Patients with non-Hodgkin lymphoma (NHL) are diagnosed at a median age of 67, with about 70% surviving over 5 years. The chemotherapy regimens for aggressive NHL includes cyclophosphamide and high doses of corticosteroids, which can result in significant bone loss. Previous studies have shown that fracture incidence approaches 30% by 10 years post-therapy and can impair quality of life of NHL survivors. However, there are no guidelines for bone health management in NHL patients, and additional screening for osteoporosis can be cumbersome and thus not routinely done. By reviewing standard of care pre- and post-treatment CT scans, we aim to evaluate the prevalence and incidence of fractures in patients treated for aggressive B-cell NHL.


In this retrospective single-center study, we reviewed patients seen at the University of California San Francisco lymphoma clinic in the last year who met the following criteria: 1) had aggressive B-cell NHL excluding primary central nervous system lymphoma, 2) received first-line therapy with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) or similar regimens such as with the addition of etoposide (R-EPOCH), 3) had baseline and post-treatment chest/abdomen/pelvis CT scans available for review. Baseline clinical characteristics and treatment information were collected from review of medical records. Axial CT images as well as sagittal and coronal reconstructions of chest/abdomen/pelvis CT images obtained at baseline and up to 2 years post-treatment were reviewed and graded by 2 radiologists to identify prevalent and incident fractures of the vertebrae, ribs, and pelvis. Exploratory analyses were performed using STATA, and comparisons between groups with or without fractures were performed with t-test or chi-squared test.


We identified 82 patients who met the inclusion criteria. Patients were diagnosed from 6/1/2007 to 11/15/2016. Median age of diagnosis was 62, with range 19-87. The most common NHL subtype was diffuse large B-cell lymphoma (93%). Patients were primarily treated with R-CHOP (63%) or R-EPOCH (34%) as first-line therapy for a median of 6 cycles, with 90% achieving complete response, 5% partial response, and 5% progressive disease. At the time of diagnosis, 13% of patients had a known diagnosis of osteopenia, osteoporosis, or prior fracture. Baseline clinical characteristics and treatment information are described in Table 1.

Of the 82 patients, 23 (28%) were found to have prevalent fractures at the time of diagnosis, with predominantly vertebral fractures. Of those with prevalent fractures, only 5 had a known diagnosis of osteoporosis or prior fracture. After treatment initiation, 2 patients developed progressive fractures and 8 (10%) patients developed new incident fractures, including 5 patients who had prevalent fractures. Compared to those who did not have prevalent fractures, patients with prevalent fractures were more likely to be older (70.9 vs 53.9, p<0.0001) and have a diagnosis of osteopenia, osteoporosis, or prior fracture at baseline (p=0.037). Patients with prevalent fractures also exhibited a trend towards more advanced stage (p=0.052) and higher international prognostic index score (p=0.060).


Using a novel approach of reviewing routinely performed pre- and post-treatment CT scans with sagittal and coronal reconstructions for the evaluation of bone health, we found that newly diagnosed patients with aggressive B-cell NHL have a 28% prevalence of fractures at diagnosis and 10% incidence of new fractures within 2 years of treatment. These rates greatly exceed the reported 15% prevalence of vertebral fractures and 0.5% annual incidence in the general population aged 65-70 (J Bone Miner Res. 2005 Jan;20(1):131-40). The high prevalence of fractures at diagnosis suggests a disease-related process that compromises bone health, and treatment-related bone loss further increases the risk of fractures. In ongoing studies, we seek to use CT-derived measures of bone mineral density and other clinical characteristics to derive a prediction score to identify patients at high risk of developing incident fractures who may benefit from preventative therapies.


Ai: Celgene: Research Funding. Andreadis: Amgen: Research Funding; Genentech Inc.: Employment, Equity Ownership, Honoraria; Cellerant Therapeutics: Research Funding; Astellas: Honoraria; Novartis Pharmaceuticals Corporation: Honoraria, Research Funding; Pharmacyclics: Research Funding; Incyte Pharmaceuticals: Research Funding; Gilead Sciences: Honoraria; Seattle Genetics: Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract

Sign in via your Institution