Key Points
COJD score combining JZmax with DMI predicted recurrence after FUAS.COJD = 2 was correlated with a 6-fold increased recurrence risk (HR = 6.004).
COJD = 2 identified patients at the highest risk of early recurrence.
Introduction
Adenomyosis refers to a common gynecological disorder characterized by the presence of ectopic endometrial tissue within the myometrium, usually presenting with dysmenorrhea and menorrhagia.1,2 Although hysterectomy remains the definitive treatment, it is not appropriate for women wishing to preserve fertility.3,4 Ultrasound-guided high-intensity focused ultrasound ablation surgery (FUAS) has emerged as a non-invasive alternative with favorable efficacy and safety.5,6 However, symptom recurrence is a clinically significant concern, typically occurs and typically occurs within 12 to 24 months post-treatment,7,8 underscoring the need for reliable prognostic markers.
The junctional zone (JZ), which refers to the innermost myometrial layer, has an essential effect on the adenomyosis pathogenic mechanism.9 Maximal junctional zone thickness (JZmax) has been the commonly used criterion for diagnosis, which is related to outcomes post-treatment like uterine artery embolization.10,11 However, its function in predicting recurrence post-FUAS remains largely unexplored.
In addition to JZmax, the depth of myometrial invasion (DMI), which refers to the greatest distance between the endometrial–myometrial interface and the deepest level of adenomyosis, is an important anatomical factor indicating disease burden.12 Beyond JZ alterations, the depth of myometrial invasion (DMI)—defined as the maximal distance from the endometrial–myometrial interface to the deepest extent of the adenomyotic lesion—serves as a key anatomical indicator of disease burden,13,14 can quantify the disease-related anatomical invasiveness.15 Consequently, integrating them in the composite score (COJD) assists in capturing more data on disease severity and providing higher prediction accuracy than a single parameter.
Given the high prevalence and recurrence rates of adenomyosis, and the varying success rates of FUAS reported globally, there is a pressing need to identify robust preoperative predictors. Therefore, the present work focused on evaluating the relationship between preoperative JZmax, DMI, as well as the combination (COJD) and symptom recurrence among intrinsic adenomyosis patients who underwent FUAS, determining the best cutoffs to facilitate risk stratification, and assessing whether COJD can be used to predict the recurrence timing.
Methods
Patients
This retrospective cohort study included 359 symptomatic adenomyosis patients treated at the Affiliated Nanchong Central Hospital of North Sichuan Medical College between June 2017 and June 2024. The study was approved by the hospital’s Ethics Committee (Approval No. 2021/104).
After applying inclusion and exclusion criteria, 326 patients with complete clinical and follow-up data were included.
The inclusion criteria were: (a) patients with presence of dysmenorrhea or menorrhagia; (b) intrinsic adenomyosis confirmed by preoperative MRI; (c) standardized JZ thickness assessment; and (d) treatment with FUAS. The exclusion criteria were: (a) patients with absence of preoperative magnetic resonance imaging (MRI) or incomplete JZ, DMI measurement; (b) concurrent uterine fibroids, endometriosis, or pelvic malignancy; (c) acute pelvic inflammatory disease; (d) inability to exclude malignancy; and (e) surgical or medical treatment for menorrhagia within 6 months prior to FUAS.
MRI Evaluation and Definitions
Pelvic MRI was performed using a 1.5T scanner with T2-weighted turbo spin-echo and diffusion-weighted sequences, following standardized protocols.6 Two radiologists, who were blinded to clinical outcomes, independently evaluated images, including MRI-based classification and evaluation of JZmax, DMI, and uterine and adenomyotic lesion volumes.6 JZmax was defined as the maximal thickness of the junctional zone.10 DMI was measured as the maximal distance from the endometrial–myometrial interface to the deepest extent of the adenomyotic lesion.13 Uterine and adenomyotic lesion volumes were calculated using the prolate ellipsoid formula: 0.52 × length × anteroposterior diameter × transverse diameter.16 In addition, an internal subtype of adenomyosis was defined as the basal endometrium invaginating into the myometrium with no involvement of the outer myometrial layer.17
FUAS Procedure, Follow-Up, and Outcome Definition
FUAS was performed using the focused ultrasound tumor therapeutic system (Model-JC200, Chongqing Haifu Technology Co., Ltd.) following standard protocols.18,19
Symptom recurrence was defined as the return of menorrhagia or dysmenorrhea after at least 3 months of symptom relief following FUAS.7 Patients were followed up via clinic visits or telephone interviews. The median follow-up period was 25 months (interquartile range (IQR): 15–46.5).
