You are here
Applying ultrasound and MRI for managing rheumatoid arthritis patients
Interview with Prof. Philip Conaghan
We are meeting today with Professor Philip Conaghan in Leeds to learn from him the ways of applying ultrasound and MRI in the management of RA patients.
We are particularly interested, Phil, in how one might use these modalities to manage patients when one is trying to refine or optimise treatment?
Professor Conaghan: I think the major lesson we have learnt from modern imaging, be it ultrasound or MRI, is that it’s much more accurate than our fingers are in detecting synovitis. We know from old studies that it detects a lot of inflammation that we don’t feel with our fingers, and it is not just small amounts of inflammation, it is inflammation that, for some reason, we can’t detect. It maybe the shape of that particular joint, or that individual’s joint. So we don’t just miss low levels of inflammation, sometimes we miss a three-out-of-three synovitis in a joint, and detect a one-out-of-three. So there is no real pattern to what we miss or detect, but that’s really the message for modern imaging is we often miss inflammation, we are not very accurate with our fingers. Knowing that information, it should help our management, but if we have been lacking good evidence I think that how would we use that in the real world?
So, quite a few studies now confirming the concept of sub-clinical inflammation, and showing that imaging-based inflammation predicts what happens to your joints in terms of erosions better than clinical-examined detected inflammation.
So the concept of what we are picking up is important and proven, but how do we use it in clinical practice? In terms of how we change our management of rheumatoid arthritis as a result of knowing about the sensitivity of these tools, we should judiciously use imaging in situations of clinical doubt. So when I am seeing a patient and starting their methotrexate, or other disease-modifying drugs, most of my clinical decisions are still made on a clinical assessment, including morning stiffness, swollen joints, and blood tests that measure inflammation. I don’t use imaging for a lot of patients because it is quite clear that they need more therapy, and I need to escalate.
Interviewer: So, Phil, I think what would be particularly helpful in this segment is to focus on managing patients who are achieving low-disease activity remission, and the consideration of optimising treatment, perhaps tapering a biologic – some examples like that.
Professor Conaghan: I think although we are lacking good clinical algorithms at present in how to use modern imaging, for me, the patient who is doing poorly, who is not doing well, is not the patient who probably needs imaging. So where there are a lot of swollen joints, a lot of morning stiffness and elevated inflammatory markers, they will go onto escalation.
Probably the greatest benefit we get from modern imaging and applying that sensitivity in detecting inflammation is in the patient who we believe is in a low-disease activity state. So in patients who we think are doing well, and who have everything else all right, my vision is that for those patients we do imaging and what type of imaging you use will probably depend on feasibility and the local circumstances, but an imaging check to see if, indeed, this patient’s disease activity is as low as you perceive.
So I think that’s a good time to consider doing a one-off imaging check is, is this patient really in low-disease activity? Yes they are. Great, I can hold my therapy for a while. Or, indeed, I am seeing three or four joints with active Power Doppler signal on ultrasound, or with active MRI synovitis. I might need to actually step up the therapy in this patient who I thought was doing well.
I think what we are lacking in real clinical practice is knowledge of how much inflammation we see with the various imaging modalities is an important amount of inflammation. What is clear, though, is that it is the patient who has discrepant values. It is the patient who has a low-disease activity state clinically and who you find a lot of inflammation on imaging, or vice versa. Those two situations are the ones where there is a big change of management warranted.
When we come back to the issue of how much inflammation is important we probably know a little bit more from ultrasound studies than from MR studies. From the ultrasound studies there have been some studies from a Spanish group led by Esperanza Naredo, and a Japanese group, both of whom looked at levels of inflammation on ultrasound using multiple joint counts, multi joint ultrasound counts, and showed that high PD signal in particular, that is Power Doppler signal, the vascular inflammatory high correlated signal, is related to your risk of flare once you reduce drug doses. So in short, if you think you have got somebody into a low-disease activity state on a particular DMARD, or biologic DMARD, if you are thinking about tapering drug, it is very worthwhile having imaging at that point to see whether this is the right candidate.
Of course, I wouldn’t do it too quickly after you have the person into that low-disease activity state. If you are wanting to reduce tapering, of course, you want to do that after somebody has been in the disease activity state for a period of time.
