Power Doppler sonography (PDS) was first introduced for cardiologic investigations in the 1980s (1, 2). Since it proved useful in cardiology, PDS was soon applied to other medical diagnostic problems (3–11). PDS characteristically encodes the amplitude of the power spectral density of the Doppler signal, rather than the mean Doppler frequency shift as in conventional color Doppler ultrasound methods (12). While conventional color Doppler sonography is well suited for evaluating high-velocity flow in large vessels, it is less effective in detecting low-velocity blood flow at the microvascular level (13–15). The value of PDS in the detection of soft tissue hyperemia was reported by Newman et al in 1994 (16). Recently, its value for estimating the fraction of moving blood in tissue was confirmed (17–19).
There are several studies on the visualization of the synovial membrane with PDS in osteoarthritis (OA) and rheumatoid arthritis (RA) (20–24), but to our knowledge, no study has been published which compares PDS findings and histopathologic findings of vascularity of the synovial membrane. The present study evaluated the correlation between PDS imaging of the vascularity of the knee joint synovial tissue and histopathologic inspection of the same tissue in order to assess the value of PDS in the imaging of synovitis.
- Top of page
- PATIENTS AND METHODS
The characteristics of the study patients, the ultrasound findings including PDS, and the joint effusion, synovial thickness, and histologic findings are shown in Table 2. The correlation between the qualitative results of PDS and the pathologists' estimation of vascularity was 0.89 by Spearman's ρ (P < 0.01) (Figure 3).
Table 2. Clinical characteristics, medications, and PDS and histologic findings in the knee joint of the study patients*
|Patient/ age/sex||Knee||Clinical diagnosis||Disease duration, years||No. of ACR criteria||Quantitative score||Qualitative score||CRP, mg/dl||BP, mm Hg||Heart rate||Cardioactive and antiinflammatory medication|
|Histology, mean ± SD||PDS||Histology||PDS||Effusion||Synovial proliferation|
|1/59/F||L||RA||12||6||36 ± 12||3.0||2–3||3||2||4||3.5||155/75||66||NSAID, MTX|
|2/80/M||L||OA||15||–||25 ± 14||2.0||3||2||2||2||0.1||140/85||75||NSAID|
|3/70/F||R||RA||25||5||29 ± 18||2.5||2–3||2||3||2||2.7||160/80||70||NSAID, MTX, corticosteroids, SSZ|
|4/29/F||R||RA||5||4||32 ± 11||2.0||3||3||3||3||1.9||140/80||68||NSAID, MTX|
|5/66/F||R||RA||23||5||41 ± 12||3.0||4||3||3||3||1.0||140/80||82||NSAID|
|6/76/M||R||OA||8||–||17 ± 7||1.5||1–2||1–2||3||1||0.0||160/100||78||NSAID, ISDN|
|7/78/F||R||RA||14||5||29 ± 11||2.5||3||3||2||3||0.7||140/60||80||NSAID|
|8/75/M||R||OA||5||–||14 ± 9||0.5||1–2||1||3||1||0.3||140/90||80||NSAID|
|9/60/M||R||OA||7||–||10 ± 8||0.5||1||1||2||2||1.1||160/90||85||NSAID|
|10/78/F||R||OA||27||–||16 ± 10||0.5||2||1||1||1||0.7||170/100||88||NSAID, ACE inhibitor|
|11/66/F||L||OA||9||–||10 ± 3||1.5||1||1||2||2||0.6||160/90||60||NSAID, ACE inhibitor|
|12/52/F||R||OA||4||–||23 ± 9||2.5||2–3||2||1||1||1.4||145/90||76||NSAID|
|13/76/F||R||RA||11||6||41 ± 11||3.0||4||3–4||4||4||5.8||150/80||79||NSAID, MTX, corticosteroids|
|14/75/F||R||OA||16||–||17 ± 10||0.5||2||2||3||1||0.5||140/70||84||ISDN|
|15/86/M||R||RA||31||7||33 ± 14||2.5||4||3||3||3||2.4||175/90||78||NSAID, calcium antagonist|
|16/77/F||L||OA||3||–||18 ± 13||0.5||2–3||1–2||1||2||0.6||125/80||80||Paracetamol, ACE inhibitor|
|17/65/F||L||OA||15||–||17 ± 6||1.0||1–2||1–2||1||2||1.4||140/60||68||No medication|
|18/79/F||L||OA||9||–||26 ± 18||3.0||3||3||3||1||0.0||160/90||76||ACE inhibitor|
|19/76/F||R||OA||8||–||33 ± 21||1.5||2||2||2||1||2.0||160/100||80||NSAID, ACE inhibitor|
|20/69/F||R||RA||22||5||45 ± 12||3.5||4||4||3||4||1.2||135/80||74||NSAID, ACE inhibitor|
|21/81/F||L||RA||13||6||42 ± 22||2.0||3–4||2–3||2||2||1.6||140/60||64||NSAID, ACE inhibitor|
|22/56/F||R||RA||17||5||54 ± 26||3.5||4||4||2||3||1.7||120/70||76||NSAID, MTX, corticosteroids|
|23/78/M||R||OA||15||–||19 ± 12||2.0||2||2||3||2||0.0||150/80||78||ACE inhibitor|
Figure 3. Correlation between the qualitative estimates of blood flow by power Doppler sonography (PDS) and vascularity in the tissue section (Spearman's ρ = 0.89, P < 0.01).