Statistical Analysis
Continuous variables were expressed as mean ± SD and compared through Student’s t-test. Categorical variables were compared by the chi-square test. Logistic regression was employed to identify factors associated with recurrence.
ROC curve analysis was performed to determine the optimal cutoffs for JZmax and DMI. The association between COJD and recurrence-free survival was assessed using Kaplan–Meier curves (K-M) and multivariate Cox regression. All tests were two-sided, with P < 0.05 representing statistical significance.
Analyses were performed using SPSS 22.0 and MedCalc 15.2.1.
Results
Patient Characteristics
Among the 326 patients, 58 (17.8%) experienced symptom recurrence. Compared with the non-recurrence group, JZmax (38.1 ± 13.0 mm vs 33.5 ± 10.8 mm, P < 0.05) and DMI apparently increased in the recurrence group, shown in Table 1.Table 1Clinicopathological FeaturesVariableTotal Cohort(n=326)Recurrence Group (n=58)Non-Recurrence Group (n=268)Test ValuePAge (years)43.3±5.442.5±5.043.4±5.5−1.1370.256BMI (kg/m2)23.6±3.023.5±2.943.4±5.50.7040.483Pre-FUAS dysmenorrhea2.1340.144 Yes (%)258 (79.1)50 (86.2)208 (77.6) No (%)68 (20.9)8 (13.8)60 (22.4)Pre-FUAS menorrhagia1.9000.168 Yes (%)293 (89.9)55 (94.8)238 (88.8) No (%)33 (10.1)3 (5.2)30 (11.2)Uterus position5.7360.057 Ante-verted (%)204 (62.6)32 (55.2)172 (64.2) Retro-verted (%)107 (32.8)20 (34.5)87 (32.5) Neutral (%)15 (4.6)6 (10.3)9(3.4)Type of adenomyotic lesions2.4700.116 Focal (%)174 (34.7)11 (19.0)78 (29.1) Diffuse (%)328 (65.3)47 (81.0)190 (70.9)Adenomyosis location2.1400.544 Anterior (%)100 (30.7)15 (25.9)85 (31.7) Posterior (%)126 (38.7)27 (46.6)99 (36.9) Fundus (%)37 (11.3)5 (8.6)32 (11.9)Anterior-Posterior (%)63 (19.3)11 (19.0)52 (19.4)Subcutaneous fat thickness (mm)18.4±6.518.8±9.119.5±9.6−0.2910.772Distance (mm)20.3±14.823.5±14.624.1±14.7−1.4250.155Uterus volume (cm3)252.2±142.1247.2±156.5225.1±122.70.2250.822Adenomyosis volume (cm3)53.6±55.778.7±68.969.6±83.00.8510.397JZmax(mm)34.5±11.438.1±13.033.5±10.83.0670.003Depth of myometrial invasion (mm)38.7±11.143.1±13.037.8±10.42.9350.004Total sonication time (s)427.9±302.3428.3±302.3427.9±300.50.0110.991Average sonication power (s/h)431.1±300.4440.1±114.7447.8±247.4−0.3560.722Abbreviations: BMI, body mass index; FUAS, focused ultrasound ablation surgery; distance, from the posterior surface of the adenomyosis lesion to the sacrum; JZmax, maximal thickness of junctional zone endometrium.
COJD Scoring System
Based on receiver operating characteristic (ROC)-derived cutoffs, a COJD was constructed as follows in Table 2: 1)COJD = 0: JZmax ≤ 23.1 mm and DMI ≤ 51.4 mm2)COJD = 1: JZmax > 23.1 mm or DMI > 51.4 mm3)COJD = 2: JZmax > 23.1 mm and DMI > 51.4 mmTable 2Prediction Scores of JZmax, DMI, and COJD of Symptom Recurrence in AdenomyosisVariableScoreJZmax (mm)≤23.10> 23.11DMI (mm)≤51.40> 51.41COJDJZmax ≤ 23.1 mm and DMI ≤ 51.4 mm0JZmax > 23.1 mm or DMI > 51.4 mm1JZmax > 23.1 mm and DMI > 51.4 mm2Abbreviations: JZmax, maximal thickness of junctional zone endometrium; DMI, depth of myometrial invasion; COJD, combined with JZmax and DMI.
