|Home | About | Journals | Submit | Contact Us | Français|
To describe (1) a technique which can detect synovial effusions not seen on static ultrasound (US) exam and (2) characterstics of patients with knee osteoarthritis (OA) for whom this technique proved useful.
We reviewed records of patients with knee OA whom we had seen for US-guided injections over a defined period. For all patients, we had recorded physical exam evidence for effusion, presence and size of effusion on US, whether quadriceps contraction was required to demonstrate effusion, and success of joint entry determined by getting synovial fluid and/or seeing an air echo or inflow of injected material.
Over the periods reviewed we saw 76 patients (112 knees) with OA for US-guided injections. 45/112 knees had no effusion on static US exam. With patients’ voluntary quadriceps contraction, we detected suprapatellar fluid in 18 knees of 14 patients. The suprapatellar synovial space seen by US was small (average maximum depth 0.38±0.04 cm). Arthrocentesis obtained 0.5 to 16 mL synovial fluid (mean+SE=2.9±0.6), which correlated with depth of effusion as seen on US with the quadriceps in maximum contraction (Spearman’s ρ = 0.5597; P=0.0157). In 4 knees from which we could not retrieve fluid, we observed inflow of injected material and a linear air echo at completion of injection.
US of the knee while the patient effects quadriceps contraction can find effusions not detectable on static US. Effusions found with quadriceps contraction provide targets for needle entry and their detection thus can improve accuracy of injection.
Using ultrasound (US) to assist arthrocentesis and injection of the knee accurately obtains fluid and places injected materials better than can be achieved using techniques guided only by external anatomic landmarks (1). Accuracy depends largely on US detection of intra-articular fluid – often missed by physical exam (2) – which provides a “target” for the US-guided needle. US-guided knee entry into the suprapatellar pouch should generally be less painful than palpation-guided methods, as the latter are more likely to hit bone while aiming for a site between the femur and patella. In knees where US cannot detect an effusion, guidance of joint penetration reverts to palpation of anatomic landmarks. An US-guided approach to the medial patellofemoral facet has been described for use in knees lacking an effusion (3), but still runs the risk of hitting bone during needle entry. While methods have been described in which US detection of injected air or other contrast can assure that a joint lacking an effusion has been entered (4,5), such techniques still carry the disadvantages of palpation-guided injections.
One of us (WJA), attempting to enhance a effusion in a manner similar to what can be accomplished with manual pressure on the gutters and posterior capsule, asked a patient to contract her quadriceps, and saw that the effusion image enlarged. We have subsequently sought to elicit this phenomenon in patients undergoing knee injection by asking the patient to contract their quadriceps while we continued to scan the suprapatellar region. We have found that this technique frequently shows an effusion where we could detect none previously, allowing us to obtain fluid and inject therapeutic material under direct guidance rather than having to resort to palpation-guided methods. To demonstrate the utility of this simple maneuver, we describe herein patients we saw over a defined period of time in whom this technique proved helpful.
We saw patients in a private practice setting (ELA, WJA) or in a clinical academic practice (RWI). All patients had knee pain that had not responded to more conservative medial management consisting of analgesics, NSAIDs, exercise, and joint protection. Protocol for retrospective data review was approved by the University of Michigan Medical School Institutional Review Board.
All gave verbal informed consent for US-guided arthrocentesis and injection. We entered joints with US guidance as has previously been described (6). With used either an HDI 4000 or GE LOGIQ e machine to perform US exams. We first surveyed the joint to find any evidence of a joint effusion, then located in the longitudinal plane the largest collection of fluid. When we could not detect any fluid, we asked the patient to contract their quadriceps muscle, noting any fluid that then appeared. We asked patients who could not contract their quadriceps on command to dorsiflex and hold against resistance their extensor hallicus longus and tibialis anterior muscles, an action which was then accompanied by quadriceps contraction. With transducer in place we marked an optimal site for joint entry. We cleansed skin then used a sterile compound as acoustic coupler (povidone-iodine gel (ELA, WJA) or sterile lubricating jelly (RWI)). After injection of local anesthetic, we penetrated the joint while visualizing the needle with US. We removed synovial fluid until the suprapatellar cavity collapsed, then delivered material (corticosteroids, hyaluronates, or platelet-rich plasma according to judgment of the treating clinician), visualizing its entry into the space. In some instances, a sharp linear echo produced by the small amount of air contained in the syringe with therapeutic material provided further assurance of proper joint entry (4). We advised patients to limit their activities for the next 24–48 hours.
