Value of Contrast-Enhanced Ultrasonography for the Detection and Quantification of Enthesitis Vascularization in Patients With Spondyloarthritis

Authors


Abstract

Objective

To evaluate if contrast-enhanced ultrasound (CEUS) can improve the detection and quantification of the vascularization of mild enthesitis in spondyloarthritis (SpA) and to evaluate the influence of nonsteroidal antiinflammatory drugs (NSAIDs) on such detection.

Methods

Fourteen patients with mildly active SpA were evaluated at 3 consecutive visits: at baseline while undergoing NSAID treatment (V1), after 1 week of stopping NSAIDs (V2), and after 1 week of resuming NSAIDs (V3). At each visit, enthesitis was evaluated clinically and by power Doppler US (PDUS). A selected enthesis with a doubtful PDUS vascularization signal was studied by CEUS in 2 steps: 1) using a dedicated technology that preserves microbubbles (Contrast Tuned Imaging technology [CEUS-CnTI]) and 2) using high PD (CEUS-PD) to destroy microbubbles. A linear mixed model statistical analysis, taking visits and contrast agent as fixed factors and the patient as a random factor, was used.

Results

Disease activity and PDUS findings increased between V1 and V2 and then decreased between V2 and V3. As compared with PDUS alone, CEUS-PD and CEUS-CnTI each detected 1 supplementary vascularized enthesis at V1, CEUS-PD detected 1 vascularized enthesis and CEUS-CnTI detected 3 vascularized entheses at V2, and CEUS-PD and CEUS-CnTI each detected 2 vascularized entheses at V3. The mean inflammation score was increased by the use of CEUS (P = 0.04). This score increased between V1 and V2 (P = 0.03 by CEUS-PD and P = 0.01 by CEUS-CnTI) and decreased between V2 and V3.

Conclusion

CEUS improved the detection of enthesitis in SpA patients by confirming all doubtful enthesitis signals and confirming the absence of enthesis vascularization. The use of NSAIDs influenced the detection of vascularization.

INTRODUCTION

Enthesitis, inflammation at the origin and insertion of ligaments, tendons, aponeuroses, and joint capsules to bone, is a widely accepted clinical, histopathologic, and imaging feature of spondyloarthritis (SpA) ([1, 2]). Enthesitis is the result of focal, destructive microscopic inflammatory lesions that evolve toward fibrous scarring and new bone formation ([3]). Enthesitis may involve synovial and cartilaginous joints, syndesmoses, and extraarticular entheses ([4]). Several studies have pointed out enthesitis as a primary lesion in SpA ([5]). In enthesitis, conventional radiography shows only structural bone changes, such as proliferation or erosions. In contrast, magnetic resonance imaging (MRI) and power Doppler ultrasound (PDUS) detect a spectrum of early and late changes, some of which are related to inflammation (i.e., adjacent bone marrow edema and abnormal vascularization at the bone–enthesis junction, respectively) ([5-9]).

Unlike MRI, US enables the examination of several entheses during the same examination session. Over the past few years, numerous studies have proven the ability of B-mode US to detect enthesitis in SpA ([10-15]). The presence of a PD signal at the cortical bone insertion, reflecting an abnormal vascularization and hyperemia of enthesitis, has not been found in healthy controls ([16]) and was shown to be specific for SpA enthesitis ([13]). PDUS has been proven to be a reliable and sensitive-to-change tool for assessing enthesitis ([17-20]). PDUS may also be helpful for diagnosing SpA early ([21, 22]). However, despite these promising results, it is sometimes difficult to detect any vascularization of an affected enthesis. This difficulty may be related to the small size of the vessels, their intraosseous location, and the presence of a slow flow and/or of Doppler movement aliasing.

