Alendronate helps the very old

​​A new observational Swedish study found that alendronate was effective in preventing hip fractures in those >80 yo who had a prior hip fracture (see doi: 10.1111/joim.12678). a publicly-funded study.

 

Details:

--90,795 men and women >80yo with history of prior fracture in a Swedish national database of people who had the standard fall risk assessment

--7844 controls and 1961 alendronate-treated patients were propensity score-matched by age, sex, weight, height, rheumatoid arthritis, alcohol-related diseases, prior steroid treatment, the individual diseases causing secondary osteoporosis (eg, hyperthyroidism, hypogonadism, chronic liver disease), Charlson Comorbidity index (includes most of the chronic medical diseases), and fracture history (time since index fracture, number of fractures, prior fall injury and osteoporosis diagnosis)

--mean age 86, 88% female, 0.4% alcohol-related disease, 5% RA, 25% prior “intense steroid use”, 4 yrs since prior fracture,70% prior fall injury, 23% prior hip fracture and 16% total hip replacement, 41% had full mobility/35% slightly limited/14% very limited/1% immobile. No significant differences between the groups for these variables (which was largely attributable to their propensity matching)

--those on alendronate had been on it a mean of 3.5 years, and were very adherent to therapy (mean medication possession ratio of 91%). These patients took more calcium/vitamin D supplements (85% vs 30%)

--end-point: incident hip fracture

 

Results:

--those on alendronate had:

    --38% decreased risk of hip fracture overall, HR 0.62 (0.49-0.79), p<0.001

    --in multivariate model, 34% decreased risk, HR 0.66 (0.51-0.86), p<0.01, controlling for the above propensity-matching risk factors along with taking calcium/vitamin d and having had total hip replacement

    --review of their graph: the hip fracture decrease with alendronate improved over time (the curves splay further apart over the 5 years on the graph)

    --the absolute risk reduction was 3.9%, leading to a number-needed-to-treat of 26 over 3 years

    --12% decreased mortality risk, HR 0.88 (0.82-0.95), with a nonsignificant trend to decreased hip fracture-related mortality

--alendronate was significantly associated  a reduction in hip fracture (36%), but only in patients with a prior fall injury

--but, alendronate was associated with a 58% increased risk of mild upper GI symptoms, HR 1.58 (1.12-2.24), not related to patient age

--No increased risk of peptic ulcers. Drug-induced osteonecrosis was rare (1 case in treated group). atypical femoral fractures not specifically assessed, but the rate of femoral shaft fractures was low and not significantly higher in alendronate users.

 

 

Commentary:

--as a perspective issue: the US population is aging (11.7 million >80yo in 2012, expected to increase to >20 million in 2030), the risk of fracture increases dramatically with age (life-time risk for women 50%, men 20%), and there is huge morbidity (fragility fractures now lead to more in-hospital days than combo of breast and prostate cancer, and has current cost of $17 billion). Hip fractures themselves have significant morbidity (40-60% of survivors do not recover pre-fracture level of mobility) and mortality (1-yr mortality increased 8-36%)

--alendronate works in post-menopausal women, with a 40% relative risk reduction, but older women (who are at much higher risk) had not been studied.

--so, this study extends these results to older patients, with similar relative risk reductions with alendronate treatment as in younger patients.

--this was not a randomized control trial, so hard to know if there were uncontrolled biases (did those put on alendronate fundamentally differ from those that their clinician elected not to prescribe alendronate?? And in ways not mathematically controlled for in the above multivariate analysis, such as smoking). Also this issue is more complex in a large data-mining study such as this one, where they had to use approximate risk factors (instead of quantifying alcohol drinking, they looked at alcohol-related diseases. Also no data on number of falls, just on fall injury). And there were no data on bone mineral density (though the prior fragility fracture would have qualified them for bisphosphonate therapy anyway)

 

So, the big finding here is that alendronate seems to work well in older patients, and with minimal adverse effects. These results were independent of sex, but there was no analysis stratifying specifically by patient sex (ie, we can't draw conclusions about men). However, these results reflect a pretty short treatment time of only 3.5 years on alendronate (and both the benefits and harms may be understated, as compared to a longer study). A prior blog reported on longer-term followup of a large Danish cohort, finding that protective effect of alendronate continued after 10 years and was quite safe (they found a 57% reduction in subtrochanteric femur fractures with >10 years of therapy!!!), see http://bit.ly/2h1TE1w . For the Am Society for Bone and Mineral Research guidelines on length of bisphosphonate therapy, see http://bit.ly/2xZ2zaR 

 

Bottom line: though not a definitive study, I think the results are quite impressive, are in line with the prior RCTs in younger women, and are especially important given the much higher absolute risk for fracture as women get older (and without major adverse events from the alendronate).  So, to me this study reinforces what I think a lot of us have been doing: putting these high risk women on alendronate as a means to decrease the well-known significant morbidity and mortality associated with hip fractures.