CT and MRI features of SAPHO syndrome: a report of 6 cases

Abstract

Objective

To investigate the CT and MRI features of synovitis, acne, pustulosis, hyperostosis, and osteitis (SAPHO) syndrome, aiming to improve the recognition of this disease by clinicians and radiologists and to reduce misdiagnosis and missed diagnosis.

Methods

Clinical and imaging data of 6 patients with confirmed SAPHO syndrome hospitalized in our hospital from November 2011 to August 2025 were retrospectively analyzed. The cohort included 4 males and 2 females, aged 26–70 years (mean 44.5 ± 19.2 years). All patients underwent both CT and MRI examinations. The distribution of lesions and the imaging features on CT and MRI were analyzed.

Results

The anterior chest wall was involved in all 6 patients, including costosternal joint involvement in 6 (100%) and sternoclavicular joint involvement in 4 (66.7%). The typical “bull head sign” was observed in 4 patients. Among a total of 144 vertebrae examined in the 6 patients, CT detected 34 affected vertebrae (23.6%) and MRI detected 51 affected vertebrae (35.4%), predominantly in the thoracic spine. Characteristic CT signs included the “bright corner sign”, “half-circle sign”, and “round sign”. Characteristic MRI signs included the “kissing sign” (present in 80.0% of continuously involved vertebrae) and bone marrow edema. CT was superior in detecting osteosclerosis, while MRI was superior in detecting bone marrow edema. Bilateral sacroiliac joint involvement was present in 3 cases, and the hip joint was involved in 1 case. Three cases had been misdiagnosed as ankylosing spondylitis, psoriatic arthritis, or spinal metastases.

Conclusion

Skeletal lesions of SAPHO syndrome predominantly involve the anterior chest wall (the “bull head sign” is characteristic) and the thoracic spine (with continuous vertebral involvement, showing the “bright corner sign” and “kissing sign”). The combined use of CT and MRI facilitates comprehensive evaluation of lesions, and identification of skin lesions can improve the recognition of this disease.

Introduction

Synovitis, acne, pustulosis, hyperostosis, and osteitis (SAPHO) syndrome is a rare autoinflammatory syndrome first named by Chamot et al. [1] in 1987. Its pathogenesis remains incompletely understood [2–4]. The most common symptoms are anterior chest wall pain (mainly involving the sternum and sternoclavicular joints), followed by cutaneous manifestations, peripheral arthritis, and sacroiliac pain [5, 6]. The diagnosis of SAPHO syndrome is primarily based on clinical and radiological findings. Due to insufficient awareness of early manifestations by clinicians, it often leads to missed diagnosis; in advanced stages, it can result in high disease burden and disability. Diagnostic delay ranges from 3 to 10 years [7, 8]. In the present series, one patient was initially misdiagnosed with metastatic tumor based on clinical and bone scan findings. The disease is frequently misdiagnosed as spinal malignancy [9–11] or infectious osteomyelitis/arthritis [12, 13]. Given its rarity and variability, diagnosis remains challenging. This study retrospectively analyzed the imaging data of 6 patients with SAPHO syndrome hospitalized in our institution from 2011 to 2025, summarizing the lesion distribution and CT and MRI features, with the aim of improving the recognition of this disease among clinicians and radiologists.

Materials and methods

Clinical data

Patients diagnosed with SAPHO syndrome in our hospital from November 2011 to August 2025 were collected. During the same period, 8 patients with SAPHO syndrome were treated; 1 was excluded due to incomplete imaging data, and 1 was lost to follow-up, leaving 6 patients with complete data for final analysis. There were 4 males and 2 females, aged 26–70 years, mean (44.5 ± 19.2) years. Disease duration ranged from 1 month to 18 years. Three patients presented with palmoplantar pustulosis as the first symptom, among whom one developed explosive palmoplantar acne triggered by seafood and beer consumption, with skin lesions appearing 5–8 years earlier than osteoarticular manifestations. One patient presented only with osteoarticular involvement without skin lesions; the patient first consulted the oncology department and was later referred to rheumatology. All patients had anterior chest wall pain and swelling, 5 had concomitant thoracic and back pain, and 2 had lumbosacral and hip joint pain. Laboratory findings: elevated erythrocyte sedimentation rate (ESR, 42–75 mm/h) in 5 patients, elevated C-reactive protein (CRP, 28.7–37 mg/L) in 5, and positive human leukocyte antigen B27 (HLA-B27) in 2. Rheumatoid factor (RF) was negative in all patients. Symptoms markedly improved after treatment with corticosteroids and immunosuppressants. Baseline patient characteristics are shown in Table 1.

