Knee Replacement After ACL Reconstruction: Rates in 111,212 Patients Over 20 Years
Key points
- I analysed NHS hospital records for 111,212 people who had a first ACL reconstruction in England between April 1997 and March 2017.
- Few went on to a knee replacement: 0.46% within 5 years, 0.97% within 10 years and 1.8% (around 1 in 55) within 15 years.
- Risk rose with age at the time of surgery: within 15 years, 0.35% of patients aged 20 to 29 had a knee replacement, against 5.5% of those aged 40 to 49.
- At ages 30 to 39, the yearly rate of knee replacement was 20 times that of the general population, but still very low at 0.04% a year.
- Everyone in the study had surgery, so it cannot show whether reconstruction itself changes the risk.
What Was This Study About?
Osteoarthritis is a recognised concern after an anterior cruciate ligament (ACL) injury, but X-ray changes do not always match a person's symptoms. A knee replacement is a clearer marker of severe arthritis, so I set out to measure how often people who have had an ACL reconstruction go on to have one.
In this study, I used the national Hospital Episode Statistics database, which records NHS hospital care across England. With my colleagues, I identified everyone who had an ACL reconstruction between 1 April 1997 and 31 March 2017. Only each patient's first reconstruction was counted, so revisions and operations on the opposite knee were not included. Multi-ligament reconstructions, ACL repairs and synthetic ligaments were excluded, as were patients who had a knee replacement at the same time and records missing age, sex, date or side.
That left 111,212 of the 124,489 patients who had ACL surgery. Their average age was 29, 77% were male, and they were followed for an average of 5.9 years (up to 20).
The main question was how many went on to a total or partial knee replacement, matched to the knee that had been reconstructed. I also looked at which patient factors were linked with a higher rate, and compared yearly rates in 2016–17 with those of the general population.
What Did We Find?
Knee replacement was uncommon. Among patients with at least 5, 10 and 15 years of follow-up, the proportions who had one were:
- within 5 years: 0.46% (212 of 46,402 patients)
- within 10 years: 0.97% (139 of 14,292)
- within 15 years: 1.8% (66 of 3,726)
Age at the time of surgery. The older a patient was when the ACL was reconstructed, the more likely a later knee replacement:
- aged 20 to 29: 0.12% within 10 years and 0.35% within 15 years
- aged 30 to 39: 0.82% and 2.2%
- aged 40 to 49: 2.7% and 5.5%
- aged 50 to 59: 7.1% and 8.3%
The 15-year figures for the two older groups rest on 403 and 60 patients, so they are less certain. After adjustment for other factors, patients aged 30 to 39 were 6.2 times as likely to have a knee replacement as those aged 20 to 29, and those aged 40 to 49 were 19 times as likely. Results for patients under 20 were suppressed because of small numbers.
Sex and other factors. Within 10 years, 1.8% of women and 0.78% of men had a knee replacement; after adjustment, women were 1.5 times as likely as men. Patients from the most deprived fifth of areas were 2.0 times as likely as those from the least deprived. Rates were lower in patients of Asian ethnicity than in those of White ethnicity, for reasons that are uncertain. Cartilage surgery at the same operation made no significant difference.
Meniscal surgery. Patients who had meniscal surgery at the same operation had a lower rate of later knee replacement (0.41 times that after ACL reconstruction alone). This was a surprise. It conflicts with studies in which people with a meniscal injury are at greater risk of arthritis progressing on X-ray, and meniscal repair and partial removal could not be analysed separately.
Comparison with the general population. For 2016–17, I compared yearly rates of knee replacement in people with and without a previous ACL reconstruction, by age in that year:
- 30 to 39: 37 against 1.9 per 100,000 (20 times as high)
- 40 to 49: 186 against 25 per 100,000 (7.5 times as high)
- 50 to 59: 384 against 151 per 100,000 (2.5 times as high)
- 60 to 69: 717 against 414 per 100,000 (not statistically significant)
Across ages 30 to 69 together, the rates were not significantly different (148 against 133 per 100,000). These findings suggest that an ACL injury is a risk factor for developing osteoarthritis early.
Why Does This Matter for Patients?
Most people in this study did not go on to a knee replacement. The rate was still higher than for other people of the same age, most clearly in the 30s and 40s. A rate 20 times as high sounds alarming, but at ages 30 to 39 it amounted to 37 knee replacements a year for every 100,000 people with a previous ACL reconstruction.
A knee replacement marks only severe, symptomatic arthritis. Surgeons are reluctant to replace a knee in a younger person, so more people will have arthritis on X-ray, or a troublesome knee, than will have a knee replacement. Earlier research cited in the paper suggests that up to half of people with a previous ACL injury show signs of osteoarthritis on X-ray within 10 to 15 years.
These results should not be used to choose between surgery and rehabilitation, because everyone in the study had a reconstruction. Where the decision is to reconstruct, they give patients and surgeons an estimate of this long-term risk by age and sex.
What Could This Study Not Tell Us?
This study cannot determine whether ACL reconstruction protects the knee against severe arthritis. It is also unclear whether arthritis after an ACL rupture is driven mainly by the original injury or by later episodes of the knee giving way. Records of the injury itself were not available, and the timing of the reconstruction was not known.
The study counted knee replacements only, with no X-ray findings or patient-reported outcomes. Some coding errors will have remained in the hospital records after data cleaning. The codes used to identify ACL reconstruction also pick up posterior cruciate ligament reconstructions, which could not be separated out but are very rare and unlikely to have materially changed the findings.
Other influences on arthritis, such as genetic, biochemical and biomechanical factors, were not measured. My colleagues and I later returned to the question in a larger study with longer follow-up, which compares each reconstructed knee with the same person's opposite knee.
Frequently asked questions
Will I need a knee replacement after ACL reconstruction?
Most people in this study did not. Among 111,212 patients who had an ACL reconstruction in England, 0.46% had a knee replacement within 5 years, 0.97% within 10 years and 1.8% (around 1 in 55) within 15 years.
Is a knee replacement more likely after ACL reconstruction than for other people my age?
At younger ages, yes. Compared with people with no history of ACL reconstruction, the yearly rate of knee replacement in this study was 20 times as high at ages 30 to 39, 7.5 times as high at 40 to 49 and 2.5 times as high at 50 to 59. The rate itself was low, at 0.04% a year at ages 30 to 39.
Does ACL reconstruction prevent arthritis or a knee replacement later in life?
This study cannot answer that. All 111,212 patients had an ACL reconstruction, so there was no group treated without surgery to compare them with, and the risk of knee replacement had they been managed without an operation is unknown.
Who is most likely to need a knee replacement after ACL reconstruction?
Patients who were older when their ACL was reconstructed. Within 15 years, 0.35% of those aged 20 to 29 at the time of surgery had a knee replacement, against 2.2% of those aged 30 to 39 and 5.5% of those aged 40 to 49. Women were 1.5 times as likely as men after adjustment for other factors.
Read the Full Paper
This paper is published open access in Acta Orthopaedica under a Creative Commons licence (CC BY 4.0). You can read the full text below, or on the journal website or on PubMed Central.
Full citation: Abram SGF, Judge A, Khan T, Beard DJ, Price AJ. Rates of knee arthroplasty in anterior cruciate ligament reconstructed patients: a longitudinal cohort study of 111,212 procedures over 20 years. Acta Orthopaedica. 2019;90(6):568–574. doi:10.1080/17453674.2019.1639360
This summary is written for general information and explains the research in patient-friendly terms. It does not replace individual medical advice. If you have questions about an ACL injury or your knee, please get in touch to arrange a consultation.