Evaluation of Pentosidine as Bone Quality Marker in Postmenopausal Breast Cancer Patients Receiving Aromatase Inhibitors

Background: Osteoporosis and fractures are important aromatase inhibitor (AI) related adverse events in postmenopausal women with hormone receptor positive breast cancer. An increment of pentosidine is associated with a deterioration of bone quality. In this study, pentosidine was evaluated in postmenopausal breast cancer patients receiving AIs. Methods: Fifty postmenopausal breast cancer patients receiving AIs were retrospectively evaluated. Sixteen patients were given a bone modifying agent (BMA) concomitant with AIs. Changes of pentosidine, bone turnover markers and bone mineral density (BMD) before and after 12 months of AI therapy were compared between BMA administered patients (BMA group) and a non-BMA group. Results: There was no signicant difference between pentosidine low and high groups in regard to age, height, weight, BMD of femoral neck and lumbar spine, and bone turnover markers including TRACP-5b and BAP. In the non-BMA group, pentosidine was increased in 18 cases (53%), and the average change of pentosidine was 21.5% (95%CI; 0.23 to 42.7%, p=0.048). In the BMA group, pentosidine was increased in 2 two cases (13%), and the average change of pentosidine was -16.6% (95%CI; -30.6 to -2.6%, p=0.023). There was a signicantly lower proportion of pentosidine-increased cases (p=0.0065) and decrease of pentosidine (p=0.021) in the BMA group compared to those in the non-BMA group. Conclusions: Pentosidine was an independent factor from risk factors of osteoporotic fracture. Pentosidine was increased with AI, however, BMA inhibits an AI-induced increase of pentosidine in postmenopausal breast cancer patients.


Background
More than 70% of early breast cancers are hormone-receptor positive, and adjuvant endocrine therapy reduces breast cancer recurrence and improves overall survival in patients with hormone receptor positive breast cancer [1]. Administration of an aromatase inhibitor (AI) for 5-10 years has become standard therapy for postmenopausal woman with hormone receptor positive breast cancer [2]. AI inhibits the peripheral conversion of androgens to estrogens, which leads to estrogen de ciency resulting in bone loss and vulnerability. Compared with tamoxifen, a selective estrogen receptor modulator, the long-term administration of AI is associated with a higher incidence of osteoporosis and bone fracture [3]. In addition to bone density, bone quality is regarded as an independent factor of bone strength [4]. Bone quality is determined by bone collagen and collagen crosslinks which determines biomechanical structures and pliability [5]. Collagen cross-links are classi ed into physiological cross-links and advanced glycation end products (AGEs) cross-links, and hyperplasia of AGEs reduces suppleness that results in hard and brittle bone. Pentosidine is an AGE and hyperplasia of bone pentosidine is associated with a deterioration of bone strength without a decrement of BMD [6]. Several studies evaluated the association between osteoporotic fractures and urinary/serum level of pentosidine. In a multicenter cohort study of osteoporosis patients under bisphosphonate treatment, a higher baseline urinary pentosidine level is associated with development of vertebral fracture [7]. In a hospital-based cohort study of postmenopausal women, increment of urinary pentosidine were signi cantly associated with incidence of vertebral fracture in multivariate Cox regression analysis [8]. The serum pentosidine level is also associated with deteriorated bone strength and incidence of vertebral fracture [9,10]. In a cohort study of patients with prostate cancer undergoing androgen deprivation therapy, serum pentosidine was increased after endocrine therapy and denosumab inhibited the increment of serum pentosidine [11]. However, there are no reports of pentosidine in breast cancer patients receiving AI therapy and the effect of a bone modifying agent (BMA) on changes from pentosidine is uncertain.
In this study, pentosidine in addition to BMD and a bone turnover marker were evaluated retrospectively in postmenopausal breast cancer patients receiving AIs with or without BMAs.

Subjects
A retrospective evaluation was done for a total of 50 consecutive postmenopausal patients that were administered AIs as adjuvant therapy for hormone receptor positive early breast cancer at Kure Medical Center and the Chugoku Cancer Center, Kure, Japan, between October 2016 and December 2019. All patients received steroidal AIs, anastrozole or letrozole, as adjuvant endocrine therapy. BMAs including bisphosphonate or denosumab were administered for 16 patients who were diagnosed as having osteoporosis or high risk for fracture based on BMD and general condition. Osteoporosis was determined as a T-score less than − 2.5 at either the lumbar spine or total hip. Patients with a T-score greater than − 2.5 but less than − 1.0 at either the lumbar spine or total hip and with clinical risk factors for fracture were also determined high risk for fracture. Among these patients, 10 received denosumab 60 mg every six months, while six received bisphosphonate. Prescribed bisphosphonate was oral sodium risedronate hydrate in two, oral alendronate sodium hydrate in two, and intravenous zoledronic acid in two patients.
Changes of pentosidine, bone turnover markers and BMD before and after 12 months of AI therapy were compared between BMA administered patients (BMA group) and a non-BMA group. Patients with a history of fracture, chronic renal failure or diabetes requiring insulin therapy were excluded from this study.

