Glossary
TNF-alpha
This is the cytokine that changed rheumatology treatment. Blocking it slows joint damage in trials, and the level in your own blood isn't something anybody follows. Those two facts get confused more often than almost anything else in this glossary.
What the research found.
One trial started 799 people with early rheumatoid arthritis who had never had methotrexate. At a year, 62 percent on adalimumab with methotrexate hit the response target, against 46 percent on methotrexate alone and 41 percent on adalimumab alone. Both comparisons came in at P under 0.001.
In that trial, joint damage progressed less on the combination. After two years it had moved 1.9 Sharp units, against 10.4 units on methotrexate alone and 5.5 on adalimumab alone. The P value was at or below 0.002, and that's damage on an x-ray rather than a symptom score.
One review pooled 21 trials and 8361 people with early rheumatoid arthritis, where adalimumab with methotrexate had the strongest effect on remission. The odds ratio was 2.90. The true value sits somewhere from 1.94 to 4.33.
One study followed 1250 patients with active ankylosing spondylitis on a TNF blocker, where everyone knew what they were taking and there was no control group. Every sleep score improved at 12 weeks, with broken sleep at an effect size of minus 0.69. That design can't tell you the drug caused it.
What it is and why you've heard of it
TNF-alpha is a cytokine, meaning a signaling protein immune cells release to affect other cells, and this one drives inflammation. Its full name is tumor necrosis factor alpha, and that name records how it was first found rather than what it does in arthritis. So it's alarming to read on a prescription while telling you nothing about your disease.
Patients meet the term for one reason. Blocking this cytokine became the first major targeted treatment in modern rheumatology, and the drugs that do it get described by what they block. So the target ends up in the name of the class.
What blocking it does to joints
One trial tested this as directly as it can be tested. It started 799 people with early, aggressive rheumatoid arthritis, none of whom had ever taken methotrexate, and split them three ways. That design is what makes the result worth reading.
At a year, 62 percent on adalimumab with methotrexate hit the response target, against 46 percent on methotrexate alone and 41 percent on adalimumab alone. Both comparisons came in at P under 0.001. After two years, 49 percent on the combination were in remission, with a disease score below 2.6, and another 49 percent had a major clinical response.
Both of those figures are roughly twice what either single drug managed on its own, which is a large enough gap to change how a first treatment gets chosen. Side effect profiles were comparable across all three arms. So the pairing won without costing anything the trial could measure.
What it does to joint damage
This is the part that separates these drugs from a painkiller, and the trial above measured it. The numbers come off x-rays rather than off a questionnaire, which is what makes them hard to argue with. They're also the reason treatment starts early.
Damage gets scored from x-rays in units named after the person who devised the scale. After one year, damage had moved 1.3 units on the combination, 5.7 on methotrexate alone, and 3.0 on adalimumab alone. After two years it was 1.9 units, 10.4, and 5.5. The P value was at or below 0.002 at both time points, which is a large and consistent separation for a structural outcome.
That's structural change to the joints themselves rather than a score of how somebody felt on the day. A larger pooled analysis points in that direction on remission too, covering 21 trials and 8361 people with early rheumatoid arthritis. Adalimumab with methotrexate had the strongest effect on remission of anything it studied, at an odds ratio of 2.90 with the true value from 1.94 to 4.33.
Which tablet goes alongside
If you take a TNF blocker, something usually goes alongside it. Which tablet turns out to be a real question rather than a formality, and one review measured it across several biologic families. The answer differed by which biologic somebody was on.
Alongside a tumor necrosis factor blocker, a tablet other than methotrexate gave a worse response at six months. The risk ratio was 0.93, from 0.87 to 1.0, at P equals 0.04, across 3843 people, and people also stayed on it less at twelve months. That's a second measure pointing in the same direction.
That review found the opposite pairing for a different drug, where leflunomide beat methotrexate alongside rituximab, at a risk ratio of 1.38 from 1.13 to 1.68. So which tablet goes with which biologic is worth asking about rather than assuming. It's a detail that rarely comes up unprompted.
