What is the paradigm change in terms of therapy for rheumatoid arthritis? In the last few years, we understand the severity and the nature of rheumatoid arthritis more than we ever have. What are those paradigm shifts?

• We're using disease modifying antirheumatic drugs (DMARDs) earlier.
• We're using aggressive methotrexate therapy.
• We're using combination therapy.
• We're using biologics.

In terms of early DMARD use, there are 2 issues that most physicians want to address. If I use a DMARD early, will it make a long-term difference to my patients? And, if I use a DMARD early, will I have a better clinical response? The concept that initiating a DMARD early may have a longer-term benefit on the patient. Are there any data to support that? There are at least 2 studies to suggest it. The first is a study by Lard. He took a group of patients who, within 2 weeks after presenting with their diagnosis, were started on a DMARD. He took another group of patients and waited a mean of 4 months and then started a DMARD. The delay, on average, is about 4 months. These populations weren't exactly matched, but they were pretty closely matched. If you look at the radiological outcome, the Sharp score, those patients who were treated early had a reasonable reduction in Sharp score, whereas those patients who were only 4 months delayed on average had quite a significant increase in Sharp score over time. That suggests earlier treatment will have a long-term benefit.

The second study has to do with disability outcomes. This study by comes out of British Columbia. They did a hydroxychloroquine (Plaquenil) trial. They did a 9-month, double-blind, placebo-controlled trial. At the end of 9 months, they put the placebo patients on Plaquenil. They watched the patients who got Plaquenil at the beginning of the study, and watched those patients receiving Plaquenil 9 months later, ie a delay, to see whether, long term, there would be a difference. If you look at improvement in those patients who got Plaquenil at the beginning, in terms of disability, there is a 57, 58, and 60 in standard deviation units. The patient population that started 9 months later, out to 33 months later, didn't match their counterparts. They still had more disability after 3 to 4 years and were not up to the base of those individuals who started on Plaquenil initially. Aggressive, early therapy could make a long-term difference. We've seen it radiologically and in terms of clinical outcomes.

What about methotrexate? What about better response to DMARDs in early versus late disease? If I use a DMARD early, will I have a better response than late? There are not a lot of data to support that. I'm going to give you 2 pieces of data that are certainly suggestive, but not definitive. One of these pieces of data comes from the trials with etanercept. They looked at the Stanford Health Assessment Questionnaire (HAQ) outcomes, the disability outcomes, in patients in the Moreland trial, which was monotherapy late disease, an average of about 10 to 12 years of disease, versus the outcomes in the Early Rheumatoid Arthritis (ERA) trial in patients who were treated with under 3 years of disease. Other than disease activity, for the baseline outcomes of tender joints, swollen joints, etc, they were comparable. Investigators had 2 comparable populations, 1 early, 1 late, and they asked the question: Using the same dose of etanercept, do you have a better response? These data show, in the ERA trial, there was a difference between these 2 populations, a better improvement in HAQ if you treated early.

The second study is a meta-analysis. This is a study by Anderson. It's probably the only study done in terms of asking the question, "Earlier treatment, better response?" This is published in Arthritis and Rheumatism. There were various DMARDs used. They looked at Plaquenil, azathioprine (Imuran), and methotrexate. They even looked at Prosorba (protein A-based extracorporeal immunoadsorption, silica). The concept is, if you treat people with a short disease duration, regardless of their treatments, in general, your sedimentation rate is improved, your swollen joint count is improved, and your tender joint count is improved. This is the proportion of those people who are improving by at least 20%. If you treat them late, if they're treated at 5 to 10 years, there is a significant difference. Suggestive, but not definitive, data indicate if you treat earlier, you're more likely to have a response.

What about aggressive methotrexate dosing? Are there any data to say that higher doses of methotrexate make a difference versus lower doses? Dan Furst did a study. He asked the question whether 20 mg of methotrexate over 14 weeks is any better than 10 mg over 14 weeks. Here is the percentage of patients with American College of Rheumatology (ACR) 50 and with improvement of at least 50%. The 20-mg dose improved more at tender joint count, swollen joint count, global joint count, and activities of daily living (ADLs) relative to 10 mg. These data say higher doses work better than lower doses, that 20 mg is better than 10 mg.

The median dose of methotrexate used about 3 years ago, when they did a pharmaceutical survey in the United States, was 11 mg. The ACR recommends that 25 mg subcutaneous injection be tried before the drug is considered "inadequate" response. Many of the so called "comparison" trials e.g.,leflunomide,early etanercept etc., trials were done at the 20 mg level. In 1999 studies determined that MTX was a "true" disease-modifying drug. The trial results indicated that 25 mg subcutaneous injection of MTX was more efficacious than oral MTX of 20 mg,with lower side effects.

