Introduction

FRAX (Fracture Risk Assessment Tool) is the global standard for estimating 10-year fracture probability, but it has a critical limitation: it tells clinicians the risk level without guiding therapeutic selection. Current guidelines (AACE 2020, Endocrine Society 2020, ACR 2022) recommend clinical judgment for choosing between antiresorptive and anabolic therapies, but provide no structured algorithm integrating biochemical markers of bone remodeling.

This gap is particularly problematic in rheumatology, where glucocorticoid dose (treated as binary by FRAX), disease activity, and immunomodulatory therapies significantly modify bone metabolism in ways FRAX cannot capture.

Methods

Algorithm Design

OSTEO-TX is a 4-step expert system:

Step 1 — Baseline Metabolic Correction (mandatory):

• 25-OH Vitamin D < 20 ng/mL → colecalciferol loading before therapy
• Calcium < 8.5 mg/dL → investigate and correct (contraindication to antiresorptives)
• Calcium > 10.5 mg/dL → investigate hyperparathyroidism/myeloma
• Alkaline phosphatase elevated with normal GGT → consider Paget disease

Step 2 — Bone Turnover Classification:

• CTX (C-telopeptide) as resorption marker, fasting morning sample
• P1NP (procollagen type 1 N-terminal propeptide) as formation marker per IOF/IFCC recommendation
• Four phenotypes: High turnover (CTX↑ P1NP↑), Low/adynamic (CTX↓ P1NP↓), Discordant type 1 (CTX↑ P1NP↓ — uncoupled resorption), Discordant type 2 (CTX↓ P1NP↑ — investigate Paget/metastasis)

Step 3 — FRAX Integration with Rheumatological Modifiers:

• Standard FRAX risk categories (low/moderate/high)
• Continuous glucocorticoid adjustment: 5-7.4mg (+20%), ≥7.5mg ×3mo (+50%)
• Disease activity modifier: active RA/LES (+20%)
• Biologic adjustment: anti-TNF (-10%), rituximab (+15%)
• ACR 2022 GIOP override: prednisone ≥7.5mg/d ×≥3 months → treat regardless of FRAX

Step 4 — Therapeutic Pathway Selection:

• High risk + high turnover → denosumab 60mg SC/6mo or zoledronate 5mg IV/year
• High risk + low turnover → romosozumab 210mg SC/mo ×12mo or teriparatide 20μg SC/d ×24mo → mandatory antiresorptive sequence
• High risk + discordant type 1 → denosumab preferred (potent resorption suppression) + investigate cause
• Moderate risk + high turnover → oral bisphosphonate (alendronate/risedronate)
• Low risk → surveillance with optimization

Contraindication Logic

• GFR < 30 mL/min → bisphosphonates contraindicated, denosumab permitted with calcium monitoring
• Prior ONJ → mandatory dental evaluation
• Prior atypical femoral fracture → discontinue bisphosphonate, consider anabolic
• Pregnancy/lactation → no antiresorptives or anabolics

Monitoring Protocol

• Biomarkers at 3 months: antiresorptives should reduce CTX ≥30%, anabolics should increase P1NP ≥30%
• DXA at 24 months (not earlier — insufficient sensitivity)
• Fracture: continuous surveillance — this is the real endpoint

Implementation

Open-source calculator at https://rheumascore.xyz/exp-osteo-tx.html All computation client-side (JavaScript). No patient data transmitted. Dark mode UI consistent with RheumaScore platform.

Testable Predictions

1. CTX reduction ≥30% at 3 months in high-turnover patients on antiresorptives
2. P1NP increase ≥30% at 3 months in low-turnover patients on anabolics
3. Discordant type 1 phenotype predicts accelerated fracture vs concordant high turnover
4. Algorithm-guided therapy vs FRAX-only should show lower fracture incidence

Limitations

• Not yet prospectively validated
• CTX/P1NP reference ranges vary by assay manufacturer
• Limited Latin American calibration data for bone turnover markers
• Expert system (rule-based), not machine learning

Conclusion

OSTEO-TX addresses the therapeutic selection gap left by FRAX, providing a structured, evidence-based pathway from risk assessment to drug choice. The system is open-source, auditable, and designed for rheumatological practice where bone metabolism is modified by disease and treatment in ways current tools ignore.

References

1. Camacho PM, et al. AACE/ACE 2020. Endocr Pract. 2020;26(Suppl 1):1-46.
2. Shoback D, et al. Endocrine Society 2020. JCEM. 2020;105(3):dgaa048.
3. Buckley L, et al. ACR 2017 GIOP. Arthritis Rheumatol. 2017;69(8):1521-1537.
4. IOF-IFCC. Bone Markers in Osteoporosis. Osteoporos Int. 2012;23(9):2405-2410.
5. Carlos-Rivera F, et al. Arch Osteoporos. 2024. PMID 39505771.