1. Top of page
  2. Abstract
  6. Acknowledgements


To assess the prevalence and severity of peripheral enthesitis among the different subtypes of spondylarthropathy (SpA) by using ultrasonography (US) in B mode with power Doppler.


One hundred sixty-four consecutive patients with SpA (according to the criteria of the European Spondylarthropathy Study Group) and 64 control patients (34 with mechanical low back pain [MBP] and 30 with rheumatoid arthritis [RA]) underwent US examination of major entheses of their limbs. Particular attention was given to the detection of vascularization at the following sites: cortical bone insertion of entheses, junction between tendon and entheses, body of tendon, and bursa.


Abnormal US findings consistent with at least one enthesitis were observed in 161 of 164 SpA patients (98%), affecting 1,131 of 2,952 entheses examined (38%). In contrast, only 132 of 1,152 entheses (11%) were found to be abnormal in 33 of 64 control patients (52%). US enthesitis was most commonly distributed in the distal portion of the lower limbs, irrespective of SpA subtype and of skeletal distribution of clinical symptoms. None of the abnormal entheses in control patients showed vascularization, compared with 916 of 1,131 abnormal entheses in SpA patients (81%), where it was always detected at the cortical bone insertion and sometimes also in the bursa. In SpA patients, the US pattern depended on the clinical presentation, with a higher prevalence of the most severe stages in those with peripheral forms.


US in B mode combined with power Doppler allowed the detection of peripheral enthesitis in a majority of SpA patients, but not in MBP or RA patients. The presence of entheseal involvement was independent of SpA subtype, but its degree of severity appeared to be greater in peripheral forms. US could be very useful for both the diagnosis and the assessment of SpA activity.

Inflammation at the insertion of ligaments, tendons, or joint capsules to bone, which is called enthesitis, is a distinctive feature of spondylarthropathy (SpA) (1, 2). It consists of focal, destructive microscopic inflammatory lesions which evolve toward fibrous scarring and new bone formation (2, 3). It may involve synovial and cartilaginous joints, syndesmoses, and extraarticular entheses (3, 4).

Peripheral enthesitis is observed in all SpA subtypes, including undifferentiated forms, and may sometimes present for a long period of time as an isolated clinical manifestation of an HLA–B27–associated disease (5). Several reports have pointed to enthesitis as a primary lesion in SpA, which may underlie all skeletal manifestations characteristic of these disorders, including synovitis (5–7). Besides SpA patients, enthesitis is also frequent among athletes as a consequence of traumatic injuries. However, in this case, it is not associated with intraarticular inflammation (i.e., synovitis). Peripheral enthesitis is usually revealed by clinical findings which lack specificity, such as localized pain, tenderness, and swelling, and there are no definite clinical criteria for the diagnosis of this manifestation. It may also be asymptomatic and detected only by imaging, such as conventional radiography, bone scintigraphy, magnetic resonance imaging (MRI), or ultrasonography (US) (4, 6–8).

In recent years, US has proved to be a highly sensitive and noninvasive tool, especially in the assessment of tendon and joint involvement (9–12). Several studies have described the use of B-mode US to identify the features of lower limb enthesitis in SpA (7, 13–15), revealing a high frequency of abnormal findings in asymptomatic entheses. More recently, power Doppler technology has allowed the visualization of abnormal vascularization and hyperemia of soft tissues in inflammatory articular diseases (16). Considering that enthesitis of SpA is underestimated and is often mistaken for sport and/or overuse pathology before the correct diagnosis is established, the aim of the present study was to use US in B mode associated with power Doppler to assess the prevalence and severity of peripheral enthesitis, including vascular changes, among the different subtypes of SpA. In addition, the prevalence of enthesitis was compared between SpA patients and control patients with mechanical low back pain (MBP) or rheumatoid arthritis (RA) to evaluate the specificity of entheseal involvement.