Where the TNF blockers differ from each other
They aren't interchangeable for everything, and eye inflammation is the clearest example of it. Two separate sources point in that direction, and the agreement between them is what gives the finding its weight. The numbers on each side are small.
One study of 301 people with spondyloarthritis compared attacks before and after treatment. Among 27 people on tumor necrosis factor antibodies, attacks fell from 1.83 per person to 0.41, at P equals 0.002, while among 19 people on etanercept they didn't fall at all, going from 0.44 to 0.79. One review reaches that conclusion too, listing infliximab, adalimumab, and certolizumab as effective at reducing how often uveitis happens and listing etanercept as not.
Those are small numbers on the trial side, so hold the size of the difference loosely. The direction has been found twice, by two different methods. It's worth raising with your rheumatologist if you've had eye attacks before, because it's one of the few places where which drug in the class you're on genuinely changes the answer.
Sleep
One study is often quoted here, and it's weaker than its effect sizes make it look, because the design is the problem rather than the numbers. It followed 1250 patients with active ankylosing spondylitis on a TNF blocker for 12 weeks, where every sleep score improved against the starting point. Broken sleep came out at an effect size of minus 0.69, sleep adequacy at 0.55, and daytime sleepiness at minus 0.52, while sleep went up from 6.2 to 6.6 hours a night, which the authors call a small effect.
The design is where it comes apart. There was no control group and no placebo, everyone knew what they were taking, the company that makes the drug paid for it, and sleep was self-reported rather than the main thing the study was built to answer. So two effects run free in it, which are people drifting back toward average and people expecting a new drug to help.
The finding fits the drug helping without being evidence that it did. A review of 36 studies in psoriatic arthritis is stronger on the problem than on the treatment, where pooled poor sleep quality reached 72.9 percent of patients, with the true value from 63 to 81.8, against 26.9 percent in healthy controls. That review reports better sleep on TNF blockers without pooling the numbers behind it.
Whether the class you're on changes your sleep
One real-world study compared classes on a sleep measure. It covered 495 ankylosing spondylitis patients in an Australian dataset and matched the two groups for their other differences. Mean sleep apnea risk was 0.4 on interleukin-17A blockers and 0.3 on TNF blockers, at a P value of 0.046.
Two cautions belong with that. The measure is a risk score rather than a diagnosis, so neither figure is a rate of apnea, and the P value sits on the border, with the difference small on a scale whose overall mean was 0.4 and standard deviation 0.3. So that's a difference worth noting and not worth switching a drug for.
That study found something much larger elsewhere in its results. Poor disease control went strongly with insomnia severity, at an odds ratio of 7.29 from 2.37 to 22.46, though it looked at people at one moment, so it says nothing about which came first. Steroid use and broken sleep came with an odds ratio of 3.98, with a lower bound of 1.03.
The level in your blood
This is the part most worth taking away, and it's the one that causes needless worry in people who have read about the biology. TNF levels aren't measured in routine rheumatology care, and no source here covers reading one in a single person. The cytokine is a drug target rather than a number to follow.
Disease activity gets tracked three ways instead, through joint counts, an inflammation marker like CRP, and your own rating of how you are. Those are cruder measures than a cytokine level sounds like it would be, and they're the ones tied to decisions about your treatment. That's a trade rheumatology has made deliberately.
If a TNF level has been measured for you, ask whoever ordered it what it changes, because our sources don't answer that question and nothing on this page can. The short form is worth remembering. A drug named for a cytokine is not an instruction to track the cytokine.
Common misconceptions.
Myth. TNF-alpha is something I should get measured.
Reality. It isn't measured in routine rheumatology care, and no source here covers reading a level in one person. Patients meet the name for a different reason, which is that a treatment class gets described by what it blocks. Disease activity gets followed three ways instead, through joint counts, an inflammation marker like CRP, and your own rating of how you're doing on the day.
Myth. Blocking it only helps how you feel.
