used to provide real-time imaging to localize the fibroid and monitor the treatment.
Every patient underwent specific bowel preparation and skin preparation before HIFU treatment.Bowel preparation included ingesting semi-liquid food or liquid food two days prior, and fasting for 12 h before HIFU.Skin preparation included shaving the hair between the lower edge of belly button and the upper edge of the pubic symphysis, then degreasing and degassing the skin with degassed water.Before HIFU treatment, a urinary catheter was inserted into the bladder, and the bladder volume was adjusted by infusing saline during the treatment to obtain a safe acoustic pathway.HIFU treatment was performed under conscious sedation.Each patient was positioned prone on the HIFU table, with the anterior abdominal wall in contact with degassed water.A degassed water balloon was placed between the abdominal wall and the transducer to compress and push the bowel away from the acoustic pathway.The sagittal ultrasound scanning mode was chosen for both pretreatment planning and sonication.The locations of the fibroids and surrounding tissues were identified on ultrasound imaging and the targeted fibroid was divided into a number of sections by using real-time ultrasound, with a distance of 5 mm between the sections.Point scan was used, and power was set between 300 and 400 W. The distance from the focal point to the endometrium was at least 1.5 cm, and the distance from the focal point to the subserosal surface of the uterus was 1 cm.During the HIFU procedure, therapeutic power was adjusted based on patient feedback and changes in grayscale on ultrasonographic imaging.The treatment was terminated when the fibroid showed increased grayscale change or there was an absence of blood supply as evaluated by contrast-enhanced ultrasound immediately after HIFU ablation.The patients' vital signs such as heart rate, blood pressure, respiration and oxygen saturation were monitored.

Post-HIFU MRI evaluation
Post-HIFU contrast enhanced MR images were used to measure the non-perfused volume (NPV) of fibroids.NPV indicates the volume of coagulative necrosis.The volume of fibroids and NPV were obtained using the software program, which was programed by the engineers from Chongqing Haifu Medical Technology Co., Ltd.(Chongqing, China), to contour the fibroids and the non-perfused region in every slice of contrast enhanced MR images, then calculated using the same program.The NPV ratio ¼ NPV/fibroid volume Â 100%.

Follow-up
In accordance with the follow-up protocol of our hospital, all patients were asked to follow up every 3 months after HIFU treatment to evaluate for symptom improvement for 3 years.The patients were requested to undergo ultrasound every 3 months after HIFU for follow-up imaging evaluation.During the follow-up period, if any patient had performed other treatment (hysterectomy, myomectomy, UAE, medication, etc.) for any reason, this was defined as a reintervention.

Statistical analysis
SPSS 26.0 (SPSS Inc., Chicago, IL) statistical software was used for data analysis.The normally distributed data were reported as a mean ± standard deviation (x±s), and the skewed distribution data were reported as a median with an interquartile range.For comparison of the variables between the two groups, independent samples t-test was used; for comparison among multiple groups, one-way ANOVA was used for normal distribution with equal variance; Kruskal-Wallis H test was used for non-normal distribution or uneven variance.The LSD t-test and Mann-Whitney U-test were used for two-way comparisons.The distribution between groups was compared by v 2 test or Fisher's exact probability method for count data.Multivariate ordered logistic regression analysis was used to figure out the main factors that affected the NPV ratio.p<.05 was considered as statistically significant difference.

