INTRODUCTION

Survival for patients who suffer estrogen receptor positive (ER+) breast cancer has improved dramatically over the years due to the addition of adjuvant hormonal therapy, especially aromatase inhibitors (AI). Letrozole, anastrozole and exemestane are third generation AIs that massively reduce circulating estrogens in postmenopausal women. Although this effect is decisive for survival and the reduction of tumor relapse, it also leads to adverse events and quality of life problems, more prominently associated with the musculoskeletal system¹.

Its use in women as adjunctive treatment for 2-5 years has been correlated with an increased risk of bone loss and fractures^2,3. Furthermore, AI administration is associated with the appearance and/or increase of arthralgia –described as joint pain– with an estimated incidence of 55% in a previous study by our group⁴. The high rate of arthralgias is of particular concern, since it is reportedly the most frequent reason for interrupting treatment^5,6. Although practical guidelines have been developed to prevent and manage IA-related bone loss⁷, effective treatment of arthralgia has yet to be addressed⁸.

Previous studies in the B-ABLE cohort, a clinical, prospective, cohort study of women diagnosed with early ER+ breast cancer, and candidates for aromatase inhibitor therapy, showed that low levels of 25-hydroxyvitamin D (25(OH)D) were associated with greater bone mass loss and worsening joint pain^9-11. Similarly, IA-related arthralgia in the B-ABLE cohort was significantly reduced in those patients who achieved serum 25(OH)D concentrations ≥40 ng/ml¹¹. Consequently, maintaining optimal 25(OH)D levels in the general population is strongly recommended to prevent not only bone loss but other nonskeletal disorders as well¹². Therefore, assessment of serum 25(OH)D levels in breast cancer patients treated with AI could be important in preventing musculoskeletal disorders, as well as other issues that affect quality of life.

To further explore the association of vitamin D status with bone loss and arthralgia, the expanded B-ABLE cohort, comprised of 927 postmenopausal women diagnosed with RE+ breast cancer and treated with IA, was evaluated. This was intended to establish target 25 (OH) D threshold levels to prevent the appearance of arthralgias associated with AI.

MATERIAL AND METHODS

RESULTS

Participants

A total of 741 of the 927 patients recruited in the B-ABLE cohort were visited at the 3-month follow-up, had data available and had baseline serum 25 (OH) D levels below 30 ng/ml (Figure 1). and, therefore, they were eligible for the present study. The baseline characteristics of the selected patients are indicated in table 1.

BMD: bone mineral density; FN: femoral neck; LS: lumbar spine; TH: total hip; VitD: serum levels of 25-hydroxyvitamin D..

Figure 1.  Flow chart showing the number of patient records included in each study point 

Table 1.  Baseline characteristics of the patients 

BMI: body mass index; BP: bisphosphonates; VAS: visual analog scale; TAM: tamoxifen; Q: quartile; CT: chemotherapy

AI-related arthralgia and vitamin D status at 3 months

At 3 months, the median [Q1;Q3] of the VAS increased from 2.00 [0.00;4.00] to 3.00 [0.00;5.00] (p<0.001), and the vitamin D increased from 15.10 [10.8;21.00] to 40.20 [30.90;52.50] (p<0.001). The change in VAS from baseline to 3 months was significantly associated with a vitamin D threshold ≥40 ng/ml (p<0.05) at 3 months of follow-up (Table 2), that is, an increase in VAS decreased 0.40 units in patients who reached a vitamin D threshold greater than 40 ng/ml with supplementation (Figure 2).

Figure 2.  VAS changes are stratified by the vitamin D threshold of 40 ng/ml at 3 months, in all patients with baseline vitamin D levels ≤30 ng/ml. VAS values are reported as median [95% CI] 

Table 2.  Linear regression between the change in the VAS from baseline to 3 months and the vitamin D threshold at 3 months (in all patients n=741) 

β: β-coefficient adjusted for age, BMI, recent chemotherapy, antiresorptive drugs and previous tamoxifen; CI: confidence interval. In bold: significant results (p<0.05).

Incident pain was assessed in a subset of 301 patients without initial pain. Of these, 117 (38.87%) developed joint pain at 3 months with a median VAS [Q1;Q3] of 3.50 [2.20;5.00]. The logistic regression between vitamin D thresholds and the appearance of joint pain at 3 months showed that those patients who achieved vitamin D le- vels ≥40 ng/ml were less likely to experience incident pain (p<0.05) (Table 3 and Figure 3).

Figure 3.  Changes in VAS in women with incident pain (n=117) and according to the vitamin D threshold of 40 ng/mL. VAS values are reported as median [95 CI) 

Table 3.  Logistic regression between incident pain and vitamin D threshold at 3 months (patients without initial pain n=301; of these, n=117 developed incident pain) 

CI: confidence interval; OR: odds ratio, adjusted for: age, BMI, recent chemotherapy, antiresorptive drugs, and previous tamoxifen. In bold: significant results (p<0.05).

