Pennsaid is only FDA-approved for knees because that is the only joint the manufacturer tested in clinical trials. The FDA can only approve a drug for conditions where clinical evidence proves it works, and every trial submitted for Pennsaid focused exclusively on knee osteoarthritis. There is no inherent biological reason the medication couldn’t work on other joints, but without trial data proving safety and effectiveness elsewhere, the FDA cannot extend the label.
The FDA Approval Process Explains the Restriction
The FDA approved Pennsaid “for the treatment of the pain of osteoarthritis of the knee(s),” and that narrow wording traces directly back to the clinical evidence package the manufacturer submitted. Seven multicenter, randomized, double-blind studies were conducted to evaluate Pennsaid, and all seven enrolled patients with knee osteoarthritis specifically. Two of those studies, each lasting 12 weeks, provided the primary proof of efficacy. Two shorter studies offered supporting evidence.
No trials were conducted on hips, shoulders, ankles, or hands. Because the FDA only approves drugs for indications backed by adequate trial data, the label was written to match exactly what was studied. This is standard practice across the pharmaceutical industry. It does not mean the drug is pharmacologically incapable of treating pain in other joints. It means no one has submitted the evidence needed for the FDA to say it does.
Why the Knee Was Chosen for Trials
The knee is the most common site of osteoarthritis and one of the easiest joints to study with a topical medication. It sits close to the skin surface with relatively little overlying muscle or fat, making it a good candidate for a drug that needs to penetrate through skin to reach joint tissue. It also provides a large, flat application area, which matters for a liquid solution that needs to be spread evenly.
Pennsaid was originally formulated as a 1.5% solution applied as 40 drops per knee, four times daily. The drops are dispensed 10 at a time and spread around the front, back, and sides of the knee until the entire joint is covered. That application method was designed around the knee’s anatomy. A later 2% formulation simplified dosing to 2 pump actuations twice daily, but the target joint remained the same. The entire dosing protocol, from volume to frequency to application technique, was built and validated for the knee.
How Pennsaid Compares to Other Topical NSAIDs
This is where the picture gets interesting. Pennsaid contains diclofenac sodium, the same active ingredient found in Voltaren Gel (now available over the counter). Voltaren Gel is approved for osteoarthritis pain in joints amenable to topical treatment, including hands and knees. The difference comes down to formulation and, more importantly, which trials each manufacturer chose to run.
Voltaren Gel’s maker submitted clinical data on both hand and knee osteoarthritis, so the FDA approved it for both. Pennsaid’s maker only submitted knee data, so that is all the label covers. The two products use different delivery systems (a gel versus a liquid solution with a carrier agent that enhances skin penetration), which also means their absorption profiles differ. But the core reason for the narrower Pennsaid label is simply the scope of the clinical trial program, not a safety signal in other joints.
Systemic Absorption and Safety Considerations
One practical reason Pennsaid’s label stays narrow is that topical NSAIDs are not purely local treatments. Some of the drug absorbs into the bloodstream. A crossover study comparing topical and oral diclofenac found that oral tablets produced blood levels 14 to 27 times higher than topical application (depending on whether one or two knees were treated). Peak blood concentrations were 87 to 161 times higher with oral dosing. So topical use delivers far less drug systemically, but it is not zero.
This matters because diclofenac carries a boxed warning about cardiovascular risk. NSAIDs as a class can increase the chance of heart attack, stroke, and blood clots, and that risk grows with longer use. The FDA label also warns about potential liver effects: liver enzyme elevations can occur during treatment, and rare but serious liver reactions, including liver failure, have been reported in post-marketing surveillance. Patients on long-term therapy are advised to have liver function monitored periodically.
Applying Pennsaid to a larger skin surface area or to multiple joints could increase systemic absorption beyond what was studied in trials. Since the safety data only covers knee application at specific doses, expanding use beyond the knee introduces unknowns the FDA has not evaluated. The prescribing information explicitly states that applying more than the recommended dose “has not been studied and is therefore not recommended.”
Can Doctors Prescribe It for Other Joints?
Yes. Doctors can legally prescribe any FDA-approved medication for uses not listed on the label. This is called off-label prescribing, and it is common throughout medicine. A physician who believes Pennsaid would help a patient’s hand or ankle osteoarthritis can write that prescription based on clinical judgment, knowledge of the drug’s mechanism, and the broader evidence for topical diclofenac in other joints.
The key distinction is between what the FDA formally approves (based on submitted trial data) and what a doctor decides is appropriate for an individual patient. Insurance coverage can be trickier for off-label uses, and pharmacies may flag it, but the prescription itself is perfectly legal. If you are interested in using Pennsaid on a joint other than the knee, that conversation with your prescriber is the logical next step.
The Short Answer
Pennsaid is “only for knees” on paper because the company behind it ran all its clinical trials on knee osteoarthritis and submitted only that data to the FDA. The regulatory system works by approving drugs for conditions where evidence has been generated. No evidence was submitted for hips, hands, or shoulders, so the label does not include them. It is a regulatory boundary, not a pharmacological one.