We computed summary statistics (means, standard errors), and examined data distributions. We used Spearman’s rank correlation coefficient to assess the magnitude and significance of association between effusion dimension and synovial fluid volume, as the latter variable was not normally distributed. We used Stata 10 software (StataCorp, College Station, TX) to analyze the data.
Combining experiences from a 4 month interval (RWI) and 9 separate clinic days (ELN, WJA) we saw 76 patients with symptomatic knee osteoarthritis (OA) for US-guided injection treatment. Of the 112 knees of these patients, 45 did not show an effusion on static US exam. With patients voluntarily contracting their quadriceps, we detected suprapatellar fluid in 18 knees of 14 patients. Characteristics of these patients are listed in Table 1. There were 13 women and one man, aged 44–88 (mean±SE = 64.7 ± 2.9). Most were obese (BMI 20.1–49.9, mean±SE = 32.7 ±2.3), with 3 classifiable as morbidly obese. OA stage as determined by X-ray tended toward the more severe, with Kellgren stage III or IV seen in 10 of the 14 knees for which xrays were available. Physical examination suggested presence of an effusion in only one knee (positive bulge sign).
The suprapatellar synovial space detected by US was small (average maximum depth 0.38 ± 0.04 cm) (Figure 1). Volume of fluid obtained at arthrocentesis ranged from 0.5 to 16 mL (mean+SE=2.9 ± 0.6) and correlated with depth of effusion as seen on US with the quadricep in maximum contraction (Spearman’s ρ = 0.5597; P=0.0157). We did not seek to evacuate all synovial fluid in every case, as in some instances contraction of the synovial cavity with aspiration threatened to reduce the space beyond reach of the needle tip. In 4 cases, we could not obtain synovial fluid despite visualizing the needle tip within the synovial space. In all these cases, we observed injected material flow into the space and also saw a linear air echo (Figure 2). No patient reported immediate complications of the procedure.
We here have described a simple technique that can detect synovial fluid in knees for which static US had failed to find evidence for an effusion. Detecting these occult effusions permitted aspiration of synovial fluid and confirmed intra-articular placement of injected material. These successes came in patients who often presented challenges of obesity and advanced arthritis. Such cases comprise an ever growing number of our patients as the general population ages and becomes more obese.
Studies analyzing pressure-volume relationships in normal and arthritic joints have interpreted findings mainly as they might contribute to pathogenesis both of symptoms and of joint damage arising from changes in synovial perfusion, supporting structures and periarticular musculature (7,8). However, some observed phenomena are applicable to clinical practice, particularly arthrocentesis. Joints accommodate their largest volume when intra-articular pressure is least, which for the knee occurs at zero degrees flexion (9), increasing very little with flexion up to about thirty degrees (10). Increased intra-articular pressure forces any synovial fluid to areas of greatest compliance (11), which for the knee are the suprapatellar recess and peripatellar gutters. Quadriceps contraction substantially increases intra-articular pressure (12), which would increase the amount of fluid in these compliant regions. Hence, we – as all rheumatologists - generally position patients for arthrocentesis with the knee extended or in slight flexion, but can produce larger amounts of suprapatellar fluid by having the patient contract their quadriceps.
It is well established that US detects many knee effusions missed by physical exam (2) and improves the accuracy of knee arthrocentesis (1). However, in some circles it remains controversial whether the improved accuracy of US-guided injections translates into a better clinical outcome (13). Such a relationship has been shown by prospective controlled trials only for the shoulder (14), and a recent study could find no difference in outcome for a number of different RA joints in which injection was performed with guidance either by palpation or US (15). An older prospective study found superior outcomes following corticosteroid injections to RA knees in the group where joint drainage preceded injection (16); at least some of this difference could be interpreted as due to the greater chance that knees first drained were more likely to have been successfully entered. Intra-articular injection of biologic compounds - such as hyaluronates, growth factors (17), and anti-TNF agents (18) – is becoming a more common clinical practice. Any differences in outcomes following palpation-guided versus US-guided injection of knees with these compounds have not been assessed. However, because these larger compounds would not likely diffuse into a joint from intra-articular placement – unlike corticosteroids – it seems intuitively evident that intra-articular treatment with these agents would be more effective if accurate placement were assured.
Hence, US has an important role in guiding needles into joints. Obtaining synovial fluid for analysis is easier if the arthrocentesis needle is localized to a fluid containing space. While not yet proven, it is likely that therapeutic injections – especially with biologic compounds – will work better if placed intra-articularly. When US of the knee does not detect a synovial space, repeat exam with the patient voluntarily contracting the quadriceps muscle can often find fluid, then amenable to an accurate aspiration and injection that would not otherwise be possible.
ECS is supported by NIH CTSA grant UL1RR024986