Following the advent of new microbubble contrast agents in sonography ([23]), an imaging technique was developed to increase the sensitivity of the Doppler examination: pulse-inversion harmonic imaging (PIHI) ([24]). PIHI exploits the properties of microbubble contrast material, which are based on the generation of harmonics from nonlinear oscillation of bubbles. US contrast agents are now routinely used in research and daily practice in oncology ([25]) and have already been applied to the rheumatology field for enhancing the vascularization signal of synovitis in rheumatoid arthritis or for studying sacroiliac joint involvement in SpA ([26, 27]). Regarding peripheral entheses, contrast-enhanced US (CEUS) has already been used to study healthy subjects ([16]), and it has been demonstrated that normal vascularization of anatomic sites such as Achilles enthesis is not detectable in spite of the enhancement of the vascularization signal. However, no study on the use of CEUS to enhance the detection of enthesitis vascularization in SpA has been published. Moreover, the influence of treatment by nonsteroidal antiinflammatory drugs (NSAIDs) on the detection of enthesitis has never been evaluated. NSAIDs remain the first treatment option for SpA patients because they are very efficacious on both axial and peripheral symptoms ([28-30]) and may even be effective to halt the structural progression of SpA as detected on radiographs ([31]). Although, to the best of our knowledge, there are no published studies regarding the influence of NSAIDs on enthesitis lesions in SpA assessed by US. The aims of this study were to evaluate if a contrast agent could facilitate the US detection and quantification of the vascularization signal of mild enthesitis, and to evaluate the influence of NSAID treatment on this vascularization.

Box 1. Significance & Innovations

  • Contrast-enhanced ultrasound (CEUS) confirmed the presence of vascularization in inflammatory enthesitis.
  • CEUS improved the detection and quantification of the vascularization of mild enthesitis in spondyloarthritis.
  • The use of nonsteroidal antiinflammatory drugs influenced the detection of vascularization of enthesitis.

PATIENTS AND METHODS

Patients

Over a period of 5 months we investigated 14 patients recruited from the outpatient rheumatology department of Ambroise Paré Hospital, near Paris. All patients fulfilled the Assessment of SpondyloArthritis international Society criteria for axial SpA ([32]), and 42.9% fulfilled the modified New York criteria for ankylosing spondylitis ([33]). The inclusion criteria were age >18 years, receiving treatment with NSAIDs at a stable dose for at least 1 month, and having moderate disease activity as evaluated by the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI; between 3 and 4 of 10) ([34]). No patients presented with any of the following: contraindication to the use of the US contrast agent, including previous allergic reaction; severe respiratory or heart disease; pregnancy; or breastfeeding. Patients currently treated with a tumor necrosis factor (TNF) blocker or disease-modifying antirheumatic drug were not included. The study was conducted in accordance with the Declaration of Helsinki and each participant gave written informed consent to stop NSAIDs during 7 days and replace them with analgesics if necessary and to undergo CEUS examinations. All types and doses of analgesic medication were allowed during the study period.

Clinical examination

Patients were examined by 1 rheumatologist (GM) who was unaware of the US results. Enthesitis was evaluated using the modified Mander index ([35]) and the presence of pain on 27 entheseal sites was recorded. The BASDAI was used to evaluate disease activity ([34]). Blood tests were performed at baseline to determine the erythrocyte sedimentation rate (mm/hour) and C-reactive protein level (mg/l).

US examination

PDUS was performed immediately after the clinical investigation by another rheumatologist (MAD) experienced in musculoskeletal US and blinded to the findings of the clinical examination. A MyLab 70 US system (Esaote) was used with 2 linear array transducers (LA435 [high-frequency probe] and LA522 [medium to low–frequency probe], operating at a frequency of 10–18 MHz and 3–9 MHz, respectively). For each patient, longitudinal and transverse scans of the 8 following entheseal sites were examined bilaterally: greater trochanter insertion of the gluteus medium tendon, patella insertions of the quadriceps femoris tendon and of the patellar ligament, Achilles tendon and plantar fascia insertions on the calcaneus, distal tibialis anterior tendon insertion, and common extensor and flexor tendons on the lateral and medial epicondyle of the elbow. Standardized machine settings both for PD and B-mode were used and remained fixed throughout the study. PDUS was performed under a constant room temperature of 20°C. All patients underwent paired clinical and US examination 3 times: at baseline while undergoing stable NSAID treatment (V1), after 1 week of stopping NSAID treatment (V2), and 1 week after resuming the same NSAID treatment as before V1 (V3).