  1. Top of page
  2. Abstract
  6. Acknowledgements


One hundred sixty-four consecutive patients presenting to our rheumatology department with a diagnosis of SpA according to the criteria of Amor et al (17) and/or the European Spondylarthropathy Study Group (18) were included in the present study. They were additionally classified as having one of the SpA subtypes, as follows: ankylosing spondylitis (AS), if they fulfilled the modified New York criteria (19); reactive arthritis (ReA), if they fulfilled the criteria of Willkens et al (20); in SpA patients without definite radiographic sacroiliitis (at least bilateral grade II or unilateral grade III), psoriatic arthritis (PsA) or arthritis-associated inflammatory bowel disease (AIBD) was diagnosed if psoriasis or defined IBD, respectively, was present; and, undifferentiated SpA (uSpA) was diagnosed if the SpA criteria were fulfilled, but no diagnosis of AS, PsA, ReA, or AIBD could be made. The diagnosis of psoriasis required the presence of typical lesions or a definite diagnosis by a dermatologist. A diagnosis of anterior uveitis was retained if it had been established by an ophthalmologist. The diagnosis of IBD required typical gut endoscopic or histologic findings of Crohn's disease or ulcerative colitis.

Current treatment was recorded in 161 patients: 111 (69%) were taking nonsteroidal antiinflammatory drugs (NSAIDs), 2 were taking corticosteroids, and 82 (51%) were taking second-line drugs (i.e., sulfasalazine, methotrexate, gold, azathioprine, leflunomide, pamidronate, thalidomide, or infliximab).

Control populations consisted of 34 patients with MBP (21) and 30 patients who fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) 1987 revised criteria for RA (22) but who did not have advanced deformities of the hand. Demographic and clinical characteristics of all study patients are shown in Table 1.

Table 1. Demographic and clinical characteristics of the study patients*
CharacteristicSpA patients
Total (n = 164)With AS (n = 104)Without AS (n = 60)Control patients
AS prim (n = 63)AS Pso (n = 33)AS IBD (n = 7)AS ReA (n = 1)uSpA (n = 30)PsA (n = 21)AIBD (n = 6)ReA (n = 3)MBP (n = 34)RA (n = 30)
  • *

    SpA = spondylarthropathy; AS = ankylosing spondylitis; AS prim = primary AS; AS Pso = AS with psoriasis; AS IBD = AS with inflammatory bowel disease (Crohn's disease or ulcerative colitis); AS ReA = AS with reactive arthritis; MBP = mechanical low back pain; PsA = psoriatic arthritis; AIBD = arthritis with IBD; uSpA = undifferentiated SpA; RA = rheumatoid arthritis; NA = not available.

  • One ReA patient also developed Crohn's disease and psoriasis.

  • Symptoms retrieved from medical history or identified by the current examination.

  • §

    Inflammatory back pain (SpA) or mechanical back pain (MBP).

  • Radiographic sacroiliitis of at least bilateral grade II or unilateral grade III.

Ratio of men to women2.112.1551.3301.51.25210.06
Age, mean ± SD years38 ± 1240 ± 1239 ± 1141 ± 103033 ± 1243 ± 1331 ± 737 ± 2549 ± 1353 ± 14
Disease duration, mean ± SD years16 ± 1216 ± 1217 ± 1116 ± 81110 ± 913 ± 136 ± 414 ± 24NA16 ± 11
HLA–B27, no. positive/no. tested (% positive)134/147 (91)48/54 (89)29/32 (91)7/7 (100)1/1 (100)26/28 (93)15/17 (88)3/5 (60)3/3 (100)NANA
Axial skeleton symptoms, no. (%)151 (92)63 (100)33 (100)7 (100)1 (100)28 (93)16 (76)5 (83)1 (33)34 (100)0 (0)
 Back pain§143 (87)57 (90)31 (94)7 (100)1 (100)26 (87)15 (71)5 (83)1 (33)34 (100)0 (0)
 Inflammatory buttock pain117 (71)52 (83)27 (82)5 (71)1 (100)18 (60)11 (52)3 (50)0 (0)0 (0)0 (0)
 Sacroiliitis on pelvic radiography104 (63)63 (100)33 (100)7 (100)1 (100)0 (0)0 (0)0 (0)0 (0)0 (0)NA
Peripheral skeleton symptoms, no. (%)122 (74)40 (63)22 (67)5 (71)1 (100)27 (90)18 (86)5 (83)3 (100)1 (3)30 (100)
 Enthesitis72 (44)15 (24)22 (67)2 (29)0 (0)20 (67)12 (57)4 (67)1 (33)1 (3)0 (0)
 Arthritis91 (55)33 (52)20 (61)4 (57)1 (100)12 (40)16 (76)2 (33)3 (100)0 (0)30 (100)
 Dactylitis37 (23)8 (13)6 (18)0 (0)0 (0)10 (33)10 (48)3 (50)2 (67)0 (0)0 (0)
Uveitis, no. (%)37 (23)17 (27)9 (27)1 (14)1 (100)4 (13)4 (19)1 (17)0 (0)0 (0)0 (0)


US examination was performed by a rheumatologist experienced in musculoskeletal US (M-ADA), and blinded to the diagnosis. Esaote (AU5 Epi; Genoa, Italy) equipment with a 13-MHz linear array transducer was used. Blinding of the US examiner was achieved as follows: SpA and control patients were positioned on the US examination table by one of the investigators (RS-N) in random order. They were asked not to communicate with the US examiner.