COJD and Recurrence Risk
The distribution of COJD scores was presented as follows: COJD = 0 in 55 patients (16.9%), COJD = 1 in 189 patients (58.0%), and COJD = 2 in 82 patients (25.1%). Multivariate Cox regression analysis identified COJD as a significant independent predictor of recurrence (hazard ratio (HR) = 6.004, 95% confidence interval (CI): 1.125–32.054, P = 0.036) (Table 3).Table 3Multivariate Cox Regression Analysis of Factors for Symptom Recurrence of AdenomyosisVariableBSEWaldExp(B)95% CIPAge (year)−0.0470.0411.3220.9540.881–1.0340.250BMI (kg/m2)0.0670.0800.7101.0700.915–1.2510.400JZmax (mm)−0.1300.0762.9120.8780.756–1.0200.088DMI (mm)−0.0120.0860.0200.9880.835–1.1690.887COJD1.7920.8554.3996.0041.125–32.0540.036Uterus volume (cm3)−0.0010.0020.2320.9990.995–1.0030.630Adenomyosis volume (cm3)0.0030.0060.1991.0030.991–1.0140.655Subcutaneous fat thickness (mm)0.0080.0360.0551.0090.940–1.0830.814Distance (mm)−0.0270.0241.2370.9740.929–1.0210.266Sonication time (s)−0.0010.0010.2750.9990.997–1.0020.600Average sonication power (sh)0.0010.0020.1291.0010.996–1.0060.719Abbreviations: BMI, body mass index; JZmax, maximal thickness of junctional zone endometrium; DMI, depth of myometrial invasion; COJD, combined with JZmax and DMI; distance, from the posterior surface of the adenomyosis lesion to the sacrum.
ROC Analysis and Cutoff Determination
ROC curve analysis for JZmax yielded an optimal cutoff of 23.1 mm for symptom recurrence (sensitivity 98.3%, specificity 19.8%, area under the curve (AUC) = 0.612, P = 0.006). In terms of DMI, the optimal cutoff was 51.4 mm (AUC = 0.589, sensitivity 55.2%, specificity 62.7%, shown in Table S1Figure 1). Figure 1ROC curve of symptom recurrence in adenomyotic patients.
Receiver-operating characteristic curve analyses of JZmax (a) depth of myometrial invasion (b) and COJD (c) in adenomyosis for symptomatic recurrence.The text describes three receiver operating characteristic (ROC) line graphs. Graph A, titled ′JZmax,′ has an x-axis labeled ′100 minus Specificity′ and a y-axis labeled ′Sensitivity,′ both ranging from 0 to 100. The curve starts near (0, 0), rises steeply to (5, 20) and progresses through points like (20, 35), (40, 50), (60, 60), (75, 75), nearing 100 sensitivity between 90 and 100 on the x-axis.
Graph B, titled ′Depth of myometrial invasion,′ follows a similar axis setup. Its curve begins at (0, 0), rises to (10, 30) and steps through (25, 40), (45, 55), (65, 60), (80, 85), reaching near 100 sensitivity between 95 and 100 on the x-axis. Graph C, titled ′COJD,′ also shares the same axis labels.
The curve starts at (0, 0), rises to (10, 35), then increases almost linearly to (80, 100), maintaining 100 sensitivity from 80 to 100 on the x-axis.Three receiver operating characteristic line graphs for JZmax, depth of myometrial invasion and COJD.Abbreviations: JZmax, maximal thickness of junctional zone endometrium; COJD, combination of JZmax and depth of myometrial invasion.
Kaplan–Meier analysis exhibited a significant gradient in median recurrence time across COJD groups: 1) COJD = 0: 53.0 months; 2) COJD = 1: 45.8 months; 3) COJD = 2: 38.7 months;Log-rank P = 0.046 (Figure 2). Figure 2Magnitude of combination of JZmax and depth of myometrial invasion for predicting symptomatic recurrence. Kaplan-Meier analysis demonstrated that patients with COJD=2 (JZmax> 23.1 mm and DMI> 51.4mm) had a significantly shorter median time to recurrence (38.7 months) compared to those with COJD=1 (JZmax> 23.1 mm or DMI> 51.4mm, 45.8 months) or COJD=0 (JZmax≤23.1mm and DMI≤51.4mm, 53.0 months), with a statistically significant difference (P = 0.046).A Kaplan Meier graph titled ′Survival Functions′ shows survival over time.
The x-axis is ′Follow up time (months)′ from 0 to 60 and the y-axis is ′Cum Survival′ from 0.0 to 1.0. The legend is ′Combined JZmax and DMI′ with series: 1) JZmax less than or equal to 23.1mm and DMI less than or equal to 51.4mm; 2) JZmax > 23.1mm or DMI > 51.4mm; 3) JZmax > 23.1mm and DMI > 51.4mm, plus censored data for each. Series 1 stays near 1.0 until 25 months, then near 0.95 until 55 months.