Table 1.

Baseline and clinical characteristics of 6 patients with SAPHO syndrome

Characteristic	Data
Number of cases	6
Male/Female	4/2
Age (years, mean ± SD)	44.5 ± 19.2
Disease duration range	1 month – 18 years
Skin lesions preceding osteoarticular involvement (n)	3
Osteoarticular involvement only (n)	1
Anterior chest wall pain/swelling (n)	6
Elevated ESR (n)	5
Elevated CRP (n)	5
HLA-B27 positive (n)	2
Misdiagnosis (n)	3

Imaging methods

All patients underwent both CT and MRI, but the examinations were not performed under a unified research protocol and were part of routine clinical workup. CT was performed using a GE Brightspeed 16-slice helical CT scanner and a Siemens Somatom Force dual-source CT scanner. MRI was performed using a GE Signa Excite 0.35 T permanent magnet MRI scanner (with a 4-channel phased-array coil) and a GE Discovery 750 3.0 T scanner (with an 8-channel phased-array body coil). Scan ranges included the cervical, thoracic, and lumbar spine. Transverse scans were obtained with reconstruction of sagittal and coronal images; the sacroiliac joints and hip joints were also scanned. Sequences included transverse, sagittal, and coronal T1WI, T2WI, and STIR/IDEAL sequences. For contrast-enhanced scans, Gd-DTPA (0.2 mL/kg) was manually injected via the antecubital vein.

Inclusion and exclusion criteria

Diagnosis of SAPHO syndrome followed the 2012 revised criteria [3, 14]; fulfillment of any one of the following four conditions was diagnostic: (1) osteoarticular manifestations associated with acne conglobata, acne fulminans, or hidradenitis suppurativa; (2) osteoarticular manifestations associated with palmoplantar pustulosis; (3) hyperostosis (involving the anterior chest wall, limb bones, or spine) with or without skin lesions; (4) chronic recurrent multifocal osteomyelitis (involving the axial or peripheral skeleton) with or without skin lesions. Exclusion criteria were infectious osteitis, bone tumors, infectious palmoplantar abscess, and non-inflammatory bony condensation lesions.

Imaging staging definition

SAPHO syndrome manifestations were staged as early: only entheseal edema and bone marrow edema (BME), with no or minimal bone erosion and no sclerosis; intermediate: definite bone erosion, focal sclerosis, without ankylosis; advanced: marked osteosclerosis and hyperostosis, joint space narrowing or obliteration, bony fusion.

Imaging sign definitions

The specific imaging signs observed in this study were defined as follows:

Bull head sign (coronal CT reconstruction): Symmetrical bone destruction, sclerosis, hyperostosis, and bony fusion of the bilateral sternoclavicular joints and first costosternal joints; the sternal manubrium resembles a bull’s head, and the bilateral joints resemble bull’s horns.

Bright corner sign (sagittal CT): Focal osteosclerosis at the superior or inferior anterior corner of a vertebral body with clear margins.

Half-circle sign (sagittal CT): Semi-circular bone defects or sclerosis at entheseal sites along the anterior margins of multiple consecutive vertebral bodies.

Round sign (sagittal CT): Osteophyte formation at the anterosuperior and anteroinferior corners of adjacent vertebral bodies that come into contact, forming a round-like image.

Kissing sign (sagittal STIR/T2WI): Continuous, symmetrical bone marrow edema or sclerosis signals at the entheseal sites on the anterior margins of adjacent vertebral bodies, appearing as “kissing” lesions.

Image evaluation

All images were independently reviewed by one chief physician (20 years of experience in musculoskeletal imaging) and one associate chief physician (15 years of experience). Both reviewers were aware of the diagnosis of SAPHO but were blinded to clinical details (disease course, skin manifestations). Differences were resolved by consensus. The consensus rate for the primary signs (bull head sign, kissing sign) was 100% (6/6), and for the bright corner sign it was 83.3% (5/6). Due to the small sample size, a formal interobserver κ test was not performed (Fig. 1).

Fig. 1 .