For each patient, the following entheseal insertion sites were examined bilaterally: greater trochanter, pubis, patella (at insertions of the quadriceps femoris and patellar tendons), Achilles tendon and plantar fascia insertions on the calcaneus, tibialis anterior tendon insertion, and medial and lateral epicondyles. US assessment was performed using the following procedure. Gel was applied to the skin to provide an acoustic interface. Each tendon was scanned in both longitudinal and transverse planes, and the scan images were registered on Digital Imaging and Communications in Medicine support. Each examination took ∼20 minutes. The processing settings were held constant during the examination, and the temperature of the room was kept stable at 20°C. For each enthesis, US was initially performed in B mode to detect morphologic abnormalities, and subsequently with power Doppler to detect abnormal vascularization.

In B mode, we searched for the following abnormal findings, and any one of them was considered to be a feature of enthesitis, as previously defined (23, 24): thickening or intratendinous focal changes of the tendon insertion, calcific deposits at the tendon insertion, or periosteal changes. In addition, an enlarged bursa was considered to be evidence of bursitis (23, 24). A comprehensive analysis of normal US features in B mode was not specifically conducted for the purpose of this study, because normal findings of the structures examined have already been extensively described by others (15, 23, 25, 26). Blood flow was examined at the entheses using power Doppler mode. Power Doppler settings were standardized with a pulse repetition frequency of 750 Hz and a power Doppler gain of 50–53 dB. Vascularization was studied at the following areas, as depicted in Figure 1 : cortical bone insertion, body of tendon, bursa, and junction between tendon and enthesis. The detection of vascularization at one of these sites was considered abnormal, since this was never found in a preliminary study of peripheral entheses in 20 normal adults (D'Agostino M-A: unpublished observations). US enthesitis was classified into 5 distinctive patterns according to the different combinations of abnormal gray-scale and/or power Doppler features, as shown in Table 2.

thumbnail image

Figure 1. A, Diagram of the patellar tendon enthesis depicting the 4 distinct areas (arrows) examined by ultrasonography in B mode combined with power Doppler for abnormal vascularization. a = cortical bone insertion; b = body of tendon; c = bursa; d = junction between tendon and enthesis; P = patella; F = femur; T = tibia. B, Ultrasonographic image of patellar tendon depicting the same 4 areas described in A.

Download figure to PowerPoint

Table 2. Classification of ultrasonography enthesitis according to B-mode and power Doppler findings
1Vascularization at the cortical junction without abnormal findings in B mode
2aVascularization associated with swelling and/or decreased echogenicity at the cortical junction in B mode
3aSame as stage 2a, plus erosions of cortical bone and/or calcification of enthesis, and optional surrounding bursitis
2bAbnormal findings in B mode as in stage 2a, but without vascularization
3bAbnormal findings in B mode as in stage 3a, but without vascularization

The intra- and interobserver agreements for the staging of US enthesitis were estimated with reference to a set of 270 images representing all varieties of gray-scale/power Doppler US patterns detected in SpA and control patients. These images were randomized and submitted to a second reading by the original US examiner (M-ADA) 6–8 months after the examination as well as a reading by a radiologist (J-LB) as an independent examiner. Agreement was assessed using the weighted kappa statistic (κw) (27). For this, the weights were calculated as follows: wi = 1 − (i2/k − 1)2, where wi = the weights assigned to images for which the readers differed by i categories and k = the number of categories. The following 6 categories were assigned: normal findings of US (stage 0); US enthesitis stages 1, 2a, 3a, 2b, and 3b. Intra- and interobserver kappa values were excellent, with κw = 0.97 and 0.83, respectively. Figure 2 shows distinctive patterns of enthesitis detected by US examination in B mode combined with power Doppler in 3 SpA patients and 1 control patient with MBP.

