Delayed and selective bone health assessment following cancer diagnosis: a real-world gap in survivorship care

Abstract

Background

Cancer treatment–induced bone loss (CTIBL) is well-recognized in hormone-sensitive malignancies such as breast and prostate cancer but is less consistently implemented in patients with other solid tumors and hematologic malignancies. Furthermore, the long-term skeletal effects of newer systemic therapies, including immune checkpoint inhibitors and targeted agents, remain incompletely characterized. Although bone mineral density (BMD) assessment by dual-energy X-ray absorptiometry (DXA) is recommended for patients at increased skeletal risk, real-world utilization patterns remain poorly characterized.

Objectives

To evaluate DXA utilization, timing, BMD results, and associated clinical characteristics following a new cancer diagnosis in a large real-world cohort.

Methods

We conducted a retrospective cohort study using the MDClone platform, including adults with a first cancer diagnosis between January 1, 2015, and December 31, 2024, treated at Assuta Medical Centers in Israel. DXA completion, timing, BMD and T-scores, and prevalence of osteopenia and osteoporosis were analyzed.

Results

Among 21,112 patients with newly diagnosed cancer (77.0% women; mean age 60.2 ± 13.1 years), 7496 (35.5%) underwent DXA at any time following diagnosis. DXA utilization was significantly lower in men than in women (15.0% vs. 41.7%). Early DXA assessment within 6 months of diagnosis was uncommon (1550 patients; 7.3% of the total cohort), and only 14.1% underwent DXA within the first 12 months. Median time from cancer diagnosis to DXA was 14.8 months. DXA utilization varied by cancer type and was highest in breast cancer (42.2%) and prostate cancer (30.0%), with substantially lower utilization among patients with other malignancies. Among patients with complete evaluable measurements, 40.2% had normal BMD, 45.2% had osteopenia, and 14.5% met criteria for osteoporosis.

Conclusions

In our real-world experience within a large private healthcare network, DXA assessment after cancer diagnosis is infrequently performed and often delayed, with marked disparities by sex and cancer type. A substantial burden of low bone mass is already present at the time of evaluation, highlighting missed opportunities for early skeletal risk identification.

Introduction

Population aging has led to a substantial rise in cancer incidence worldwide, with older adults accounting for most new diagnoses and cancer-related mortality. Advances in early detection and oncologic therapies have significantly improved survival, resulting in a rapidly growing population of cancer survivors. As malignancy increasingly becomes a chronic condition, survivorship care must extend beyond tumor control to address long-term treatment-related complications, including skeletal health [1, 2].

Skeletal morbidity represents a major but often underrecognized component of cancer survivorship. Osteoporosis and fragility fractures are common age-related conditions, and their risk is further amplified in patients with cancer due to both the malignant process itself and cancer-directed therapies [3–6]. Cancer treatment–induced bone loss (CTIBL) is a multifactorial phenomenon driven by systemic inflammation, reduced physical activity, nutritional deficiencies, and exposure to bone-depleting treatments such as chemotherapy, radiotherapy, endocrine therapies, glucocorticoids, and hematopoietic stem cell transplantation [6–10]. These mechanisms may act synergistically to accelerate bone turnover, reduce bone mineral density (BMD), and increase fracture risk. Bone health complications are well-established in hormone-sensitive malignancies, particularly breast and prostate cancer, where aromatase inhibitors and androgen deprivation therapy are strongly associated with accelerated bone loss and increased fracture risk [11–18]. Consequently, clinical awareness and guideline recommendations regarding bone health assessment are more clearly defined in these populations. However, accumulating evidence indicates that skeletal fragility is not confined to hormone-sensitive cancers. Population-based studies have demonstrated an increased fracture risk across a broad range of malignancies, with risk persisting for years after cancer diagnosis, underscoring the broader relevance of bone health in oncology [19–22]. This point was reinforced by a large population-based cohort study of over half a million cancer survivors across 20 tumor types, which found an increased risk of fracture, including major osteoporotic fracture, that extended well beyond multiple myeloma, prostate cancer, and breast cancer, prompting an accompanying editorial to call for greater clinical awareness of skeletal health across the broader oncology population [23]. These data support the rationale for evaluating DXA utilization in an unselected, real-world cancer cohort rather than restricting attention to malignancies with the most established bone health guidelines.

Despite this elevated risk, osteoporosis risk assessment tools frequently underestimate fracture risk in cancer survivors, and a history of malignancy is not consistently incorporated into fracture prediction models [21]. International clinical practice guidelines from oncology and endocrine societies, including the European Society for Medical Oncology, the American Society of Clinical Oncology, and the Endocrine Society, recommend bone health evaluation using dual-energy X-ray absorptiometry (DXA) in patients at increased skeletal risk, particularly those exposed to bone toxic therapies [24–26]. Nevertheless, real-world studies consistently demonstrate low rates of DXA utilization and osteoporosis management among oncology patients, even within well-defined high-risk groups [27–31].

