American Society of Hirudotherapy

Knee Osteoarthritis

Off-label use; several small RCTs in knee OA report pain and function improvement with leeches versus NSAID gel, sham leech or TENS, though results against sham and TENS are mixed, and a 2022 trial against physiotherapy found no therapeutic effect.

Tier B — RCT-supported off-labelStudied off-label: Off-label use means application outside the FDA-cleared indication. Clinical evidence (RCTs or systematic reviews) supports this use, but the FDA has not cleared it.Last Updated: November 12, 2026 · ASH Editorial · based on the books

Patient Summary

Is this FDA-cleared for this use?
No — investigated off-label. The FDA cleared medicinal leeches in 2004 only for venous congestion in microsurgical reconstruction (K040187). Use for knee osteoarthritis is supported by published research but not FDA-evaluated.
What evidence exists?
Several published randomized trials (Michalsen 2003, Stange 2012, Andereya 2008, İşık 2017) and a 2014 meta-analysis (Lauche et al.) suggest a short-term reduction in knee pain compared with a topical NSAID gel, TENS, or a sham leech, although the İşık 2017 trial found leeches about as effective as TENS and a 2022 trial (Talebi, 98 patients) found no benefit and better results with physiotherapy. Effects have been measured out to about 3 to 6 months. The trials are mostly from research groups in Germany, all are small (under 130 patients), and longer-term outcomes (12+ months) are not well studied. This is not a substitute for proven treatments such as exercise therapy, weight management, or surgery when indicated.
Main risks
  • Bleeding and oozing from each bite site for several hours, sometimes up to 24 hours
  • Itching, redness, and irritation at the bite sites lasting days to weeks (15 to 20 percent of patients)
  • Blood loss: usually small, but stronger bleeding occasionally lowers hemoglobin; a leech therapy handbook (Michalsen and colleagues) describes a treatment for knee osteoarthritis after which a blood transfusion was needed
  • Local skin infection or, rarely, Aeromonas infection from leech gut bacteria
  • Allergic reaction to leech saliva (uncommon)
  • Small bite marks that usually fade over weeks to months (rarely a pale spot stays longer)
Who should not consider this
  • Patients taking blood thinners such as warfarin, apixaban, rivaroxaban, dabigatran, or heparin
  • Patients with hemophilia or other inherited bleeding disorders
  • Patients with anemia
  • Patients with a weakened immune system, especially severe neutropenia
  • Patients with an active joint infection (septic arthritis) or skin infection at the planned bite area
  • Patients within 6 months of a knee replacement surgery
  • Pregnant patients (leech therapy is not done during pregnancy)
What to ask your clinician
  • Have I tried first-line treatments such as exercise therapy, weight loss, and topical or oral pain medicines, and what does my orthopedist think?
  • What is the realistic chance leech therapy will help me, and for how long?
  • What evidence does the practitioner rely on, and have they treated patients like me before?
  • Are leeches from an FDA-registered supplier and used only once?
  • What antibiotic, if any, will I be given for Aeromonas prevention?
  • What is the cost and is it covered by insurance? (typically not covered as it is off-label)
  • What is the plan if symptoms come back — repeat session, escalation to injections, or surgery referral?
When to seek urgent care
  • Bleeding from a bite site lasting more than 24 hours or soaking through dressings
  • Spreading redness, warmth, pus, red streaks, or increasing pain around the bite sites
  • Fever above 38.0 C / 100.4 F or chills
  • Sudden severe knee pain, swelling, or inability to bear weight
  • Hives, facial or tongue swelling, throat tightness, or breathing difficulty

What this does NOT mean

  • It does not mean leech therapy is FDA-cleared for knee osteoarthritis — it is not.
  • It does not mean leeches cure osteoarthritis — the underlying cartilage wear continues; treatment may only reduce symptoms temporarily.
  • It does not mean leech therapy is better than exercise therapy, weight management, physiotherapy, or surgery for appropriate candidates — those remain the standards.
  • It does not mean every patient improves — most trials show improvement on average, but one found none, and individual responses vary widely.
  • It does not mean a single session is enough forever — the controlled trials followed patients for no longer than about 6 months.

Clinical Profile

Category
Musculoskeletal
ICD-10
M17.0, M17.10, M17.11, M17.12, M17.5
Safety tier
Low

Evidence Summary

Randomized trials of leech therapy in knee OA are mostly positive but small, short-term and heterogeneous, and one is negative. Michalsen 2003 (n=51) showed WOMAC pain reduction of ~64% at day 7 vs. ~18% for topical diclofenac (P<0.001); the pain difference was no longer significant after day 7, but function, stiffness and total-symptom scores stayed better with leeches through day 91. Stange 2012 (n=52, unblinded crossover) found leeches superior to TENS in the first period (days 0-21); Andereya 2008 (n=113, blinded, with an artificial-leech control) found improvement in all three groups, significant throughout the 26-week follow-up only in the two leech groups, and could not separate a specific effect from placebo; İşık 2017 (n=90) found leeches about as effective as TENS; and Talebi 2022 (n=98, leeches vs physiotherapy) found no therapeutic effect of leeches, with physiotherapy performing better at 90 days. A 2014 systematic review and meta-analysis (Lauche et al., Clin J Pain; 3 RCTs + 1 controlled trial, 237 patients) found strong evidence for immediate pain reduction (SMD -1.05, P<0.01) and moderate evidence for long-term pain effects (SMD -0.45); a 2018 meta-analysis of 4 RCTs (Wang et al., 264 patients) also favoured leeches at 1 to 7 weeks, rated the evidence quality moderate, and reported more adverse events with leeches. Lauche et al. found no serious adverse events, whereas Wang et al. reported a higher overall incidence of adverse events with leeches. Controlled trials followed patients for up to about 6 months; the only 12-month data identified in the journal literature come from a retrospective series of 24 patients with mixed arthritic joints (Loeser 2020), and the book by Michalsen et al. (2007) also reports preliminary retrospective clinic data on long-term effects (see “What the books describe” below).