thumbnail image

Figure 2. Distinctive patterns of enthesitis detected by ultrasound (US) examination in B mode combined with power Doppler in patients with spondylarthropathy (SpA) and control patients. A, Normal features of the Achilles tendon enthesis in an SpA patient. Tendon thickness is 3.6 mm. C = calcaneum. B, Stage 1 plantar fascia enthesitis in an SpA patient. Abnormal vascularization at the cortical junction is seen, without other US abnormalities, in B mode. Fascia thickness is 3.5 mm. C, Stage 3a Achilles tendon enthesitis in an SpA patient. Abnormal vascularization associated with marked swelling and decreased echogenicity of the enthesis at the cortical junction can be seen, as well as erosions of cortical bone. Tendon thickness at the level of the enthesis is 9.2 mm. D, Stage 2b Achilles tendon enthesitis in a patient with mechanical low back pain. Swelling and decreased echogenicity at the cortical junction are seen, without an abnormal vascularization pattern. Tendon thickness at the level of the enthesis is 7.9 mm.

Download figure to PowerPoint

Clinical examination of entheses.

All patients were routinely clinically assessed by experienced rheumatologists before US examination. Thirty-four of the SpA patients were randomly selected to undergo an additional careful clinical examination of the same entheses that had been evaluated by US. An independent examiner who was unaware of the US results (CH-B) performed this clinical examination. This examiner recorded spontaneous pain, tenderness elicited by pressure, mobilization and contraction against resistance of the corresponding tendons, and local swelling of the enthesis. The presence of at least one of these findings was taken as evidence of enthesitis.

Statistical analysis.

Differences between groups were tested by chi-square test (categorical data) or Student's unpaired t-test (numerical data). A chi-square test for trend was also used to compare the distribution of severity on power Doppler sonography among different groups of patients. P values less than 0.05 were considered significant. In the case of multiple comparisons, crude P values were corrected for the number of comparisons by the Bonferroni method (Pcorr). A positive correlation between current treatment and the number of sites discordant between US and clinical examination was tested by using Spearman's 1-tailed test.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Abnormal entheseal involvement identified by US.

One hundred sixty-one of the 164 SpA patients (98%) had at least 1 abnormal enthesis by gray-scale combined with power Doppler US examination. The 3 patients without US enthesitis had uSpA, PsA, and ReA. In contrast, only 15 of the 34 patients with MBP (44%) and 18 of the 30 patients with RA (60%) had any abnormal entheses, which was statistically significantly fewer than in the SpA group (P < 0.0001 for both comparisons, by chi-square test). Furthermore, of 2,952 entheses examined by gray-scale/power Doppler US in SpA patients, a total of 1,131 (38%) were considered abnormal, without significant variation in the frequency of enthesis involvement among the different subtypes of SpA (Table 3), while only 59 of 612 entheses (10%) and 73 of 540 entheses (14%) were found to be abnormal in the MBP and RA groups, respectively (Table 3).

Table 3. Distribution of abnormal entheses by B-mode US combined with power Doppler US in SpA patients and control patients*
Site of enthesisSpA patients (n = 164)Controls with MBP (n = 34)P, SpA vs. MBPControls with RA (n = 30)P, SpA vs. RA
TotalWith AS (n = 104)Without AS (n = 60)
  • *

    Except where indicated otherwise, values are the number (%) of abnormal entheses by ultrasonograpy (US) in the corresponding group of patients. See Table 1 for other definitions.

  • By chi-square test, except where indicated otherwise.

  • By Student's unpaired t-test.

  • §

    Only the insertion on the patellar apex was studied.

  • Insertion on the base of the first metatarsal head.