Reality. It also slows what happens to the joints themselves. In one trial of 799 people, joint damage after two years had moved 1.9 Sharp units on adalimumab with methotrexate, against 10.4 units on methotrexate alone and 5.5 on adalimumab alone. That's damage measured on an x-ray rather than a symptom score somebody filled in.
Myth. The drug works best on its own.
Reality. It worked best paired with methotrexate in the trial that tested all three arms against each other. At a year, 62 percent on the combination hit the response target, against 46 percent on methotrexate alone and 41 percent on adalimumab alone. Side effect profiles were comparable across all three arms, which is worth knowing before anybody argues for the simpler option.
Myth. All TNF blockers behave alike.
Reality. Not for everything they treat. One study compared attacks of eye inflammation before and after treatment, where people on tumor necrosis factor antibodies went from 1.83 attacks to 0.41, at P equals 0.002, while people on etanercept went from 0.44 to 0.79. One review lists three antibody drugs as effective at cutting those attacks and lists etanercept as not.
Myth. The sleep improvement on that treatment proves the drug caused it.
Reality. That study design can't support it, because there was no control group and no placebo, everyone knew what they were taking, the company that makes the drug paid for it, and sleep was self-reported as a second-line measure. So two effects run free in it, which are people drifting back toward average and people expecting a new drug to help. The finding fits the drug working, and it isn't evidence that it did.
Related terms.
- Biologics, the drug family this target belongs to.
- Cytokines, the category this belongs to and why measuring one isn't measuring a disease.
- Interleukin-6, the other cytokine with a major drug class named after it.
- Uveitis, where the TNF blockers differ most from each other.
Questions patients ask.
What is TNF-alpha?
It's a cytokine, meaning a signaling protein immune cells release to affect other cells, and this one drives inflammation. Its full name is tumor necrosis factor alpha. That name records how it was first found rather than what it does in arthritis, and it became the first major drug target in modern rheumatology, which is why the name reaches patients at all.
Why is a drug class named after it?
Because those drugs work by blocking this one cytokine, so the target ends up in the name of the treatment. That's a handy shorthand for clinicians and a confusing one for patients, because it makes a signaling protein sound like something to monitor. It's really a description of how a drug works.
What does blocking it do to joints?
One trial started 799 people with early disease who had never taken methotrexate. At a year, 62 percent on adalimumab with methotrexate hit the response target, against 46 percent on methotrexate alone and 41 percent on adalimumab alone. After two years, 49 percent on the combination were in remission.
Does it slow joint damage?
In that trial it did. Damage is scored on x-rays using Sharp units, and after two years it had moved 1.9 units on adalimumab with methotrexate, against 10.4 on methotrexate alone and 5.5 on adalimumab alone, at a P value at or below 0.002. That's structural change rather than how somebody felt on the day.
Will my TNF level be measured?
Not in routine care, and no source here covers reading one in a person. Routine care follows disease activity three ways instead, through joint counts, an inflammation marker like CRP, and your own rating. So if a level has been measured for you, ask whoever ordered it what it changes, because our sources don't answer that.
Do these drugs help sleep?
One study points that way and it's weaker than it looks. It followed 1250 people with active ankylosing spondylitis for 12 weeks, where every sleep score improved, with broken sleep at an effect size of minus 0.69 and sleep going from 6.2 to 6.6 hours a night. There was no control group and everyone knew what they were taking, so the finding fits the drug working without being evidence that it did.
Is one TNF blocker better than another?
For eye inflammation, one study suggests so. Attacks fell from 1.83 to 0.41 per person on tumor necrosis factor antibodies, at P equals 0.002, while on etanercept they went from 0.44 to 0.79. One review lists infliximab, adalimumab, and certolizumab as effective at cutting those attacks, and lists etanercept as not.
Which tablet goes alongside it?
Methotrexate, on the evidence here. One review looked at which conventional drug pairs best with which biologic, and alongside a tumor necrosis factor blocker a tablet other than methotrexate gave a worse response at six months. The risk ratio was 0.93, from 0.87 to 1.0, and people stayed on it less.