Unenhanced PDUS

PDUS was first performed with the LA435 probe to detect morphologic abnormalities and/or abnormal vascularization (Figure 1A). In B-mode, US examination was performed as previously described ([13]). The following parameters were noticed: thickening, hypoechogenicity of the enthesis, presence of bone erosion, and calcification or enthesophyte. PD assessment of the enthesis was performed on a region of interest (ROI) that included the entheseal attachment site and its surrounding bony margins and soft tissues. The settings were as follows: Doppler frequency 9.1 MHz, pulse repetition frequency (PRF) 500 Hz, gain setting was just below the noise level (usually ∼50% gain), medium wall filter, and persistence of 3. PDUS examination allowed for the identification of enthesitis vascularization and a normal nutritive vessel located closely to the enthesis insertion. True enthesitis vascularization was defined by the presence of a PD signal at the cortical bone insertion and flow confirmed by pulse wave Doppler spectrum to exclude artifacts. An uncertain or doubtful enthesitis vascularization signal was defined by the presence of a real vascular flow (i.e., confirmed by pulse wave Doppler), which was difficult to follow for >2 seconds. A video clip of 8 seconds (in B-mode and PD) of each examined enthesis was recorded.

Figure 1.

Longitudinal scan of the lateral epicondyle enthesis in a 31-year-old patient. A, B-mode and power Doppler using unenhanced ultrasound (LA435 probe) demonstrates an active enthesitis (grade 3) with vascularization adjacent to the cortical bone and large confluent marked vessel covering >50% of the examined area. B, Contrast-enhanced ultrasound examination using the dedicated Contrast Tuned Imaging technology (LA522 probe) confirms abnormal vascularization arising from periosteal artery (arrow), shows enhancement of the signal intensity (grade 2) (arrowheads), and shows movement of bubbles in capillaries (asterisks) 20 seconds after the injection of the contrast agent.

CEUS

CEUS examination was done using a medium to low–frequency probe (LA522 probe) with the following settings: Doppler frequency 4.2 MHz, PRF 500 Hz, and a gain setting just below the noise level (usually ∼60% gain). Because of the half-time of the contrast-enhanced agent, only 1 enthesis was studied by CEUS. The selected enthesis (lateral epicondyle) was always the same in all patients and during all 3 visits. The choice of the enthesis was made independently of the clinical symptoms and before the study start. This a priori choice was based on the accessibility of the site to the US examination, the possibility to visualize the enthesis and the feeding vessel by the same scan, and the frequency of such involvement in SpA ([13]).

In order to limit the variability of the detection of the vascular signal due to the position and inclination of the probe, our protocol was standardized as follows: once the enthesis was examined and the feeding vessel was detected in a longitudinal scan, an 8-second video clip in B-mode and PD (i.e., without contrast agent) was recorded. A bolus injection of 4.8 ml of a second-generation US contrast agent (SonoVue [Bracco]), composed by sulfur hexafluoride, was given intravenously.

CEUS was therefore performed in 2 steps with the LA522 probe: first using the dedicated Contrast Tuned Imaging technology (CnTI [Esaote]), a PIHI operating at a low mechanical index of 0.15 (Figure 1B), and second using PD with a high mechanical index (0.4) to destroy microbubbles when they passed under the probe (CEUS-PD).

Using CnTI, a longitudinal scan of the selected enthesis was recorded on a 90-second video clip and then evaluated by a time/intensity analysis (see Supplementary Figure 1, available in the online version of this article at http://onlinelibrary.wiley.com/doi/10.1002/acr.22195/abstract). At the end of the examination, another 8-second video clip of the selected enthesis was saved with B-mode and PD that was enhanced by the permanence of the contrast agent (CEUS-PD). Each examination (unenhanced PDUS and both CEUS examinations) took 45 minutes.

Video sequence analysis

At the end of the study, all sequences were randomly scored separately by 2 examiners (GM and MAD) blinded to the identity of the subjects, the time of evaluation (V1 through V3), and whether the sequence was before or after contrast injection. Disagreements were resolved by consensus during a second examination. The disagreement was <10% (17 sequences of 210 examined). The PD signal in video clips taken with unenhanced PDUS and CEUS-PD was graded from 0–3, where 0 = no signal, 1 = mild signal (i.e., single-color–marked vessel), 2 = medium signal (i.e., covering <50% of the examined area), and 3 = strong signal (i.e., large confluent marked vessel covering >50% of the examined area).