Another study that suggests the possibility that methotrexate works well is the Early Rheumatoid Arthritis trial. We think about etanercept as one of the gold standards

Can we take methotrexate, which is our gold standard, raise the dose in 2 months to 20 mg, and compete with etanercept? In the first 4 months, you don't compete. To an extent, etanercept worked quicker. By 12 months, the 2 look almost identical. There is only a 5% difference between etanercept and methotrexate in terms of clinical outcomes at 1 year. These are very good outcomes. Most of the other disease modifiers and biologics aren't going to beat methotrexate. So why use it? Or, why not use it? Here is the concept. Radiologically, at 2 years, there was a minor difference between etanercept and methotrexate. The Sharp score point differences were about 2. However, it's clear that, over time, radiological outcomes are linear. It was thought to be only 6 years, but now we're talking 15 to 20 years. The concept is, if these 2 curves continue to diverge, by 5 years, there is going to be a 5 Sharp score point difference and at least that from then on.

How long does it take to notice a difference, radiologically, in terms of change in Sharp score? Nobody knows, but I'll give you a number anyway. About a 50 Sharp score point difference and your grandmother can tell the difference between those 2 x-rays. Therefore, if it's 5 years, it doesn't take very long for those 2 x-rays to be different. If you continue methotrexate, you will have different x-rays than if you were on etanercept. Over time, the tolerability isn't as good with methotrexate, so that 74% of the patients stay on etanercept for 24 months, and 59% of the patients stay on methotrexate. There is a difference between the two. Methotrexate makes some people feel sick. Etanercept makes many patients feel good. If the cost would come down, everybody would be switching to etanercept.

What about combination therapy? There are 3 ways to combine DMARDs. You can add a DMARD in those partially responsive. That's called step up. We can use them in parallel from the beginning, or we can use step down. We can use multiple DMARDs at the beginning and, as the patient responds, keep removing DMARDs until you're back down to baseline, maybe with monotherapy. That's the induction and maintenance concept. You've seen lots of add-on therapies, because that's the way the trials are being designed. The step-up therapy here is leflunomide. In patients who are partially responsive to methotrexate, the combination is a 51 ACR 20 compared with methotrexate alone, which means you continue on it. There is quite a difference. There appears to be a clinical added benefit to adding leflunomide in patients who are partially responsive to methotrexate. These are some data that suggest combination therapy is not bad.

Two studies recently asked the following question: At the beginning, if I initiate methotrexate and sulfasalazine, will it be better than either alone? Two studies say no. Initiating sulfasalazine and methotrexate at the same time is not better than either methotrexate or sulfasalazine alone. These are data using the Disease Activity Score (DAS) and the ACR 20 scores. What about triple therapy? Jim Ardell added triple therapy at the beginning and asked whether it was better than Plaquenil, sulfasalazine, or methotrexate alone. He had striking results. The triple therapy at the beginning was much better than the others. There were a couple of problems with the study. It was pretty long. There were long-duration patients, about 10 years. The number of patients who were methotrexate naïve was quite considerable, and most of the patients failed about .9 DMARDs.

It's an unusual population, but these studies were done a number of years ago when methotrexate wasn't on everybody's "radar screen." It's still an unusual population. This study is critical to be reproduced in early rheumatoid arthritis because it says that the toxicity wasn't any different whether you were on triple, double, or single therapy. There are a couple of open-label trials. There is a triple therapy trial -- methotrexate, sulfasalazine, and Plaquenil versus methotrexate and sulfasalazine, or methotrexate and Plaquenil. It was triple therapy versus double therapy versus single therapy. The single therapy was methotrexate. They showed triple therapy worked quite well. The problem in this study is they used methotrexate at low doses up to 15 mg. I don't know that triple therapy is better than high-dose methotrexate. I don't think we'll ever see the study.

This is another trial. In this trial they used combination therapy along with corticosteroids versus corticosteroids and sulfasalazine. It was actually quadruple therapy versus sulfasalazine, plus or minus cortisone. You didn't have to add cortisone. The results were very good. The combination therapy was excellent. The problem is their comparator was sulfasalazine and I'm not sure this combination is any better than high-dose methotrexate.