Data describing bone health assessment patterns across unselected cancer populations in routine care remain limited. In particular, there is a paucity of large real-world analyses evaluating the timing and uptake of DXA assessment following cancer diagnosis, as well as the extent to which bone loss is already present at the time of evaluation. Understanding these patterns is essential for identifying gaps in care and informing strategies for more integrated bone–oncology and survivorship pathways.

The present study aimed to evaluate DXA utilization following a new cancer diagnosis within a large real-world oncology cohort treated in a private healthcare network in Israel. Specifically, we examined the frequency and timing of DXA assessment, BMD outcomes, and patient characteristics associated with DXA referral, in order to identify disparities in bone health evaluation and highlight opportunities for earlier skeletal risk identification.

Methods

Study design and data source

We conducted a retrospective cohort study using the MDClone platform, which allows generation of fully anonymized yet statistically accurate datasets. Adults (≥ 18 years) with a first documented cancer diagnosis in Assuta Medical Centers between January 1, 2015, and December 31, 2024 were included. Assuta Medical Centers is a leading private healthcare provider in Israel, operating four hospitals and multiple diagnostic centers across the country. The network provides a broad range of medical services, including oncology, surgery, and advanced imaging, and serves over two million patients annually. Assuta is a privately owned hospital network and is owned by Maccabi Healthcare Services, the second-largest of Israel’s four health maintenance organizations. It functions as a preferred, in-network provider for Maccabi members, who are referred there for oncologic and other treatments and whose care is reimbursed by Maccabi as part of their standard HMO coverage. Under Israel’s National Health Insurance Law, all Israeli residents are entitled to mandatory membership in one of four nonprofit HMOs, each of which is required to provide an identical, government-defined basic basket of services, including oncologic care, irrespective of the member’s socioeconomic status. Maccabi covers approximately one-quarter of the Israeli population, drawn broadly across the country’s socioeconomic and demographic spectrum. Care delivered at Assuta to Maccabi-referred patients therefore reflects standard, HMO-reimbursed oncologic care.

Population and inclusion criteria

Patients were eligible if they were aged 18 years or older, had a new diagnosis of any malignancy during the study period, and had complete demographic data. Patients with missing essential demographic information were excluded.

Data collected

For all subjects with a newly diagnosed malignancy treated in Assuta Medical Centers, we collected demographics age, sex, BMI, and clinical characteristics: smoking status, cancer type, and the treatment received. Cancer diagnosis and site were defined according to the recorded organ of origin at the time of diagnosis, based on structured electronic medical record documentation. Data on oncologic treatments were extracted from structured electronic medical records using the MDClone platform and included documented exposure to chemotherapy, endocrine therapy, biologic agents, radiotherapy, and surgical procedures. Bone mineral density was assessed by dual-energy X-ray absorptiometry (DXA), performed using GE Lunar DXA (GE Healthcare, Madison, WI, USA) systems. We extracted DXA parameters: lumbar spine, total hip, and femoral neck BMD and T-scores. We calculated the time from cancer diagnosis to DXA completion and defined bone status categories: osteoporosis (T-score ≤ –2.5) and osteopenia (T-score between −1.0 and −2.5 according to WHO criteria using the lowest T-score obtained from lumbar spine, femoral neck, or total hip measurements.

Outcomes

Primary outcome is the proportion of cancer patients undergoing DXA.

Secondary outcomes were the time between the oncologic diagnosis and DXA scan performance, BMD/T-score distribution, prevalence of osteoporosis and osteopenia, and the distribution of DXA testing by cancer type.

Statistical analysis

Continuous variables were summarized as mean ± standard deviation and categorical variables as counts and percentages. Comparisons between patients who underwent DXA assessment and those who did not were performed using Student’s t-test or Mann–Whitney U test for continuous variables, as appropriate, and χ² test for categorical variables. Time from cancer diagnosis to DXA assessment was evaluated descriptively, and cumulative incidence curves were used to illustrate differences in timing according to sex. Multivariable logistic regression analysis was performed to identify factors independently associated with DXA assessment following cancer diagnosis. Covariates included age, sex, cancer type (grouped by hormone sensitivity), and documented exposure to chemotherapy and radiotherapy.

Results are presented as adjusted odds ratios with 95% confidence intervals. All analyses were descriptive and exploratory in nature, aimed at characterizing patterns of bone health assessment and identifying disparities in care within the oncology population. Statistical analyses were conducted using IBM SPSS Statistics software (version 29), and a two-sided p value < 0.05 was considered statistically significant.

