SAPHO syndrome associated spondylitis

Abstract

The concept of synovitis, acne, pustulosis, hyperostosis, osteitis (SAPHO) syndrome has been well clarified, after Chamot et al. suggested this peculiar disorder in 1987. The most commonly affected site in SAPHO syndrome is the anterior chest, followed by the spine. However, the clinical course and taxonomic concept of SAPHO spinal lesions are poorly understood. This study was performed to analyze: (1) the detailed clinical course of spinal lesions in SAPHO syndrome, and (2) the relationship between SAPHO syndrome with spinal lesions and seronegative spondyloarthropathy. Thirteen patients with spondylitis in SAPHO syndrome were analyzed. The features of spinal lesions were a chronic onset with a slight inflammatory reaction, and slowly progressing non-marginal syndesmophytes at multi spinal levels, besides the coexistence of specific skin lesions. SAPHO syndrome, especially spinal lesions related to palmoplantar pustulosis, can be recognized as a subtype of seronegative spondyloarthropathy.

Introduction

The relationship between palmoplantar pustulosis (PPP) and bony or arthritic lesions has long been understood. In 1967, Sasaki first described a case of aseptic hyperostosis of the clavicles associated with PPP [24]. In 1981, Sonozaki proposed the concept of pustulotic arthro-osteitis (PAO) [26]. In 1987, Chamot et al. suggested the concept of synovitis, acne, pustulosis, hyperostosis, osteitis (SAPHO) syndrome to collectively describe peculiar bone, joint, and skin lesions including PAO [4]. Nowadays, the concept of SAPHO syndrome has been widely accepted. SAPHO syndrome is defined as the presence of any one of the following four: osteoarticular manifestation of severe acne, osteoarticular manifestation of PPP, hyperostosis, or chronic recurrent multifocal osteomyelitis [11]. The spine is one of the most commonly affected sites, whose incidence is 32–52% in SAPHO syndrome [11, 21, 29]. However, the detailed clinical course of spinal lesions in SAPHO syndrome is not clearly understood. Spinal lesions in SAPHO syndrome generally show a good prognosis and rarely cause neurological deterioration [11, 21, 28, 29]. SAPHO syndrome may be assumed as a subtype of seronegative spondyloarthropathy (SNSA) regarding its relation with axial involvement, inflammatory bowel disease, and HLAB27. However, the taxonomic concept is still controversial. The purpose of this study was to analyze: (1) the clinical and radiological course of spondylitis in SAPHO syndrome, and (2) the relationship between spondylitis in SAPHO syndrome and SNSA.

Patients and methods

Thirteen patients with spondylitis in SAPHO syndrome were analyzed. All were treated in our hospital from 1984 to 2007, met the inclusion and exclusion criteria of SAPHO syndrome [11], and were complicated with spondylitis. Suspicious cases with infectious condition, e.g., infectious osteomyelitis, arthritis, or spondylitis were excluded. Two cases of SAPHO syndrome reported previously as destructive spondylitis were included in 13 cases [28]. We collected clinical information from the medical records and analyzed these 13 patients retrospectively. For each patient, the following were recorded: background data regarding sex, age at SAPHO diagnosis, time from initial symptom to diagnosis, family history, and presence of HLAB27. As for spinal lesions, the initial symptom of spondylitis, onset pattern (classification of Guri [10] and Kulowski [19]), laboratory data including erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), the white blood cell (WBC) count, and rheumatoid factor (RF), imaging findings including Xp (classification of McEween et al. [22]), magnetic resonance imaging (MRI), computerized tomography (CT), and bony scintigraphy were studied. Whether these patients met the criteria of ESSG [6] or Amor [2] regarding SNSA was analyzed. We also studied extra-spinal lesions: the onset and course of the skin lesions and extent of osteoarthritic lesions using Xp, CT, or ^99mTc-MDP bony scintigraphy. The therapy for spinal lesions and clinical course were also investigated.

Results

Patients’ backgrounds

The characteristics of the 13 patients are summarized in Table 1. Twelve female and one male were included. The mean age at diagnosis was 51.8 years old. The duration between the initial symptom and diagnosis was 6.4 years (4 months to 13 years). The mean follow-up period after diagnosis was 2.4 years (3 months to 9 years). No patient had a family history of skin disease, SAPHO syndrome, or SNSA. A family history of collagen disease was only identified in case 7, whose father had systemic lupus erythematosus (SLE). HLA examinations performed in eight patients showed that all were HLA-B27 negative.