Total no. of entheses examined2,9521,1345941261854037810854612540
All sites examined1,131 (38)417 (37)267 (45)61 (48)6 (33)197 (36)128 (34)42 (39)13 (24)59 (10)<0.000173 (14)<0.0001
Abnormal entheses per patient, mean ± SD7 ± 37 ± 38 ± 39 ± 567 ± 36 ± 37 ± 24 ± 42 ± 4<0.00012 ± 6<0.0001
Great trochanter143 (44)50 (40)37 (56)9 (64)2 (100)19 (31)18 (43)7 (58)1 (17)8 (10)<0.00015 (8)<0.0001
Pubis6 (2)3 (2)1 (2)0 (0)0 (0)1 (2)1 (2)0 (0)0 (0)0 (0)<0.00350 (0)<0.0005
Quadriceps tendon91 (28)32 (25)23 (35)6 (43)0 (0)14 (23)11 (26)3 (25)2 (33)6 (8)<0.00014 (7)<0.0001
Patellar tendon§193 (59)72 (59)50 (76)11 (79)0 (0)27 (45)25 (60)6 (50)2 (33)18 (17)<0.000119 (32)0.9
Tibialis anterior tendon33 (10)11 (9)11 (17)1 (7)0 (0)5 (8)2 (5)1 (8)2 (33)0 (0)<0.00016 (10)<0.0001
Achilles tendon260 (79)92 (73)56 (85)13 (93)2 (100)51 (85)32 (76)12 (100)2 (33)12 (19)<0.000122 (37)<0.0001
Plantaris fascia243 (74)100 (79)48 (73)12 (86)2 (100)44 (73)24 (57)9 (75)4 (67)12 (19)<0.000116 (37)<0.0001
Lateral epicondyle80 (24)32 (25)18 (27)5 (36)0 (0)17 (28)8 (19)0 (0)0 (0)1 (2)<0.00011 (3)<0.0001
Medial epicondyle82 (25)25 (20)23 (35)4 (29)0 (0)19 (32)7 (17)4 (33)0 (0)2 (4)<0.00010 (0)<0.0001

In SpA, the sites most commonly affected were the distal portions of the lower limbs (i.e., of Achilles tendon, plantar fascia, and patellar tendon origin, of which 79%, 74%, and 59%, respectively, were abnormal by US) (Table 3). The pattern of distribution appeared to be largely independent of SpA subtype. Among MBP and RA patients, the hierarchy of affected entheses was similar to that observed in SpA patients. However, the frequency of affected entheses was statistically significantly lower than that in SpA for all sites examined, except for the patellar tendon among RA patients (Table 3).

In SpA patients, 916 of 1,131 entheses (81%) found to be affected on US examination presented with abnormal vascularization (Table 4). Only 3 SpA patients (2%) did not present with any abnormal vascularization of their entheses. As shown in Figure 1, abnormal vascularization was always detected at the cortical bone insertion and sometimes also in the bursa (14% of abnormal entheses). The distribution of enthesis involvement according to the 5 distinctive patterns is shown in Table 4. Notably, no difference was observed in the distribution of enthesis pattern between distinct SpA subtypes. In contrast to these results, in MBP and RA patients, all entheseal abnormalities were detected only in B mode and corresponded to stage 2b, without vascularization on power Doppler (Table 4).

Table 4. Classification of abnormal peripheral enthesis features by B-mode US combined with power Doppler US in SpA and control patients*
Abnormal enthesis patternSpA patients (n = 164)Control patients
TotalWith AS (n = 104)Without AS (n = 60)
AS primAS PsoAS IBDAS ReAuSpAPsAAIBDReAMBP (n = 34)RA (n = 30)
  • *

    US = ultrasonography. See Table 1 for other definitions.

  • See Table 2 for classification of ultrasound features of enthesitis.

  • The proportion of abnormal entheses with vascularization was significantly greater in SpA patients than in patients with MBP or patients with RA (both P < 0.0001 by chi-square test).

  • §

    Refers to vascularization at the insertion of entheses on the cortical bone.

Total no. of abnormal entheses1,13141726761619712842135973
With vascularization, no. (%)§916 (81)342 (82)201 (75)51 (84)6 (100)164 (83)103 (80)37 (88)12 (92)0 (0)0 (0)
 Stage 1547 (48)211 (51)136 (51)31 (51)2 (33)90 (46)48 (38)25 (60)4 (31)0 (0)0 (0)
 Stage 2a255 (23)93 (22)42 (16)14 (23)1 (17)57 (29)35 (27)8 (19)5 (38)0 (0)0 (0)
 Stage 3a114 (10)38 (9)23 (9)6 (10)3 (50)17 (9)20 (16)4 (10)3 (23)0 (0)0 (0)
Without vascularization, no. (%)§215 (19)75 (18)66 (25)10 (16)0 (0)33 (17)25 (20)5 (12)1 (8)59 (100)73 (100)
 Stage 2b109 (10)38 (9)34 (13)3 (5)0 (0)25 (13)7 (5)1 (2)1 (8)59 (100)73 (100)
 Stage 3b106 (9)37 (9)32 (12)7 (11)0 (0)8 (4)18 (14)4 (10)0 (0)0 (0)0 (0)

US identification of enthesitis according to clinical presentation of SpA.

