The data for this retrospective pharmacovigilance study were obtained from FAERS, a global spontaneous reporting system that collects safety information on approved drugs and therapeutic biologic products from various sources including manufacturers, healthcare professionals, and consumers. FAERS is the primary source of post-marketing safety monitoring and evaluation for the FDA, and it provides signal detection and quantification of the association between drugs and AEs (Tang et al., 2022). The database contains seven categories of data including demographic and management information, drug information, adverse events, patient outcomes, report sources, treatment start and end dates, and indication.

To extract adverse event (AE) reports from the FDA Open-FDA program, we utilized the online tool OpenVigil 2.1 (http://openvigil.sourceforge.net/). Individual safety reports (ISRs) for Dmab and ZA were extracted from the FAERS database. ISRs are the count of raw data extracted by OpenVigil 2.1 and an ISR code represents an AE report.

The study retrieved data from FAERS covering the period between January 2004 and November 2022. The search for Dmab included its generic name “DENOSUMAB” and commodity names “Xgeva,” “Ranmark,” and “Prolia,” while for ZA, the search included its generic drug name “ZOLEDRONIC ACID,” “ZOLEDRONATE,” and trade names “ACLASTA,” “RECLAST”, and “ZOMETA.” Drugs irrelevant to the study and those with uncertain names were excluded. Only drugs listed as the “primary suspect” were included in the analysis as they were most likely associated with the AEs (Verden et al., 2018; Omar et al., 2021).

Disproportionality analysis was conducted to identify potential safety signals for the drugs, with RORs as measures of association (van Puijenbroek et al., 2002; Hauben, 2003; Tang et al., 2022). The analysis of the association between drug exposure and adverse events (referred to as “signals”) in OpenVigil relies on the use of a 2 × 2 contingency table (Böhm, 2018; Noguchi et al., 2021) (Refer to Table 1), which can be effortlessly generated within the platform. The higher the ROR values, the stronger the correlation between the drug and target AE. Significant signals were identified based on criteria including AE reports >3, ROR and PRR >2.0, ROR lower bound of 95% confidence interval (CI) value exceeds 1.0, and χ ² > 4 (Böhm, 2018; Shao et al., 2021; Tang et al., 2022). The equations and criteria for the three algorithms are shown in Table 2. Data processing was carried out using Microsoft Excel 2016 and GraphPad Prism 9.

The FAERS database contained 385,327 reports of primary suspect drugs from its inception until October 2022, with 297,896 AEs associated with Dmab and 87,431 AEs related to ZA. After removing duplicates, a total of 154,735 reports were included, consisting of 117,857 AEs for Dmab and 36,878 AEs for ZA. Process flowchart is shown in Figure 1.

The characteristics and clinical information are summarized in Table 3. The majority of the reports for both drugs were from females (76.19% for Dmab and 67.09% for ZA), and the median age of the reports was 68 and 73 years for Dmab and ZA, respectively, with a focus on the elderly population.

In order to make the changes more intuitive, we visualized the AEs metric data of each year with a line chart, as shown in Figure 2A. The chart shows an increasing trend in AEs for both drugs year by year, but a decline in 2012 and 2018 for ZA and Dmab, respectively. In addition, we also visualized the serious AE outcome metric data for the two drugs, as shown in Figure 2B. Serious AEs were mainly attributed to death (15.42% for Dmab and 22.29% for ZA) and hospitalization (9.22% for Dmab and 19.39% for ZA). Furthermore, ZA had slightly higher proportions of life-threatening (2.05% vs. 0.44%) and disability (4.69% vs. 0.92%) according to the reports from the database compared to Dmab.

We then conducted a disproportionality analysis using ROR to detect AE signals, which led to the identification of 333 significant AE signals related to Dmab and 1379 associated with ZA. Interestingly, 243 new AE signals and 528 off-label use for Dmab that were not registered in the FDA-approved specification were found, along with 1139 new signals and 206 off-label uses for ZA.