Table 1.

Clinical features

Case	Age of diagnosis (years)/sex	Symptom of spinal lesion	Age of initial spinal symptom (years)	Onset of spinal lesion	Level of spinal lesiona	Related skin lesion	Age of initial skin lesion (years)	Other bony lesion	Amor criteria (score)	ESSG criteria
1	61 F	Lumbago, sciatica	55	Insidious	L1, L4/5	PPP	49	SCCJ	No (4)	No
2	42 F	Lumbago	29	Subacute	C2-7, L1/2/	PPP	42	SCCJ	No (5)	No
3	50 F	Lumbago	45	Insidious	T11-L3	PPP	46	SCCJ, hip joint	No (5)	No
4	63 F	Nape pain, paresis	63	Insidious	C4-7, T/5/6, L4/5	PPP	61	SCCJ, sacroiliac joint, wrist joint	Yes (6)	Yes
5	64 F	Nape pain	64	Subacute	C/5/	PPP	64	–	No (3)	No
6	67 F	Nape pain	67	Subacute	C/4, C/7, T8	PPP	57	SCCJ	No (5)	No
7	30 F	Lumbago	26	Insidious	L5/S1	–	–	SCCJ, sacroiliac joint	Yes (7)	Yes
8	24 F	Lumbago	23	Insidious	L4/5	PPP	23	SCCJ	No (5)	No
9	39 M	Nape pain, radial hypesthesia	37	Insidious	C5/6	PPP	35	SCCJ	No (5)	No
10	69 F	Back pain	69	Insidious	T7-12	–	–	SCCJ, sacroiliac joint	Yes (6)	Yes
11	55 F	Lumbago	50	Insidious	L2-S1	PPP	45	SCCJ	No (4)	No
12	53 F	Nape pain, radial hypesthesia	44	Insidious	C7-L3	PPP	38	SCJ	No (5)	No
13	61 F	Nape pain	55	Insidious	C5, T9/T10	–	–	SCCJ, knee joint	No (5)	No

PPP palmoplantar pustulosis, SCCJ sternocostoclavicular joint. Slash indicates the adjacent disc level, and dash indicates the consecutive lesions

^aSpinal lesion: number indicates the vertebral lesion level (C cervical, T thoracic, L lumbar, S sacral)

Spinal lesions

Initial symptom and onset pattern. The initial symptom of spondylitis was local pain in all 13 patients. Numbness and radial pain due to radiculopathy was observed in four patients. One showed destruction of the vertebrae followed by severe kyphotic deformity (Case 10). Another rapidly became quadriplegic due to destruction of the vertebrae followed by compression of the spinal cord (Case 4). Low-grade fever around 37°C was observed in three patients during their first clinical phase. Ten patients showed an insidious or chronic onset pattern, and three patients revealed a subacute onset pattern. No case showed an acute onset pattern.

Laboratory data. Nine patients exhibited an increased ESR (>30 mm/h), four of these showed an increased CRP level (>1 mg/dl), and one of these revealed an increased WBC count (>10,000/μl). All 13 cases were RF negative.

Imaging findings. MRI was the most sensitive method to identify spinal lesions. Spondylitis was frequent at the lumbar and cervical spine level. Abnormal signal intensity was most frequently noted in the vertebral endplate in low-intensity T1 and high-intensity T2 images. Non-consecutive, multi-level spinal lesions were observed in five (38%) patients, and consecutive spinal level lesions at more than three vertebral levels were observed in five (38%) patients (Fig. 1, Case 2). Obvious bone changes in Xp were observed in 11 patients during the follow-up period. Five cases presented non-marginal syndesmophytes (Fig. 2, Case 3), four cases presented non-specific osteophytes, and the other two demonstrated severe destruction. There was no case of marginal syndesmophyte, which is specific to ankylosing spondylitis.

Fig. 1.

Case 2. A 42-year-old woman had a past history of receiving antibiotics due to unidentified spondylitis at 29 years old. There was no particular symptom afterwards. At the age of 42, lower back and nape pain with low-grade fever appeared. Simultaneously, a cutaneous lesion (PPP) appeared for the first time. a, b Cervical MRI showing consecutive multi-level lesions from C2–C7. c, d Lumbar MRI showing an abnormal intensity of L1/2 and L2/3 discs as well as L1 and L2 vertebrae

Fig. 2.

