59-year old previously healthywoman visiting New York Citywas riding a subway train one morningwhen the train jolted. She shifted all herweight to one leg, felt a bone snap, andfell to the floor of the train. A hip x-rayin a local emergency department revealeda comminuted spiral fracture involvingthe upper half of the right femur (see fig-ure). The woman was transferred to anorthopedic hospital, where she was notedto be 5’9” tall, 155 pounds, and in a greatdeal of pain. She had no significant med-ical problems aside from osteoarthritisof the knees and thumbs. Her medica-tions consisted of hormone replacementtherapy and alendronate, 70 mg/week.She had been taking alendronate for ap-proximately seven years.The patient had experienced earlymenopause at age 42 and began hormonereplacement therapy (HRT) at that time.At age 52 she had a dual energy x-ray ab-sorptiometry bone density (DXA) scan,which showed a T-score at the lumbarspine of –2.0. This finding, along with astrong family history of osteoporosis, ledto the addition of alendronate to her HRTregimen. Follow-up DXA scans showedsignificant improvementin her bone density. (ADXA scan done 3 monthsafter the fracture showeda spinal mean T score atL2-L4 of –0.6, and a totalfemoral T score of –0.9,with the femoral neck showing a T score of –1.4)Her only prior fracturehistory was a skiing acci-dent at age 24 resulting infractures of her left tibia,fibula, and wrist, all of which healed unevent-fully with casting. Oneday, three months prior tothe subway accident,however, she had begunto experience moderatepain in her right thigh withevery step. There was nopreceding trauma, nor anyrecent increase in herphysical activity. An x-rayof her right femur wasread as “There is slightthickening of the cortex of the right fe-mur laterally, significance uncertain. Sus-pect this is simply normal variation.” Be-cause the pain persisted, a bone scan wasdone, and it was read as “Intense focusof radionuclide uptake in the proximalright femur correlating with a focal areaof cortical thickening. This finding isvery suggestive of a possible underlyingosteoid osteoma. Radiographs reveal noother changes of lytic disease or blasticdisease that would suggest primary orsecondary neoplastic disease.” The bonescan was done one week prior to the oc-currence of the fracture. An MRI of thefemur was ordered for two weeks hence;however, because of a family emergency,the patient flew to New York before thedate of the scheduled MRI.Because the fracture preceded the pa-

Case ReportCase Report

Dr. Schneider

is certified in inter-nal medicine,addiction medicine,and pain management; she prac-tices in Tucson,AZ

Disclosure:

Dr. Schneider reports noconflict of interest.

Jennifer P. Schneider, MD, PhD

Should bisphosphonates be continued indefinitely?

An unusual fracture in a healthy womanon long-term alendronate

www.geri.com

January 2006 Volume 61, Number 1

Geriatrics

The case patient’s hip x-ray showing a comminutedspiral fracture involving the upper half of the right femur.

X-ray courtesy of the author.

tient’s fall, it was thought likely that shehad a pathologic fracture, perhaps sec-ondary to some metastatic lesion. Shewas therefore placed in traction and un-derwent extensive CT scanning of thechest, abdomen, pelvis, lumbar spine,and the femur, as well as plain x-rays,which revealed no evidence of patholog-ical disease consistent with metastaticor primary lesion suggestive of carci-noma. Three days later, an intramedullarytitanium rod was placed. The opinion of the orthopedic surgeons after they re-viewed the out-of-state x-ray and bonescan was that it was typical of a stressfracture. The jolt in the moving subwaytrain completed the fracture.In the months following, it becameclear that the fracture was not uniting.Physical therapy and an extensive trialof an external electrical bone stimulatordid not result in significant union of thefracture. An orthopedic trauma special-ist recommended the patient undergo a“revision intramedullary rodding proce-dure with use of a recon-type nail to aidin fixation of the proximal fragment.”The surgery was done nine months afterthe initial fracture.At the time of her initial hospitalization,the patient was told to stop her HRT be-cause of risk of deep-vein throm-bophlebitis (DVT) related to her immo-bilization. She asked about continuingthe alendronate, since she was concernedthat its suppression of bone turnovermight inhibit healing of the fracture. Shewas told that although this was a theoret-ical possibility, there was no evidence tothat effect, so that there was no reason tostop the drug. However, after months of delayed healing, the patient chose to stopthe alendronate. After the second proce-dure, there was some delay in healing,but by 6 months it was clear that the frac-ture was uniting. Two years after her firstsymptoms of a stress fracture of the fe-mur, she was finally able to get back toher usual level of physical activity.After more than two years off alen-dronate therapy, a DXA scan showedsome decrease in bone density, and thepatient was advised to resume taking thedrug. One year later, she awoke to findthat she had moderate pain in her rightfoot with every step. There was no pre-ceding trauma nor any increase in activ-ity. The possibility of another nontrau-matic stress fracture was considered andagain the bisphosphonate therapy wasstopped. Two months later, a bone scanshowed intense uptake in the secondmetatarsal bone, consistent with a stressfracture. The patient continued takingcalcium supplements, 500 mg/bid, and anestrogen/progesterone combination,1 mg/d, and walking one mile daily wear-ing sturdy shoes to support her foot. Af-ter several months, the fracture healed.

Discussion

This case report describes a previouslyhealthy woman who experienced twonontraumatic stress fractures, four yearsapart, while on alendronate therapy, andalso nonunion of the spiral femoral frac-ture that resulted from the stress frac-ture. A spontaneous stress fracture of thefemur is so unusual that neither her or-thopedic surgeon nor the radiologist whoread the bone scan even considered thatdiagnosis in their differential. Conse-quently, the patient was never cautionedabout her vulnerability to sustain a com-pleted fracture.Bisphosphonates–such as alendronate,risedronate, and ibandronate—are in-hibitors of bone resorption. Extensivestudies have shown that therapy with bis-phosphonates improves bone density anddecreases fracture risk.

1–5

These drugs,especially the oldest one, alendronate,are used by large numbers of post-menopausal women, as well as smallernumbers of men with idiopathic, steroid-induced, hypogonadal, or other causesof osteoporosis. Combined use of bis-phosphonates and estrogen gives evengreater improvement in bone density.

6–8

Increased bone density does not nec-essarily equate with good bone quality,however. Bone turnover is a natural partof maintaining bone health. By decreas-ing osteoclast activity and bone resorp-tion—and therefore bone formation aswell—microdamage that occurs regu-larly in bone but is normally repairedmight accumulate after long-term use.There have long been concerns about thepotential oversuppression of boneturnover during long-term use of bispho-sphonates and therefore their long-termsafety.

9–14

The concern is increased whenthe bisphosphonate is taken concurrentlywith another agent that may inhibit boneturnover, such as estrogen. The currentpatient package insert (PPI) for Fosamax(alendronate) states, “The long-term ef-fects of combined Fosamax and HRT onfracture occurrence and fracture healinghave not been studied.” Clearly, suchstudies are needed.Recently, Odvina and colleagues

re-

Case Report

Geriatrics

January 2006 Volume 61, Number 1

Studies show thattherapy withbisphosphonatesimproves bonedensity, decreasesfracture risk

TableBisphosphonates approved for osteoporosis

Agent dose