We conducted a 16-week prospective, randomized, open-label study comprising 64 postmenopausal women with a BMD T-score <− 1. Participants were randomized 1:1 to two groups of 32 women each. One group received branded ALN and the other ALN-EFF 70 mg weekly. In addition, all women received daily supplementation of 800 mg calcium and 38 µg vitamin D.

We conducted the trial according to the Helsinki Declaration, the Danish Health Law, and GCP guidelines. The Danish Data Protection Agency (1-16-02-306-21), The Regional Ethics Committee (Eudract No. 2020-005040-35) and The Danish Health and Medicine Authority (2020-120512) approved the study. The local Good Clinical Practice (GCP) at Aalborg and Aarhus University Hospital monitored the trial. The trial is registered at clinicaltrials.gov (identifier NCT05325515).

We included postmenopausal women (2 years since last menstrual bleeding) with a BMD T-score <−1. Moreover, as the primary endpoint was decrease in CTx (see below), we included women with CTx > 420 ng/L which is the median value for postmenopausal women at our laboratory. We excluded participants who had ever received treatment for osteoporosis, had contraindications for alendronate (according to the SPC), fulfilled the criteria for teriparatide treatment in Denmark, received treatment with systemic glucocorticoids within the past 12 months, received hormone replacement therapy, had rheumatoid arthritis, inflammatory bowel disease, untreated thyroid disease, primary hyperparathyroidism, diabetes mellitus, unstable liver disease, cancer within the past 2 years (except basal cell carcinoma of the skin), major GI disease within the past 12 months, estimated glomerular filtration rate (eGFR) < 60 mL/min, or p-25-hydroxy-vitamin D (D3 + D2) < 50 nmol/L.

We randomized the participants to one of the two treatment arms by 1:1 randomization. We did the randomization in blocks of 8 and randomization was performed using REDCap [9]. The randomization algorithm was generated by the local REDCap support. Study data were collected and managed using REDCap electronic data capture tools hosted at Aarhus University, Denmark.

We recruited participants via the dual-energy x-ray absorptiometry (DXA) service at the Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Denmark. All participants were referred from their general practitioner or other hospital department for evaluation by DXA. We identified women with osteopenia or osteoporosis (BMD T-score <− 1 at lumbar spine or total hip) and reviewed their medical records for any exclusion criteria. Women who were still eligible at this stage were sent a letter with written information about the project and contacted the investigator by telephone if they were interested in further information and participation in a screening visit.Written informed consent was obtained from all participants included in this study before the first procedures were performed.

At the screening visit, we acquired information on medical history, prescription medication, calcium intake, age at menopause, smoking history and performed a medical examination. Furthermore, a screening blood test was analyzed to rule out presence of the exclusion criteria mentioned above. Bone turnover marker CTx was measured and women with CTx > 420 ng/L were included if they did not meet any exclusion criteria.

We collected fasting blood samples at baseline and at weeks 4, 8, and 16. Information on adverse events (AEs) was obtained at each visit. Participants were provided with tablets for the entire study period at baseline and were instructed regarding the administration of the medication. The ALN group should swallow the pill with a glass of water and the ALN-EFF group had to dissolve the effervescent tablet in water and drink it immediately. Both groups were instructed to take the medication fasting in the morning and allow 30 min before consuming food/drinks and that they should not lay down but remain upright. At week 8 and week 16 participants were instructed to bring the remaining tablets and compliance was assessed by tablet count (Fig. 1). Furthermore, they were provided with Unikalk Forte and were instructed to take one tablet twice a day.

Blood samples were collected between 7.30 and 10.00 am following an overnight fast. Samples were then left to clot for 30 min and afterwards centrifuged. The serum and plasma were stored at − 80 °C until analysis at the end of trial. BTM were analyzed using EDTA plasma. All BTMs were analyzed in one single batch to reduce variation in the analysis. We measured BTMs using an electrochemiluminescent immunoassay on a Roche COBAS 8000 reader (Roche, Rotkreuz, Switzerland). At our facility the least significant change (LSC) for PINP is 22.9% and for CTx 33.0% and the coefficiency of variation is 10% for CTx and 7.4% for PINP.

Our primary endpoint was percent change in CTx from baseline to end of the study (after 16 weeks). The secondary endpoints were percent change in PINP from baseline to end of study (EOS), rate of decline of CTx and PINP, compliance and adherence to the treatments, AEs, and serious adverse events (SAEs).

The sample size was determined based on the primary endpoint. With a minimum expected decrease in CTx of 60% with ALN, a non-inferiority margin of 12% (80% of efficacy retained), an SD of 15% on change in CTx, and a significance level of 5%, 54 participants should be included in the study in order to obtain statistical power of 90%. We adjusted this to 64 participants (32 in each group) to account for dropouts. Baseline variables were checked for normality using qq-plots and log-transformed as appropriate. Data are displayed as means ± SD for normally distributed data and median with CIs for non-normally distributed data. We calculated percent changes in BTMs from baseline to EOS, checked normality of data using qq-plots and compared differences between groups using independent samples t test. For the calculation of change in BTMs we used the value from the baseline visit to minimize variation. We furthermore assessed differences in changes in BTMs over time using a general linear model with repeated measures. Differences in AEs and SAEs in the two treatment groups were assessed using non-parametric statistics.