A semiquantitative scale from 0–2, where 0 = no enhancement, 1 = mild enhancement, and 2 = strong enhancement, was used to analyze contrast medium enhancement measured as slope values with the CnTI 90-second video clips. This video sequence was also exported on a PC (audio video interleave format) and analyzed using dedicated quantification software (Qontrast, version 4.0). This software performed a full map parametric analysis of the enthesis perfusion and also allowed for the segmental analysis of enthesis perfusion inside 2 ROIs (entheseal insertion site and its nutritive vessel). The time/intensity analysis provided a signal intensity–averaged trend, measured on the CnTI images for a period of 90 seconds (see Supplementary Figure 1, available in the online version of this article at http://onlinelibrary.wiley.com/doi/10.1002/acr.22195/abstract).

Statistical analysis

Mean, median, and dispersion parameters (SD and interquartile range) were calculated for quantitative variables. Proportions were used to describe qualitative variables. Comparisons of clinical activity scores between visits were performed with Friedman's nonparametric test for repeated measures with post hoc test if the global test was significant. The PD rank-transformed scores at each visit and before/after contrast agent injection were analyzed by using a linear mixed model taking time (V1/V2/V3) and contrast agent (yes/no) as fixed factors, but the patient as a random factor. P values less than 0.05 were considered significant. Statistical analysis was performed with R software, version 2.10 (R Foundation for Statistical Computing).

RESULTS

Patient characteristics

Of the 14 enrolled patients, 1 withdrew after V1 because of a disease flare 2 days after stopping their NSAID, which led the patient to resume taking their NSAID. Demographic and clinical characteristics of the studied patients are shown in Table 1. Several NSAIDs were used up to V1 and after V2, including meloxicam (n = 5), diclofenac (n = 2), ketoprofen (n = 2), piroxicam, nabumetone, naproxen, flurbiprofen, and phenylbutazone (n = 1).

Table 1. Characteristics of the study population (n = 14)*
 Value
  1. Clinical characteristics correspond to symptoms present at baseline or retrieved from medical history. SpA = spondyloarthritis; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; NSAIDs = nonsteroidal antiinflammatory drugs; DMARDs = disease-modifying antirheumatic drugs; TNF = tumor necrosis factor.
  2. aDefined by an improvement of ≥50% of symptoms in ≤48 hours.
Age, mean ± SD years41.2 ± 46
Men, no. (%)11 (78.6)
Age at first symptoms, mean ± SD years27 ± 7.9
Disease duration, mean ± SD years14.2 ± 38.1
SpA classification, no. (%) 
Axial SpA14 (100)
Ankylosing spondylitis with definite radiographic sacroiliitis (grade II bilateral or III unilateral)6 (42.9)
Family history, %57.1
HLA–B27, %92.8
Articular manifestations, % 
Inflammatory back pain100
Peripheral arthritis50
Heel enthesitis71.4
Extraarticular manifestations, %42.8
Uveitis14.2
Psoriasis21.4
Inflammatory bowel disease14.2
Definite radiographic sacroiliitis (grade II bilateral or III unilateral), %42.9
Acute-phase reactants at baseline 
ESR, median (range) mm at first hour8.5 (3–33)
CRP level, median (range) mg/l7.5 (3.1–16)
Previous treatment 
Number of NSAIDs, median (range)4 (2–4)
Classic DMARDs, %42.8
TNF blockers, %7.1
Efficacy of NSAIDs, %a92.8

No adverse event related to the contrast agent injection was noted during the study. Table 2 shows the clinical disease activity at baseline (V1) and during followup (V2 and V3). Disease activity increased 1 week after NSAIDs were stopped and decreased 1 week after resuming the treatment.