In parallel with the subdivision of their disease into classic subtypes, SpA patients were classified as having axial, peripheral, or mixed presentation forms of enthesitis according to the clinical distribution of skeletal symptoms. No major variation in the proportion of abnormal entheses identified by US examination was observed between these different forms (Table 5). However, a greater proportion of affected Achilles tendons was detected in the mixed form than in the axial form (P < 0.001) (Table 5).

Table 5. B mode combined with power Doppler ultrasonographic (US) involvement of peripheral entheses among SpA patients according to the clinical distribution of skeletal symptoms*
Site of enthesisSpA patients (n = 164)P/Pcorr, axial vs. mixed vs. peripheral
Axial (n = 37)Mixed (n = 111)Peripheral (n = 16)
  • *

    Except where indicated otherwise, values are the number (%) of abnormal entheses by US in the corresponding group of patients. In the axial group, 23 patients presented with primary AS, 8 with AS with psoriasis, 2 with AS IBD, and 4 with uSpA; in the peripheral group, 2 patients presented with primary AS, 1 with AS with psoriasis, 3 with uSpA, 6 with PsA, 2 with AIBD, and 2 with ReA; the remaining 111 patients had mixed (axial and peripheral) clinical presentation. See Table 1 for other definitions and descriptions of patients.

  • By chi-square test Pcorr = P value corrected for the number of comparisons by the Bonferroni method.

  • No statistical analysis was possible because of the small sample size.

  • §

    Only the insertion on the patellar apex was studied.

  • Insertion on the base of the first metatarsal head.

Total no. of entheses examined6661,998288
All sites examined248 (37)789 (39)94 (33)0.07
 Great trochanter35 (47)96 (43)12 (37.5)0.64
 Pubis3 (4)2 (1)1 (3)NA
 Quadriceps tendon20 (27)65 (30)6 (19)0.46
 Patellar tendon§55 (74)122 (55)16 (50)0.013/0.12
 Tibialis anterior tendon10 (14)20 (9)3 (9)0.57
 Achilles tendon46 (62)191 (86)23 (72)<0.0001/<0.001
 Plantaris fascia48 (65)169 (76)26 (81)0.09
 Lateral epicondyle18 (24)58 (26)4 (12.5)0.24
 Medial epicondyle13 (18)66 (30)3 (9)0.01/0.8

In contrast, the vascularization pattern was dependent on the clinical presentation of SpA (Table 6). Thus, the distribution of vascularized entheses was uneven between subgroups (P = 0.006 by chi-square test). In particular, patients with the mixed form presented with significantly more vascularized entheses than did patients with the axial form (Pcorr = 0.005 by chi-square test). In addition, among vascularized entheses, there was an uneven distribution of gray-scale morphologic severity between the different clinical forms (P < 0.0001 by chi-square test for trend). More specifically, entheseal alterations tended to be more pronounced in the peripheral form (fewer stage 1 patterns and more stage 3a patterns) than in the axial or mixed forms (P < 0.0001 for both comparisons, by chi-square test). No statistically different distribution of severity pattern was observed for nonvascularized enthesitis between the different forms of SpA (stages 2b and 3b). Notably, the proportion of patients receiving each class of treatment (i.e., NSAID or second-line drug) was not different between clinical subgroups.

Table 6. Classification of abnormal peripheral enthesis features by B-mode ultrasonography (US) combined with power Doppler US in spondylarthropathy (SpA) patients, according to the clinical distribution of skeletal symptoms*
Abnormal enthesis patternSpA patients (n = 164)
Axial (n = 37)Mixed (n = 111)Peripheral (n = 16)
  • *

    See Tables 1 and 5 for description of patients. See Table 2 for description of patterns.

  • Refers to vascularization at the insertion of entheses on the cortical bone.

Total no. of abnormal entheses24878994
With vascularization, no. (%)184 (74)657 (83)75 (80)
 Stage 1117 (47)402 (51)28 (30)
 Stage 2a53 (21)179 (23)23 (24)
 Stage 3a14 (6)76 (10)24 (26)
Without vascularization, no.  (%)64 (26)132 (17)19 (20)
 Stage 2b35 (14)69 (9)5 (5)
 Stage 3b29 (12)63 (8)14 (15)

Comparison between findings of US and clinical examination of entheses.