The most common AEs associated with Dmab were death, osteonecrosis of the jaw, back pain, tooth disorder, bone pain, and hypocalcemia. For ZA, the most frequent adverse events were osteonecrosis, death, pain, osteonecrosis of the jaw, and tooth extraction. Among these, death and fall were not mentioned in the drug labels for either Dmab or ZA. Top l0 significant AE signals sorted by frequency for both drugs are presented in Table 4.

As there were numerous shared AEs between Dmab and ZA, we conducted a further comparison of the overlapping AE signals. Out of the 189 identical positive AE signals between the two drugs, 29 AEs of Dmab exhibited stronger correlation than ZA, while 160 AEs of Dmab had weaker correlation than ZA, as determined by the ROR value. Table 5 presents the AE signals with significant differences in intensity between the two drugs. The AE signals of Dmab with stronger correlation than ZA (d > 20) included exostosis of jaw (ROR: 182.66 vs. 5.769), atypical fracture (ROR: 55.589 vs. 9.123), and atypical femur fracture (ROR: 49.824 vs. 4.968), while the AE signals of ZA with stronger correlation than Dmab (d > 50) were related to abscess jaw (ROR: 84.119 vs. 11.12), gingival ulceration (ROR: 74.125 vs. 4.827), increased bone formation (ROR: 69.344 vs. 3.218), and bone disorder (ROR: 68.221 vs. 3.189), among others.

While analyzing the data, we found off-label use was also a significant signal. Therefore, we further analyzed the data on off-label use. As we observed mixed reports of different specifications for each drug, for example, Dmab had a 60 mg specification for giant cell tumor and hypercalcemia of malignancy, while the 120 mg specification was used for postmenopausal osteoporosis. Similarly, ZA had two different specifications with mixed reports. Therefore, we combined the FDA-approved indications for both specifications of each drug and compared them with the indications in the database to identify off-label uses. We found 528 types of off-label use for Dmab and 206 types for ZA. Table 6 shows the top 10 off-label uses not mentioned in the drug instructions for both drugs, which are frequently used for treating various tumors. Breast cancer (1.03% and 3.21%) and prostate cancer (1.04% and 2.48%) were the most commonly off-label use for both drugs in the database. Other off-label uses for Dmab included arthritis (0.42%), vitamin D deficiency (0.26%), spinal compression fracture (0.25%), gastroesophageal reflux disease (0.24%), plasma cell myeloma (0.19%), rheumatoid arthritis (0.19%), and chronic kidney disease (0.10%). Off-label uses for ZA included neoplasm malignant (0.52%), renal cancer (0.45%), lung cancer (0.40%), plasma cell myeloma (0.30%), plasma cytoma (0.19%), colon cancer (0.08%), and osteoarthritis (0.07%).

We conducted a comparison of the AEs associated with Dmab and ZA in off-label use for breast cancer and prostate cancer. After comparing with the FDA-approved instructions and removing similar AEs, we found 451 AEs in Dmab and 848 AEs in ZA for breast cancer treatment. For prostate cancer treatment, we found 341 AEs in Dmab and 583 AEs in ZA that were not mentioned in the drug instructions.

In breast cancer treatment, the top AEs associated with Dmab were death (11.6%), disease progression (3.3%), breast cancer metastatic (2.7%), neutropenia (2.7%), emotional disorder (2.3%), and pyrexia (1.7%). For ZA, the most frequent AEs were death (19.8%), emotional disorder (12.9%), osteomyelitis (11.4%), neoplasm progression (10.7%), cardiac disorders (10.6%), and impaired healing (9.5%). Cardiac disorders in patients with breast cancer treated with ZA included tachycardia (3.04%), congestive heart failure (1.35%), arrhythmia (1.01%), and palpitations (1.01%). Both drugs were associated with varying degrees of mental illness such as emotional distress, depression, and personality disorder, particularly in treating breast cancer, even leading to suicidal ideation. Tables 7, 8 display the six most frequent AEs in breast cancer and prostate cancer treatments, respectively, which were not registered in the drug specifications.