Patients and lesions
Among the 401 patients, 100 patients had extremely hypointense fibroids, 133 patients had hypointense fibroids, 104 patients had isointense fibroids and 64 patients had hyperintense fibroids.In patients with extremely hypointense fibroids, 66 patients were heterogeneous and 34 patients were homogeneous.In patients with hypointense fibroids, 89 were heterogeneous and 44 were homogeneous.In patients with isointense fibroids, 83 were heterogeneous and 21 were homogeneous.In patients with hyperintense fibroids, 56 were heterogeneous and eight were homogeneous.
As shown in Table 1, the median fibroid volume in the group of patients with extremely hypointense fibroids was 113.1 (interquartile range: 59.4-199.8)cm 3 , 84.0 (interquartile range: 47.6-169.9)cm 3 in the group of patients with hypointense fibroids, 161.6 (interquartile range: 80.7-261.5)cm 3 in the group of patients with isointense fibroids, and 148.8 (interquartile range: 82.3-298.2) cm 3 in the group of patients with hyperintense fibroids.The fibroid volume in the group with hypointense fibroids was significantly lower than that of patients with isointense and hyperintense fibroids (p<.001).The median abdominal wall thickness was 22.6 (interquartile range: 18.6-26.4)mm in patients with extremely hypointense fibroids, 21.7 (interquartile range: 17.3-26.2) mm in patients with hypointense fibroids, 19.7 (interquartile range: 15.1-24.3)mm in patients with isointense fibroids, and 18.8 (interquartile range: 15.6-28.8)mm in patients with hyperintense fibroids.The abdominal wall thickness in the group with extremely hypointense fibroids was significantly thicker than that of patients with isointense fibroids (p¼.015).The median distances from the posterior surface of the fibroid to the sacrococcyx were 19.7 (interquartile range: 9.7-35.4)mm in patients with extremely hypointense fibroids, 18.8 (interquartile range: 9.8-37.0)mm in patients with hypointense fibroids, 14.2 (interquartile range: 8.1-24.1)mm in patients with isointense fibroids, and 15.2 (interquartile range: 9.0-27.2) mm in patients with hyperintense fibroids.The distance from the posterior surface of the fibroid to the sacrococcyx in the isointense group was significantly shorter than that of patients in the hypointense group (p¼.023).The median distance from the anterior surface of the fibroid to the skin was 40.2 (interquartile range: 29.2-57.2) mm in the group of patients with extremely hypointense fibroids, 43.4 (interquartile range: 29.1-58.7)mm in patients with hypointense fibroids, 37.0 (interquartile range: 26.4-53.6)mm in patients with isointense fibroids, and 35.5 (interquartile range: 22.4-46.6)mm in patients with hyperintense fibroids.The distance from the anterior surface of fibroid to the skin in the group of patients with hypointense fibroids was significantly higher than that of patients with hyperintense fibroids (p¼.018).Subserosal fibroids were more common in the groups with extremely hypointense and hypointense fibroids than the groups with isointense and hyperintense fibroids, while the proportion of intramural fibroids was greater in the groups with isointense and hyperintense fibroids than that in the groups with extremely hypointense and hypointense fibroids (p<.001).No other significant differences in baseline characteristics were observed.

Comparison of baseline characteristics between patients with homogeneous and heterogeneous fibroids
The comparative results in patient baseline characteristics are shown in Table 2.In patients with extremely hypointense fibroids, the median fibroid volume in patients with fibroids with heterogeneous signal intensity was significantly higher than that of patients with homogeneous signal intensity on T2WI (p<.05).No other significant differences were observed between the two subgroups in terms of the distance from the posterior surface of the fibroid to the sacrococcyx, the distance from the anterior side of the fibroid to the abdominal wall, and the types of fibroids.
In patients with hypointense fibroids, the median fibroid volume in patients with fibroids with heterogeneous signal intensity was significantly higher than that of patients with homogeneous signal intensity on T2WI (p<.05).The distance from the anterior surface of the fibroid to the abdominal wall was significantly shorter for patients with fibroids with heterogeneous signal intensity than for patients with homogeneous signal intensity on T2WI (p<.05).No other significant differences in patient baseline characteristics between the two subgroups were observed.
In patients with isointense fibroids, the median fibroid volume in patients with fibroids with heterogeneous signal intensity was significantly higher than in patients with homogeneous signal intensity on T2WI (p<.05).The distance from the anterior side of the fibroid to the abdominal wall was significantly shorter in patients with fibroids with heterogeneous signal intensity than in patients with homogeneous signal intensity on T2WI (p<.05).No any other significant difference in baseline characteristics was observed between the two subgroups.
Since the number of cases with homogeneous hyperintense fibroids was small, the intergroup comparison between the two subgroups was not performed.

