Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Quinolones and Achilles tendon rupture

Study
Results
Comment

Quinolone antibiotics have been associated with tendon problems since they were introduced in the 1980s. Case reports have indicated that Achilles tendon problems, and tendon rupture, may be a particular concern. How common these adverse effects may be, and whether they are associated with quinolone use, is answered by a case-control study using the UK general practice database [1].

Study

The UK general practice database is an electronic database with detailed information from a large number of practices and patients, with information on diagnoses, prescriptions, demographics, and hospital episodes. The population for this study was patients aged 18 to 95 years over the study period of 1989 to 1998.

Cases were people with a first time rupture of the Achilles tendon, with over 18 months of history in the database, and without a history of cancer, drug abuse, alcoholism, or hospital admission in the month before the event. Rupture due to major trauma (falling, car accident) was also an exclusion criterion. Indirect trauma associated with exercise, for instance, was not excluded. Controls were 50,000 randomly chosen people, with similar exclusion criteria applied.

For each case and control exposure to oral or parenteral quinolones was ascertained, defined as current (0-1 month), recent (2-6 months) or past (7-18 months). Type of quinolone and fraction of defined daily dose were recorded. Potential risk factors for Achilles tendon rupture were also recorded, including transplantation, arthritis, gout, systemic corticosteroid use, and other factors. Risk factors found to be associated with Achilles tendon rupture were used to adjust the final odds ratio of association with use of quinolones.

Results

There were 1,367 cases of Achilles tendon rupture after a blinded review to ensure that inclusion and exclusion criteria were applied, and diagnosis appropriate. Between the cases and controls the major differences were the proportion of men (69% cases, 48% controls), corticosteroid use (11% cases, 4.6% controls), gout (3% cases, 1.5% controls), and transplant or dialysis (0.2% cases, 0.03% controls).

Most cases and controls had not used quinolones (Table 1), and any exposure to quinolones in the previous 18 months occurred in 4.5% of cases and 2.0% of controls. Only 14 cases were current quinolone users, and these were all aged over 60 years (Table 1).



Table 1: Number of cases with Achilles tendon rupture and controls, and use of quinolones



Use of quinolines
Cases
Controls
Total patients
1367
50000
Non-use
1305
48981
Past use (7-18 months)
24
605
Recent use (2-6 months)
24
315
Current use (0-1 months)
14
100
Of which
Age <60 years
0
50
Age 60-79 years
11
41
Age ≥80 years
3
9


Four different quinolones were used (ofloxacin, ciprofloxacin, norfloxacin and nalidixic acid). Numbers of cases using individual drugs were small (six or fewer), but the highest adjusted odds ratios compared with nonusers were found for ofloxacin (OR 28) and norfloxacin (OR 14). For use of quinolones within three months, odds ratios were considerably higher with increasing DDD of quinolones (Figure 1), though only three cases and three controls used quinolones at the highest dose above 1.25 DDD.



Figure 1: Adjusted odds ratio of association between Achilles tendon rupture and dose





Nine cases and five controls were currently using quinolones and were exposed to corticosteroids, and three were currently using quinolones with recent exposure to corticosteroids. For these the adjusted odds ratio for Achilles tendon rupture was about 18 compared with no quinolone exposure, while for the four cases and 39 controls currently using quinolones and with no exposure to corticosteroids it was 5.3.

Comment

This is a detailed study, looking at a large population with excellent recording of patient details, and with a large number of Achilles tendon ruptures. Only 14 of these events occurred in people using quinolones, all aged over 60 years, and in nine of those there was concurrent use of corticosteroids, also known to be associated with Achilles tendon rupture. For any further analysis, by type of drug, by dose of drug, or by duration of use of drug, the numbers of cases mainly fall to single figures.

What we can be reasonably sure of is that quinolone use in older people also taking corticosteroids has a large increased risk of Achilles tendon rupture. Larger doses probably carry a bigger risk. Most of the other analyses contain very few cases or controls, and there must be a risk that they will be wrong just by the random play of chance.

Reference:

  1. PD van der Linden et al. Increased risk of Achilles tendon rupture with quinolone antibacterial use, especially in elderly patients taking corticosteroids. Archives of Internal Medicine 2003 163: 1801-1807.

previous or next story