There is step-down. The key trial for step-down was the COBRA trial (comparison of combined step-down prednisolone, methotrexate and sulfasalazine with sulfasalazine alone in early rheumatoid arthritis). It was high-dose steroids for a short period of time and then eventually tapered off. They used a methotrexate combination, and a sulfasalazine combination. All triple therapy was compared with sulfasalazine alone. Notice the clinical results. Patients on triple therapy did extremely well until you stopped the prednisone. Then they came back to the baseline. They came back to what sulfasalazine alone looked like. After you dropped the steroids, they looked clinically alike.

The only difference was a significant difference in the radiologic outcomes when you added low-dose methotrexate along with sulfasalazine, and the prednisone versus sulfasalazine alone. This is triple therapy slowing progression relative to sulfasalazine. At 1 year, the patients on sulfasalazine were allowed to go onto any therapy they wanted. Over the next 5 years, those patients on sulfasalazine never caught up. That may have to do with the DMARDs they went on, but the sulfasalazine group never caught up to the triple-therapy group. At this point, remember, you're only on sulfasalazine. You've already stopped the prednisone and the methotrexate. It's an interesting concept that with early aggressive therapy, tapering off "the bad guys," and leaving a simple drug on, you may end up with a significant difference radiologically.

Biologic therapy says that if you fail methotrexate and you add a disease modifier, like a biologic, you have good data. There are good ACR 20s and 50s in combination: etanercept versus methotrexate. With infliximab you have to use methotrexate, good data. Anakinra (Kineret) just came on the market recently. What you see, again, are reasonable data. These time courses are a little different. You can't compare study with study but, in general, it says if doctors add a biologic to those patients partially responsive to methotrexate, they have additional clinical benefit. This is not a positively established fact.

In these trials, you never had the biologic arm alone. You had the methotrexate arm. You had the combination arm. But, you never had the biologic arm. How do doctors know that you really need methotrexate? There is an uncanny concept with the ACR 20 outcomes. If you used etanercept as monotherapy, the ACR 20 was 60. If you used it and added methotrexate, the ACR 20 outcome was 71, not very different.

Leflunomide as monotherapy, the ACR 20 outcome was 46. In those patients who were partially responsive to methotrexate, it was 46. If you look at anakinra (Kineret), it was 38 by monotherapy, and adding methotrexate, it was 37. You've got to wonder whether you really needed the methotrexate if those numbers look alike. Those are the data. It's food for thought.

What's new? Scientists have made great gains in terms of the clinical and radiological outcomes with tumor necrosis factor (TNF) antagonists. The new agent to come out in terms of TNF-Humira (Adalimumab). This is a fully human anti-TNF monoclonal antibody with a half-life of about 2 weeks. It fixes complement. It's able to use cells,like infliximab. It kills those cells that actually make TNF. It was given subcutaneously once a week as monotherapy. The results were quite good. The ACR 20s were comparable to anything else we've seen, around 55%. The ACR 20s were about 25% or 30%. That is good. This is once-weekly subcutaneous administration in a severe population of patients.

What happens if I have a partial response patient population on methotrexate? Using an average dose of about 16.8 mg, patient disease duration was 12.3 years. What happens when you add subcutaneous Humira (DE 27-formerly )every second week in those patients who were partially responsive? ACR 20 is about 65%. Reasonable. ACR 50 at the 40-mg dose was quite high at 53%, and ACR 20 was 26.9%. This value was higher than we see in most studies, but the 80-mg dose was slightly less, in terms of efficacy, than the 40-mg for reasons that are unclear. This study was not powered to tell the difference between doses, so we still don't know what the minimally effective dose is. At least we can say there is good benefit if you give subcutaneous D2E7 every second week.

Cortisone is such an important drug that we have, but it needs to be used very carefully because it can cause a lot of problems. Using intermittent injections of cortisone into the joints can be a real godsend and can really improve people's function. For example, if someone's shoulder is limited in its ability to move, an injection of cortisone in the shoulder can provide rapid relief that can really provide benefits while you're waiting for the disease modifying drugs to take effect.

Gold is an old treatment, and in some ways that's really nice because physicians know a lot about it. They know what's good about it, and the bad about it. It is an effective treatment in about 55 percent of people who use gold. Gold suppresses the disease partially, but it's very slow. You need to use it with weekly injections for 20 weeks often to see clinical benefit. And sometimes people can have problems with rash or low blood counts. Others can have oral sores or even problems with protein or blood leaking from the kidneys. While gold can be used safely and there may be particular patients that's a really good drug to use, it is something that is sometimes a troublesome drug, too. One of the problems with gold is that patients and doctors do have to wait often the 20 weeks or more to make a decision whether this is an appropriate therapy, and also,may have lost six months of ability to treat the patients. Gold is probably not a very good first choice of a second-line agent or of a disease modifying drug unless there's some reason that the physician didn't want to use methotrexate. Patients has to get a urine test and a blood test once a week to be sure that they don't have the kidney and bone marrow troubles that some patients have.