Measurement of pentosidine, bone turnover marker and BMD
Pentosidine and bone turnover markers, including TRACP-5b and BAP, were evaluated using blood samples. Fasting blood samples were collected before and 12 months after administration of AI and were stored at 4 °C for measurement of BAP and − 80 °C for measurement of TRACP-5b and pentosidine. TRACP-5b was measured using enzyme immunoassays by Osteolinks enzyme immunoassays (SB bioscience). Intra-assay CV was less than 15%. Bone alkaline phosphatase (BAP) was measured by Access Ostase immunoassay (Beckman Coulter). Intra-assay CV was 7.4%. Pentosidine was measured using enzyme-linked immunosorbent assays by microplate reader Bench mark (SAKURA). Intra-assay CV was 5.5%. BMD was measured in the lumbar spine and femoral neck before and 12 months after administration of AI using dual-energy X-ray (DXA) absorptiometry with the Horizon DXA system (Hologic, Japan, Inc). The quality control was performed using a spine phantom. Intra-assay CV was 0.6%.

Statistical analysis
A chi-square test was used to compare the proportion of patients between the BMA and non-BMA group.
Patient characteristics and changes in bone markers and BMD between the BMA and non-BMA group were analyzed with Student's t-tests. A regression analysis was used to evaluate the relation between changes of pentosidine and bone turn over markers. Statistical analyses were performed using SPSS software (version 11 for Windows; 5 SAS Institute, Tokyo, Japan). A P value < 0.05 was considered as statistically signi cant.

Relationship between serum pentosidine and clinical factors
The median age of the subjects was 67 years, and 21, 23 and 6 subjects were classi ed as normal, bone loss and osteoporosis, respectively, according to the BMD results. Table 1 shows the relationship between pentosidine and other clinical factors. The subjects were classi ed into a pentosidine high or pentosidine low group according to the median value of pentosidine 0.555 µg/ml. There was no signi cant difference between the pentosidine high and pentosidine low group in regard to age, height, weight, BMD of the lumber spine and femoral neck and bone turnover markers including TRACP-5b and BAP. No patient experienced breast cancer recurrence, non-traumatic fractures or osteonecrosis of the jaw during this study period. Changes in serum pentosidine according to BMA Table 2 shows the change of pentosidine in the BMA-group and BMA + group. In the non-BMA group, pentosidine increased in 18 cases (53%), and the average change of pentosidine was 21.5% (95%CI; 0.23 to 42.7%, p = 0.048). In the BMA group, pentosidine increased in two cases (13%), and the average change of pentosidine was − 16.6% (95%CI; -30.6 to -2.6%, p = 0.023). A signi cantly lower proportion of pentosidine-increased cases (p = 0.0065) and decrease of pentosidine (p = 0.021) were recognized in the BMA group compared with the non-BMA group.  Changes in bone turnover markers and BMD according to BMA Table 3 shows changes in bone turnover markers and BMD in the femoral neck and lumbar spine. After 12 months treatment with AIs, the mean changes of TRACP-5b (-54.4% vs. 10.9%, p < 0.001) and BAP (-44.8% vs. 19.3%, p < 0.001) were signi cantly lower in the BMD + group compared with BMA-group, respectively. In regard to BMD, the mean changes of femoral neck (3.9% vs. -1.9%, p < 0.001) and lumbar spine (7.4% vs. -0.93%, p < 0.001) were signi cantly higher after 12 months of treatment with AIs in the BMD + group compared with the BMA-group, respectively. The association between changes of pentosidine and bone turnover markers, including TRACP-5b and BAP was evaluated using regression analysis. Neither change of TRACP-5b (R 2 = 0.027, p = 0.25) nor BAP (R 2 = 0.026, p = 0.26) was signi cantly associated with change of pentosidine.

