Multivariable logistic regression models were used to identify risk factors associated with the occurrence of bone patin of grade ≥2 or any grade in cycle 1 and across cycles 1-6. Data collected in the trials and included in the regression model include: baseline patient characteristics (gender, region, race, age, body surface area, and baseline absolute neutrophil count), disease characteristics (primary tumor type, tumor stage), treatment characteristics (chemotherapy, radiotherapy, taxane vs no taxane treatment, and dose of pegfilgrastim), and medical history (osteoporosis/osteopenia, hypercholesteremia, anemia, neutropenia, osteomyelitis/osteonecrosis, arthritis, peripheral neuropathy, chronic fatigue syndrome, gout, bone fracture, and bone pain).