Comparison of therapeutic results between different types of fibroids
As shown in Table 3, there were statistically significant differences in treatment time, sonication time, treatment intensity, total treatment dosage, treatment efficiency, energy-efficiency factor (EEF) and NPV ratio among the groups with extremely hypointense, hypointense, isointense and hyperintense (p<.05).Treatment time was prolonged with the increase of signal intensity of uterine fibroids on T2WI, with the following pattern: extremely hypointense group < hypointense group < isointense group < hyperintense group (p<.001).Sonication time was also prolonged with the increase of signal intensity of uterine fibroids on T2WI (p<.001).The treatment intensity for uterine fibroids in the extremely hypointense and hypointense groups was significantly lower than that in the isointense and hyperintense groups (p<.001).The total energy used for treatment was also increased with the increase of signal intensity of uterine fibroids on T2WI (p<.001).The EEF for extremely hypointense fibroids was significantly lower than that of hypointense fibroids, isointense fibroids and the hyperintense fibroids, but the EEF for hypointense fibroids was significantly higher than the other three groups (p<.001).The NPV ratio achieved after HIFU treatment in each group was 75.2 ± 14.6%, 71.1 ± 15.6%, 68.2 ± 17.3% and 67.8 ± 16.6%, respectively.The highest NPV ratio was seen in the group with extremely hypointense fibroids, and the NPV ratio achieved in both the extremely hypointense and hypointense fibroids were .436Sm: submucosal; I: intramural; Ss: subserosal; A: anterior wall of the uterus; P: posterior wall of the uterus; L: left wall of the uterus; R: right wall of the uterus; F: fundus.a A significant difference between the group with extremely hypointense fibroids and each of the other group, p<.05.b A significant difference between the group with hypointense fibroids and the group with isointense fibroids, as well as the group with hyperintense fibroids, p<.05.

Comparison of therapeutic results between fibroids with homogeneous signal intensity and heterogeneous signal intensity
As shown in Table 4, sonication time, treatment intensity and total energy for uterine fibroids with heterogeneous signal intensity were significantly greater than that of uterine fibroids with homogeneous signal intensity in the group with extremely hypointense fibroids (p<.05).In contrast, there was no significant difference between the two subtypes in the group with extremely hypointense fibroids in treatment time, treatment efficiency, EEF and NPV ratio (p>.05).
In the group with hypointense fibroids, the treatment efficiency for uterine fibroids with heterogeneous signal intensity was significantly greater than that of fibroids with homogeneous signal intensity, which indicated that the ablated volume of uterine fibroids with heterogeneous signal intensity per unit time was significantly greater than that of the fibroids with homogeneous signal intensity (p<.05).The EEF of uterine fibroids with heterogeneous signal intensity was significantly smaller than that of fibroids with homogeneous signal intensity (p<.05).There were no significant differences in treatment time, sonication time, treatment intensity, treatment energy and NPV ratio between the two subtypes in the group with hypointense fibroids.
In the group with isointense fibroids, treatment time for uterine fibroids with heterogeneous signal intensity was significantly longer, while the NPV ratio was significantly lower than that for uterine fibroids with homogeneous signal intensity (p<.05).No other significant differences were found between the two subtypes in this group.
As the number of patients with homogeneous hyperintensity fibroids was small, comparisons between the two subtypes in the hyperintense fibroids group were not performed.

Analysis of the factors that affected the NPV ratio
The NPV ratio was used as the dependent variable.The other variables that may affect the NPV ratio were used as independent variables.In the independent variables, the median value was used to classify the continuous numerical variables into two groups.Five independent variables related to the NPV ratio were first identified by the univariate analysis: ablation volume of fibroids, distance from anterior side of fibroids to abdominal wall, type of fibroids, treatment time and sonication time (p<.05) (Table 5).Then multivariate ordered logistic regression analysis was used to investigate the factors that independently affect NPV ratio.The analysis results showed that the risk factors affecting the NPV ratio were ablation volume of fibroids (p<.001,OR ¼ 0.404, 95% confidence interval 0.260-0.628)and treatment time (p¼.002,OR ¼ 2.391, 95% confidence ratio 1.366-4.186)(Table 6).

Adverse events
All patients tolerated the HIFU treatment well.Some patients experienced treated area pain, sciatic or buttock pain, skin burning sensation, groin pain and transient leg pain during treatment, while the pain relieved after sonication was terminated.As shown in Table 7, no significant difference was observed in treated area pain, sciatic or buttock pain, skin burning sensation, groin pain among the patients with extremely hypointense, hypointense, isointense and hyperintense fibroids.The leg pain was more often seen in patients with isointense fibroids than that in patients with extremely hypointense and hypointense fibroids (p<.05).
All patients went home one day after HIFU treatment.Nurses called the patients daily during the first three days after HIFU treatment, but no one reported any clinically important symptoms.One hundred and forty-four patients who had surgical scars received full energy treatment, no skin burn occurred.All patients returned to day-to-day activities one day after the procedure and no major complications occurred in any patient in this study.

