Discussion
This is the first report of clinical application of MRgFUS for knee OA. Real-time monitoring of the sonication area and temperature elevation enabled performance of safe and accurate treatment. Even though the inclusion was restricted to most severe OA in this pilot study, 75% of patients showed successful pain relief. Similar to previous reports of bone metastases[13–15] or lumbar facet OA,[16] the pain alleviation was rapid and long-lasting. Unlike a conventional transducer integrated with MRI table, a newly developed conformal sonication device was a good fit for extremities and enabled easier treatment of knee OA. Intravenous sedation and opioid administration were not necessary for our treatment, which were applied in the previous series of bone metastases[13–15] or low back pain.[16] Local anesthesia with ropivacaine around the periosteum was enough to reduce pain associated with sonication. Patients were able to relax throughout the procedure and to walk soon after treatment.
The mechanism of pain alleviation is most likely local denervation caused by the heat denaturation of the treated area. However, no previous studies have suggested an assessment method to estimate the denervated area of MRgFUS. In this regard, pressure algometry is a quite simple and useful tool for quantitative evaluation after denervation treatment. Reliable repeated PPT measurements around knee joint have been documented by means of locating the assessment sites in relation to bone landmark.[20] In the present study, all sites in medial knee were easily identified based on the location of joint space, medial collateral ligament and tibial osteophyte, enabling to retest PPTs in a reproducible manner. In our patients who responded to the treatment with pain reduction, PPTs on sonication area were significantly increased after treatment, which means that patients felt less pain by pressure stimulation after denervation. The patients who did not respond to the treatment did not show increased pressure pain thresholds which may suggest that an PPT increase would be a necessary condition of successful treatment. Future studies will be needed to verify if earlier follow-up assessments may be used to predict treatment success.
In this series, all sonications were applied to bone surface just below the rim osteophyte of medial tibia plateau. From a pathophysiological perspective it has been reported that sensory nerve invasion containing substance P and calcitonin gene related peptide was seen in tibial osteophyte in human OA patients.[21] Because surface area of the tibial rim osteophyte itself was a bit narrow to plan sonication, the base of the osteophyte was treated instead. Furthermore, lower PPTs were observed in this area at pre-treatment in all patients and this is also a general finding in OA knees.[17] In other words, hypersensitivity of nociceptive nerve terminals against pressure stimulation was seen in this area, which was preferable for denervation treatment. From a practical perspective there were other reasons to select the treated area. Firstly, bone is a better indication for MRgFUS than soft tissues. Lower thermal conductivity and higher ultrasound absorption rate of cortical bone allows the denervation treatment safe and efficient, which had been demonstrated in previous reports.[13–16] Secondly, tibial rim osteophyte was a good landmark for reproducible planning, treatment and assessment.
The osteosclerotic change after the treatment was interesting. In our patients, the temperature elevation of bone surface was aimed at 60°C because protein denaturation occurred above the temperature of 57–60°C for a few seconds.[22,23] The goal temperature was almost same as previous publications, and some authors found similar new bone formation after the treatment of bone metastases.[13,14] The mechanism of osteosclerotic change in the treated area is unknown. Although it cannot be excluded that minor thermal or non-thermal bone damage occurred, new bone formation might be an encouraging radiological finding of this therapy.[24] Including its relation to the long-lasting pain relief, further basic research of treated bone marrow would be necessary to assess this phenomenon.
According to the pain alleviation mechanism and results of this study, a good candidate for MRgFUS treatment is patient presenting with localized medial pain, lower PPTs around tibial osteophyte, and no bone marrow lesion or osteonecrosis. In this series, two patients did not respond the treatment. One patient complained spreading medial knee pain and the other had small bone marrow lesion in medial femoral condyle and tibia plateau. Detailed assessment of pain distribution, pressure pain sensitivity, and MRI examination before the treatment might be essential to achieve satisfactory results.
Percutaneous radiofrequency treatment has been reported as a beneficial local denervation therapy for knee OA.[4,25] Comparing with radiofrequency, MRgFUS treatment has some advantages. Closed-loop, real-time spatial and thermal monitoring enables the treatment safer and more accurate. Identifying target nerve is not trivial and the outcome is highly technique-dependent in radiofrequency.[25] MRgFUS treatment is not a technique-dependent procedure and low inter-operator variability is expected. MRgFUS treatment does not cause widespread hypoesthesia which often observed in radiofrequency treatment,[25] because MRgFUS treats most peripheral zone of the sensory nerve. On the other hand, there are some obvious disadvantages of MRgFUS. Enormous initial cost of the treatment is most critical. In addition, patients of contraindications for MRI cannot undergo the treatment. Required time for the set up and treatment is also longer than radiofrequency.
This study has some limitations. First, the most important weakness is that it was a case series including small number of patients without control group. Hence, it is difficult to be sure that there were no placebo effects. However, 75% of patients showed successful pain relief along with significant increase of PPTs. Although further study with blinded and randomized controlled trial is required for constructing evidence, our initial results suggested the safety and efficacy of the treatment. Second, the inclusion was restricted most severe medial knee OA because this is a pilot study so that the patients should be salvaged by TKA conversion. Based on this study and the mechanism of pain relief, medial OA in earlier stage or lateral OA might possibly become a candidate for the treatment. Third, this study did not have a long follow-up period. Two patients underwent total knee arthroplasty and one non-responder dropped out one month after the treatment. Long-term effectiveness of MRgFUS treatment including ADL and QOL assessment should also be carried out in a continuing study.
BMC Musculoskelet Disord. 2013;14(267) © 2013 BioMed Central, Ltd.
© 1999-2006 BioMed Central Ltd