Among the 34 SpA patients who underwent a careful clinical examination, 88 of 612 entheses (14.4%) were found to be clinically abnormal in 21 patients (62%), while 220 entheses (36%) were considered abnormal by US examination in 32 patients (94%) (P < 0.0001 for both comparisons, by chi-square test) (Table 7). Clinical examination most frequently detected enthesitis at the patellar tendon insertion on the patellar apex (17 of 68 sites; 25%), at insertions of tendons on the great trochanter (16 of 68 sites; 24%), and at the plantar fascia on the calcaneus (11 of 68 sites; 16%). The comparison of clinical and US findings showed a broad discrepancy in abnormal entheses detected for all sites examined (Table 7). No positive correlation was found between the number of sites discordant between US and clinical examination on the one hand and treatment with NSAIDs, second-line drugs, or any other agent on the other hand (r2 = 0.009, P = 0.29; r2 = 0.013, P = 0.26; and r2 = 0.01, P = 0.27, respectively, by Spearman's correlation test).

Table 7. Comparison of abnormal findings on clinical and B-mode/power Doppler ultrasonography (US) in spondylarthropathy patients*
Site of enthesisPositive clinical finding (88/612)Negative clinical finding (524/612)
Abnormal by USNormal by USAbnormal by USNormal by US
  • *

    Values are the number of entheses in the corresponding category (% of all entheses examined at the corresponding site).

  • Number of entheses in the corresponding category/number of entheses examined.

  • Only the insertion on the patellar apex was evaluated.

  • §

    Insertion on the base of the first metatarsal head.

All sites examined36 (6)52 (8)184 (30)340 (56)
 Great trochanter12 (18)4 (6)24 (35)28 (41)
 Pubis0 (0)15 (22)3 (4)50 (74)
 Quadriceps tendon2 (3)9 (13)18 (26)39 (57)
 Patellar tendon8 (12)9 (13)42 (62)9 (13)
 Tibialis anterior  tendon§0 (0)4 (6)6 (9)58 (85)
 Achilles tendon5 (7)4 (6)43 (63)16 (24)
 Plantaris fascia9 (13)2 (3)31 (46)26 (38)
 Lateral epicondyle0 (0)3 (4)8 (12)57 (84)
 Medial epicondyle0 (0)2 (3)9 (13)57 (84)


  1. Top of page
  2. Abstract
  6. Acknowledgements

Recently, the importance of peripheral enthesis involvement among the manifestations of SpA has been emphasized by several groups of investigators (2, 14, 28, 29) and is best reflected by its inclusion as a classification criterion for the diagnosis of SpA (18). However, peripheral enthesitis is commonly mistaken for other pathology and is most certainly underdiagnosed during the course of SpA.

To our knowledge, this is the first report of a systematic analysis of peripheral entheses by US examination in B mode combined with power Doppler in a large group of patients with the different SpA subtypes. There were several important findings in this study. First, the frequency of abnormal peripheral entheses was very high among SpA patients compared with controls. Second, abnormal entheses were uniformly found among SpA patients, irrespective of the disease subtype. Third, abnormal vascularization of affected entheses was detected exclusively in SpA patients. Finally, there was a trend toward a more severe pattern of US involvement in the peripheral forms of SpA.

In this study, the high prevalence of peripheral entheseal abnormalities detected by gray-scale combined with power Doppler US in SpA patients further outlines the primary significance of this finding among SpA manifestations. The greater sensitivity of US compared with clinical examination for the detection of SpA enthesitis was previously reported (13, 30) and was largely confirmed in the present study. In addition, we have shown that US assessment of inflammation at entheseal insertions could be dramatically improved by combining power Doppler with B mode, resulting in the visualization of a pattern highly specific for SpA. Thus, 98% of the SpA patients had at least one vascularized enthesitic site, compared with none of the controls.

So far, MRI has remained the imaging technique most widely applied in the assessment of musculoskeletal inflammatory processes. The MRI pattern of SpA enthesitis has been described as a diffuse bone edema adjacent to entheses, associated with surrounding soft tissue edema (31). However, this technique lacks sensitivity and specificity (8). This is because changes in the fibrous part of the enthesis, where fibroblasts are tightly cross-linked with little capacity for the accumulation of water, cannot easily be detected with MRI (8, 32). Since MRI is not appropriate and histologic assessment of enthesitis is not practical because of limited access, there is no “gold standard” for validating the US features of enthesitis.

