Abstract
Purpose
This study aimed to establish a novel acne scoring system (GAGS-HFUS) that combined the Global Acne Grading System (GAGS) with high-frequency ultrasound (HFUS), and to evaluate its predictive value for unfavorable treatment response in patients who have moderate-to-severe acne.
Methods
This prospective study recruited consecutive patients with mild-to-severe acne vulgaris. The GAGS-HFUS was developed based on the HFUS-detected morphological characteristics of acne and the GAGS facial region classification. Acne severity was assessed using both ultrasonic scales (the GAGS-HFUS, Sonographic Scoring System for Acne [SSSA], and Sonographic Scoring of Acne [SOS-Acne]) and a clinical scale (the Investigator Global Assessment [IGA]). Three months of followup were conducted in a subgroup of 34 patients with moderate-to-severe acne treated with oral isotretinoin. Univariate and multivariate Poisson regression analyses were conducted to determine the predictive value of GAGS-HFUS for unfavorable treatment response in this subgroup.
Results
Overall, 60 patients with mild-to-severe acne (mean age, 24 years; range, 16 to 34 years; 25 males) were enrolled. The GAGS-HFUS demonstrated substantial agreement with SOS-Acne (Spearman rank correlation coefficient [Rs], 0.864; kappa, 0.713) and IGA (Rs, 0.837; kappa, 0.660). In the follow-up cohort of 34 patients with moderate-to-severe acne, GAGS-HFUS score was identified as an independent predictor of unfavorable treatment response (risk ratio, 1.125; P=0.002). Its predictive performance (area under the receiver operating characteristic curve, 0.936; P<0.001) was superior to those of IGA and SOS-Acne.
Conclusion
GAGS-HFUS is an independent predictor of unfavorable treatment response in patients with moderate-to-severe acne, underscoring its value in the strategic management of acne.
Keywords: Dermatologic ultrasound, Ultra-high-resolution ultrasound, Acne vulgaris, Global Acne Grading System with high-frequency ultrasound (GAGS-HFUS)
Graphic Abstract
Introduction
Acne vulgaris is a chronic inflammatory disorder that affects the pilosebaceous follicles of the skin. It is the eighth most prevalent disease worldwide, with a prevalence of 9.4% [1]. As a superficial inflammatory skin disease, acne imposes a substantial burden due to the risk of deep, disfiguring scars, which are associated with higher rates of depression and suicidal ideation [2-4]. Systemic treatments, including antibiotics and retinoids, are the standard of care for moderate-to-severe acne. However, regardless of the specific antibiotic or retinoid prescribed, the recurrence rate remains as high as 30% [5,6].
A key reason for this high recurrence rate is that acne severity is often assessed based on dermatologist experience [1,7]. Such subjective visual inspection is insufficient to detect subtle or deeper follicular involvement. Clinical evaluation has been reported to consistently underestimate disease severity [8]. Although histopathological analysis is viewed as the gold standard for assessing the severity of acne lesions, performing such an invasive procedure on the face is challenging. Non-invasive imaging modalities, such as dermoscopy [9], reflectance confocal microscopy [10], and optical coherence tomography [11], have been introduced for diagnosing and managing acne. Nonetheless, these techniques are limited to the papillary dermis and do not yield information on the deeper infiltration of lesions. The absence of appropriate medical imaging devices and modalities complicates effective management of the disease.
As a convenient, non-invasive imaging modality for skin lesions, high-frequency ultrasound (HFUS) offers a favorable balance between penetration and resolution [12,13]. HFUS has been found to facilitate in vivo visualization of various subclinical anatomical changes in the skin of patients with acne, including widening of the hair follicles, heterogeneous echogenicity of the dermis, anechoic or hypoechoic pseudocysts, fistulous tracts, and fluid collections [14]. The presence of fistulas is a key predictor of suboptimal therapeutic efficacy and an increased risk of scar development [15], underscoring its pivotal role in disease prognosis. HFUS enables the effective localization, measurement, and tracking of these lesions.
Due to major advancements in HFUS technology, researchers have developed several sonographic scoring systems for the management of patients with acne. For instance, Wortsman et al. [14] proposed the Sonographic Scoring of Acne (SOS-Acne), which involves counting all detectable lesions on the face using HFUS. However, this method is impractical for daily clinical use due to time constraints. More recently, Wang et al. [16] introduced the novel Sonographic Scoring System for Acne (SSSA), which evaluates only the most severe lesion in each patient. Although this approach significantly reduces examination time, it is insufficient to represent the patient’s overall condition using a single lesion. Therefore, an accurate and practical sonographic scoring system is urgently needed to overcome current clinical limitations, enabling precise classification of acne and providing reliable imaging evidence for clinical management.
To prioritize both accuracy and practicality in acne scoring, the present authors integrated the Global Acne Grading System (GAGS) [17] with HFUS, proposing a new quantitative scoring system—termed the GAGS-HFUS—for the assessment of acne severity. Specifically, the GAGS evaluates five distinct facial regions (forehead, right cheek, left cheek, chin, and nose) as well as the chest/upper back [18], while HFUS assesses the most severe lesion in each region individually. Overall, the GAGS-HFUS not only reduces inspection time compared to the SOS-Acne, but also provides a better objective assessment of the patient’s overall condition compared to SSSA.