Download figure to PowerPoint
The visual estimation of vascularity based on PDS images and on tissue sections was compared with the results of digital image processing by Spearman's rank correlation test. The correlation between visual and digital interpretation of PDS was 0.89 by Spearman's ρ (P < 0.01) (Figure 4). The correlation of visual and digital analysis of the factor VIII–stained sections (immunohistochemistry) was 0.88 by Spearman's ρ (P < 0.01) (Figure 5). Both values support the reliability of the visual interpretation of the tissue sections and the PDS images by the examiner.
Figure 4. Correlation between the visual (qualitative) interpretation of power Doppler sonography (PDS) images and PDS digital image processing (Spearman's ρ = 0.89, P < 0.01).
Download figure to PowerPoint
Figure 5. Correlation between the visual and digital image analysis of factor VIII–stained tissue sections (Spearman's ρ = 0.88, P < 0.01).
Download figure to PowerPoint
Pearson's correlation between the digital analysis of the PDS images and the digital analysis of the factor VIII–stained tissue sections was 0.81 (P < 0.01) (Figure 6).
Figure 6. Correlation between the digital analysis of power Doppler sonography (PDS) images and digital analysis of factor VIII–stained tissue sections (Pearson's correlation coefficient r = 0.81, P < 0.01).
Download figure to PowerPoint
We found a correlation between the thickness of the synovial membrane and the PDS signal (Spearman's ρ = 0.69, P < 0.01). Spearman's ρ for correlation of the thickness of the synovial membrane and vascularity in the tissue section was 0.64 (P < 0.01). There was no correlation between synovial proliferation and effusion (Spearman's ρ = 0.223, P = 0.31).
There was a significant increase in the power Doppler signal for the numerical PDS score (P < 0.01) as well as the qualitative grade (P < 0.01) in patients with RA. Digital (P < 0.01) and visual (P < 0.01) analysis of the immunohistochemistry sections showed a higher degree of vascularity in patients with RA, and hypertrophic synovium was more frequently found in patients with RA (P < 0.01). Effusion was found in both groups, but there was no significant between-group difference.
- Top of page
- PATIENTS AND METHODS
Synovial proliferation is the fundamental event in rheumatoid joint lesions (35). Growth of fibroblast-like synoviocytes and metalloproteinase production by fibroblast-like synoviocytes contribute to cartilage and bone destruction associated with hypervascularized tissue containing fibroblastic elements referred to as pannus (36, 37). A joint affected by inflammatory tissue as seen in RA requires a specific treatment, such as disease-modifying drugs, local injection of corticosteroids, radionucleotide synoviorthesis, or synovectomy. Visualizing inflammatory tissue is therefore an important element in the diagnosis and monitoring of disease activity in RA.