Methotrexate is actually a therapy that's been around for a very long time. It was used for cancer patients, and it's been used in rheumatoid arthritis for well over 20 years now. The way it actually works is still not perfectly clear. Physicians know it works by decreasing inflammation, but it also works by decreasing immune cells. A very large study that was published in 1999 showed that, finally, that it is not only a medicine that makes the patients' signs and symptoms or their joint pain and swelling better, but it slows the x-ray progression down, so it truly is one of our medicines that modifies the disease. With reference to side effects,it's important also to understand that there are risks with not treating rheumatoid arthritis,that's the perspective or the frame in which you've got to look at the possible risks of methotrexate.

Methotrexate is very well tolerated by most people, and even in relatively high doses, more than 85 percent of people will be able to continue it for at least a year. The other signs of its safety is that at five years, more people tend to stay on methotrexate than any other drug. It's important to note that generally it is very well tolerated by most people. There are two types of side effects. One is the relatively common kind of minor side effects. Occasionally people are a little tired or may have a little bit of nauseousness or loose stools after they take their weekly dose. And then there are probably the less common or more sporadic type of problems which can be more serious, which are methotrexate-associated serious liver scarring, which occurs approximately in one in 1000 people who take it for five years, or serious lung scarring which can occur in around one and a half or two percent of people. Methotrexate can be administered either by pill, which is the most common way that it's used in the United States, but it also can be administered by injection. The injections offer possibly a little more safety, a little more regulation of the dose, and in higher doses doctor's are able to have to switch their patients to injectable methotrexate.

Leflunomide ( Arava ) is a drug which is similar in some ways to methotrexate in that it impairs the ability of inflammatory cells to reproduce rapidly. It is a bit expensive, about 280 dollars a month, and so that's actually somewhat of a barrier to people who don't have health insurance, but it is generally well tolerated by people. The side effect seen most commonly is loose stools associated with it, but for most people that's not too much of a problem. It has some benefits in that it does seem to have some benefit within the first six to eight weeks of use, and it appears that its ability to suppress swelling and pain and to prevent joint damage is perhaps close to or similar to methotrexate. It would be an alternative to patients who either did not respond or could not take methotrexate.

Plaquenil has been used for a long time, and there are some really nice features about it which doctors include that it really doesn't need frequent blood monitoring, and it does not have a significant toxicity to liver. However, at least in patients with recent-onset rheumatoid arthritis, there has never been a published study that has shown that it slows or prevents joint damage. Therefore,for someone who has significant and functionally limiting rheumatoid arthritis, Plaquenil is not a good first choice. And also I think the fact that it does take many months to see a response, we may lose our window of opportunity.

In Europe and outside of the United States, sulfasalazine is very commonly used. It's less commonly used in the United States. Sulfasalazine can be a very helpful drug. It is relatively inexpensive. The mode of onset is not as fast as methotrexate, but certainly faster than gold. The side effect profile, unless it causes stomach upset, is generally pretty good. In the research studied, it may not be as effective as methotrexate or leflunomide, but it is certainly an effective treatment. For some people, it is a very useful drug. In people with the most severe arthritis, though, again that a faster-acting and more potent drug would be a choice of physicians.

D-Penicillamine, is rarely used now,especially with the newer drugs available. With the advent of the biologics, doctors don't need to use it because it is a troublesome medicine. Now that we have a number of better-tolerated, faster and more effective treatments, these are clearly treatments physicians don't use as often.

Azathioprine is a drug that is effective to some degree in reducing joint swelling and pain. The data supporting the idea that it prevents joint damage is really not as robust, and it's really quite slow. It's largely been supplanted in its place by Arava.

Treatment with cyclophosphamide was tested in the '60s, late '60s, early '70s, and it's certainly a very potent agent, but unfortunately it does carry with it a substantial risk of developing leukemia in the ten years after being treated with it. Cyclophosphamide really is not,or rarily used for rheumatoid arthritis.

Summary: In someone who has either very mild rheumatoid arthritis or maybe you're not sure it's rheumatoid, rheumatologists might consider Plaquenil early, but choices such as sulfasalazine, Arava or methotrexate would be their first choices.