In prostate cancer treatment, the top AEs associated with Dmab were death (8.9%), prostate cancer metastatic (1.6%), and renal impairment (1.7%), while for ZA, the most common AEs were death (34.4%), general physical health deterioration (19.9%), and hemoglobin decreased (18.9%). Additionally, ZA was also associated with increased prostatic specific antigen (18.4%) and cardiac disorders (13.6%), while Dmab was associated with emotional disorder (1.6%) and cardiac disorders (1.3%).

In this study, we performed a pharmacovigilance analysis using FAERS to investigate suspected AEs and off-label uses associated with Dmab and ZA. The data covers a substantial timeframe from 2004 to 2022, during which these two medications were administered in clinical practice at different time periods. Notably, the reporting rate for AEs can differ not only among various drugs but also for the same drug as time progresses (Moore et al., 2007; Alatawi and Hansen, 2017). Additionally, media attention, regulatory measures, Risk Evaluation and Mitigation Strategy, new indications, formulation changes, or shifts in marketing approaches can impact the adverse events profiles (Chhabra et al., 2013). Furthermore, both drug reporting trends exhibit a Weber-like effect (Hoffman et al., 2014; Noguchi et al., 2021), where AEs increase prior to marketing approval and subsequently decrease. Consequently, these variations in usage timelines may have led to different adverse event profiles, potentially impacting the results of our data analysis.

Nonetheless, the current understanding of these drugs is not yet fully comprehensive, and many AEs still require adequate attention. To better understand the AE profile of these drugs, it is recommended to collect as much clinical data as possible and conduct more in-depth analysis and evaluation.

The most frequent AEs of Dmab were death, osteonecrosis of jaw, back pain, tooth disorder, bone pain and hypocalcaemia and those for ZA were osteonecrosis, death, pain, osteonecrosis of jaw, and tooth extraction. The AEs identified in this analysis were generally in line with the known AEs of these drugs, indicating the validity of the study and suggesting that the findings may accurately reflect real-world clinical practices.

It is known that Dmab and ZA share many similar AEs. In our study, we conducted a comparative analysis to assess the signal strength of AEs between these two drugs. Among the signals of Dmab stronger than ZA (d > 20), the significant signals were exostosis of jaw, atypical fracture, and atypical femur fracture, suggesting that Dmab may be more prone to these AEs than ZA. Exostosis of jaw may be associated with the widely recognized osteonecrosis of the jaw (ONJ), which is a rare but serious side effect of anti-bone resorption inhibitors. Although a study demonstrated that patients with bone metastases treated with Dmab or ZA had similar incidences of ONJ (Nicolatou-Galitis et al., 2019), a meta-analysis of patients with solid tumors found that the use of Dmab was linked to a significantly higher risk of ONJ compared to ZA (Boquete-Castro et al., 2016). It is important to note that the incidence of ONJ may also be related to the dosage and duration of drug exposure (Khan et al., 2015). Thus, long-term and high-dose use of Dmab or ZA requires vigilance against ONJ. In contrast, the signals of ZA stronger than Dmab (d > 50) were mostly related to oral problems, which may also have potential implications for ONJ. Regular dental examinations should be conducted when using Dmab and ZA.

Dmab and ZA, have been approved for preventing bone metastases associated with solid tumors. However, our research has found that these drugs are also frequently used in bone metastasis-free cancer. It should be emphasized that in some reports, cases of non-bone metastatic cancers may have been reported ambiguously without clear indication of the presence or absence of bone metastasis, thereby posing a limitation to the study. The theory of cancer treatment may primarily base on preventing cancer treatment-induced bone loss. Furthermore, some studies have shown that both drugs have potential anti-cancer properties (Dedes et al., 2012; Ubellacker et al., 2017; de Groot et al., 2018), but whether they have a positive effect on fighting cancer remains a matter of debate.