In recent years, the capacity of power Doppler techniques to detect inflammation of musculoskeletal tissues, such as the synovial membrane, was reported (16). The performance of power Doppler in detecting low-velocity blood flow at the microvascular level in several tissues has been demonstrated (33, 34). In this study, the use of power Doppler US allowed us to detect abnormal vascularization of entheses in almost all SpA patients. Such abnormalities were primarily detected at the insertion of entheses to cortical bone and less frequently in bursae. We have previously reported such vascularization of enthesis insertion in 2 patients with SpA, with disappearance of this abnormality following efficacious therapy with anti–tumor necrosis factor α (anti-TNFα) (35). Abnormal vascularization was also previously described in the calcaneal bursa of 1 PsA patient (36). Nevertheless, this latter localization is not specific for SpA, since it can also be present in patients with mechanical enthesitis (37) or other inflammatory arthritides, such as RA. In contrast, vascularization of the body of the tendon and/or the junction between the tendon and the enthesis, which reflects tendinitis, was not detected in SpA patients in the present study (38).

The higher prevalence of enthesitis in the lower limbs, which was seen both in SpA and control patients in this study, has repeatedly been reported (13, 14, 39). This result is unlikely to reflect an overestimation of pathologic findings of US. Thus, normal US features and dimensions of entheses have been properly described by others, and we have used reported criteria for identification of abnormal features on US in B mode (15, 22, 23, 25, 26). It was suggested that the length, anatomy, and physiology of entheses may each contribute to such preferential distribution (13, 40). Thus, enthesitis, which could partly result from biomechanical stress forces, may happen where such forces predominate (i.e., at the distal entheseal sites of the lower limbs), both in inflammatory and mechanical diseases. Furthermore, such events might be amplified in the context of SpA (8). However, even in SpA, the US features of enthesitis were quite variable, suggesting that the importance of enthesitis could be graded according to the results of this imaging technique.

We adopted a 5-stage classification system that reflected our observations. Stage 1 presumably represented isolated vascular changes occurring during the initial phase of enthesitis. Stages 2a and 3a probably reflected increased vascular alterations associated with graded signs of morphologic alterations. Stages 2b and 3b, which were devoid of vascularization, were suggestive of inactive lesions. The relevance of this interpretation is supported by the monitoring of enthesitis showing 1) progression from stages 2a or 3a to stages 2b and 3b, respectively, during efficacious treatment of SpA with anti-TNFα, and 2) sequential progression from normal to stage 1, then to stage 2a, and finally to stage 3a during relapse after discontinuation of treatment (ref. 35 and D'Agostino M-A, Said-Nahal R, Dougados M, Breban M: unpublished observations).

We detected no major differences in the frequency or categorized localization of peripheral enthesitis by US between the different SpA subtypes, even when according to the clinical distribution of skeletal symptoms (i.e., axial, peripheral, or mixed). In contrast, a higher severity of US-identified enthesitis seemed to be associated with a predominantly peripheral clinical presentation. Taken together, these observations are consistent with the concept that different subtypes of SpA represent phenotypic variations of a unique disease (41, 42). Furthermore, the poor reliability of clinical findings in detecting enthesitis could explain some degree of the inadequacy inherent in SpA classification based only on clinical presentation.

We have reported here the results of a cross-sectional study performed on patients with established diagnoses. Since the recruitment was hospital based, the patients included here may have had more severe disease than that observed in the general SpA population. This could have contributed to the high prevalence of inflammatory enthesitis detected with US. To evaluate the diagnostic value of gray-scale combined with power Doppler US techniques, it will be necessary to conduct a longitudinal study including patients with early disease and still-uncertain diagnoses. It is also well established that the performance of power Doppler US can be influenced by the examiner, the machine, and the acoustical conditions involved in image processing. In the present study, all parameter settings were kept stable to optimize the intraobserver reliability of the technique. Indeed, intra- and interobserver reliability assessed on saved images was satisfactory. However, to validate this technique as a potentially useful tool for both the diagnosis and followup of SpA, it will also be necessary to assess the interobserver reliability of examination.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We gratefully acknowledge the contribution of Dr. Guy-Robert Auleley to the statistical analysis of intra- and interobserver reliability.


  1. Top of page
  2. Abstract
  6. Acknowledgements