SAPHO syndrome in a 28-year-old male with sudden onset of back pain and hand pustulosis. a Coronal CT reconstruction of the sternoclavicular joints shows bilateral bone blurring and a small focal bone destruction in the left clavicle (arrow). b Sagittal CT of the sternal angle shows bone blurring and small focal bone destruction at the manubriosternal junction (arrow). c-d Sagittal T1WI (c) and T2WI (d) of the lumbar spine show patchy long T1 and long T2 signal at the anterosuperior corner of L4 (short arrow) and in the T10 vertebral body and posterior elements (long arrows). e STIR image shows high signal (SME, bone marrow edema) at the anterosuperior corner of L4 (short arrow) and in the T10 body and posterior elements (long arrows). f Follow-up DR at 3 years and 7 months shows a “bright corner sign” at the anterosuperior corners of T12 and L1 (long arrows), and a bone bridge between the anteroinferior corner of L3 and the anterosuperior corner of L4, i.e., the “half-circle sign” (double arrowheads). g STIR image shows patchy high signal (SME) in the bilateral ilia (arrows). h Follow-up STIR image at 2 years and 5 months shows patchy high signal (SME) in the bilateral sacral bones (arrows)

Results

Anterior chest wall changes

All 6 patients showed variable involvement of the anterior chest wall (on a per-patient basis): sternoclavicular joints in 4/6 (66.7%), costosternal joints in 6/6 (100%), and sternal angle in 4/6 (66.7%). Early CT showed blurring of the sternoclavicular joints, manubriosternal junction, costoclavicular ligaments and their entheses with small-focal bone erosion and adjacent sclerosis; MRI showed ligamentous edema, BME, and surrounding soft tissue swelling at the above sites. In intermediate/late stages, CT showed irregular bone destruction, uneven margins, sclerosis, joint space narrowing, and eventual bony fusion at the sternoclavicular joints, costosternal joints, and manubriosternal junction. Four patients (4/6, 66.7%) had involvement of both costosternal joints and the manubrium; coronal CT reconstructions showed bone destruction and sclerosis at the superior margin of the manubrium and bilateral first costosternal joints, forming the classic bull head sign (Fig. 2a). On MRI, early lesions showed long T1, long T2 signals and high signal on STIR/IDEAL (BME); in the hyperostotic/sclerotic stage, they showed long T1, short T2 signals and low signal on STIR/IDEAL.

Fig. 2 .

SAPHO syndrome in a 65-year-old female patient. The patient presented with vertebral abnormalities at T4–6 and T8–10 and was admitted to the oncology department with suspected metastatic disease. a Coronal CT reconstruction of the chest shows sclerosis, mild destruction, and joint space narrowing of the bilateral first costosternal and sternoclavicular joints, forming the classic “bull head sign” (arrows). b Sagittal CT reconstruction of the thoracic spine shows sclerosis of multiple vertebral corners, the “round sign” at the anterior margins of T5–6 (short red arrows), the “bright corner sign” and “half-circle sign” at the anteroinferior corner of T6 (long red arrows), narrowing of the T5–6 disc space, and the “kissing sign” at the anterior margins of T8–9 and T9–10 (double short red arrows). c–f Sagittal MRI of the thoracic spine shows sclerosis of multiple vertebral corners and vertebral bone marrow edema (SME). Patchy long T1 and long T2 signal is seen at T4–6 c–d. IDEAL images show high signal (SME) with the “kissing sign” at the anterior margins of T4–5 and T5–6 (red arrows in e). On the IDEAL image, SME at the anterior margin of T8–9 shows the “half-circle sign” (red arrow in f)

Spinal imaging findings

For the 6 patients, each patient had 7 cervical, 12 thoracic, and 5 lumbar vertebrae, total 24 vertebrae per patient, i.e., 6 × 24 = 144 vertebrae. The following counts are intended only to describe lesion distribution and not to treat vertebrae as independent analytic units.

CT findings

A total of 34 affected vertebrae were detected (34/144, 23.6%), including thoracic in 25 (25/34, 73.5%), lumbar in 8 (8/34, 23.5%), and cervical in 1 (1/34, 2.9%). CT features: Among the 34 affected vertebrae, only the vertebral corners were involved in 23 (23/34, 67.6%), both corners and adjacent endplates in 11 (11/34, 32.4%) – overlapping with the former. Continuously involved vertebrae (≥ 2 consecutive vertebrae) were present in 4/6 patients (66.7%), accounting for 15 vertebrae (15/144, 10.4%; 44.1% of affected vertebrae). Characteristic signs: bright corner sign in 17 vertebrae (17/34, 50.0%), half-circle sign in 9 (9/34, 26.5%), round sign in 10 (10/34, 29.4%). Additionally, bone defects were seen in 19 (19/34, 55.9%) and osteosclerosis in 31 (31/34, 91.2%).