Therefore, the objective of this study was to develop the GAGS-HFUS acne scoring system and to determine the predictive value of the GAGS-HFUS for unfavorable treatment response in patients with moderate-to-severe acne, thus evaluating the potential need for modifications in disease management.
Materials and Methods
Compliance with Ethical Standards
This study was ethically approved by the Shanghai Jiao Tong University School of Medicine's Ethics Committee (Ethical Number 2024-KY-148(K)). Written informed consent was obtained from all participants after a thorough explanation of the study protocol. Additionally, written informed consent for publication of clinical and imaging photographs was obtained from the patients.
Study Cohorts
This prospective study included consecutive patients with mild-to-severe acne vulgaris who visited the university hospital dermatology department and underwent HFUS between June 2024 and January 2025. The exclusion criteria were as follows: (1) destructive treatments (photodynamic therapy, surgery, laser therapy, etc.) or medication treatment within 3 months prior to HFUS imaging; (2) incomplete clinical information; and (3) insufficient image quality (e.g., scab barrier impeding effective ultrasound transmission) (Fig. 1).
Fig. 1. Flowchart of patient selection.
According to the Global Acne Grading System with high-frequency ultrasound (HFUS) criteria, patients with mild acne were defined as those with a score below 19 points. n, number of patients.
Clinical Evaluation
Demographic and clinical characteristics, including age, sex, body mass index (BMI), alcohol consumption, smoking status, family history of acne, history of acne scars, and concomitant diseases (depression/anxiety, hypertension, hyperlipidemia, diabetes mellitus, and insulin resistance) were collected from clinical records. Clinical grading for each patient was assessed using the Investigator Global Assessment (IGA) scale for acne vulgaris [19]. All assessments were conducted independently by two senior dermatologists. In cases of disagreement, consensus was achieved through discussion.
Ultrasound Equipment, Examination, and Interpretation
All examinations were conducted at baseline by a radiologist with 5 years of experience in dermatological ultrasound. HFUS was performed using a linear array transducer (Aplio i900, Canon, Tokyo, Japan) operating at a frequency of 33 MHz. Two experienced physicians independently assigned ultrasound grades for the lesions, utilizing different ultrasound criteria (including the GAGS-HFUS, SSSA, and SOS-Acne scoring systems). In cases of disagreement between the two physicians, consensus was reached through joint review. To facilitate understanding of the structural and methodological differences among the three acne severity scoring systems, a comparative summary of their key features was provided in Supplementary Table 1.
GAGS-HFUS scoring system
The GAGS score was calculated by summing six regional subscores. Specifically, each subscore was determined by multiplying the factor for each region (chest and upper back, 3; chin and nose, 1; forehead and each cheek, 2) by the most heavily weighted lesion within that region (Supplementary Fig. 1). The regional factors were established based on the density, surface area, and distribution of pilosebaceous units [18]. The GAGS-HFUS retained these six regions and their respective factors; however, the most heavily weighted lesion in each region was identified using HFUS rather than clinical evaluation (Fig. 2). Severity was then classified into three GAGS-HFUS categories: scores of 1-18 were defined as mild, 19-30 as moderate, and >30 as severe [18] (Fig. 3). Treatment was prescribed accordingly; mild acne was either untreated or managed with topical treatments alone, while moderate-to-severe cases were treated with oral isotretinoin (Fig. 1).
Fig. 2. Various grades of acne classified by high-frequency ultrasound (HFUS).
HFUS (33 MHz) images of each lesion were delineated (yellow lines). Five grades of acne severity were observed and categorized by HFUS. Grade 1 corresponds to comedo, presenting as a widened hypoechoic band on HFUS, indicating swelling of the hair follicle. Grades 2 and 3 represent papules and pustules, respectively. Initially, acne presents as a localized hypoechoic dermal lesion with an oval shape, suggesting local inflammatory infiltration and tissue liquefaction. As the disease progresses, inflammation extends deeper and peripherally, forming a vase-shaped hypoechoic structure; this indicates severe infection of the hair follicle beyond its anatomical boundaries, resulting in perifolliculitis. Grade 4 denotes nodules or cysts, characterized by hypoechoic structures larger than 5 mm, with or without epidermal elevation. Grade 5 indicates fistula, defined as a pathological connection between tissues. On HFUS, fistulas appear as anechoic or hypoechoic band-like structures within the dermis and/or hypodermis. Grades 1-4 were adapted from the original Global Acne Grading System (GAGS) [18], with modifications to reflect HFUS morphological features [8]. Grade 5 was added based on characteristic HFUS findings of deep, communicating lesions suggestive of higher clinical risk [14].
Fig. 3. Global Acne Grading System with high-frequency ultrasound (GAGS-HFUS).
Lesion grades (1-5) were defined based on HFUS morphological features, as detailed in Fig. 2.
Other ultrasound scoring systems
One ultrasound scoring system, the SOS-Acne [14], involved counting all detectable lesions on the face using HFUS. SOS-Acne was divided into three categories (stages 1-3): cases with fewer than five pseudocysts without fistulas were defined as mild (stage 1); those with five to nine pseudocysts without fistulas were defined as moderate (stage 2); and those with more than 10 pseudocysts, with or without fistulas, were defined as severe (stage 3).
