Predicting Post-Zoledronic Acid Musculoskeletal Pain: A New Risk Nomogram for Osteoporosis Patients
Introduction
Musculoskeletal pain is the most common side effect of zoledronic acid injection, affecting up to 36.46% of patients and can be severe. The pain can significantly reduce medication compliance due to fear and reluctance to continue bisphosphonate treatment. Identifying risk factors and predicting pain occurrence are crucial for prevention and improving patient adherence.
Aim
To construct a new prediction nomogram to predict the risk of musculoskeletal pain in patients with primary osteoporosis who receive zoledronic acid intravenously for the first time.
Patient Profile
Patients with primary osteoporosis who receive the first intravenous (IV) injection of zoledronic acid.
Methods
- N=368 patients
- Patients were divided into a musculoskeletal pain group (n = 258) and a non-musculoskeletal pain group (n = 110) based on the presence or absence of musculoskeletal pain three days after IV zoledronic acid injection.
- Statistically significant predictors were screened by logistic regression analysis and the minimum absolute contraction and selection operator (LASSO) to construct a nomogram.
Results
- The independent predictors of Musculoskeletal Pain were:
- Age, serum vitamin D, prior vitamin D intake, Body mass index (BMI), Nonsteroidal anti-inflammatory drugs (NSAID) use.
- Identified using Multivariate logistic regression analysis (Table 1)
Table 1: Multivariate logistic regression table
|
Characteristics |
B |
SE |
OR |
CI |
Z |
P |
|
Age |
-3.403 |
1.024 |
0.03 |
0-0.25 |
-3.324 |
0.001 |
|
BMI |
2.134 |
0.846 |
8.45 |
1.61–44.36 |
2.523 |
0.012 |
|
NSAIDs |
-1.979 |
0.933 |
0.14 |
0.02–0.86 |
-2.121 |
0.034 |
|
25(OH)D |
-4.614 |
0.966 |
0.01 |
0-0.07 |
-4.777 |
0.000 |
|
prior Vitamin D intake |
2.389 |
0.847 |
10.9 |
2.07–57.35 |
2.821 |
0.005 |
BMI: Body mass index, NSAIDs: Nonsteroidal anti-inflammatory drugs
Nomogram Development and Performance
- The nomogram for predicting musculoskeletal pain risk performed well in patients randomized to both training and validation sets
- ROC curve showed AUC: 0.980 in the training set, and 0.979 in the validation set, showing excellent predictive performance (Figure 1).
Figure 1: AUC of muscle soreness model
- Calibration curve showed good agreement between predicted and actual outcomes
- Decision curve analysis indicated net benefit of the pain prediction nomogram was 0.20–0.49
Conclusion
- Five variables were identified as independent risk factors for pain: age, BMI, vitamin D level, NSAID use, and prior vitamin D intake.
- A nomogram constructed from the above predictors can be used to predict risk of musculoskeletal pain after the first zoledronic acid injection.
Reference
BMC Muscul Dis. 2023; 24:841