MRI findings

A total of 51 affected vertebrae were detected (51/144, 35.4%), including thoracic in 32 (32/51, 62.7%), lumbar in 14 (14/51, 27.5%), and cervical in 5 (5/51, 9.8%). MRI features: Among the 51 affected vertebrae, only vertebral corners were involved in 17 (17/51, 33.3%), both corners and adjacent endplates in 22 (22/51, 43.1%), fatty deposition signal in 14 (14/51, 27.4%); among active lesions, BME was seen in 25 (25/51, 49.0%), cortical defect in 23 (23/51, 45.1%), osteosclerosis in 22 (22/51, 43.1%). Continuous involvement: 4/6 patients (66.7%) with 15 continuously involved vertebrae, among which 12/15 (80.0%) showed the kissing sign.

Comparison between CT and MRI

It must be emphasized that CT and MRI were not performed under a standardized protocol (different equipment, time periods, sequences) and reflect different pathological substrates (CT: cortical bone/sclerosis/destruction; MRI: bone marrow edema/fatty deposition). The numerical differences above should not be directly interpreted as superiority of one modality over the other, but rather reflect their complementary roles.

Intervertebral disc changes

Abnormalities were found in 16 intervertebral discs (denominator 138 disc levels, 16/138, 11.6%). All affected disc levels showed mild narrowing. On MRI, early stages showed high signal on T2WI and fat-suppressed images; intermediate/late stages showed low signal.

MRI findings of sacroiliac joints and hip joints

Among the 6 patients, 3 (3/6, 50.0%) had normal sacroiliac joints; 3 (3/6, 50.0%) had bilateral sacroiliac joint involvement. In one patient, initial examination showed BME on the iliac side; 3-year follow-up MRI showed resolution of iliac BME but new BME on the sacral side. Hip joint: 5/6 (83.3%) were uninvolved, and 1/6 (16.7%) showed involvement of the left greater trochanter (Table 2).

Table 2 .

Summary of main imaging signs in 6 patients with SAPHO syndrome

Imaging sign	Number of patients	Proportion (%)
Anterior chest wall involvement	6	100
- Costosternal joint involvement	6	100
- Sternoclavicular joint involvement	4	66.7
- Sternal angle involvement	4	66.7
Bull head sign	4	66.7
Spinal involvement	6	100
- Continuous vertebral involvement (≥ 2 consecutive vertebrae)	4	66.7
- Kissing sign (MRI)	4	66.7
- Bright corner sign (CT)	4	66.7
Sacroiliac joint involvement	3	50.0
Hip/greater trochanter involvement	1	16.7

Discussion

Clinical and epidemiological features

SAPHO syndrome is a rare chronic autoimmune non-specific inflammatory disease whose underlying causes include immune dysfunction, infection [15], and genetic susceptibility [16], involving both skin and osteoarticular tissues [4, 6, 17–19], with an estimated prevalence of approximately 1/10,000 [6, 20]. The peak age of onset is 40–60 years; cases older than 60 years are rare, and the female-to-male ratio is about 2.2:1 [21]. The disease course is often chronically relapsing and remitting. Approximately 40%–68% of patients have skin manifestations that precede osteoarticular symptoms, with an interval that can reach 35–40 years; in 32%–60% of patients, osteoarticular symptoms appear first [6, 22]. In the present series, the male-to-female ratio was 4:2, skin lesions preceded osteoarticular involvement in 3 patients, and one patient had only osteoarticular symptoms without skin manifestations, which is generally consistent with the literature.

Imaging features of the anterior chest wall

SAPHO syndrome frequently involves the axial skeleton; the prevalence of anterior chest wall involvement ranges from 60 to 95% [23–25]. All 6 patients in this series had anterior chest wall bone changes, with involvement of the anterior superior chest wall being the most characteristic. Typical affected structures include the sternoclavicular joints, costosternal joints, and costoclavicular ligaments. Bone and joint changes in this region can be divided into three phases: soft tissue swelling, bone erosion, and new bone formation accompanied by hyperostosis and osteosclerosis [26]. The disease initially presents as enthesitis of the costoclavicular ligament; plain radiographs are often negative, while CT can show thickening of the costoclavicular ligament and blurring with small focal destruction at its entheses; MRI can demonstrate ligamentous edema, soft tissue swelling, and BME even earlier. When bilateral sternoclavicular and costosternal joints show bone destruction, hyperostosis, sclerosis, and ankylosis, the manubrium appears like a bull’s head, and the joints like bull’s horns – the bull head sign [27, 28]. This sign was observed in 4 of our patients and represented the most characteristic imaging finding in this series.