Series 2 drops from 1.0 to 0.80 by 25 months and to 0.75 by 55 months. Series 3 falls to 0.80 by 12 months, 0.75 by 18 months, 0.65 by 33 months and 0.53 by 36 to 55 months. Censor marks are shown along each curve at various times.A Kaplan Meier line graph showing survival functions by combined JZmax and DMI groups.Abbreviations: JZmax, maximal thickness of junctional zone endometrium; COJD, combination of JZmax and depth of myometrial invasion.
Discussion
This study indicated that the combination of JZmax with DMI—quantified as the COJD score—was a significant independent predictor of symptom recurrence after FUAS for intrinsic adenomyosis. Patients with a COJD score of 2 (both JZmax > 23.1 mm and DMI > 51.4 mm) suggested the shortest median time to recurrence (38.7 months), which was shown to be significantly shorter than that observed in patients with lower scores. These findings suggested that the COJD score may serve as a clinically useful tool for postoperative risk stratification.
Compared to previous studies that primarily focused on individual parameters, our study introduces a novel, integrated scoring system that provides superior prognostic accuracy by combining functional and anatomical disease assessments.
The function of COJD in prognosis prediction may be ascribed to the fact that it integrates JZmax and DMI. In this work, patients developing relapsed symptoms showed markedly increased preoperative JZmax and DMI relative to those not developing recurrence, indicating that these two parameters indicate the associated disease severity. Notably, JZ thickening suggests tissue injury and repair (TIAR), chronic inflammation, and estrogen-driven hyperperistalsis, and these jointly establish a microenvironment good for ectopic endometrial survival and persistent disease.20–22 Furthermore, DMI suggests the myometrium’s anatomical involvement level as well as residual disease burden post-ablation.13 JZmax can reflect dysfunction at the endometrial–myometrial interface, and DMI can quantify adenomyosis invasiveness.15 Consequently, integrating these two parameters helps comprehensively assess their disease severity from the function and anatomy perspectives.14
Apparently, COJD performed well relative to single parameter, as revealed by multivariate Cox regression, and the HR was 6.004 (P = 0.036). Such a result suggests that the composite score is superior in prognosis modeling. Furthermore, the progressively increased recurrence risk due to growing COJD score, which were 53.0 and 38.7 months in the low- and high-risk groups separately, suggests that it may be used to guide follow-up intensity and inform decision-making concerning adjunctive treatment in high-risk patients.
In FUAS, such mechanistic knowledge assists in elucidating the relationship between higher preoperative JZmax and DMI and poor prognosis. FUAS can relieve symptom by thermally ablating ectopic endometrial tissue in adenomyosis, thus decreasing disease burden while mitigating relevant symptoms.6 Nonetheless, FUAS mainly targets the predominant lesion seen on imaging but can not totally remove microscopic ectopic foci, especially those situated deep in the myometrium or within the dysfunctional JZ microenvironment.
Although the predominant lesion is successfully ablated, the residual JZ shows a continuous dysfunctional state, characterized by dysregulated contractility, persistent angiogenesis, and chronic inflammation, which probably promotes microscopic disease regrowth, finally giving rise to symptom recurrence. Besides, the JZ thickening-related aberrant uterine peristalsis perpetuates mechanical irritation as well as tissue microtrauma, which activates the tissue injury and repair (TIAR) cycle and creates a pro-inflammatory environment good for the proliferation and survival of ectopic endometrial cells.13,23 Simultaneously, the angiogenic microenvironment enhanced by vascular endothelial growth factor (VEGF) upregulation contributes to recurrent or residual lesion revascularization, thus maintaining disease activity.24
DMI, which represents the myometrial infiltration depth, adds a new dimension to the pathophysiological condition. An increased invasion depth suggests the further extension of adenomyotic lesions from the endometrial–myometrial interface to the myometrium, which adds to the technical challenge in complete ablation.14 Lesions situated deeper are probably more proximal to key structures like uterine vasculature and serosa, which may limit the safety margin for delivering thermal energy in FUAS and increase the risk of residual viable tissue following treatment. In addition, extended myometrial involvement indicates an aggressive disease phenotype and an increased risk of regrowth.
For patients with a COJD score of 2, who were at the highest risk of early recurrence, alternative conservative techniques could be considered. Although FUAS yielded favorable outcomes for a large number of patients, the elevated recurrence risk observed in this specific subgroup indicated that hysterectomy or more aggressive combined medical and surgical approaches offered superior long-term symptom control. In contrast, for patients with COJD scores ranging from 0 to 1, FUAS demonstrated distinct advantages over alternative treatment modalities.
