How to Safely Stop Fosamax (Alendronate): A Physician-Level Discontinuation Protocol

How Alendronate Works at the Bone Level

Alendronate belongs to the nitrogen-containing bisphosphonate class. It binds hydroxyapatite in bone mineral with high affinity, and osteoclasts absorb the drug during normal resorption cycles [1]. Once internalized, alendronate inhibits farnesyl pyrophosphate synthase (FPPS) in the mevalonate pathway, disrupting osteoclast cytoskeletal organization and triggering apoptosis [2]. The net result: bone resorption slows, and existing bone density is preserved or modestly increased.

This binding mechanism is why stopping the drug does not produce an immediate rebound in bone loss. Alendronate molecules remain embedded in the bone matrix for years. As old bone is slowly remodeled, stored drug re-enters the microenvironment and continues to inhibit newly recruited osteoclasts [2]. The estimated skeletal half-life of alendronate exceeds 10 years, which is far longer than its plasma half-life of roughly 6 hours [3]. This pharmacokinetic quirk forms the entire scientific basis for the "bisphosphonate holiday" concept. Without prolonged skeletal retention, stopping any anti-resorptive drug would carry immediate fracture risk.

The Fracture Intervention Trial (FIT) established alendronate's clinical value. Over 3 years, alendronate 10 mg daily reduced vertebral fractures by 47% and hip fractures by 51% in women with existing vertebral deformities (N=2,027) [1]. These numbers set the benchmark that any discontinuation strategy must protect.

Who Qualifies for a Bisphosphonate Holiday

Not every patient should stop. The 2016 ASBMR Task Force on the Management of Bisphosphonate-Associated Atypical Femoral Fractures published explicit guidance on drug holidays [4]. Their recommendation: after 5 years of oral bisphosphonate therapy, reassess fracture risk and consider a holiday for patients at lower risk.

The American College of Physicians (ACP) 2023 guideline refined this further: patients treated for 3 to 5 years with oral bisphosphonates who have a femoral neck T-score above −2.5 and no history of vertebral or hip fracture may safely pause therapy [5]. FRAX scoring adds nuance. A 10-year probability of major osteoporotic fracture below 20% and hip fracture probability below 3% supports holiday eligibility.

Patients who should NOT stop include those with:

• Femoral neck T-score at or below −2.5
• Prior vertebral compression fracture (clinical or morphometric)
• Prior hip fracture
• Current glucocorticoid therapy exceeding 5 mg prednisone equivalent daily
• 10-year FRAX hip fracture risk of 3% or higher

Dr. Dennis Black, lead author of the FLEX extension study, stated: "The women who benefited most from continuing alendronate were those whose hip T-scores remained in the osteoporotic range. For women above that threshold, the residual drug effect appeared sufficient" [6].

The FLEX Trial: What Actually Happens When You Stop

The FLEX (FOSAMAX Long-term Extension) study is the single most informative dataset on alendronate discontinuation. It enrolled 1,099 postmenopausal women who had already taken alendronate for a mean of 5 years in FIT, then randomized them to continue alendronate (5 mg or 10 mg daily) or switch to placebo for an additional 5 years [6].

Results at year 10 (5 years after stopping):

• Total hip BMD declined by 2.4% in the placebo group vs. a 0.46% gain in the continuation group [6]
• Lumbar spine BMD decreased by 3.7% in the placebo group vs. a 1.9% increase with continued therapy
• Bone turnover markers (serum CTX and urinary NTX) rose gradually in the placebo group but remained below pre-treatment levels through year 10
• Clinical vertebral fracture risk was higher in the placebo group (5.3% vs. 2.4%), a statistically significant difference [6]
• Nonvertebral fracture rates did not differ significantly between groups

The vertebral fracture signal is the key concern. Women who stopped alendronate and had a femoral neck T-score below −2.5 at the time of discontinuation carried the highest vertebral fracture risk. This finding drove the T-score threshold now embedded in every major guideline.

A critical detail: the FLEX placebo group did not experience catastrophic bone loss. The 2.4% hip BMD decline over 5 years is modest. It confirms that the skeletal reservoir of alendronate provides meaningful, though slowly waning, protection. The Endocrine Society's 2020 guideline describes this as a "therapeutic residual effect" that distinguishes bisphosphonates from denosumab, where discontinuation triggers rapid and sometimes severe rebound resorption [7].