BMD and vitamin D status

Data from 473 patients who were not exposed to any antiresorptive treatment and who had BMD data at 12 months of follow-up were analyzed. In these patients, the BMD of the LS, FN and TH decreased significantly after 12 months of treatment with AI (p<0.001) (Table 4).

Table 4.  Mean BMD values and its percentage change from baseline to 12 months of AI treatment 

No association was detected between any of the vitamin D thresholds analyzed (≥20 ng/ml, ≥30 ng/ml, ≥40 ng/ml, or ≥50 ng/ml) at 3 months and the relative changes in BMD of LS, FN and TH at 12 months. However, each 10 ng/ml increase in serum vitamin D at three months was associated with a lower loss of BMD in LS (unadjusted β = +0.194% [95% CI: 0.028 to 0.359] and adjusted β = +0.177% [95% CI: 0.014 to 0.340]; p<0.05). No significant associations were observed between vitamin D levels and BMD of FN and TH.

DISCUSSION

An observational, prospective, and real-life study of postmenopausal women treated with aromatase inhibitors included in the B-ABLE cohort was carried out. AI treatment in ER+ early breast cancer patients is strongly associated with musculoskeletal side effects. However, vitamin D supplementation early in AI appears to attenuate one of the main risk factors for treatment interruption: AI-related arthralgia. Our results suggest that AI-induced joint pain is vitamin D dependent, and that 40 ng/ml is the effective target threshold for serum 25 (OH) D levels to reduce the risk of both joint pain incidence and its worsening. However, this threshold is not significantly related to changes in BMD at one year of follow-up. However, vitamin D supplementation was inversely correlated with bone loss of CL, as each 10 ng/ml increase in serum 25 (OH) D at 3 months resulted in a reduction in blood pressure and 0.177% bone loss.

Vitamin D is known to play an important role in musculoskeletal tissues in addition to bone¹⁴, including muscle¹⁵, cartilage¹⁶, and synovium¹⁷. Previous studies carried out in women with ER+ early breast cancer receiving AI treatment, who also frequently present vitamin D deficiency¹⁸, provide evidence of the possible effects of vitamin D status on musculoskeletal health^9,19. In our cohort study, the main musculoskeletal effect of vitamin D supplementation was found in AI-related arthralgia, consistent with a previous study by Prieto-Alhambra et al. in 2011¹¹. Similarly, another observational study showed that a high dose of vitamin D (50,000 IU weekly of vitamin D3 orally) improved arthralgia values in patients who achieved mean concentrations of 25 (OH) D higher than the mean of 66 ng/ml²⁰. In our case, the threshold was defined as ≥40 ng/ml, which was reached after 3 months of vitamin D supplementation in approximately 50% of patients. Clinically, the fact of containing the increase in pain related to AI at 3 months helps to improve the patients' quality of life, as well as avoiding treatment discontinuity²¹.

Unlike pain, changes in BMD usually take longer to notice. Bone remodeling is a progressive process that results in long-term changes in BMD. Therapeutic interventions on BMD are evaluated annually in routine clinical practice. For this reason, in our study, BMD was assessed after 1 year of follow-up. Associations with vitamin D intake were only detected in the lumbar spine, which is not surprising given that bone remodeling is more active in this area and the pharmacological effects are more visible in trabecular bone compared to other skeletal locations with greater cortical content. We observed that increases in serum 25 (OH) D at 3 months were inversely correlated with AI-related bone loss at 1 year, therefore, this increase in 25 (OH) D could predict bone behavior at 1 year, but only visible in column. This coincides with a previous study by Prieto-Alhambra et al.²², although they found greater reductions in bone loss of 1.70%, in patients who achieved serum vitamin D levels ≥40 ng/ml.

This study has several limitations. First, this is not a randomized control trial, so the efficacy of high-dose vitamin D supplementation compared to a placebo group could not be assessed. Furthermore, compliance with vitamin D supplementation was not strictly controlled. This could explain the variability of 25 (OH) D levels between patients after 3 months of treatment. Finally, the current assumption that circulating 25(OH)D concentrations are a measure of vitamin D functional status may be incorrect. However, measuring 25(OH)D levels is the easiest and most reliable assessment of vitamin D status currently available.

Vitamin D supplementation administered to patients in specified doses increases levels from 15.10 [10.8;21.00] to 40.20 [30.90;52.50], thus reaching adequate levels of vitamin D at 3 months in most of the patients. The goal of therapy is to treat vitamin D insufficiency/deficiency rather than increase to supranormal concentrations, so we believe the risk of harm from administered doses is very low.

Our results suggest that optimal levels of vitamin D are associated with a reduced risk of joint pain related to AI treatment. A target threshold of 25OHD serum levels was set at 40 ng/ml to significantly reduce the increase in joint pain. It should be noted that this threshold is well above the goal of 20 ng/ml recommended by the 2010 Institute of Medicine (IOM) report²³. Therefore, vitamin D supplements at the specified doses could be protective against arthralgia and AI-induced spinal bone loss. As a final observation, vitamin D supplements are inexpensive, safe, and easily accessible, making these drugs easy to use on a wide scale.