When methotrexate alone does not control rheumatoid arthritis, patients and physicians usually choose between two paths: add two older, inexpensive oral drugs, sulfasalazine and hydroxychloroquine (together called "triple therapy"), or add an injectable or infused biologic medicine. This article explains what randomized trials have found when these strategies were compared directly. In brief, several well-conducted trials show that triple therapy controls symptoms and disease activity about as well as adding a tumour necrosis factor (TNF) inhibitor for many patients. Biologics hold a modest edge in slowing joint damage on X-ray and in tolerability for some people. Cost, side effects and family plans weigh heavily in the choice.
What hydroxychloroquine is and why it belongs in the combination
Hydroxychloroquine is a close chemical relative of quinine, the compound found in the bark of the cinchona tree and used for centuries against malaria. It is a modest but useful gift from the natural world. Researchers refined it in the mid-twentieth century, and it was later observed to calm inflammatory joint and skin disease. It is approved in the United States for rheumatoid arthritis, lupus and malaria.
Its mechanism is only partly understood. Laboratory and cell-culture studies show that hydroxychloroquine accumulates inside acidic compartments of immune cells, called lysosomes, and raises their pH. This appears to interfere with how those cells process and present antigens. It also dampens signalling through certain toll-like receptors, which are part of the body's early-warning system against infection. These findings come mostly from cell and animal work. In people, the clinically relevant fact is simpler: on its own, hydroxychloroquine is a mild disease-modifying drug, but it seems to add meaningfully to methotrexate and sulfasalazine. It has a long safety record, does not suppress the bone marrow in ordinary doses, and does not appreciably raise the risk of serious infection.
The foundational trials of triple therapy
The modern case for triple therapy began with James O'Dell and colleagues at the University of Nebraska. Their randomized, double-blind trial, published in the New England Journal of Medicine in 1996, enrolled about 100 patients with established rheumatoid arthritis who had responded poorly to earlier treatment. Over two years, 77 percent of those on all three drugs achieved and sustained a 50 percent improvement in disease measures. The figure was 33 percent with methotrexate alone and 40 percent with sulfasalazine plus hydroxychloroquine. Triple therapy was not more toxic than the alternatives.
A follow-up trial by the same group, published in Arthritis & Rheumatism in 2002, compared triple therapy with two-drug combinations. Triple therapy again performed best, with roughly 78 percent of patients reaching a standard 20 percent improvement threshold (ACR20) at two years. Methotrexate plus hydroxychloroquine reached about 60 percent, and methotrexate plus sulfasalazine about 49 percent.
In Finland, the FIN-RACo trial (Möttönen and colleagues, The Lancet, 1999) studied early rheumatoid arthritis. It found that a combination of methotrexate, sulfasalazine, hydroxychloroquine and low-dose prednisolone roughly doubled the two-year remission rate compared with a single drug, at about 37 versus 18 percent.
These trials established that the combination works better than its parts. They did not test it against biologics, which arrived in the late 1990s. Several later trials did.
Head-to-head trials against biologic add-on therapy
Four lines of evidence matter most.
- SWEFOT (Sweden, The Lancet, 2009 and 2012). Ronald van Vollenhoven and colleagues took patients with early rheumatoid arthritis who were not adequately controlled after three to four months of methotrexate. They randomized them to add sulfasalazine and hydroxychloroquine or to add infliximab, a TNF inhibitor. The trial was open-label. At 12 months, a good response by European criteria was more common with infliximab, at about 39 versus 25 percent. By 24 months the difference in clinical response was no longer statistically significant. However, X-ray progression of joint damage remained somewhat less with infliximab.
- TEAR (United States, Arthritis & Rheumatism, 2012). Larry Moreland and colleagues randomized 755 patients with early, poor-prognosis disease. They received immediate methotrexate plus etanercept, immediate triple therapy, or methotrexate alone with a step-up to either combination if disease remained active at six months. Over two years, disease activity scores with triple therapy and with methotrexate plus etanercept were essentially equivalent. Radiographic progression was low in all groups, and slightly lower with etanercept, a difference the authors judged of uncertain clinical importance. Patients who started on methotrexate alone and stepped up when needed did about as well by the end. That finding matters because roughly a third responded adequately to methotrexate and never needed a second drug.
- RACAT (Veterans Affairs and Canadian sites, New England Journal of Medicine, 2013). This double-blind trial, led again by O'Dell, enrolled 353 patients with active disease despite methotrexate. It is probably the strongest single piece of evidence. Patients added either sulfasalazine and hydroxychloroquine or etanercept, with a switch to the other strategy permitted at 24 weeks if there was no improvement. Triple therapy met the prespecified test for non-inferiority. Improvement in disease activity scores at 48 weeks was very similar, at roughly 2.1 versus 2.3 points on the DAS28 scale. Similar proportions of patients in each arm needed to switch, and those who switched did equally well whichever direction they went. Radiographic progression was marginally lower with etanercept, but the difference was small.