Spinal imaging features

The spine is the second most commonly involved site, with reported involvement in approximately 32%–52% of patients [23, 29, 30]; the thoracic spine is the most frequent location [31, 32]. All 6 patients in this series had thoracic spine involvement, consistent with the literature. Spinal lesions develop in stages: early, MRI shows vertebral BME; as inflammation spreads, CT may show the bright corner sign; with continuous vertebral involvement, CT may show the half-circle sign and round sign, while MRI may show the kissing sign. In our series, continuous BME at the anterior margins of vertebral bodies with round or semicircular involvement and sclerosis along ligamentous attachments were important clues to SAPHO syndrome.

Value of CT and MRI

In this study, MRI detected more abnormal vertebrae (51) than CT (34). This may be due to the higher sensitivity of MRI for bone marrow edema, but more importantly to the non-standardized examination conditions, different disease stages, and the fact that the two modalities reflect different pathological aspects (edema vs. sclerosis). CT is excellent for evaluating bony changes (osteosclerosis, hyperostosis, bony fusion), while MRI is superior for assessing the extent of active inflammation (bone marrow edema, enthesitis, soft tissue swelling) and follow-up. Combined use of CT and MRI provides more comprehensive information for clinical management [6, 33–35].

Differential diagnosis

Three patients in this series were misdiagnosed as ankylosing spondylitis, psoriatic arthritis, or spinal metastases. Key differentiating points: SAPHO typically affects patients aged 40–60 years, often involves the sternoclavicular joints and thoracolumbar spine early, with mild or absent sacroiliitis; characteristic imaging signs (bull head sign, bright corner sign, kissing sign) and skin manifestations help distinguish it. Ankylosing spondylitis predominantly affects young males (16–25 years), almost always begins in the sacroiliac joints, ascends the spine, and leads to “bamboo spine” in late stages, without sternoclavicular joint involvement. Psoriatic arthritis is heterogeneous, with spinal involvement mainly in the cervical spine. Spinal metastases usually have a history of primary malignancy, tend to involve the posterior vertebral elements, and do not involve the sternoclavicular joints.

Treatment

Treatment options for SAPHO syndrome are diverse, including nonsteroidal anti-inflammatory drugs (NSAIDs), bisphosphonates, and conventional or biologic disease-modifying antirheumatic drugs [21]. There is no consensus on the treatment of SAPHO syndrome, and research on biologic agents targeting specific molecular targets is increasing [36]. In our patients, NSAIDs as well as treatments promoting blood circulation and removing stasis, regulating bone metabolism, traditional Chinese medicine, and acupuncture significantly improved symptoms.

Limitations

Limitations (1) Equipment heterogeneity: This study spanned 14 years, with different CT/MRI equipment and magnetic field strengths (0.35 T to 3.0 T; 16-slice to dual-source CT); scanning parameters were not strictly standardized. This may overestimate or underestimate the detection of lesions by MRI (higher field strengths more sensitively detect BME). Nevertheless, all MRI examinations in patients with BME were clinically useful, and sclerotic lesions were mainly seen on CT in early cases, so qualitative comparisons are still informative. Prospective validation with uniform equipment is recommended. (2) Selection bias: Only 6 patients with complete data were included; this is a single-center retrospective case series. (3) Reader reproducibility: Although consensus rates for the main signs were high, a formal κ test was not performed.

Summary and outlook

SAPHO syndrome is a chronic, relapsing disease. Our data show that skeletal lesions predominate in the anterior chest wall (bull head sign is characteristic), the spine (mainly thoracic, with continuous vertebral involvement, bright corner sign, and kissing sign), and the sacroiliac joints (bilateral, mild). Familiarity with these imaging findings, together with recognition of skin lesions, facilitates appropriate consideration of this entity in the differential diagnosis. In patients with unexplained back pain and continuous vertebral involvement, SAPHO should be suspected, and early MRI of the sternoclavicular joints and sacroiliac joints is recommended. Future multicenter prospective blinded diagnostic studies are needed to rigorously evaluate the true diagnostic performance of these signs.

Authors’ contributions

Kaixi XU and Yun MENG: Co-authored Paper Writing Yun MENG: data analysis and data collection Kaixi Xu and Taosheng ZUO: experimental design.

Funding

Management Measures for Science and Technology Awards of Jiangsu Provincial Hospital of Traditional Chinese Medicine (Lianyungang Hospital). Document No.: Lian Zhong Yi [2024] No. 20.

Data availability

The data that support the findings of this study are available from the authors or the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This retrospective study adhered to the Declaration of Helsinki, was approved by the Ethics Committee of Lianyungang Affiliated Hospital of Nanjing University of Chinese Medicine, and obtained informed consent from participants. Approval number: Ethics Review (KY)-12.

Consent for publication

All authors and the patient have approved the manuscript for publication.

Competing interests

The authors declare no competing interests.