Biologics: Enbrel, which is etanercept, and infliximab, the trade name is Remicade. They both are targeted towards the inflammatory messenger called TNF or tumor necrosis factor, and these are similar in many ways. Enbrel is actually the receptor that TNF binds to on the inflammatory cell, and research-scietists figured out how to make that little receptor and how to help it circulate in the blood, and so what it acts as a sponge to take this out of circulation.

Infliximab or Remicade is a little different in that instead of just a receptor floating around, infliximab is an antibody, and that is it binds specifically to TNF and "kills" or eliminates it out of the blood. With Enbrel individuals can administer that to themselves, and so that gives them a measure of control and independence. It's also not really associated with a meaningful immune response, so it can be taken all by itself. Remicade is a little different in that it needs to be given as an intravenous treatment, and because one third of Remicade comes from a mouse, they have to take a drug like methotrexate with it to suppress making anti-mouse antibodies

Currently,these drugs are limited because of cost and the requirements are that patients if they fail on an initial agent such as methotrexate need to be considered for a biologic,unless one has private health coverage and most plans have the same limitations. Infliximab or Remicade, because of the mouse protein in it, one has to continue the methotrexate whereas with Enbrel or etanercept, it can be used as a single agent with very good success.

The other important point is that not all of these medicines work for every patient, and rheumatologists have to make decisions which are go, no-go decisions whether this is the right medicine and the right dose of medicine for the patient. Being on one of these medications is not enough. The rheumatologist and the physician taking care of the patient has to make a decision whether it's really doing the job that it needs to be done. Therefore,rheumatologists will look at the patient's joints and they'll do a joint count so they find out how many of the joints are tender and how many are swollen, and they will follow x-rays on a periodic basis to be sure that there is no damage that's progressing to make decisions whether or not this is an effective medicine, whether it be a biologic such as Enbrel or a synthetic such as methotrexate.

One of the things that we as a society have to come to grips with is that the cost of biologics is much more than methotrexate is. And so, while there are individuals that without any question should be started on the biologics first, like people with significant liver disease or someone who has moderately severe kidney disease, in most individuals who do not have those processes, probably an initial trial of a potent disease-modifying pharmaceutical like methotrexate is probably from a financial standpoint the most sensible thing, and probably from a society stand.

The actual cost of the biologics is higher than methotrexate. Also, many of the patients with rheumatoid arthritis often come already on methotrexate. Many internal medicine doctors have already started them on methotrexate, and a decision needs to be made is that the right medicine and if so, is it the right dose, and often it's too low, and one pushes the dose. A comparative study has been shown that higher doses of methotrexate work better. This is a very individual decision with the patient and the physician.

But what doctors would focus the most on isn't necessary what is the first drug that is used but that these biologics are not something that should be saved until there are no other options. These are drugs that should be used easily in the first six months to a year of disease if traditional therapies don't very effectively suppress the arthritis. People should not wait because, the pace and the progression of joint damage is a continuous and rapid process, and if we can in a rapid fashion suppress people's disease,and over many years will reduce the amount of joint damage which should translate into better work, better health, and better social function.

The new COX-2 selective non-steroidal anti-inflammatory drugs (NSAID's)celecoxib (Celebrex). They differ from the regular NSAID's because they are associated with a lower incidence of peptic ulcers and their complications and they do not inhibit platelets to prolong bleeding. They are as equally effective as the regular NSAID's for arthritis and pain-killing.

In the large CLASS study, there did not appear to be an increase in heart attacks or strokes in patients on Celebrex compared to those on ibuprofen and diclofenac: 13 per 1000 on Celebrex versus 12 per 1000 on ibuprofen and diclofenac.

These drugs do not affect the kidneys or blood pressure adversely in those who have no problems with their kidneys or blood pressure. In those that do have such problems, all NSAID's including the coxibs are hazardous. They can worsen the blood pressure, fluid retention (edema and heart failure) and kidney function. Celebrex appears to be no different than ibuprofen and diclofenac in these matters but Vioxx seems to be worse for heart failure and high blood pressure.

If you have no risk factors for peptic ulcers, heart attacks or strokes, high blood pressure, fluid retention or kidney disease, then you can use NSAID's or coxibs. If you only have risk factors for peptic ulcers, then use coxibs. If you only have risk factors for heart attacks or strokes, then take NSAID's (naproxen may be one of the best) and probably Celebrex but not Vioxx. If you have high blood pressure, fluid retention or kidney disease or use diuretics, then all NSAID's and coxibs could make these problems worse but Vioxx is probably more of a problem than the others.Try some other treatment if possible. If you have problems in more than one of these groups, then try some other treatment if possible. If not possible, choose the NSAID or coxib of least risk for the the greatest problem that you have.