- Meta-analysis (BMJ, 2016). Glen Hazlewood and colleagues at the University of Calgary pooled many trials in a network meta-analysis, a method that combines direct and indirect comparisons. For methotrexate-naive patients and those with an inadequate response to methotrexate, they found that triple therapy was not statistically inferior to methotrexate plus a biologic for the ACR50 response. Biologic combinations may offer a small additional benefit in limiting X-ray damage, but the authors judged its clinical significance uncertain.
The evidence for comparable symptom control is reasonably strong. It comes from multiple randomized trials, including a double-blind non-inferiority trial and a large pooled analysis, with broadly consistent results. The evidence on long-term joint damage favours biologics slightly. Most trials ran only one to two years, so it is not known whether those small X-ray differences translate into meaningful disability decades later.
Where triple therapy falls short: tolerability and real-world persistence
In the trials, the main weakness of triple therapy was not efficacy but tolerability. Sulfasalazine commonly causes nausea, stomach upset and headache, and occasionally rash or blood-count changes. Taking several tablets a day is also a real burden. Observational studies from routine practice, including analyses of United States veterans and insurance databases, suggest that many patients stop triple therapy within the first year, more often than patients started on a biologic. Some of this is avoidable: slow dose titration, enteric-coated sulfasalazine and good communication help. But it explains why triple therapy can look better in tightly managed trials than in everyday use.
Each drug carries its own monitoring needs:
- Methotrexate: regular blood tests for liver function and blood counts, limits on alcohol, and folic acid supplementation.
- Sulfasalazine: blood counts and liver tests, especially in the first months. It should be avoided by people with sulfa allergy.
- Hydroxychloroquine: a baseline eye examination and annual retinal screening after five years of use, or sooner if risk factors are present. Long-term use can damage the retina. Current ophthalmology guidance recommends a daily dose no higher than about 5 mg per kilogram of actual body weight to keep that risk low. Hydroxychloroquine can also, rarely, affect heart rhythm, especially when combined with other drugs that do so.
Biologics have different trade-offs. They raise the risk of serious infections and can reactivate latent tuberculosis or hepatitis B, so screening is required before starting. They are given by injection or infusion, and many patients find them easy to tolerate day to day.
Cost, family planning and the decision itself
The cost difference is large. In the United States, all three triple-therapy drugs are inexpensive generics. Biologics, even with newer biosimilar versions, cost many times more each year. Economic analyses, including a 2017 modelling study in the Annals of Internal Medicine by Nick Bansback and colleagues using RACAT and related data, concluded that trying triple therapy before a biologic is the more cost-effective strategy for most patients with an inadequate response to methotrexate. For a family managing its own budget and insurance, this is not a trivial consideration. Good stewardship of household resources is part of caring for those who depend on us.
Professional guidance is more cautious than some advocates of triple therapy. The American College of Rheumatology's 2021 guideline conditionally favours adding a biologic or a targeted synthetic drug over triple therapy for patients not at target on methotrexate. The guideline cites the somewhat better radiographic outcomes. It also acknowledges that triple therapy remains a reasonable choice, particularly where cost or patient preference weighs heavily. "Conditional" means the evidence leaves room for informed patients and their physicians to decide differently.
Family plans deserve special attention, because every child conceived during treatment is a life worth protecting from the outset. Methotrexate can cause miscarriage and serious birth defects. It must be stopped before conception, generally one to three months ahead according to current rheumatology reproductive-health guidance, and effective contraception is advised while taking it. By contrast, hydroxychloroquine and sulfasalazine (with folic acid supplementation) are generally considered compatible with pregnancy. Some biologics can also be used during pregnancy with planning. Couples hoping for children should raise this with their rheumatologist early, before any change in therapy, so that disease control and the safety of a future child can be planned together.
What this means for patients
The trials offer an unusually clear message for an informed conversation with one's own physician. For a person whose arthritis is not controlled on methotrexate, adding hydroxychloroquine and sulfasalazine is a well-tested option. It produces disease control broadly similar to adding a TNF inhibitor, and patients who do not respond can move on to a biologic without having lost ground, as RACAT showed. A biologic may be preferred by those who cannot tolerate sulfasalazine, those with aggressive, erosive disease where every degree of joint protection matters, or those for whom simplicity of dosing is decisive. Neither path is right for everyone. The patient who understands the evidence is best placed to weigh it, alongside a doctor who knows his or her history.
Key takeaway: Randomized trials show that triple therapy with methotrexate, sulfasalazine and hydroxychloroquine controls rheumatoid arthritis about as well as adding a TNF-inhibitor biologic for many patients, at far lower cost, while biologics offer a modest edge in limiting X-ray joint damage and are often easier to tolerate.