Table 2. Clinical disease activity at each visit
CriteriaVisit 1Visit 2Visit 3Visit 2– visit 1Visit 3– visit 2Visit 3– visit 1General test, P
  1. aMorning stiffness duration was evaluated on the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI).
Morning stiffness duration by Friedman's post hoc test, median (first and third quartiles) minutesa25 (10, 60)60 (5, 120)30 (5, 60)0 (0, 50); P < 0.001−15 (−30, −2); P < 0.0010 (−15, 0); P = 0.11P < 0.001
BASDAI by Friedman's post hoc test, median (first and third quartiles) minutes3.14 (1.88, 4.36)3.74 (2.47, 5.44)1.91 (0.98, 2.98)1.6 (−6.5, 16.3); P = 0.03−6.8 (−23 9, −2.1); P < 0.001−1 (−19.1, 1.4); P = 0.27P < 0.001
Mander index (range 0–90), median (first and third quartiles)9 (4, 16)8 (2, 18)6 (2, 19)0 (−2, 5)0 (−4, 0)−2 (−4, 1)P = 0.53
Number of painful entheses (range 0–27), median (first and third quartiles)2.5 (1.25, 3.75)3 (1, 6)1 (0, 2)0 (9, 2); P = 0.24−2 (−3, 0); P < 0.001−1 (−2, 0); P = 0.08P < 0.001
Analgesics use, no. (%)5 (35.7)10 (69.2)4 (30.8)P = 0.002P = 0.002P = 1P = 0.002

Unenhanced PDUS results

All PDUS vascularized entheses (LA435 probe) and the relationship between PDUS and clinical findings site by site across the time points are shown in Supplementary Table 1 (available in the online version of this article at http://onlinelibrary.wiley.com/doi/10.1002/acr.22195/abstract). Nine certain PD vascularized entheses were detected at V1, 10 were detected at V2, and 9 were detected at V3 (P not significant). Certain PD-positive vascularization was detected more frequently in the lateral epicondyle site. Six PD vascularized lateral epicondyle entheses were detected at V1, 7 were detected at V2, and 8 were detected at V3 (P not significant). In addition, we detected at baseline 6 doubtful PD vascularization of 6 entheses in 6 patients. All doubtful vascularized entheses were detected at the lateral epicondyle (3 at the right lateral epicondyle and 3 at the left lateral epicondyle). When no certain or doubtful PD vascularization was detected, the right lateral epicondyle was chosen. B-mode and PD abnormalities of the selected enthesis with unenhanced PDUS are shown in Table 3.

Table 3. Unenhanced PDUS: B-mode and PD abnormalities of 1 selected enthesis (lateral epicondyle) per patient*
Patient no.B-mode abnormalities at V1PD vascularization (semiquantitative scale)
ThickeningHypoechogenicityBone erosionCalcification/ enthesophyteV1V2V3
  1. PDUS = power Doppler ultrasonography; V = visit; NA = not available.
1NoNoNoNo011
2YesNoYesYesNA31
3YesNoYesYes121
4NoNoNoNo011
5YesNoYesNo333
6YesYesYesYes332
7NoNoNoNo000
8NoNoNoYes211
9NoNoNoNo011
10NoNoNoNo000
11NoNoNoNo000
12YesYesYesYes232
13NoNoNoYes001
14YesNoNoYes2NANA
All, %       
Yes42.914.335.750   
No57.185.764.350   
0    53.830.823.1
1    7.730.853.8
2    23.17.615.4
3    15.430.87.7
Median (range)    0 (0–3)1 (0–3)1 (0–3)

CEUS results

The use of CEUS permitted us to confirm as positive all doubtful enthesitis vascularization at each visit and to show some additional vascularized entheses (1 additional at V1 and V2 and 2 additional at V3 with CEUS-PD as compared to PDUS alone). The distribution of PDUS and CEUS-PD semiquantitative scores at each visit is shown in Figures 2A and B. The mean PD semiquantitative score increased after contrast injection, regardless of which visit (from 0.69 to 0.86 at V1, from 1 to 1.23 at V2, and from 0.85 to 1.08 at V3; P = 0.04). By using the CnTI technology, we detected the enthesitis signal already seen by CEUS-PD at V1 and V3 plus 2 additional vascularized entheses at V2 (in total: 1 at V1, 3 at V2, and 2 at V3).