Results

Study population

The study included 21,112 adults with a first diagnosis of malignancy between 01.01.2015 and 31.12.2024. Women comprised 77.0% of the cohort (n = 16,244), while 22.9% were men (n = 4838). The mean age of the cohort was 60.2 ± 13.1 years, and mean BMI was 26.7 ± 5.4 kg/m². Documented current or former smoking was present in 11,059 patients (52.4%). Documented chemotherapy exposure was recorded in 4512 patients (21.4%) and radiotherapy in 1473 (7.0%) (Table 1).

Table 1.

Baseline characteristics of the study cohort

Characteristic	Total cohort (n = 21,112)
Age, years	60.2 ± 13.1
Female sex, n (%)	16,244 (77.0%)
Body mass index (BMI), kg/m2	26.7 ± 5.4
Current or former smoking†, n (%)	11,059 (52.4%)
Breast cancer, n (%)	8851 (41.9%)
Prostate cancer, n (%)	1411 (6.7%)
Other malignancies, n (%)	10,850 (51.4%)
DXA performed after cancer diagnosis, n (%)	7496 (35.5%)
DXA within 6 months of diagnosis, n (%)	1550 (7.3%)
Documented chemotherapy exposure†, n (%)	4512 (21.4%)
Documented radiotherapy exposure†, n (%)	1473 (7.0%)

^†As documented in the medical record

Utilization of bone density assessment

Overall, 7496 patients (35.5%) underwent DXA assessment following cancer diagnosis, of whom 1550 (7.3%) were evaluated within 6 months of diagnosis. Thus, nearly two-thirds of oncology patients did not receive objective skeletal evaluation, despite known cancer- and treatment-related bone risks. Patients who underwent DXA differed significantly from those who did not, being older, more frequently female, and more likely to have hormone-sensitive malignancies. Documented smoking status and oncologic treatment exposures were substantially more frequent among DXA performers (all p < 0.001) possibly reflecting more comprehensive clinical documentation rather than true differences in exposure prevalence (Table 2).

Table 2.

Characteristics of patients with and without DXA assessment

Characteristic	DXA performed (n = 7496)	No DXA performed (n = 13,616)	p value*
Age, years	61.6 ± 11.2	59.4 ± 13.9	< 0.001
Female sex, n (%)	6771 (90.3%)	9473 (69.6%)	< 0.001
BMI, kg/m2	27.0 ± 5.3	26.4 ± 5.4	< 0.001
Current or former smoking†, n (%)	6372 (85.0%)	4687 (34.4%)	< 0.001
Documented chemotherapy exposure†, n (%)	4512 (60.2%)	1906 (14.0%)	< 0.001
Documented radiotherapy exposure†, n (%)	1473 (19.6%)	681 (5.0%)	< 0.001
Breast cancer, n (%)	3782 (49.5%)	5067 (37.2%)	< 0.001
Prostate cancer, n (%)	440 (5.8%)	971 (7.1%)	< 0.01

p values derived from χ² tests for categorical variables and Student’s t-test or Mann–Whitney U test for continuous variables, as appropriate

^†Smoking status and oncologic treatment exposure were incompletely captured in the electronic health record and are reported as documented exposure only

Sex disparities in DXA assessment

DXA utilization differed markedly by sex. Women were significantly more likely to undergo DXA than men (41.7% vs. 15.0%). Male sex was associated with a substantially lower likelihood of DXA assessment.

DXA assessment by cancer type and treatment

DXA utilization varied significantly by cancer type. The highest rates were observed in breast cancer (42.2%) and prostate cancer (30.0%), accounting for approximately half of all DXA examinations. In contrast, DXA assessment was infrequent in lung, gastrointestinal, ovarian, and pancreatic cancers, with utilization rates generally below 15%. Using an organ-based classification, hematologic malignancies accounted for approximately 6–8% of the overall cohort and 4–5% of patients who underwent DXA assessment (Fig. 1). Among patients who underwent DXA, 4512 (≈60%) had documented exposure to chemotherapy and 1473 (≈20%) had received radiotherapy.

Fig. 1.

Proportion of patients undergoing DXA assessment following cancer diagnosis, by cancer type. ^*Hematologic malignancies were identified using an organ-based classification and may be under-ascertained

Timing of DXA assessment

Early skeletal evaluation was uncommon. Only 1550 patients (7.3% of the total cohort) underwent DXA within 6 months of cancer diagnosis, and only 14.1% underwent DXA within the first 12 months. Among those who underwent DXA, the median time to assessment was 14.8 months, indicating that bone health evaluation was generally delayed and not integrated into early oncologic care.