Another ultrasound scoring system was the SSSA [16], which evaluated only the most severe lesion in each patient. Under the SSSA, acne was divided into three grades. Grade I (mild) was defined as the presence of a widened hypoechoic band, with or without surrounding heterogeneous dermal echogenicity; grade II (moderate) was characterized by a focal area of dermal hypoechogenicity; and grade III (severe) represented a pseudocyst with or without apparent epidermal bulging.
Other HFUS features
Other features included epidermal and dermal thickness (the vertical distance of the epidermis and dermis around the lesion), layer involvement (epidermis, dermis, or subcutaneous tissue), echogenicity (hypoechogenicity, isoechogenicity, or hyperechogenicity), superb microvascular imaging (SMI) signals, and mean vascular index (VI). The VI was defined as the ratio of color (detected flow) pixels to all pixels within a defined region of interest [20]. SMI and VI represented a valuable method for tracking dynamic disease changes [21,22]. The mean VI value was calculated as the arithmetic average of the cumulative VI from six discrete anatomical regions (Supplementary Fig. 2).
Clinical Follow-up
A 3-month clinical follow-up was conducted on a cohort of patients with moderate-to-severe acne (GAGS-HFUS ≥19, n=34) treated with oral isotretinoin (Fig. 1). These patients were re-evaluated according to the IGA grading system. The effective treatment group was defined as those with a downgrade in IGA grade compared to baseline. The primary endpoint of this study was the predictive value of GAGS-HFUS for unfavorable treatment response in patients who had moderate-to-severe acne.
Statistical Analyses
Statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). The Kolmogorov-Smirnov test was used to assess the normality of data distribution. Continuous variables were reported as mean±standard deviation (SD) for normally distributed data, or as median and 25th-75th percentiles for non-normally distributed data. Categorical data were presented as counts and percentages. Spearman rank correlation was used to assess the continuous associations between clinical and ultrasound grading.
The consistency between clinical and ultrasound grading was measured using kappa values, which were interpreted as follows: (1) kappa <0, poor agreement; (2) 0<kappa<0.20, slight agreement; (3) 0.20<kappa<0.40, fair agreement; (4) 0.40<kappa<0.60, moderate agreement; (5) 0.60<kappa<0.80, substantial agreement; (6) 0.80<kappa<1, perfect agreement.
To evaluate inter-observer agreement, two experienced raters independently scored all cases using the GAGS-HFUS, SSSA, SOS-Acne, and IGA systems. For GAGS-HFUS, the intraclass correlation coefficient (ICC) was calculated using a two-way random-effects model for consistency. For SSSA, SOS-Acne, and IGA, inter-rater reliability was assessed using the Cohen kappa. These values were reported with 95% confidence intervals (CI) and corresponding P-values.
Factors associated with unfavorable treatment response were analyzed using univariate analysis (two-sample t-test, chi-square test, or Mann-Whitney U-test) and multivariate Poisson regression analysis. Receiver operating characteristic (ROC) curves were constructed, and the area under the ROC curve (AUROC) was calculated to evaluate the predictive value of GAGS-HFUS for unfavorable treatment response. P-values less than 0.05 were considered to indicate statistical significance.
Results
Patient Clinical Features
The prospective cohort comprised 60 patients with mild-to-severe acne according to the pre-specified criteria (Fig. 1). The mean age was 24 years (interquartile range [IQR], 4 years), and 41.7% of the patients were men. The mean BMI was 20.69 kg/m2 (IQR, 4.50 kg/m2). Most patients (75.0%) reported no family history of acne vulgaris; however, a substantial proportion (63.3%) had a history of acne scars. Patients were assessed using the IGA acne grading system, with six (10.0%), 26 (43.3%), 25 (41.7%), and three (5.0%) patients classified as grades I (slight), II (mild), III (moderate), and IV (severe), respectively. A complete description of patient characteristics is provided in Supplementary Table 2.
GAGS-HFUS Scoring System and Other HFUS Morphological Findings
All 60 patients with mild-to-severe acne underwent HFUS examination at baseline and were sequentially scored using the GAGS-HFUS criteria (Figs. 2, 3).The mean GAGS-HFUS score was 19.62 (SD, 10.08), with the most severe case receiving a score of 46 (Fig. 4).
Fig. 4. Diagram depicting the application of Global Acne Grading System with high-frequency ultrasound (GAGS-HFUS) in an 18-yearold male patient with severe acne vulgaris.
The most severe lesion in each facial region was identified (white circles). Corresponding high-frequency ultrasound (33 MHz) images of each lesion were delineated (yellow lines). (1) GAGS-HFUS (46 points): Lesions on the nose and chin (grade 3) exhibited a characteristic vase-like appearance on HFUS, indicating severe infection of the hair follicle beyond its anatomical limits, resulting in perifolliculitis. Lesions on the chest and right cheek (grade 4) presented as pseudocysts (>5 mm) within the dermis, with well-defined borders, uneven internal echogenicity, and epidermal elevation. Lesions on the forehead and left cheek (grade 5) demonstrated fistulas, indicative of pathological connections between tissues. HFUS revealed hypoechoic band-like structures within the dermis. Based on the GAGS-HFUS criteria, the global score was determined to be 46. (2) Sonographic Scoring System for Acne (SSSA) (grade III): Assigned based on the presence of a pseudocyst as the most severe lesion [8], characterized by a large hypoechoic structure with well-defined margins and visible epidermal elevation. (3) Sonographic Scoring of Acne (SOS-Acne) (stage 3): Assigned based on the identification of 17 pseudocysts, five of which were accompanied by fistulous tracts, on HFUS. According to SOS-Acne criteria [14], this lesion burden corresponds to stage 3, indicative of severe disease.