References.
- Breedveld FC; Weisman MH; Kavanaugh AF et al. The PREMIER study: A multicenter, randomized, double-blind clinical trial of combination therapy with adalimumab plus methotrexate versus methotrexate alone or adalimumab alone in patients with early, aggressive rheumatoid arthritis who had not had previous methotrexate treatment. Arthritis Rheum. 2006;54:26-37. 10.1002/art.21519PREMIER: two-year
- Xu H; Li C; Ding R et al. Efficacy and safety of non-conventional synthetic disease-modifying antirheumatic drugs in early active rheumatoid arthritis: a network meta-analysis. BMC Rheumatol. 2025;10:5. 10.1186/s41927-025-00603-xSystematic review and network meta-analysis of randomised controlled trials to February 2025
- RUDWALEIT M; GOOCH K; MICHEL B et al. Adalimumab Improves Sleep and Sleep Quality in Patients with Active Ankylosing Spondylitis. The Journal of Rheumatology. 2010;38:79-86. 10.3899/jrheum.100213OPEN-LABEL
- Grant C; Woodbury M; Skougaard M et al. Sleep Problems in Patients With Psoriatic Arthritis: A Systematic Literature Review and Metaanalysis. The Journal of Rheumatology. 2023;50:1594-1609. 10.3899/jrheum.2022-1169Systematic literature review and meta-analysis of 36 studies
- Tymms K; Butcher B; Sletten T et al. Prevalence of sleep disturbance and the association between poor disease control in people with ankylosing spondylitis within the Australian clinical setting (ASLEEP study): a real-world observational study using the OPAL dataset. Clinical Rheumatology. 2021;41:1105-1114. 10.1007/s10067-021-05953-8Retrospective
- Frantz C; Portier A; Etcheto A et al. Acute anterior uveitis in spondyloarthritis: a monocentric study of 301 patients. Clin Exp Rheumatol. 2019;37:26-31. PMID 30620268Cross-sectional single-centre observational study
- Sharma SM; Jackson D. Uveitis and spondyloarthropathies. Best Pract Res Clin Rheumatol. 2017;31:846-862. 10.1016/j.berh.2018.08.002Review of uveitis in the spondyloarthropathies
- Decarriere G; Barnetche T; Combe B et al. Most Appropriate Conventional Disease-Modifying Antirheumatic Drug to Combine With Different Advanced Therapies in Rheumatoid Arthritis: A Systematic Literature Review With Meta-Analysis. Arthritis Care Res (Hoboken). 2021;73:873-884. 10.1002/acr.24195Systematic literature review with meta-analysis comparing methotrexate against other conventional synthetic disease-modifying drugs when combined with advanced therapies in rheumatoid arthritis
- Youssef P; Ciciriello S; Tahir T et al. Real-World Persistence and Effectiveness of Upadacitinib versus Other Janus Kinase Inhibitors and Tumor Necrosis Factor Inhibitors in Australian Patients with Rheumatoid Arthritis. Rheumatol Ther. 2025;12:173-202. 10.1007/s40744-024-00736-4Retrospective non-interventional study of electronic medical records from the Australian OPAL dataset
- Peper SM; Lew R; Mikuls T et al. Rheumatoid Arthritis Treatment After Methotrexate: The Durability of Triple Therapy Versus Etanercept. Arthritis Care Res (Hoboken). 2017;69:1467-1472. 10.1002/acr.23255Open-label extension of a 48-week double-blinded noninferiority randomised trial
This page gathers the published research on this subject into one place. The studies behind it were published between 1989 and 2026, and every figure links to the paper it came from. Those studies were peer reviewed. This summary of them was not. Dr. Sarah Luebker is reviewing these pages one at a time and has not reached this one yet, so it carries no medical review date and nothing here is her opinion or her advice to you. Each page gets updated as she reaches it. It is here in the meantime because the science is worth having in one organized place that is easy to find and easy to read. Talk to your own clinician before acting on any of it.