Long-term follow-up results
Of the 401 patients in this study, 215 patients completed 3year follow-up and 30 patients received re-interventions during the follow-up period due to the recurrence of symptoms.Among them, eight patients had recurrence within 1 year after HIFU, eight patients had recurrence in 2 years after HIFU, and 14 patients reported recurrence in 3 years after HIFU.The re-interventions included myomectomy in 20 cases, HIFU in six cases, hysterectomy in one case, two patients were given medications and one patient did not want to disclose the specific treatment.
In order to reduce the impact of lost visits on re-intervention results, we calculated the re-intervention rate using the re-intervention rate correction formula.The overall cumulative re-intervention rate at 36 months after HIFU was 9.7%.The cumulative re-intervention rates for patients with extremely hypointense fibroids, hypointense fibroids, isointense fibroids and hyperintense fibroids were 8.4%, 10.3%, 12.5% and 6.1%, respectively.There was no significant difference in the 3-year cumulative re-intervention rate among the groups (p>.05) (Table 8).

Discussion
The present study showed significant differences in treatment time, sonication time, treatment intensity, total treatment dosage, treatment efficiency, EEF and NPV ratio among the four groups of patients with extremely hypointense, hypointense, isointense and hyperintense fibroids.Sonication time, treatment intensity and total energy for heterogeneous fibroids were greater than that for homogeneous fibroids in patients with extremely hypointense fibroids.Treatment time The NPV ratio was used as the dependent variable and the patients were divided into three groups according to NPV ratio.In the independent variables, the median value was used to classify the continuous numerical variables.The NPV ratio was used as the dependent variable and the patients were divided into three groups according to NPV ratio.In the independent variables, the median value was used to classify the continuous numerical variables.
for heterogeneous fibroids was significantly longer than that for homogeneous fibroids in patients with isointense fibroids.
It is generally believed that a large NPV ratio achieved is significantly correlated with long-term symptom relief after HIFU treatment.In the past, some studies have compared the difficulty of ablation of uterine fibroids based on the characteristics of magnetic resonance signals [8][9][10][11][12][13].The signal intensity of fibroids on T2WI is a reliable predictor when assessing the difficulty level of HIFU ablation [8][9][10][11].In this study, we found that the NPV ratio achieved was correlated with the signal intensity of fibroids on T2WI, which is consistent with the findings of et al. [8].Although more treatment time was spent and more acoustic energy delivered, we still failed to achieve a high NPV ratio in hyperintense fibroids as opposed to the hypointense or extremely hypointense fibroids.Several studies have demonstrated that tissue characteristics of fibroids are key factors affecting the efficacy of HIFU treatment [9,10].The signal intensity on T2WI is positively correlated with the number of uterine smooth muscle cells and negatively correlated with fibrous tissue [10].Hyperintense fibroids are generally rich in cellular components and less in fibrous tissue.Therefore, the absorption coefficient of ultrasound energy is low and acoustic energy deposition in hyperintense fibroids is difficult.In addition, this type of fibroid is often rich in blood supply, and blood flow takes away part of the ultrasound energy, which also affects its absorption of ultrasound energy and often requires more energy deposition to improve the NPV ratio [14,15].In this study, we found the NPV ratio in fibroids with heterogeneous signal intensity was significantly lower than that in fibroids with homogeneous signal intensity, suggesting that both T2WI signal intensity and signal homogeneity of fibroids are important factors affecting the NPV ratio.This phenomenon may be due to heterogeneous fibroids have larger density variability, which may lead to higher scatter, reflection and refraction of ultrasound.In addition, there may be larger blood vessels in heterogeneous fibroids, and the blood flow may also take away the energy, which reduces the energy deposition at the treated area and affects the ablation effect [10].
Previous studies have shown that various factors including abdominal wall thickness, location of fibroids, and size of fibroids can also affect treatment efficacy of HIFU [14,[16][17][18].Therefore, we first compared the baseline characteristics of patients in these four groups, and the results showed that there were no significant differences in age, BMI, abdominal wall fat thickness, and location of fibroid among the four groups.However, the volume of fibroid in the group with hypointense fibroids was significantly smaller than that of patients with isointense and hyperintense fibroids.Meanwhile, the abdominal wall fat thickness, the distance from the posterior surface of the fibroid to the sacrococcyx, and the distance from the anterior surface of the fibroid to the abdominal wall in the groups of patients with isointense or hyperintense fibroids were significantly thinner or shorter than those of patients with extremely hypointense or hypointense fibroids.In this study, we found the ratio of the number of subserosal fibroids over all fibroids was the highest in the groups of patients with extremely hypointense and hypointense fibroids.The number of intramural fibroids was the highest in the group of patients with hyperintense fibroids.These factors may affect the results of HIFU ablation for uterine fibroids and should be considered when comparing the treatment dosage and long-term outcome of different types of fibroids.In this study, we further performed multivariate ordered logistic regression analysis and found that ablation volume of the treated fibroids and treatment time were correlated with NPV ratio, but no correlation was found between T2WI signal intensity and NPV ratio (Table 6).This phenomenon may be explained by that a large NPV ratio in hyperintense fibroids could be achieved by delivering more treatment energy.
EEF, defined as the energy required for ablating 1 mm 3 of uterine fibroid tissue, is considered a quantitative index for evaluating the dosage of HIFU treatment.Several studies have demonstrated that hyperintense fibroids are difficult to treat with HIFU, and EEF for hyperintense fibroids was higher than that for isointense and hypointense fibroids [8][9][10].The present study showed that the EEF of fibroids with extremely low signal intensity was significantly smaller than that of hypointense fibroids, isointense fibroids and hyperintense fibroids.However, the EEF of hypointense fibroids was larger than that of isointense and hyperintense fibroids.This phenomenon can be explained by the significantly smaller size of uterine fibroids and the greater distance from the anterior surface of fibroid to the abdominal wall in the hypointense fibroids group than those of other groups.A study has The cumulative reintervention rate was calculated using the correction formula: CI¼I/(N -W/2), where I¼number of re-interventions during the observation period, W¼number of lost visits during the observation period and N¼number of originating observations.