In addition to the composite GAGS-HFUS score, a total of 214 acne lesions from 60 patients were further characterized, with each representing the most severe lesion within an affected facial or truncal region as identified by HFUS. The mean epidermal and dermal thicknesses were 0.19±0.03 mm and 1.30±0.16 mm, respectively. Regarding layer involvement, 79 lesions (36.9%) were confined to the dermis, while 135 lesions (63.1%) extended into the subcutaneous tissue. Echogenicity patterns indicated that 24 lesions (11.2%) were anechoic, 128 (59.8%) hypoechoic, 29 (13.6%) isoechoic, and 33 (15.4%) mixed. The mean VI was 36.17%±16.60%, with increased VI values generally observed in inflammatory lesions.
Comparison of Clinical Versus Ultrasonic Evaluations
The 60 patients with acne were evaluated at baseline using four different scoring systems to assess disease severity: the IGA, GAGS-HFUS, SOS-Acne, and SSSA grading systems.
Spearman rank correlation tests were performed between pairs of these scoring systems to determine their level of agreement. The results demonstrated high levels of correlation among the four systems, with a Spearman rank correlation coefficient (Rs) of 0.837 between IGA and GAGS-HFUS (P<0.001), an Rs of 0.864 between SOS-Acne and GAGS-HFUS (P<0.001), and an Rs of 0.737 between SSSA and GAGS-HFUS (P<0.001) (Fig. 5).
Fig. 5. Heatmap integration of the Global Acne Grading System with high-frequency ultrasound (GAGS-HFUS), Investigator Global Assessment (IGA), Sonographic Scoring System for Acne (SSSA), and Sonographic Scoring of Acne (SOS-Acne) grading systems.
Spearman correlation analysis was performed among these four scoring systems. The Spearman rank correlation coefficient (Rs) indicated high levels of correlation between the GAGS-HFUS and the IGA, SSSA, and SOS-Acne scales (all P<0.001).
Under the GAGS-HFUS, 26 cases were classified as mild, 25 as moderate, and nine as severe acne. A complete description of patient distributions according to the different acne grading systems was provided in Supplementary Table 3. The kappa test results confirmed substantial agreement between GAGS-HFUS and SOS-Acne (kappa, 0.713; P<0.001) (Table 1) and substantial agreement between GAGS-HFUS and IGA (kappa, 0.660; P<0.001) (Supplementary Table 4), but poor agreement between GAGS-HFUS and SSSA (kappa, -0.073; P=0.202) (Supplementary Table 5).
Table 1.
Comparison between GAGS-HFUS and SOS-Acne evaluations at baseline
| SOS-Acnea) | GAGS-HFUS |
Total | ||
|---|---|---|---|---|
| Mild | Moderate | Severe | ||
| Mild | 26 | 2 | 0 | 28 |
| Moderate | 0 | 16 | 2 | 18 |
| Severe | 0 | 7 | 7 | 14 |
| Total | 26 | 25 | 9 | 60 |
GAGS-HFUS, Global Acne Grading System with High-Frequency Ultrasound; SOS-Acne, Sonographic Scoring of Acne.
Kappa, 0.713; P<0.001.
In addition, inter-observer agreement was evaluated across the four scoring systems. The GAGS-HFUS displayed excellent reproducibility, with an ICC of 0.960 (95% CI, 0.946 to 0.974; P<0.001). For the categorical systems, the Cohen kappa values were 0.791 for SSSA (95% CI, 0.632 to 0.950; P<0.001), 0.868 for SOS-Acne (95% CI, 0.757 to 0.979; P<0.001), and 0.870 for IGA (95% CI, 0.745 to 0.996; P<0.001). These findings collectively confirmed good-to-excellent inter-rater consistency across all scoring systems, with the highest reproducibility observed for the GAGS-HFUS framework.
Clinical Follow-up
A prospective longitudinal study was conducted on a cohort of 34 patients with moderate-to-severe acne (GAGS-HFUS ≥19). Of these, 21 patients displayed a reduction in their IGA acne grade from baseline, classifying them as the treatment-effective group. In contrast, 13 patients showed no change in IGA grade, identifying them as the treatment-ineffective group (that is, those who experienced unfavorable treatment responses).
The effective group demonstrated a mean GAGS-HFUS score of 23.00 (SD, 3.02), whereas the ineffective group displayed a mean score of 33.08 (SD, 6.50). Univariate analysis (two-sample t-test, chi-square test, or Mann-Whitney U-test) indicated that GAGS-HFUS score (P<0.001), SOS-Acne stage (P=0.001), and IGA grade (P=0.005) were all predictors of unfavorable treatment response; these variables were subsequently included in the multivariate Poisson regression analysis. After adjusting for age, BMI, and sex, the multivariate analysis demonstrated that GAGS-HFUS score (risk ratio, 1.125; P=0.002) remained an independent predictor of unfavorable treatment response. Details of the univariate and multivariate Poisson regression analyses are presented in Tables 2-4.
Table 2.