Discussion
In this cohort study of postmenopausal breast cancer patients receiving AIs as adjuvant therapy, the bone quality marker pentosidine was increased by AI and reduced with BMAs. Bone turnover markers and BMD also showed signi cant improvement with BMAs. These results indicate that administration of BMA exerts not only an inhibition of bone loss, but also an improvement of bone quality in patients receiving AIs.
Fracture and osteoporosis are important adverse events in postmenopausal breast cancer patients receiving adjuvant AIs, and the prophylactic administration of BMA is recommended for patients with a high risk of bone-related adverse events [12,13]. Because long-term prescription of BMA can cause serious adverse events, such as osteonecrosis of the jaw and atypical femoral fracture in breast cancer, an appropriate evaluation of the risk of AI related fracture is desirable [14,15]. Although low BMD in femoral neck or lumbar spine is regarded as an independent risk factor for fracture, a certain proportion of patients with a non-osteoporotic range of BMD experience nontraumatic fractures [16,17]. In a cohort study of postmenopausal women, more than half of osteoporotic fractures occur in cases with a nonosteoporotic range of BMD [18]. The Fracture Risk Assessment Tool (FRAX) is used to estimate the risk of osteoporotic fracture using femoral neck BMD and clinical risk factors, which include age, prior nontraumatic fracture, glucocorticoid use, low body mass index, family history of osteoporosis, smoking, and excess alcohol intake, which are also associated with incidence of fracture. However, FRAX is not designed to estimate the risk of AI-related fracture, and the estimated fracture risk by FRAX tends to be lower than that of the actual fracture incidence in breast cancer patients receiving AI [19]. Thus, the evaluation of other risk factors is warranted to determine appropriate administration of BMA to prevent AI-related fracture in postmenopausal breast cancer patients.
Independent of BMD, bone quality is regarded as an important factor of bone strength when considering the risk of osteoporotic fracture. The trabecular structure, including trabecular number, separation and homogeneity of the trabecular network, is also deteriorated with exemestane. A follow-up study of 29,407 postmenopausal women aged 50 years and older showed that patients with bone loss by BMD and a low Trabecular Bone Score (TBS) were at the same risk of fracture as those with osteoporosis by BMD and high TBS [20]. In an accompanying study of a randomized control trial evaluating exemestane, an AI, for prevention of breast cancer in healthy postmenopausal women, high-resolution peripheral quantitative CT showed a signi cant decrease in volumetric BMD and cortical thickness at the distal radius and distal tibia with exemestane [21]. These ndings suggest that deterioration of bone quality is also an important AI-related bone adverse effect.
Serum and urine pentosidine have been investigated as bone quality markers. Pentosidine is positively correlated with deteriorated collagen crosslinks AGEs [6], and elevated pentosidine is shown as an independent risk factor of fracture in postmenopausal women [22]. In this study, we evaluated the clinical signi cance of pentosidine in postmenopausal breast cancer patients receiving adjuvant AIs. First, pentosidine was shown to be independent from other risk factors of osteoporotic fracture. When patients were divided into a pentosidine low and high group according to median value of pentosidine, there was no signi cant difference between the two groups in regard to age, height, weight, BMD of lumber spine and femoral neck and bone turnover markers including TRACP-5b and BAP. This nding indicates that measurements of pentosidine may lead to a more accurate assessment of fracture risk based on conventional BMD and physical factors. Second, approximately half of subjects who were administered adjuvant AIs had an elevation of pentosidine. AI treatment is associated with not only a decrease of BMD, but also a deterioration of bone quality. Finally, in this study, the administration of BMA signi cantly suppressed an AI-induced pentosidine increase and BMD decrease. Our ndings are consistent with a previous study evaluating the effects of denosumab on prostate cancer patients treated with hormonal therapy. BMA therapy can maintain BMD and prevent a deterioration of bone quality in patients treated with hormonal therapy.
This study has several limitations. First, this is a cohort study with a relatively small number of cases and a short duration of follow up. A prospective study with a larger number of cases with a long-term period is warranted to con rm our ndings. Second, BMAs were essentially prescribed for patients with a high risk of osteoporotic fracture determined by low BMD and/or clinical factors in this study. The effect of BMA on preventing AI-induced bone quality deterioration is uncertain in patients with a normal range of BMD. Finally, no bone quality assessment other than pentosidine was done in this study. Imaging studies evaluating bone structure should be done in future study to con rm our ndings.
In summary, this study demonstrated that serum pentosidine signi cantly increased with AI therapy in postmenopausal women with hormone receptor positive breast cancer, and administration of BMAs inhibit an AI-induced increase of pentosidine. A prospective long-term study is warranted to con rm the relationship between pentosidine and non-traumatic fracture during adjuvant AI therapy.