revealed that the size of uterine fibroids is related to EEF [10].EEF was negatively correlated with the volume of the fibroids which can be explained by 'damage-damage' interference effects.The necrotic area in the fibroid changed the 'acoustic environment' of the fibroid and thus contributed to the ultrasonic energy deposition [19].During HIFU treatment, ultrasound energy attenuation occurs as ultrasound beams penetrate through the skin and other intermediate tissue layers before reaching the fibroid.Every interface in the acoustic pathway may cause reflection and scattering of ultrasound beams.Thus, if the distance from fibroid to the abdominal wall is longer, more energy is lost, and there is a need to deliver more energy to ablate the same volume of fibroid than that for treating a fibroid with a short distance from the fibroid to the abdominal wall [16].On the other hand, degenerative changes are more common in large fibroids, and degenerative fibroids are not rich in blood supply and are more prone to energy deposition than fibroids with smaller size [15].We further compared the characteristics of fibroids with homogeneous signal intensity and heterogeneous signal intensity in the extremely hypointense, hypointense and isointense fibroid groups.The results showed that the mean volume of fibroids with homogeneous signal intensity was significantly smaller than that of fibroids with heterogeneous signal intensity.Also, the distance from the anterior surface of the fibroid to the abdominal wall was significantly greater in isointense fibroids with homogeneous signal intensity and in hypointense fibroids with homogeneous signal intensity in comparison with their corresponding fibroids with heterogeneous signal intensity.Although the mean EEFs for the extremely hypointense and the isointense fibroids with homogeneous signal intensity were greater than those for the fibroids with heterogeneous signal intensity in the group with extremely hypointense fibroids, as well as in the group with isointense fibroids.However, no statistical difference was observed except the EEF for fibroids with homogeneous signal intensity was significantly higher than that of fibroids with heterogeneous signal intensity in the hypointense fibroid group.We speculated that this may be related to the small sample size.Therefore, the present study also demonstrated that EEF was significantly correlated with the signal intensity on T2WI, and the size and location of fibroids also affected the EEF.The present study also demonstrated that ultrasoundguided HIFU ablation of uterine fibroids is safe [6,7,20].During the procedure of HIFU treatment, patients experienced treated area pain, sciatic or buttock pain, skin burning sensation, groin pain and transient leg pain, while the pain relieved after sonication was terminated.We did not observe any significant difference in the rates of treated area pain, sciatic or buttock pain, skin burning sensation, groin pain except rate of transient leg pain.Leg pain was related to the location of fibroids, not caused by the difference of signal intensity of fibroids.The patients with isointense fibroids had a higher rate of transient leg pain may be due to the relatively short distance between the fibroids and the sacrococcygeal region (Table 1).After HIFU treatment, patients often complained of mild lower abdominal pain because of inflammation and uterine contraction; sciatic or buttock pain was also a frequent adverse effect when the lesions were located at the posterior wall of the uterus, but such pain was mild.Skin burn is a complication unique to HIFU, especially for those patients with abdominal scars [21].However, in present study, no patients had skin blisters immediately after HIFU.In accordance with the institutionally approved protocol, all patients were required to have routine bowel preparation before the procedure; as a result, no bowel injury occurred in any patient.There were no reports of abdominal cramping or nerve injury.
The re-intervention rate after HIFU treatment has been a major clinical concern.Several studies have shown that the long-term re-intervention rate after HIFU treatment ranged from 3.2% to 13.7% [14,15,19,22].Previous studies have also shown that the NPV ratio is related to long-term re-intervention rate.Quinn et al. reported their results from 280 patients with uterine fibroids who were treated with MRgFUS, the overall 5-year re-intervention rate was 58.64%, while in patients with NPV ratio more than 50%, the 5-year re-intervention rate was 50%, which was lower than the overall reintervention rate [22,23].Recently, Liu et al. retrospectively analyzed 629 patients with a solitary uterine fibroid treated with HIFU, of which 536 completed long-term follow-up with a median follow-up time of 69 (interquartile range: 48-89) months and 110 (20.5%) were found to have local recurrence and 77 (14.4%) received re-interventions.They also revealed a median NPV ratio of 73% (interquartile range: 58-87%) in the recurrence group compared to 89% (interquartile range: 79-100%) in the non-recurrence group [24].In this study, the NPV ratios achieved were 75.2 ± 14.6%, 71.1 ± 15.6%, 68.2 ± 17.3% and 67.8 ± 16.6% in the fibroids with extremely low signal intensity, low signal intensity, isointensity and hyperintensity, respectively.Considering the effect of lost patients during follow-up, we used a correction formula for cumulative incidence to calculate the re-intervention rate in this study.The re-intervention rates at 36 months after HIFU were 8.4%, 10.3%, 12.5% and 6.1%, respectively (Table 8).Although a significant difference was observed in NPV ratio among the four groups, no statistically significant differences were found in the re-intervention rate between any of them.We further compared the re-intervention rate of patients with homogeneous versus heterogeneous signal intensity fibroids in each group, and there was no statistically significant difference between any two subtypes in the three groups.
This study is limited because it is a retrospective study.The patients were not randomized and some bias may occur.This study is also limited because we only enrolled patients with a solitary fibroid.Patients with multiple fibroids may have different results.In addition, many patients were lost to follow-up, which may affect the results.Finally, the relatively small number of patients with some subtypes of fibroids limited the statistical analysis.Therefore, a large-scale multicenter prospective study with standard protocol can be performed to confirm the findings.