Case 3. A 50-year-old woman had marked lumbago since the age of 45 years old, showing repeated remission and exacerbation. a, b Plain radiographs showing ivory vertebrae and non-marginal syndesmophyte

Spondyloarthropathy

Only three patients (23%; Cases 4, 7, and 10) met the criteria of SNSA regardless of whether ESSG [6] or Amor [2] criteria were used (Table 1).

Extra-spinal lesions

Twelve patients presented some skin lesions. Ten patients had PPP, which is one of the skin-lesion criteria for SAPHO syndrome (Fig. 3, Case 9). The other two skin lesions, which were atopic dermatitis and keratosis, were unrelated to SAPHO syndrome. One patient presented no skin lesion. The first skin manifestation (PPP) preceded the first clinical spinal manifestation in eight patients, whereas the opposite occurred in two patients. Sacroilitis was identified in three patients. Sternocostoclavicular hyperostosis, by where ^99mTc-MDP bony scintigraphy showed an abnormally high uptake, was seen in 12 patients (92%) (Fig. 4, Case 6). All these patients experienced from anterior chest wall pain and swelling. An abnormal high uptake on bony scintigraphy was seen in the hip, wrist, and knee joint in one patient each, although there were no accompanying symptoms.

Fig. 3.

Case 9. A 39-year-old man had suffered from intractable palmar (a) and plantar (b) pustulosis since 35 years old. He sometimes felt nape and anterior chest pain from 37 years old, and hypesthesia of his left upper extremity at 39 years old

Fig. 4.

Case 6. A-67-year-old woman experienced nape and back pain with low-grade fever. She developed PPP and sometimes felt anterior chest pain from 57 years old. a, b AP and PA views, respectively. Bony scintigraphy (^99mTc-MDP) shows an abnormally high uptake of the bilateral sternocostoclavicular joint and C4, C7, and T8

Therapy for spinal lesions

All cases complained of spinal pain. Ten cases were treated and cured with nonsteroidal anti-inflammatory drugs. One case required additional morphine due to persistent pain. Two cases were treated surgically due to destructive spondylitis: one case became quadriplegic and the other case showed progressive kyphotic spinal deformity (Case 4 and 10, respectively) [28].

Discussion

The diagnosis of spondylitis in SAPHO syndrome is often difficult, even though the disease concept is well-understood. The most common feature of SAPHO syndrome is the accompanying skin lesion, in which PPP, acne, and psoriasis are well-recognized [27]. However, concurrent skin disease is not essential to make a diagnosis of SAPHO syndrome. Hayem et al. reported that the mean interval between the onset and diagnosis of SAPHO syndrome was 9.1 years [11]. Our cases also revealed that a long duration (mean: 6.4 years) was needed to make a definite diagnosis. This long interval occurs in cases whose cutaneous lesion appears a long time after the osteoarticular lesion develops, and in those that do not develop cutaneous lesions. The absence of definitive examination findings makes the diagnosis of SAPHO syndrome rather complicated. It is not difficult to tentatively diagnose spondylitis in SAPHO syndrome when a typical skin manifestation exists. Of course, the exclusion of various manifestations, such as infectious diseases, tumorous diseases, diffuse idiopathic skeletal hyperostosis, and retinoid therapy related bony deformity, is needed before a final diagnosis [11]. In cases where cutaneous lesions appear after osteoarticular lesions, a long follow-up period is needed. Kahn et al. described two cases of SAPHO syndrome in which a more than 20-year interval existed between the onset of skin lesions and bone involvement [15]. When frequent biopsies result in non-specific inflammation, diagnosis should be made based on the clinical course and imaging findings [5, 18]. The purpose of biopsy in SAPHO syndrome is mainly to exclude other diagnoses. Clinicians should minimize this procedure to avoid unnecessary invasion. There were many variations of disease progression. In cases not showing skin lesions at first, long-term course observations may be necessary.

We demonstrated that spondylitis in SAPHO syndrome showed a distinctly different pattern compared to pyogenic spondylitis: chronic onset without a marked fever, a slightly elevated inflammatory reaction, and multi-level spinal lesions. These findings correspond those of a previous study [29].