Demographic and clinical characteristics of the treatment-ineffective versus treatment-effective groups (univariate analysis, 34 patients with moderate-to-severe acne)
| Variable | Ineffective group (n=13) | Effective group (n=21) | P-value |
|---|---|---|---|
| Age (year) | 24.00±4.76 | 24.95±3.57 | 0.511 |
| Sex | |||
| Female | 5 (38.5) | 14 (66.7) | 0.160 |
| Male | 8 (61.5) | 7 (33.3) | |
| Body mass index (kg/m2) | 22.81±3.51 | 21.61±4.18 | 0.393 |
| Weight (kg) | 67.31±11.03 | 61.76±14.47 | 0.246 |
| Height (cm) | 172.00±10.37 | 168.57±8.96 | 0.315 |
| Smoking | |||
| No | 11 (84.6) | 19 (90.5) | 0.627 |
| Yes | 2 (15.4) | 2 (9.5) | |
| Alcohol consumption | |||
| No | 10 (76.9) | 18 (85.7) | 0.653 |
| Yes | 3 (23.1) | 3 (14.3) | |
| Family history of acne | |||
| No | 12 (92.3) | 13 (61.9) | 0.107 |
| Yes | 1 (7.7) | 8 (38.1) | |
| History of acne scars | |||
| No | 4 (30.8) | 10 (47.6) | 0.477 |
| Yes | 9 (69.2) | 11 (52.4) | |
| Depression/anxiety | |||
| Mild | 0 | 2 (9.5) | 0.584 |
| Moderate | 12 (92.3) | 17 (81.0) | |
| Severe | 1 (7.7) | 2 (9.5) | |
| IGA score | |||
| 2 (mild) | 0 | 6 (28.6) | 0.005* |
| 3 (moderate) | 10 (76.9) | 15 (71.4) | |
| 4 (severe) | 3 (23.1) | 0 |
Values are presented as mean±SD or number (%).
IGA, Investigator Global Assessment; SD, standard deviation.
P<0.05 in univariate analysis.
Table 3.
High-frequency ultrasound features of the treatmentineffective versus treatment-effective groups (univariate analysis, 34 patients with moderate-to-severe acne)
| Variable | Ineffective group (n=13) | Effective group (n=21) | P-value |
|---|---|---|---|
| GAGS-HFUS, points | 33.08±6.50 | 23.00±3.02 | <0.001* |
| SOS-Acne | |||
| 1 (mild) | 0 | 2 (9.5) | 0.001* |
| 2 (moderate) | 3 (23.1) | 15 (71.4) | |
| 3 (severe) | 10 (76.9) | 4 (19.0) | |
| SSSA | |||
| I (mild) | 0 | 0 | 0.431 |
| II (moderate) | 0 | 1 (4.8) | |
| III (severe) | 13 (100) | 20 (95.2) | |
| Vascular index (%) | 32.74±14.21 | 43.62±13.17 | 0.073 |
| Epidermal thickness (mm) | 0.19±0.03 | 0.18±0.03 | 0.296 |
| Dermal thickness (mm) | 1.36±0.18 | 1.31±0.09 | 0.350 |
Values are presented as mean±SD or number (%) unless otherwise indicated.
GAGS-HFUS, Global Acne Grading System with High-Frequency Ultrasound; SOSAcne, Sonographic Scoring of Acne; SSSA, Sonographic Scoring System for Acne.
P<0.05 in univariate analysis.
Table 4.
Factors associated with treatment ineffectiveness (multivariate analysis, 34 patients with moderate-to-severe acne)
| Variable | RR | 95% CI | P-value |
|---|---|---|---|
| Age (year) | 1.058 | 0.960-1.165 | 0.254 |
| Sex | 1.524 | 0.549-4.229 | 0.419 |
| Body mass index (kg/m2) | 1.016 | 0.878-1.175 | 0.834 |
| GAGS-HFUS, points | 1.125 | 1.045-1.211 | 0.002* |
| IGA | 0.634 | 0.249-1.614 | 0.339 |
| SOS-Acne | 2.895 | 0.933-8.978 | 0.066 |
RR, risk ratio; CI, confidence interval; GAGS-HFUS, Global Acne Grading System with High-Frequency Ultrasound; IGA, Investigator Global Assessment; SOS-Acne, Sonographic Scoring of Acne.
P<0.05 in multivariate analysis.
The predictive performance of the GAGS-HFUS, SOS-Acne, and IGA grading systems for unfavorable treatment response in patients with moderate-to-severe acne was further evaluated. The AUROC for the GAGS-HFUS system was 0.936 (P<0.001), outperforming both the IGA system (AUROC, 0.725; P=0.029) and the SOS-Acne system (AUROC, 0.800; P=0.004). In addition, a GAGS-HFUS score above 28 was associated with an increased risk of unfavorable treatment response, with a sensitivity of 0.769, a specificity of 0.952, and a Youden index of 0.721 (Fig. 6).
Fig. 6. Receiver operating characteristic (ROC) curve for different acne scoring systems.
ROC curves were constructed to evaluate the predictive performance of various acne scoring systems. The area under the ROC curve for the Global Acne Grading System with highfrequency ultrasound (GAGS-HFUS) system was superior to that of both the Investigator Global Assessment (IGA) and Sonographic Scoring of Acne (SOS-Acne) grading systems.