Postmenopausal women with breast cancer have a higher risk of osteoporosis due to the decrease in estrogen, as well as the effects of chemotherapy, radiotherapy, endocrine therapy, and the tumor itself (Guise, 2006; Chen et al., 2009; Gralow et al., 2013; Shapiro, 2020). Endocrine therapies such as tamoxifen and aromatase inhibitors have been shown to increase bone loss or fracture risk in both pre- and postmenopausal women with early-stage breast cancer (Powles et al., 1996; Sverrisdóttir et al., 2004; Aihara et al., 2010; Zaman et al., 2012; Tseng et al., 2018). Dmab 60 mg is approved for aromatase inhibitor-induced bone loss in women with breast cancer regardless of whether there is bone metastasis, while ZA did not receive such approval. Interestingly, ZA is reported to be used for preventing bone loss or decreasing fracture in premenopausal women with breast cancer (Gnant et al., 2015; Wilson et al., 2018). There is no clinical evidence that Dmab is suitable for use this population. Evidence suggests that Dmab or ZA could be applied as adjuvant therapy to improve bone density in postmenopausal women with early-stage breast cancer (Brufsky et al., 2009; Waqas et al., 2021). Note that one phase 3 trial shows that Dmab did not improve disease-related outcomes and did not support a role as an antitumor agent in early-stage breast cancer for women with high-risk early breast cancer, in addition to the benefits of delaying cancer bone-related events (Coleman et al., 2020).

Antihormonal treatments for prostate cancer can also cause bone loss. The FDA has approved Dmab (60 mg) for the treatment of bone loss or preventing fracture in non-metastatic prostate cancer, while ZA currently lacks FDA approval. Several small randomized trials have shown that bisphosphonates can increase BMD in patients with non-metastatic prostate cancer (Smith et al., 2001; Smith et al., 2003; Klotz et al., 2013). Note that no benefit has been shown among bisphosphonates in preventing fractures among patients with nonmetastatic prostate cancer (Strum et al., 2018).

As for the treatment of osteoarthritis (OA), a study in rabbits with experimental knee osteoarthritis showed that ZA had protective effect on articular cartilage and subchondral bone (She et al., 2017). An initial trial showed that ZA may be effective in treating osteoarthritis (Aitken et al., 2018). However, we have not yet found strong evidence that osteoarthritis can benefit from ZA. Markers of bone turnover are increased in patients with progressive OA, similar to those in patients with postmenopausal osteoporosis (Bingham et al., 2006). Based on that mechanism, ZA may have a prospective benefit for osteoarthritis. Regarding Dmab, it has rarely been reported in osteoarthritis, but evidence suggests that Dmab may be a potential new therapeutic option for treating rheumatoid arthritis (Hu et al., 2021; Tanaka et al., 2021).

In conclusion, mining new indications from the database has the potential to expand drug application range, promote drug research and development, and improve clinical practice. However, it is crucial to conduct further real-world research to validate these new indications and ultimately benefit patients.

According to the reports, disease progression was observed more frequently in the treatment of breast cancer and prostate cancer with Dmab or ZA. However, current evidence does not establish a definitive link between tumor progression and drug exposure. Our study also found a high frequency of neutropenia among breast cancer patients treated with Dmab, which is consistent with reports of neutropenia in a phase III study of multiple myeloma patients treated with both Dmab and ZA (Raje et al., 2018). Mental problems were also reported in breast cancer patients treated with either drug, although drug-induced mental disorders on Dmab or ZA are currently poorly documented. A case report indicated that extreme anxiety and hypocalcemia after denosumab treatment for cancer-related bone metastasis may have contributed to depressive mood (Lin et al., 2015). Although atrial fibrillation is a known AE to Dmab, our study also found a high frequency of heart problems in ZA-treated patients. Previous studies have reported an increased rate of heart failure in zoledronate-treated patients (Black et al., 2007; Rubin et al., 2020), suggesting that more clinical trials are needed to confirm the safety of ZA. Renal toxicity is a potential AE of ZA treatment, although Dmab is considered relatively safe for the kidneys. However, renal toxicity has been observed in the treatment of multiple myeloma using Dmab (Raje et al., 2018). From the pharmacokinetics perspective, Dmab is not metabolized by the kidneys and theoretically has minimal damage to the kidneys, it is still relatively safe.