Earwaker et al. reported that the spinal lesions in SAPHO syndrome begin at vertebral end plates [7]. Toussirot et al. reported that they begin at the vertebral body and spread to the disc [29]. Maugars et al. proposed that they begin with enthesis and lead to osteolysis, erosion, synovitis, hyperostosis, and finally synostosis followed by ankylosis and reduction of hyperostosis [21]. Laredo et al. stated that vertebral corner erosion is important in MRI findings [20]. In our study, the origin of spinal lesions was not clarified because the number of our cases was small and various stages of spondylitis were mixed, including the case showing insidious onset and partial progression. However, spondylitis in early stage tended to lead to abnormal signal intensity (T1 low and T2 high) in the end plate. Furthermore, the progressed stage did not show marginal syndesmophytes, which is characteristic in ankylosing spondylitis, but non-marginal syndesmophytes or an osteophyte pattern, which is characteristic in psoriatic spondylitis. Non-marginal syndesmophytes do not simply extend from one vertebral angle to the other, but may start from the middle of one vertebral body and extend to the same area of the adjacent body [16]. It is impossible to distinguish definitely between SAPHO syndrome related spinal lesions and degenerative spinal deformity (non-specific osteophytes). However, SAPHO syndrome has a high tendency of ossification in the progressed stage. Sonozaki et al. reported that 3 of 18 cases of PPP-related spondylitis were actually ankylosing spinal hyperostosis [26]. One of our 13 cases presented ankylosing hyperostosis. Ankylosing spinal hyperostosis may be the terminal stage of spondylitis in SAPHO syndrome. Some cases involved in this study may also develop ankylosing hyperostosis in the future.

Taxonomically, it is still controversial whether SAPHO syndrome belongs to SNSA or not [11, 12]. Sonozaki proposed that PAO is one subtype of SNSA [25]. Maugars et al. reported that 43 and 33% of 19 SAPHO syndrome patients fulfilled the criteria of ESSG and Amor, respectively [21]. Hukuda et al. reported that 4.7 and 0.3% of Japanese 990 SNSA patients showed PPP and SAPHO syndrome, respectively [13]. In this study, even though each case had spinal lesions, the rate of fulfilling the SNSA criteria was low (23% for either ESSG [6] or Amor [2]). This was because only a few cases were positive for symptoms other than spinal lesions or arthritis-related items. All our 3 cases that met the criteria of SNSA had sacroilitis. Another possible reason might be that the total number of cases was small and our coexistence rate of sacroilitis (23%) was lower than that of past studies (52% [21], 40% [11]). A strong correlation between PPP and psoriasis has been recognized [3]; however, Sugimoto et al. stated that psoriasis should not be included as a skin lesion of SAPHO syndrome [27]. If PPP is added as a skin lesion to the criteria of spondyloarthropathy, 100 and 92% of our cases would fulfill the diagnostic criteria of ESSG [6] and Amor [2], respectively. We propose that SAPHO syndrome, especially spinal lesions related to PPP, should be recognized as a subtype of reactive spondyloarthropathy in SNSA.

The spondylitis in SAPHO syndrome generally have a good prognosis and treated conservatively [29]. However, it may cause severe destruction rapidly which needs surgery. Clinicians should carefully follow the course especially in the initial phase of the disease. Some papers indicate that antibiotics are effective because of the isolation of Propionibacteriumacnes from biopsies [8, 9, 17]. And others demonstrated that the detection rate of acnes is low and antibiotics are ineffective [4, 14]. In our two cases treated surgically, acnes were not detected. And antibiotics were not helpful to prevent the progression of lesion. We thought that the administration of antibiotics including a prophylactic meaning do not have a dramatic effect. NSAIDs, corticosteroids, methotrexate, anti-TNFα agent, and biosphosphonate may be more effective because the main cause of the lesions seems to be inflammation and the following bony reaction [1, 11, 23].

Conclusion

The features of spondylitis of SAPHO syndrome are: (1) coexistence of specific skin lesions, (2) chronic onset with a slight inflammatory reaction, and (3) slowly progressing non-marginal syndesmophytes at multiple spinal levels. SAPHO syndrome, especially spinal lesions related to PPP, may be recognized as a subtype of SNSA.