Biodegradable Polymer Safety Analysis

What this research found

Biodegradable screws and plates are meant to dissolve as bone heals, but which polymer to pick for load-bearing hardware has rarely been settled with real-world failure data. Six years of FDA adverse-event reports on bioresorbable orthopedic devices were linked to the polymer each device is made from, and mechanical failure turned out to affect 32.6% of poly(lactic-co-glycolic acid) devices against 0.7% of poly-L-lactic acid devices — a 47-fold gap the analysis reports as significant at p = 1.37 × 10⁻¹². The finding was turned into a material-selection decision framework.

• Devices made from poly(lactic-co-glycolic acid), or PLGA, failed mechanically in 30 of 92 reports (32.6%), against 1 of 153 (0.7%) for poly-L-lactic acid, or PLLA — a 47-fold difference with a chi-square statistic of 50.23 and p = 1.37 × 10⁻¹².
• Mechanical failure dominated the linked reports at 31 of 245 (12.65%), followed by incomplete resorption at 7 (2.86%). No linked report was classified as premature degradation or inflammatory reaction.
• Matching reports to polymers by commercial trade name rather than chemical name raised the linkage rate from 0.5% to 61.25%, converting 245 of 400 reports into usable structure-performance records.
• The two polymers differ sharply on molecular descriptors: PLLA at 90.08 g/mol molecular weight and 57.5 Ų topological polar surface area, against 260.20 g/mol and 105.0 Ų for PLGA — 189% and 83% higher respectively.
• Molecular weight, hydrophobicity and polar surface area all correlate with mechanical failure at exactly the same value (r = 0.4655, p = 1.41 × 10⁻¹⁴). The analysis attributes this to having only two polymers with internally uniform properties, so the correlation restates polymer identity rather than showing a dose-response relationship.
• The resulting framework recommends PLLA for load-bearing applications, treats roughly 175 g/mol as a molecular weight threshold, and flags polymers above 200 g/mol as warranting additional mechanical testing.

How it was done

FDA's Manufacturer and User Facility Device Experience database was queried for bioresorbable orthopedic devices — screws, pins, plates, anchors and scaffolds — reported between January 2019 and December 2024, returning 400 device reports. Report narratives were classified into four failure modes (mechanical failure, premature degradation, inflammatory reaction, incomplete resorption) using keyword pattern matching with manual review of ambiguous cases. Because reports name commercial products rather than chemistries, a curated map from trade names to constituent polymers was assembled from manufacturer specifications, FDA 510(k) clearance documents and published literature. Molecular weight, XLogP and topological polar surface area were taken from PubChem and standardized clinical-grade values, then chi-square tests and point-biserial correlations tested polymer type and properties against each failure mode. The output was an 18-page manuscript with a material-selection decision flowchart.

Data sources

• FDA MAUDE — 400 adverse event reports for biodegradable orthopedic devices, January 2019 to December 2024
• PubChem — molecular properties for PLLA, PLGA, polyglycolic acid and polycaprolactone
• Manufacturer product specifications and FDA 510(k) clearance documents — trade name to polymer mapping

Figures from this analysis

Outputs produced

• 01_data_acquisition_v2.py

  Improved data acquisition script with multiple fallback strategies and validation

• fda_maude_events.json

  FDA MAUDE adverse event reports for biodegradable stents (400 records, 3.2 MB)

• polymer_properties.csv

  Chemical properties for 4 biodegradable polymers from PubChem (PLA, PGA, PLGA, PCL)

• 02_failure_mode_classification.py

  Text classification script for categorizing adverse events by failure mode and polymer-relevance

• classified_events.csv

  Classified FDA adverse event reports with binary flags for 5 categories (400 records, 379 KB)

• failure_mode_counts.txt

  Summary statistics for failure mode classification

• 03_structure_property_integration.py

  Polymer identification and data merging script using synonym matching with word boundary detection

• structure_performance_matrix.csv

  Unified Structure-Performance Matrix with adverse events and chemical properties (400 records, 387 KB, 18 columns)