So far, visualization of synovial tissue and synovial vascularity has been dominated by magnetic resonance imaging (MRI). The ability to differentiate between effusion and synovial proliferation by MRI has been greatly improved by the introduction of the intravenous application of paramagnetic contrast agents (38–44). Van Dijke et al (45) described a technique for measuring abnormal capillary permeability in synovial tissue of arthritic knees of rats, by use of dynamic MRI with an intravenous gadolinium-based blood pool agent. In that study, MRI-derived microvascular characteristics correlated positively with histologic findings. O'Byrne et al (46) compared MRI and histologic findings in a rabbit model of OA and immune arthritis. Those investigators reported that MRI can be used to observe the therapeutic effects of disease-modifying drugs on synovial inflammation and cartilage degradation in rabbit knees. Since enhancement of the synovium is time dependent, the sensitivity of MRI can be reduced by chemical shift artifacts (47). Consequently, for a detailed assessment of disease progression based on serial MRI examinations, precise timing and standardization of the MRI protocol are required (48, 49).
The value of PDS in detecting low-velocity blood flow at the microvascular level in several tissues has been demonstrated by Newman et al (16). Strouse et al (50) did not consistently find increased flow in their study of septic arthritis of the hip in children. Their investigations in a rabbit model confirmed their previous findings in children (24). In the latter study, only 23 of 45 examinations of infected knees were unequivocally positive by PDS examinations performed 1–6 days after inoculation. However, several groups of investigators have reported that ultrasound imaging of the synovial tissue is an objective and reliable technique (16, 21, 23, 28, 51–53).
Fiocco et al (21) compared conventional ultrasound imaging of synovial proliferation with arthroscopic visualization before and after synovectomy in patients with knee joint synovitis and found a significant correlation between clinical and ultrasound indices. Those authors concluded that sonography can be used as an objective method for monitoring response to therapy in patients with inflammatory knee joint diseases. Silvestri et al (54) used color Doppler to monitor the activity of RA in the knee joint and found a correlation between vascular findings and clinical symptoms.
In the present studies, effusion was found in patients with OA as well as in patients with RA, with no significant between-group differences. The correlation between the thickness of the synovial membrane, the PDS signal, and tissue vascularity is consistent with clinical experience; however, a thick synovial membrane or effusion does not necessarily mean that inflammation is present. There was no correlation between synovial proliferation and the presence of an effusion. Tissue debris, blood clots, and fibrin are known to mimic some ultrasound features of synovial proliferation (21), and these features can be excluded by PDS.
PDS reflects the movement of blood cells within a vessel; however, it does not always indicate increased vascularity or hyperemia of the synovium, which is still a problem in interpreting PDS images (55). PDS findings are influenced by the examiner, the machine, and the acoustical conditions involved in image processing. A sonographic evaluation of the arthritic joint should also be performed. PDS should be used to assist the clinician in determining whether the region of interest shows increased blood flow compared with other tissues (29). This information can be important in distinguishing between hypervascular and fibrous pannus. The value of PDS in the assessment of therapeutic response to the treatment of synovitis of the knee joint was shown by Newman et al (23), who reported a qualitative decrease in synovial perfusion visualized by PDS after intraarticular administration of steroids.
Our study is the first to correlate PDS findings with histologic findings in the synovial tissue, supporting the value of PDS in clinical practice, where the degree of vascularity can be graded according to PDS images. The best correlation was found when a 1–4 grading scale was used by both the sonographer in assessing PDS signals and the pathologist in assessing the degree of vascularity.
Although techniques of digitizing the images and analyzing blood vessels and PDS signals are automatically assumed to be a more accurate approach, this method can be influenced by technical artifacts. Fat cells with intensely colored cell membranes can mimic a high vascularity in the segmentation algorithms of the image analyzing software. Besides technical considerations, the cost of PDS investigations is of increasing concern in our health care systems. The results of the present study, however, indicate that PDS is a powerful tool in the evaluation and objective assessment of synovial tissue perfusion. PDS represents a readily available, easy to handle, and cost effective alternative to dynamic MRI with intravenous gadolinium-based contrast agent. In the future, the use of ultrasound contrast agents and harmonic imaging, both of which greatly enhance color flow sensitivity, may be shown to increase the potential utility of PDS.
In conclusion, PDS provides a reliable and accurate method for visualizing blood flow in the synovial tissue. The highly significant correlation between the PDS findings and the histologic findings supports the value of this technique. Since ultrasound is an easy to handle, safe, inexpensive, and noninvasive procedure that is available in most departments of rheumatology, PDS may continue to play an important role in the diagnosis and monitoring of musculoskeletal disorders.