Interviewer: That is quite clear. I am wondering how we advise practising rheumatologists around the world? How does one acquire the skill of ultrasound, for example, which obviously can be used in the office if one has the appropriate equipment?
Professor Conaghan: I think that is a big question. I think, first of all, let me back up and answer that in a couple of ways. How do we use ultrasound, for example, in clinical practice? As I alluded to, we are still not quite sure how much inflammation is important. Indeed, if you had a big lump of Power Doppler three-out-of-three scored synovitis in one joint, is that more important than having three joints with a grade one Power Doppler signal? So we still need to understand how much is acceptable, and I don’t think we are quite there yet with understanding those levels of inflammation.
Also, a lot of this literature is based around the Power Doppler signal, and whereas that is more related to inflammation in the sense it’s highly related to vascularity, grey scale is not a benign feature, and in people, histologically, who have grade one to grade three grey scale can have quite a lot of inflammatory infiltration and potential on-going damage risk. So I am concerned that we do need to understand more the levels of inflammation that are critical to driving dose.
How do people learn skills like ultrasound? I think it differs from country to country. A lot of countries around the world have national programmes, and others have local programmes that are looking at teaching. ULA offers an online basic course now that anybody can do at a distance, but it is good to have hands-on training and it is not good enough to just buy a machine and start doing it. You do need some tuition along the way, and everybody gets skills over time. We know that even with careful teaching it’s probably three months before you’re useful at looking at small-joint synovitis in the hands, and it can be two years of doing quite a lot of scans before you are any good at scanning a shoulder, which is a much more complex joint to scan. So there is a learning curve, it is dependent on joints, and it might be dependent differently on identifying synovitis to identifying a joint line so that you can do a joint aspiration or injection. There are slightly different skill sets involved. It all comes down to adequate training, and I think ACR has courses in America. A number of regional groups have courses in America, and there are a number of groups around every country we know that are doing training.
Interviewer: So if we just take a broad overview and consider the way that ultrasound is increasingly being applied in patient management, is this a skill that the average rheumatologist should be acquiring?
Professor Conaghan: I wrote an editorial, I think, back in 2005, where I referred to ultrasound as our stethoscope, as being the extra tool that we needed for looking at joints, and I think there is no doubt that that is true today. We have learnt a lot more about how it helps with diagnosis, and in this interview we have just been talking about management of RA, but there is a whole lot of issues that it helps with.
For patients, for example, and this will be a common example to clinicians who see a lot of RA patients, they are not divorced from having osteoarthritis. They have the same tendonitis and mechanical joint pain as everybody else, and it’s quite comforting to be able to look at a joint that’s swollen and tender in the presence of normal sed rates and normal CRPs, and see those osteophytes and reduced joint space that tell you the reason that joint isn’t getting better with therapy is because it has osteoarthritis in it.
So I think helping in differential diagnosis, both at the joint and the patient level is one of the obvious examples, and guiding therapies and difficult aspirations is a useful thing for a lot of clinicians.
Interviewer: That is a really helpful perspective, Phil. Thanks very much for sharing.
Thinking about where we are with the studies that have looked at tapering, Georg ?Shevers is really very firm on requiring patients to be in remission before starting tapering. He doesn’t talk about ultrasound remission so much, but the clinical remission. What is your current perspective?
Professor Conaghan: I think clinical remission is such a mixed bag of active disease, or not, that whether you are in low-disease activity or remission is probably not a big difference in terms of the presence of ultrasound inflammation. I think, and there might be some data in an old paper - maybe not that old - a 2010 paper from Benazir Saleem in our group, where she looked at different DAS remission states and amount of grey scale ultrasound and PD, and it was pretty much the same across the low-level disease states.
Interviewer: Oh, okay.
Professor Conaghan: So that would argue against it mattering whether it was in low disease or remission. You probably need to check.
Interviewer: Okay. So the big question that is looming at this time is whether one needs another domain to define remission, which would be an imaging domain?
Professor Conaghan: Yes, and I think what we called it in OMERACT at one stage was an acceptable state. So I think that imaging remission makes people think of zero. I would rather think of it as acceptable disease activity state since it is pretty well impossible to get anybody who has had RA for more than a year or two, especially if they are over 50, to have no synovitis in their joints because they all have a bit of OA.
Interviewer: Sure, yes.
Professor Conaghan: So I think an imaging acceptable disease activity state is where we are heading.