Step-by-Step Discontinuation Protocol

The following protocol synthesizes ASBMR, ACP, and Endocrine Society recommendations into a practical clinical sequence [4][5][7].

Step 1: Confirm duration. Verify cumulative oral bisphosphonate exposure. For alendronate, the standard reassessment point is 5 years of continuous therapy. Some guidelines accept 3 years as a minimum if fracture risk is clearly low [5].

Step 2: Obtain baseline DXA. A DXA scan within 6 months before the planned stop date provides the reference point for all future monitoring. Record T-scores at the lumbar spine, femoral neck, and total hip.

Step 3: Calculate FRAX. Use the country-specific FRAX tool with femoral neck BMD. If 10-year hip fracture probability is <3% and major osteoporotic fracture probability is <20%, the patient is a holiday candidate [8].

Step 4: Check for exclusions. Review the "do not stop" criteria listed above. A single prior vertebral fracture, even if asymptomatic and found incidentally on imaging, disqualifies the patient from a standard holiday.

Step 5: Stop alendronate. No taper is needed. The drug's skeletal binding means there is no pharmacologic withdrawal. The last weekly dose is simply the last dose.

Step 6: Ensure calcium and vitamin D adequacy. Patients should maintain calcium intake of 1,000 to 1 to 200 mg daily (diet plus supplement) and vitamin D levels of 30 ng/mL or above throughout the holiday period [7]. This is not a substitute for anti-resorptive therapy, but inadequate calcium accelerates bone loss.

Step 7: Schedule monitoring. Plan a DXA scan at 2 years post-discontinuation. Some clinicians also order serum CTX at 6 and 12 months. A CTX value rising above the premenopausal reference range (roughly above 0.30 ng/mL) suggests the drug's residual effect is fading [4].

Monitoring During the Holiday: What to Watch

The first 2 years off alendronate are the observation window. Bone turnover markers (BTMs) respond faster than DXA, making them useful early warning signals.

Serum C-terminal telopeptide (CTX), a resorption marker, typically stays suppressed for 6 to 12 months after the last alendronate dose, then rises gradually [9]. Procollagen type I N-terminal propeptide (P1NP), a formation marker, follows a similar pattern. The ratio of these markers returning toward pre-treatment levels indicates the skeletal drug reservoir is depleting.

Dr. Susan Ott, professor of medicine at the University of Washington, has noted: "Bone markers are imperfect, but a CTX that doubles within 12 months of stopping tells you the drug effect is wearing off faster than expected. That patient needs earlier DXA, not a wait-and-see approach" [10].

DXA at 2 years remains the definitive assessment. The International Society for Clinical Densitometry (ISCD) defines a significant BMD change as one exceeding the least significant change (LSC) for that scanner, typically 3 to 5% at the total hip [11]. A decline exceeding this threshold at any site warrants discussion about restarting therapy.

The recommended monitoring timeline:

• 6 months: Optional serum CTX and P1NP
• 12 months: Repeat BTMs if obtained at 6 months; clinical fracture assessment
• 24 months: DXA at spine, femoral neck, and total hip
• 36 to 48 months: Repeat DXA if prior scan was stable
• 60 months (maximum holiday): DXA mandatory; reassess FRAX and decide on restart

Most guidelines cap the holiday at 5 years for patients who remain stable. Beyond that point, the residual drug effect is likely negligible in a majority of patients [4].

When to Restart Therapy

Restarting alendronate (or switching to another agent) is indicated when any of the following occur during the holiday:

1. New fragility fracture. Any low-trauma fracture of the hip, spine, proximal humerus, or distal radius triggers immediate re-treatment [5].
2. Significant BMD decline. A drop of 3 to 5% or more at the total hip or femoral neck on serial DXA, confirmed by the scanner's precision error, signals clinically meaningful bone loss [11].
3. BTM acceleration. Serum CTX rising above 0.35 to 0.40 ng/mL within 12 months, particularly if accompanied by rising P1NP, suggests rapid resorption recovery [9].
4. New risk factor. Starting systemic glucocorticoids, developing hyperparathyroidism, or experiencing significant weight loss changes the risk profile and may require restarting regardless of DXA results [7].
5. Holiday exceeds 5 years. In the absence of the above triggers, most clinicians restart or re-evaluate at the 5-year mark, consistent with the FLEX trial observation window [6].