What the books describe

Below is a summary of what books on hirudotherapy say about this condition. These are the authors’ observations, not a proven effect and not a promise of results. Hirudotherapy has contraindications; the decision is made by a physician.

  • Baskova and Isakhanyan list arthroses among the indications for hirudotherapy; regarding leech treatment of diseases of the skin, connective tissue and joints, they write that there are few publications on this in the literature[B1].
  • They report observations by Sulim (1998): 162 patients with arthrosis of the shoulder, wrist, knee and hip joints had leeches applied as part of combined treatment together with manual therapy and herbal medicine; after treatment the pain had gone in 148 of them[B1].
  • Michalsen et al. call painful (symptomatic) knee arthrosis one of the best-studied indications[B3].
  • In their randomized study of 51 patients, leech treatment was compared with diclofenac gel and the patients were followed for three months: for pain, the result favored leeches, with a clear advantage in the first week, and later the difference became less convincing; the study was open (without blinding), but according to the authors' check, the patients' expectations did not significantly influence the outcome[B3].
  • The authors call the retrospective data from their clinic (about 400 patients) preliminary: pain decreased noticeably in most but not all, and in some patients the relief lasted many months[B3].
  • Before treatment, Michalsen et al. advise confirming the diagnosis of arthrosis with imaging: in their experience, knee pain from other causes, especially after a meniscus injury, responds poorly to leech treatment[B3].
  • They advise continuing therapeutic exercise, but not overloading the joint in the first days of improvement so as not to weaken the effect of treatment; moderate itching lasting 2–3 days is often seen after the procedure[B3].

Treatment specifics

How many leeches, where they are placed, how long a session lasts, and whether to repeat are clinical decisions made by a qualified provider under institutional protocol — not something to self-administer. Discuss the specifics with a clinician experienced in medicinal leech therapy. (Clinicians: where the society's reference books describe a protocol, it is shown in the clinician view — switch the audience selector in the top bar to “Clinician”.)

Key Trials

  1. Michalsen A et al. (2003), n=51 · Ann Intern Med · PMID 14597456 · ASH analysis →
  2. Stange R et al. (2012), n=52 · Complement Ther Med · PMID 22305242 · ASH analysis →
  3. Lauche R et al. (2014), n=237 · Clin J Pain · PMID 23446069 · ASH analysis →
  • Michalsen A et al. 200351
  • Stange R et al. 201252
  • Lauche R et al. 2014237
Participants per key trial (n). Larger trials generally carry more statistical weight; case series with unspecified counts are omitted.
Study-design composition of the evidence base for Knee Osteoarthritis3studies
  • RCT: 2 (67%)
  • Meta Analysis: 1 (33%)
Composition of the cited evidence base by study design, from strongest (meta-analysis, systematic review) to weakest (case report). A base weighted toward RCTs and reviews is more reliable.

Detailed Trial Entries

10 trials indexed in the ASH RCT Library with full Study Profile, GRADE rating, and clinical implications:

Contraindications

  • Hemophilia or another bleeding disorder
  • Taking anticoagulants (blood thinners such as warfarin or heparin)
  • Anemia or suppressed blood formation
  • Low blood pressure, shock, collapse or fainting
  • Severe wasting (cachexia)
  • Intolerance of leeches or a history of severe allergic reactions
  • Pregnancy
  • Stomach or duodenal ulcer, erosive gastritis
  • Acute infectious disease or high fever
  • Severe unstable illness, markedly weakened immunity (for example during chemotherapy) or HIV infection
  • Drinking alcohol at the time of treatment
  • No consent from the patient
  • Acute septic arthritis
  • Knee replacement (within 6 months)

Related Conditions

References

  • [B1]

    Баскова И.П., Исаханян Г.С. Гирудотерапия. Наука и практика [Hirudotherapy: Science and Practice]

    Москва: Монолит(2004)https://search.rsl.ru/ru/record/01007959553

    Monograph in Russian, 506 pp., ISBN 5-89527-058-1. Chapter 15: sensation at the bite, length of the session; secretion kininases and pain.

  • [B3]

    Michalsen A., Roth M., Dobos G. Medicinal Leech Therapy

    Stuttgart — New York: Thieme(2007)https://books.google.com/books?vid=ISBN9781588905635

    ISBN 978-3-13-143581-1 / 978-1-58890-563-5. Chapter 9, p. 107: perceptions of the bite — a local dragging pain lasting around one to five minutes.

The information on this page is based on the hirudotherapy books listed under References and reflects the experience their authors describe. Sensations and results differ from person to person. This information is not medical advice, a diagnosis or a prescription, and it does not replace an examination by a physician. Do not stop or change any prescribed treatment without your physician. Hirudotherapy has contraindications.

This website provides educational information and does not constitute medical advice, diagnosis, or treatment recommendations. Medicinal leech therapy carries clinically meaningful risks and should be performed only by qualified clinicians under institutionally approved protocols. FDA 510(k) clearance for medicinal leeches is limited to specific indications; investigational and off-label discussions are labeled accordingly. For patient-specific guidance, consult a qualified healthcare provider.