Discussion
This study developed the GAGS-HFUS acne scoring system and demonstrated that GAGS-HFUS was a strong independent predictor of unfavorable treatment response in patients who had moderate-to-severe acne.
Acne vulgaris, as a superficial yet often occult inflammatory disease, poses a significant challenge in both diagnosis and management [23] due to the discrepancy between its clinical appearance and underlying pathology. Historically, surface morphology was considered the standard for diagnosing acne vulgaris [24]. However, clinicians have frequently overlooked internal alterations and deeper pathological changes within lesions, contributing to high failure rates and frequent relapses. With recent advancements in HFUS, several ultrasound-based scoring systems have been developed to more accurately assess the severity of acne. For example, Wortsman et al. [14] proposed the SOS-Acne system, which counted all facial lesions detectable by HFUS, while Wang et al. [16] proposed the SSSA classification, which evaluated only the most severe lesion. Nevertheless, both systems have limitations in clinical practice: SOS-Acne is too time-consuming for routine use, and SSSA may not fully represent the overall patient condition.
Given these considerations, we integrated the GAGS [17] with HFUS and introduced a novel quantitative ultrasound-based scoring system—the GAGS-HFUS—to assess acne severity. The GAGS-HFUS not only reduces examination time compared to SOS-Acne, but also offers a more comprehensive and objective assessment than SSSA.
Notably, the GAGS-HFUS differs from the GAGS in multiple ways. First, in the GAGS, a papule is graded as 2 points and a pustule as 3 points. In contrast, in GAGS-HFUS, a localized hypoechoic dermal lesion with an oval shape is assigned 2 points. As the condition progresses, lesions exhibit deeper and more extensive heterogeneous dermal echogenicity, typically taking on a vase-like shape, and are graded as 3 points. This reflects severe infection extending beyond the hair follicle’s anatomical boundaries, resulting in perifolliculitis. Second, GAGS-HFUS grades fistula as 5 points. In previous research, limitations of imaging technology often led to the fistula type being overlooked. However, fistula is a significant risk factor for suboptimal clinical outcomes and scar formation [15]. With advancements in HFUS, it is now possible to detect subclinical fistula [25]. Therefore, fistula has been incorporated into GAGS-HFUS to assign appropriate clinical importance.
In the 3-month clinical follow-up of 34 patients with moderate-to-severe acne treated with oral isotretinoin, GAGS-HFUS score was confirmed as a strong independent predictor of unfavorable treatment response. When both clinical variables and imaging parameters were considered, multivariate Poisson regression analysis demonstrated that GAGS-HFUS—and not IGA—served as an independent predictor of unfavorable treatment response. This finding may be attributed to the advantages of HFUS, which improves the detection of subclinical fistulas and nodules or cysts that cannot be identified through physical examination. Compared to SOS-Acne, the integration of the GAGS facial region into the GAGS-HFUS significantly reduces examination time, making routine clinical application feasible. Furthermore, the AUROC of the GAGS-HFUS was significantly higher than those of the IGA and SOS-Acne grading systems, further demonstrating the superior predictive performance of the GAGS-HFUS as an independent predictor of unfavorable treatment response.
Considering these findings, the clinical implications of this study are as follows. First, this study developed the GAGS-HFUS as a novel ultrasound-based acne scoring system. Leveraging HFUS, it can detect subclinical lesions that are not identifiable through physical examination or visual inspection. By incorporating the GAGS facial subregion classification, the GAGS-HFUS achieves a balance between accuracy and practicality for clinical application. Second, this study confirmed that GAGS-HFUS score was a strong independent predictor of unfavorable treatment response in patients with moderate-to-severe acne, helping guide appropriate therapeutic strategies and reduce the risk of poor outcomes.
However, this study has certain limitations. First, although the GAGS-HFUS provides a more comprehensive evaluation of acne, it may not sufficiently capture disease severity when multiple lesions are confined to a few specific locations. Second, "gold standard" histological studies are of limited utility for acne due to the potential for scarring in the facial area, which is highly visible. As a result, HFUS imaging findings could not be reliably compared with pathology results. Third, clinical follow-up was limited to patients with moderate-to-severe acne; a larger and longer-term research cohort is needed to further demonstrate the predictive value of the GAGS-HFUS for unfavorable treatment response in patients with acne.
This study proposes the GAGS-HFUS acne scoring system. Integrating the GAGS facial subregion classification with HFUS improves the system’s accuracy while maintaining practicality. Additionally, GAGS-HFUS score was identified as a strong independent predictor of unfavorable treatment response in patients with moderate-to-severe acne. Overall, the GAGS-HFUS represents a valuable tool for objective clinical decision-making and treatment optimization.
Key point
This study was the first to develop the Global Acne Grading System with high-frequency ultrasound (GAGS-HFUS) acne scoring system, which balances accuracy and practicality in clinical application. The GAGS-HFUS serves as an independent predictor of unfavorable treatment response in patients with moderate-to-severe acne, highlighting its indispensable role in the strategic management of acne. The GAGS-HFUS holds significant promise for guiding appropriate therapeutic strategies, reducing the high recurrence rate of acne, and preventing the formation of deep disfiguring scars.