Figure 2.

Distribution of semiquantitative scores in selected lateral epicondyle entheses at each visit (V; n = 14). A, Power Doppler ultrasound semiquantitative score (LA522 probe before injection). B, Power Doppler ultrasound enhanced by contrast agent semiquantitative score (LA522 probe after injection). C, Contrast-enhanced ultrasonography–Contrast Tuned Imaging technology semiquantitative score (after injection).

Influence of NSAIDs on vascularization detection

By using CEUS-PD, we observed a statistically significant increase in the mean PD semiquantitative score between V1 and V2 (from 0.86 to 1.23; P = 0.03), which decreased when NSAID treatment was resumed (1.08 at V3), although the latter difference (V2 versus V3) was not statistically significant. By using CEUS-CnTI, the global test was significant (P = 0.01). The mean enthesitis enhancement semiquantitative score increased between V1 and V2 (from 0.79 at V1 to 1.15 at V2; P = 0.01) and decreased between V2 and V3 (to 0.85 at V3; P = 0.037). The distribution of CEUS-CnTI semiquantitative scores at each visit are shown in Figure 2C. The number of vascularized entheses with no enhancement decreased from 7 (50%) of 14 at baseline to 3 (23%) of 13 at V2. These data were also confirmed by the dedicated software Qontrast. Parameters such as the peak of intensity and regional blood flow extracted from the time/intensity analysis for the selected enthesitis site were quantitatively more important at V2 than at V1 and V3, although these differences were not statistically significant. Consistently, the mean transit time (MTT) was shorter at V2 (26.7 seconds) as compared with V1 and V3 (28.1 and 32.2 seconds, respectively [P = 0.01]), reflecting a faster enthesis vascularization circulation (Table 4). The time to peak varied in the same direction as the MTT between V1 and V2, but without reaching statistical significance.

Table 4. Contrast-enhanced ultrasound: quantitative results obtained with Qontrast software with the Contrast Tuned Imaging dedicated technology at each visit*
 Nutritive vesselEntheses
V1V2V3V1V2V3
  1. Values are the median (first and third quartiles). V = visit; TTP = time to peak.
  2. aMaximum signal intensity reached during the transit of the bolus at time t equals TTP.
  3. bProportional to the area under the curve.
  4. cRegional blood flow is obtained by the ratio of regional blood volume to mean transit time.
Peak, %a15.7 (14.5–17.6)13.7 (11.6–20.6)14.5 (11.5–22.2)10.2 (7.7–11.3)10.6 (9.4–12.1)10.2 (7.3–12.5)
TTP, seconds14.1 (12–15.4)13.6 (12.0–16.5)13.37 (12.9–14.4)18.4 (15.1–19.7)15.7 (14.9–17.9)15.3 (14.2–16.8)
Regional blood volume, mm2b540 (432.6–618.7)576.7 (287.9–640)416.6 (382.9–759.7)335.3 (273.23–408.9)333.7 (300.5–441.3)404.7 (283.9–562.5)
Mean transit time, seconds25.68 (24.2–28.8)23.4 (21.1–27.8)27.7 (21.7–30.9)28.1 (27.1–31.1)26.7 (22.2–30.1)32.21 (25.1–37.1)
Regional blood flow, mm2/ secondc19.79 (18.1–22)18.2 (14.8–26)18.8 (13.4–27)11.8 (8.7–13.4)12.9 (11.7–15.1)12.5 (9.3–15.6)

DISCUSSION

To our knowledge, this study demonstrated for the first time that a US contrast agent can be used to enhance enthesitis vascular signal detection, and that the detected signal could be modulated by NSAID use. The modification of the PD signal under NSAID therapy was detected either by unenhanced PDUS or by CEUS in both modalities (CEUS-PD and CnTI). In fact, semiquantitative values and quantitative data obtained with posttreatment software were concordant with the use of PDUS alone, which reinforces our results.