Such benefits included lower complication rates, shorter recovery periods, and potentially reduced overall healthcare costs, which stemmed from its non-invasive properties and outpatient application setting. These inherent advantages rendered FUAS a particularly optimal first-line conservative treatment choice for patients with a low recurrence risk.
Taken together, such parameters obtain additional knowledge regarding disease chronicity and recurrence risk. To be specific, JZmax indicates dysregulation of function and microenvironment at the endometrial–myometrial interface, marking the “soil” condition fostering disease persistence, whereas DMI contributes to quantifying the anatomical invasion degree, which is the “seed” burden determining baseline disease volume as well as technical resectability. Consequently, integrating these two parameters in the COJD score can holistically assess the terrain for adenomyosis development and the entrenching depth inside the myometrium.
Such dual perspective interprets the superior performance of their combination to a single parameter in predicting recurrence following FUAS, and highlights the importance of preoperative COJD evaluation in the identification of patients probably benefiting from close monitoring and adjunctive treatment post-ablation.
Based on these findings, we propose a management algorithm that stratifies patients by COJD score to guide clinical decision-making (Figure 3). For patients with COJD 0, routine FUAS followed by standard follow-up is recommended. For COJD 1, closer surveillance with imaging and symptom assessment at 6-month intervals is advised.
For COJD 2, given the high recurrence risk, a multidisciplinary discussion should be considered to evaluate the potential benefits of adjunctive medical therapy or alternative surgical interventions. Figure 3Proposed clinical management algorithm for adenomyosis patients based on COJD score stratification following FUAS. COJD = 0 (Low risk): patients with JZmax ≤ 23.1 mm and DMI ≤ 51.4 mm were recommended to undergo standard FUAS, followed by routine annual clinical follow-up; COJD = 1 (Intermediate risk): patients with either JZmax > 23.1 mm or DMI > 51.4 mm were advised to receive FUAS with intensified surveillance, including symptom and imaging assessments every 6 months; COJD = 2 (High risk): patients with both JZmax > 23.1 mm and DMI > 51.4 mm should be discussed in a multidisciplinary team (MDT) setting.
Individualized decisions may include FUAS combined with adjunctive medical therapy (eg, GnRH-a, LNG-IUS), alternative conservative surgical options, or hysterectomy for those without fertility desire. Post-treatment monitoring should be more frequent (every 3–6 months) to detect early recurrence.The flowchart details adenomyosis management for FUAS patients, starting with preoperative MRI to assess JZmax and DMI. The COJD score determines three pathways: COJD 0, 1 and 2.
COJD 0 (JZmax less than or equal to 23.1 mm, DMI less than or equal to 51.4 mm) suggests standard FUAS with annual symptom checks. COJD 1 (JZmax > 23.1 mm or DMI > 51.4 mm) requires FUAS with enhanced surveillance every 6 months. COJD 2 (JZmax > 23.1 mm and DMI > 51.4 mm) involves a multidisciplinary team discussion, considering FUAS with adjunct therapy, conservative surgery, or hysterectomy if no fertility desire.
Follow-up includes frequent surveillance every 3-6 months, monitoring symptom recurrence to decide on re-intervention or maintaining the current protocol.A flowchart of adenomyosis patient management based on COJD score stratification following FUAS.Abbreviations: FUAS, focused ultrasound ablation surgery; MRI, magnetic resonance imaging; JZmax, maximal junctional zone thickness; DMI, depth of myometrial invasion; COJD, combination of JZmax and DMI; GnRH-a, gonadotropin-releasing hormone agonist; LNG-IUS, levonorgestrel-releasing intrauterine system.
Certain limitations should be noted in the present work. Firstly, considering the retrospective design, it is impossible to establish causal inferences concerning the relation of COJD with symptom recurrence, despite the potent relation observed. Secondly, patients with insufficient follow-up data were excluded, probably introducing selection bias, although great efforts were made to ensure sufficient data acquisition.
Thirdly, symptom recurrence was self-reported by patients with the binary (yes/no) classification but not the well-established graded scale like the Numerical Rating Scale, probably affecting the symptom severity assessment accuracy. Fourthly, our single-center nature limits the result’s generalizability. Multicenter prospective studies that incorporate standard symptom metrics and verified outcome measures should be conducted for validating the COJD scoring system and refining risk prediction models.
Conclusions
To conclude, the COJD score, integrating JZmax and DMI, independently predicted symptom recurrence post-FUAS for adenomyosis. Typically, a score of 2 may identify patients with the greatest risk of early recurrence, which could instruct individualized postoperative monitoring and management.