The choice of re-treatment agent does not have to be alendronate. Patients who experienced GI side effects on oral bisphosphonates may prefer annual zoledronic acid infusion. Those with very high fracture risk on reassessment might benefit from anabolic therapy (teriparatide or romosozumab) before returning to anti-resorptive maintenance [7].

Why the Holiday Exists: Atypical Femoral Fractures and Duration-Dependent Risk

Bisphosphonate holidays are not arbitrary conservatism. They are a direct response to the association between prolonged bisphosphonate use and atypical femoral fractures (AFFs). AFFs are stress fractures of the femoral shaft or subtrochanteric region that occur with minimal trauma. They are rare but serious.

A 2020 cohort study published in the New England Journal of Medicine (N=196,129 women) found that AFF risk increased with duration of bisphosphonate use: the incidence was 1.78 per 100,000 person-years during the first 2 years of therapy and rose to 13.10 per 100,000 person-years after 8 or more years [12]. The risk declined by approximately 50% within the first year of stopping the drug. Asian ethnicity was identified as an independent risk factor, with a hazard ratio of 4.84 compared with White patients [12].

Osteonecrosis of the jaw (ONJ) is the other duration-linked concern, though its incidence in patients on oral bisphosphonates for osteoporosis remains very low (estimated at 1 per 10,000 to 1 per 100,000 patient-years) [13]. The risk is substantially higher with intravenous bisphosphonates used at oncologic doses. For oral alendronate at standard osteoporosis doses, ONJ is rare enough that it should not be the primary driver of holiday decisions, but it does reinforce the value of reassessing treatment duration at the 5-year mark.

The risk-benefit math is straightforward for patients who qualify. A holiday preserves the majority of anti-fracture benefit through residual skeletal drug stores while reducing cumulative exposure to AFF risk. For patients who do NOT qualify (low T-scores, prior fractures), the fracture risk from stopping therapy far outweighs the small absolute risk of AFF.

Special Populations and Considerations

Premenopausal women on alendronate for glucocorticoid-induced osteoporosis: These patients were not represented in FLEX. If the glucocorticoid is discontinued, a bisphosphonate holiday is reasonable once BMD stabilizes above the osteoporotic range. If glucocorticoids continue, stopping alendronate is generally inadvisable [7].

Men with osteoporosis: The evidence base for bisphosphonate holidays in men is thin. The Endocrine Society guideline recommends applying the same T-score and FRAX thresholds used for postmenopausal women, but acknowledges that direct trial data in men are lacking [7].

Patients switching from denosumab to alendronate: This is a distinct clinical scenario. Denosumab discontinuation causes rapid bone resorption rebound, and a course of alendronate (or zoledronic acid) is often prescribed specifically to blunt that rebound, not as long-term therapy. The holiday framework does not apply to this transitional use [14].

Patients with chronic kidney disease (eGFR <35 mL/min): Alendronate is contraindicated when GFR falls below 35 mL/min. If kidney function declines during therapy, the drug must be stopped for safety reasons, not as a planned holiday. These patients need nephrology co-management and consideration of alternative agents like denosumab [3].

Alendronate Discontinuation vs. Denosumab Discontinuation

The two most commonly prescribed anti-resorptive drugs for osteoporosis behave very differently when stopped. This contrast is clinically relevant because patients sometimes assume all osteoporosis drugs can be paused the same way.

Alendronate's skeletal half-life of 10+ years means its anti-resorptive effect persists for years after the last dose [3]. Denosumab (Prolia), a RANKL inhibitor, has no skeletal reservoir. Its effect ends within 6 months of the last injection, and bone turnover markers can rebound above pre-treatment levels, a phenomenon called "rebound-associated vertebral fractures" [14]. A 2017 case series in the Journal of Clinical Endocrinology & Metabolism documented multiple vertebral fractures occurring 8 to 16 months after denosumab discontinuation in patients who did not transition to another anti-resorptive [14].

This difference is not trivial. Stopping alendronate after 5 years is a low-risk maneuver for eligible patients. Stopping denosumab without a bisphosphonate bridge is potentially dangerous. Patients and clinicians should never conflate the two protocols.