Footnotes
Author Contributions
Conceptualization: Yang L, Hu B, Su Y, Sun D. Data acquisition: Yang L, Su Y. Data analysis or interpretation: Yang L, Guo J, Sun D. Drafting of the manuscript: Yang L. Critical revision of the manuscript: Hu B, Guo J, Su Y, Sun D. Approval of the final version of the manuscript: all authors.
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
This study has received funding by National Natural Science Foundation of China (Grant No. 82371973, 82171954), International Cooperation Program of Shanghai Municipal Committee for Science and Technology (Grant No. 23410713000), the research project of Shanghai sixth people's hospital (Grant No. Ynyq202303, Ynhg202128), Shanghai Key Clinical Research Center Fund (Grant No. 2023ZZ02006).
Supplementary Material
Key comparative features of the three acne severity assessment systems: GAGS-HFUS, SOS-Acne, and SSSA (https://doi.org/10.14366/usg.25077).
Demographic patient characteristics (https://doi.org/10.14366/usg.25077).
Patients distribution across various acne grading systems (https://doi.org/10.14366/usg.25077).
Comparison between GAGS-HFUS and IGA evaluation at baseline (https://doi.org/10.14366/usg.25077).
Comparison between GAGS-HFUS and SSSA evaluation at baseline (https://doi.org/10.14366/usg.25077).
The Global Acne Grading System (GAGS) (https://doi.org/10.14366/usg.25077).
The superb microvascular imaging (SMI), three-dimensional SMI (3D SMI) and vascular index (VI) of acne lesions on high-frequency ultrasound image (https://doi.org/10.14366/usg.25077).
References
- 1.Reynolds RV, Yeung H, Cheng CE, Cook-Bolden F, Desai SR, Druby KM, et al. Guidelines of care for the management of acne vulgaris. J Am Acad Dermatol. 2024;90:1006. doi: 10.1016/j.jaad.2023.12.017. [DOI] [PubMed] [Google Scholar]
- 2.Witkam W, Dal Belo SE, Pourhamidi S, Raynaud E, Moreau M, Aguilar L, et al. The epidemiology of acne vulgaris in a multiethnic adolescent population from Rotterdam, the Netherlands: a cross-sectional study. J Am Acad Dermatol. 2024;90:552–560. doi: 10.1016/j.jaad.2023.10.062. [DOI] [PubMed] [Google Scholar]
- 3.Vallerand IA, Lewinson RT, Parsons LM, Lowerison MW, Frolkis AD, Kaplan GG, et al. Risk of depression among patients with acne in the U.K.: a population-based cohort study. Br J Dermatol. 2018;178:e194–e195. doi: 10.1111/bjd.16099. [DOI] [PubMed] [Google Scholar]
- 4.Hughes O, Bewley A. Is it really ever 'just acne'?: considering the psychodermatology of acne. Br J Dermatol. 2023;189:i11–i16. doi: 10.1093/bjd/ljad251. [DOI] [PubMed] [Google Scholar]
- 5.Ibrahim S, Osman B, Awaad RM, Abdoon I. Acne vulgaris relapse in sudanese patients treated with oral isotretinoin: rate and predictive factors. J Multidiscip Healthc. 2023;16:839–849. doi: 10.2147/JMDH.S405509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Blasiak RC, Stamey CR, Burkhart CN, Lugo-Somolinos A, Morrell DS. High-dose isotretinoin treatment and the rate of retrial, relapse, and adverse effects in patients with acne vulgaris. JAMA Dermatol. 2013;149:1392–1398. doi: 10.1001/jamadermatol.2013.6746. [DOI] [PubMed] [Google Scholar]
- 7.Picardo M, Cardinali C, La Placa M, Lewartowska-Bialek A, Lora V, Micali G, et al. Efficacy and safety of N-acetyl-GED-0507-34-LEVO gel in patients with moderate-to severe facial acne vulgaris: a phase IIb randomized double-blind, vehicle-controlled trial. Br J Dermatol. 2022;187:507–514. doi: 10.1111/bjd.21663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Luo Y, Wang J, Gao Y, Wang Y, Zhang S, Liu Z, et al. Value of high-frequency ultrasound in the treatment of moderate and severe acne vulgaris. Skin Res Technol. 2022;28:833–839. doi: 10.1111/srt.13208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Errichetti E, Pietkiewicz P, Bhat YJ, Salwowska N, Szlazak P, Stinco G. Diagnostic accuracy of ultraviolet-induced fluorescence dermoscopy in non-neoplastic dermatoses (general dermatology): a multicentric retrospective comparative study. J Eur Acad Dermatol Venereol. 2025;39:97–108. doi: 10.1111/jdv.19795. [DOI] [PubMed] [Google Scholar]
- 10.Cretu S, Papachatzopoulou E, Dascalu M, Salavastru CM. The role of in vivo reflectance confocal microscopy for the management of acne: a systematic review. J Eur Acad Dermatol Venereol. 2023;37:2428–2439. doi: 10.1111/jdv.19327. [DOI] [PubMed] [Google Scholar]
- 11.Fuchs CS, Andersen AJ, Ardigo M, Philipsen PA, Haedersdal M, Mogensen M. Acne vulgaris severity graded by in vivo reflectance confocal microscopy and optical coherence tomography. Lasers Surg Med. 2019;51:104–113. doi: 10.1002/lsm.23008. [DOI] [PubMed] [Google Scholar]