Table 1 .
Comparison of baseline characteristics of patients with different types of uterine fibroids based on T2-weighted MR images.

Table 2 .
Comparison of baseline characteristics of patients with heterogeneous uterine fibroids or homogeneous fibroids based on T2-weighted MR images.
Fibroid type (Sm: submucosal; I: intramural; Ss: subserosal); fibroid location (A: anterior wall of the uterus; P: posterior wall of the uterus; L: left wall of the uterus; R: right wall of the uterus; F: fundus).a A significant difference between the heterogeneous group and the homogeneous group in patients with extremely hypointense fibroids, p<.05.b A significant difference between the heterogeneous group and the homogeneous group in patients with hypointense fibroids, p<.05.c A significant difference between the heterogeneous group and the homogeneous group in patients with isointense fibroids, p<.05.

Table 4 .
Comparison of HIFU treatment results between heterogeneous uterine fibroids and homogeneous fibroids in different groups based on T2-weighted MR images.

Table 5 .
Univariate analysis related to NPV ratio.

Table 6 .
Multivariate ordered logistic regression analysis for NPV ratio.

Table 7 .
Comparison of adverse effects among different types of uterine fibroids based on T2-weighted MR images.

Table 8 .
Comparison of re-intervention rate between different types of uterine fibroids based on T2-weighted MR images.