Twenty-year follow up study comparing operative versus nonoperative treatment of anterior cruciate ligament ruptures in high-level athletes

Key points

1. 01Knee osteoarthritis rates at 20 years were 80% (operative) vs 68% (non-operative), a non-significant difference (p=0.508).
2. 02Subjective functional scores (Lysholm, IKDC, KOOS, Tegner) were not significantly different between groups at any time point.
3. 03Objective knee stability was significantly better after surgery: 84% of the operative group had a normal IKDC score vs 20% of the non-operative group (p<0.001).
4. 04Meniscectomy rates over 20 years were similar: 72% (operative) vs 76% (non-operative), with no significant difference (p=0.057).
5. 05OA rates roughly doubled in both groups between the 10-year and 20-year follow-ups, from 48% to 80% (operative) and 28% to 68% (non-operative).

How it was conducted

Design

Retrospective pair-matched cohort study with 20-year follow-up

Participants

50 high-level athletes with ACL rupture: 25 treated operatively (bone-patellar-tendon-bone reconstruction, 1994-1996), 25 treated non-operatively (physiotherapist-led rehabilitation, 1992); matched by age, gender, and Tegner score

Primary outcome

Radiological knee osteoarthritis (Kellgren and Lawrence grade >=2)

Secondary outcomes

Functional scores (Lysholm, IKDC subjective, KOOS, Tegner), knee stability (KT-1000, pivot shift, Lachman, one-leg-hop test), meniscectomy rates

Follow-up

Median 21.2 years (operative group) and 24.1 years (non-operative group)

Response rate

100% - all 50 patients from the 10-year study included

What they found

• Knee OA at 20 years: 20/25 (80%) operative vs 17/25 (68%) non-operative, p=0.508 (not significant).
• Lysholm score at 20 years: median 86.0 (IQR 75.5-91.0) operative vs 89.0 (IQR 75.5-95.5) non-operative, p=0.851.
• IKDC subjective at 20 years: median 81.6 (IQR 59.8-89.1) operative vs 78.2 (IQR 61.5-92.0) non-operative, p=0.679.
• Normal IKDC objective score (A+B) at 20 years: 21/25 (84%) operative vs 5/25 (20%) non-operative, p<0.001.
• Negative pivot shift at 20 years: 17/25 (68%) operative vs 3/25 (13%) non-operative, p<0.001.
• Negative Lachman at 20 years: 12/25 (48%) operative vs 1/25 (4%) non-operative, p=0.002.
• KT-1000 side-to-side difference >3mm at 20 years: 10/25 (40%) operative vs 19/23 (82.6%) non-operative, p=0.013.
• One-leg-hop test at 20 years: median 85.9% (IQR 68.1-101.9) operative vs 95.1% (IQR 70.8-104.7) non-operative, p=0.449 (not significant).
• Additional meniscectomies over 20 years: 4 in the operative group vs 10 in the non-operative group, p=0.057 (not significant).
• Knee OA in patients who had a meniscectomy: 81%; in those without: 54%.

Limitations

• Retrospective non-randomized design with pair-matching used to reduce allocation bias, but confounding cannot be excluded.
• The transtibial bone-patellar-tendon-bone technique used in the 1990s is now considered outdated; modern anatomical femoral placement may yield different results.
• The observer was not blinded to treatment group, introducing potential assessment bias for clinical stability measures.
• Small sample size (n=25 per group) limits statistical power, particularly for detecting moderate differences in osteoarthritis rates.

Why it matters

For patients

Athletes who rupture their ACL can reasonably consider non-operative rehabilitation as a long-term strategy, since surgery does not appear to prevent osteoarthritis or improve function over 20 years.

For clinicians

The decision to reconstruct the ACL in high-level athletes should be based on symptom burden and instability complaints rather than the assumption that surgery prevents cartilage degeneration, as both approaches lead to similar OA rates by 20 years.

For readers

This study challenges the presumption that ACL reconstruction is inherently superior for active patients, while acknowledging that knee stability is objectively better after surgery even if that does not translate into functional or structural benefit.