- 12.Wortsman X, Wortsman J. Clinical usefulness of variable-frequency ultrasound in localized lesions of the skin. J Am Acad Dermatol. 2010;62:247–256. doi: 10.1016/j.jaad.2009.06.016. [DOI] [PubMed] [Google Scholar]
- 13.Shan D, Wu N, Wang Q, Ren W, Zhu A, Wang L, et al. Value of pseudopod sign on high-frequency ultrasound in predicting the pathological invasion of extramammary Paget's disease lesions. J Eur Acad Dermatol Venereol. 2022;36:1235–1245. doi: 10.1111/jdv.18104. [DOI] [PubMed] [Google Scholar]
- 14.Wortsman X, Claveria P, Valenzuela F, Molina MT, Wortsman J. Sonography of acne vulgaris. J Ultrasound Med. 2014;33:93–102. doi: 10.7863/ultra.33.1.93. [DOI] [PubMed] [Google Scholar]
- 15.Krajewski PK, Jfri A, Ochando-Ibernon G, Martorell A. Ultrasonographic railway sign in tunnels as a new independent risk factor of adalimumab failure in hidradenitis suppurativa. J Am Acad Dermatol. 2023;88:732–734. doi: 10.1016/j.jaad.2022.08.064. [DOI] [PubMed] [Google Scholar]
- 16.Wang J, Luo Y, Liu J, Zhu Q, Wang Y, Jin H. High-frequency ultrasonography and scoring of acne at 20 and 50 MHz. J Eur Acad Dermatol Venereol. 2020;34:e743–e745. doi: 10.1111/jdv.16584. [DOI] [PubMed] [Google Scholar]
- 17.Doshi A, Zaheer A, Stiller MJ. A comparison of current acne grading systems and proposal of a novel system. Int J Dermatol. 1997;36:416–418. doi: 10.1046/j.1365-4362.1997.00099.x. [DOI] [PubMed] [Google Scholar]
- 18.Barrea L, Donnarumma M, Cacciapuoti S, Muscogiuri G, De Gregorio L, Blasio C, et al. Phase angle and Mediterranean diet in patients with acne: two easy tools for assessing the clinical severity of disease. J Transl Med. 2021;19:171. doi: 10.1186/s12967-021-02826-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Thiboutot DM, Dreno B, Abanmi A, Alexis AF, Araviiskaia E, Barona Cabal MI, et al. Practical management of acne for clinicians: an international consensus from the Global Alliance to Improve Outcomes in Acne. J Am Acad Dermatol. 2018;78:S1–S23. doi: 10.1016/j.jaad.2017.09.078. [DOI] [PubMed] [Google Scholar]
- 20.Fu Z, Zhang J, Lu Y, Wang S, Mo X, He Y, et al. Clinical applications of superb microvascular imaging in the superficial tissues and organs: a systematic review. Acad Radiol. 2021;28:694–703. doi: 10.1016/j.acra.2020.03.032. [DOI] [PubMed] [Google Scholar]
- 21.Spiesecke P, Schmidt J, Peters R, Fischer T, Hamm B, Lerchbaumer MH. Assessment of quantitative microflow vascular index in testicular cancer. Eur J Radiol. 2024;176:111513. doi: 10.1016/j.ejrad.2024.111513. [DOI] [PubMed] [Google Scholar]
- 22.Liu FJ, Ci WP, Cheng Y. Clinical study of carotid superb microvascular imaging in evaluating the activity of Takayasu's arteritis. Front Cardiovasc Med. 2023;10:1051862. doi: 10.3389/fcvm.2023.1051862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wang Y, Feng Z, Ma B, Li S, Guo W, Song J, et al. Diagnosis of lupus mastitis via multimodal ultrasound: case description. Quant Imaging Med Surg. 2024;14:3235–3238. doi: 10.21037/qims-23-1279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Daniele SG, Kim SR, Grada A, Moore AY, Suozzi KC, Bunick CG. Truncal acne and dcarring: a comprehensive review of current medical and cosmetic approaches to treatment and patient management. Am J Clin Dermatol. 2023;24:199–223. doi: 10.1007/s40257-022-00746-4. [DOI] [PubMed] [Google Scholar]
- 25.Martorell A, Alfageme Roldan F, Vilarrasa Rull E, Ruiz-Villaverde R, Romani De Gabriel J, Garcia Martinez F, et al. Ultrasound as a diagnostic and management tool in hidradenitis suppurativa patients: a multicentre study. J Eur Acad Dermatol Venereol. 2019;33:2137–2142. doi: 10.1111/jdv.15710. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Key comparative features of the three acne severity assessment systems: GAGS-HFUS, SOS-Acne, and SSSA (https://doi.org/10.14366/usg.25077).
Demographic patient characteristics (https://doi.org/10.14366/usg.25077).
Patients distribution across various acne grading systems (https://doi.org/10.14366/usg.25077).
Comparison between GAGS-HFUS and IGA evaluation at baseline (https://doi.org/10.14366/usg.25077).
Comparison between GAGS-HFUS and SSSA evaluation at baseline (https://doi.org/10.14366/usg.25077).
The Global Acne Grading System (GAGS) (https://doi.org/10.14366/usg.25077).
The superb microvascular imaging (SMI), three-dimensional SMI (3D SMI) and vascular index (VI) of acne lesions on high-frequency ultrasound image (https://doi.org/10.14366/usg.25077).