Imaging techniques are of great importance and help to detect asymptomatic enthesitis, which accounts for ∼50% of all cases of enthesitis in SpA ([10]). There is currently no gold standard imaging technique to detect enthesitis ([4]). Considering entheses involvement is critical in SpA, it would be useful to develop reliable tools to assess enthesitis in order to improve the management of SpA. Indeed, US seems to be the imaging modality of choice because of its capability to study several entheses at the same examination time and its capability to detect active inflamed enthesitis. However, the detected vascular signal varies depending on the Doppler system and settings used and can be absent because of the difficulty to detect vessels of small size.

Contrast agent injection improves the sensitivity of the Doppler examination and provides additional and more precise information concerning the perfusion of soft tissues ([36]). We demonstrated here that the presence of a doubtful PD signal could be confirmed by CEUS. Furthermore, CEUS confirmed the absence of enthesitis if no signal was detected by PDUS. This study therefore underlines the potential utility of CEUS for the evaluation of inflammatory activity (i.e., vascularization) at the entheseal level. However, despite these promising results, it is difficult at the moment to suggest a wide use of US contrast agent. The relatively short half-life of the contrast agent and the necessity to use a low-frequency probe to favor oscillation of bubbles are real limitations for using the contrast enhancement technique for superficial structures. Therefore, such structures are better evaluated with a high-frequency probe to improve the resolution of tissues. The contrast enhancement technique could benefit in the future from technological advances in the development of contrast agents composed of smaller-diameter microbubbles ([23]). Because of these objective limitations, only 1 enthesis after each bolus can be examined appropriately, which implies to select the right one. In our study, we decided to evaluate the common extensor tendons on the lateral epicondyle because of its accessibility. We could have chosen the heel enthesis, which is also an important and frequently affected enthesis in patients with SpA. However, this site does not permit visualizing the enthesis and the feeding vessel by the same scan. As an alternative, the most painful enthesis or a clinically asymptomatic enthesis could have been examined, but in that case, the US examiner would become unblinded. In order to increase the potential detection of vascularization, we made the hypothesis that NSAIDs can influence the detection of vascular signals at the entheseal level because NSAIDs are an efficacious treatment in SpA. To date, there have been only a few longitudinal studies of the response to therapy of PD entheseal abnormalities in SpA ([18, 20, 37]). These studies were performed in patients treated with anti-TNF therapy. In this study, we showed for the first time that NSAIDs can have an influence on entheseal vascularization. The impact of NSAIDs on the detection of PD inflammation has only been studied in rheumatoid arthritis before. A randomized study has recently shown that NSAIDs may have a significant effect on gray-scale US and PD assessment of synovitis in rheumatoid arthritis ([38]). In our study, we did not demonstrate an effect of NSAIDs on the Mander index; however, previous studies have demonstrated that the clinical examination of entheses is not sensitive enough to readily demonstrate a response to treatment ([15, 39]). In our study, the change in the BASDAI, which increased after stopping NSAIDs and decreased after resuming NSAIDs, was consistent with previously published studies ([20, 29, 30]).

Some limitations of our study should be noted. First, the small sample size and the absence of a control group limited our study. We already know that CEUS is not able to increase the detection of a normal vascular signal of the enthesis in healthy subjects ([16]). Further longitudinal studies evaluating patients with chronic low back pain or other degenerative diseases are warranted to compare their enthesis vascularization by CEUS. Second, the question on the setting for the diagnosis in patients who only have 1–2 entheses showing questionable findings on PDUS is not answered by our data, even if CEUS might be helpful for the identification of doubtful cases.

In summary, the utility of CEUS was proven by its ability to confirm a doubtful PD signal or conversely its absence. The use of CEUS improved the detection of inflammatory enthesitis in SpA patients with a moderate disease activity. Moreover, we showed that enthesitis vascularization was influenced by NSAIDs and increased after 1 week of stopping NSAIDs. Because of CEUS, we could visualize the microcirculation and perfusion of soft tissue, and this led us to improve our knowledge of this major feature of SpA.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. D'Agostino had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Mouterde, Aegerter, Correas, Breban, D'Agostino.

Acquisition of data. Mouterde, D'Agostino.

Analysis and interpretation of data. Mouterde, Aegerter, Correas, Breban, D'Agostino.

Ancillary