Although women were significantly more likely than men to undergo DXA, timing among those evaluated was broadly comparable between sexes (Fig. 2a). Median time from cancer diagnosis to DXA differed by malignancy type. Among breast cancer patients (n = 3516), median time to DXA was 12.9 months (IQR 5.8–28.0). In prostate cancer (n = 390), median time was 12.7 months (IQR 4.4–28.8). In contrast, patients with other malignancies (n = 3019) experienced a substantially longer delay, with a median time of 17.9 months (IQR 8.3–35.1) (Fig. 2b). Patients without DXA were administratively censored at the end of the study period. Notably, among patients who underwent DXA, the cumulative curves for men and for prostate cancer patients rose slightly faster than those for women and breast cancer patients during the first year, before being overtaken from approximately 12–18 months onward; this describes the speed of DXA uptake conditional on undergoing DXA at all and does not contradict the much lower overall proportion of men and prostate cancer patients who ever underwent DXA, as noted above. Because prostate cancer is diagnosed exclusively in men and breast cancer almost exclusively in women, this early crossover in timing reflects the same underlying pattern viewed through two overlapping stratifications and is consistent with earlier initiation of baseline bone density testing around the start of androgen deprivation therapy relative to aromatase inhibitor–related testing in breast cancer.

Fig. 2.

Cumulative proportion of DXA performers by time since cancer diagnosis, shown over the first 18 months. a Stratified by sex. b Stratified by cancer type; vertical dashed lines mark 6 and 12 months after diagnosis, with the cumulative proportion of each group reaching DXA by 6 months annotated on the plot

Bone density findings

Among patients who underwent DXA assessment (n = 5570 with complete evaluable measurements), bone mineral density at the time of evaluation was frequently already abnormal. Using the lowest T-score across lumbar spine, femoral neck, and total hip, 2520 patients (45.2%) met criteria for osteopenia and 809 (14.5%) met criteria for osteoporosis, whereas only 2241 (40.2%) had normal BMD. Among patients with osteopenia, the mean lowest T-score was − 1.72 ± 0.39, suggesting established low bone mass rather than marginal diagnostic values. In those meeting criteria for osteoporosis, the mean lowest T-score was − 2.95 ± 0.47, indicating clinically significant skeletal compromise well below the diagnostic threshold (Table 3). Site-specific analysis confirmed that bone loss was particularly evident at the femoral neck: mean femoral neck T-score was − 1.14 ± 1.05 (BMD 0.86 ± 0.13 g/cm²), compared with − 0.83 ± 1.12 at the total hip (BMD 0.90 ± 0.15 g/cm²) and − 0.54 ± 1.50 at the lumbar spine (BMD 1.09 ± 0.22 g/cm²) (Table 4). Osteopenia was most frequent at the femoral neck (50.3%), while osteoporosis, though present at all three sites, was most common at the lumbar spine (10.2%). As the Table 3 composite classification requires only one of three sites to meet the osteoporosis threshold, it necessarily identifies a larger proportion of patients as osteoporotic than any single site considered individually. These findings suggest that DXA assessment was often performed after substantial bone loss had already occurred, and that femoral neck measurements—of greater clinical significance for hip fracture risk than spine or forearm values—showed the least favorable T-score distribution.

Table 3.

Bone mineral density results among DXA performers, by composite classification and skeletal site (composite classification (lowest T-score across all sites), n = 5570^*)

Diagnostic category (lowest T-score)	n	%	Mean lowest T-score ± SD
Normal (T ≥  − 1.0)	2241	40.2%	− 0.16 ± 0.71
Osteopenia (− 1.0 > T >  − 2.5)	2520	45.2%	− 1.72 ± 0.39
Osteoporosis (T ≤  − 2.5)	809	14.5%	− 2.95 ± 0.47

*Includes patients with complete evaluable DXA measurements across skeletal sites

Table 4.

Bone mineral density results among DXA performers, by composite classification and skeletal site (site-specific results)

Skeletal site	N	Mean BMD (g/cm2) ± SD	Mean T-score ± SD	Mean Z-score ± SD	Normal, %	Osteopenia, %	Osteoporosis, %
Lumbar spine (L1–L4)	5457	1.09 ± 0.22	− 0.54 ± 1.50	− 0.06 ± 1.44	58.3	31.6	10.2
Femoral neck	5536	0.86 ± 0.13	− 1.14 ± 1.05	− 0.34 ± 0.95	40.7	50.3	8.9
Total hip	5536	0.90 ± 0.15	− 0.83 ± 1.12	− 0.35 ± 1.10	52.0	41.8	6.3

T-score and Z-score reflect the lower value between the left and right side where both were available, consistent with standard clinical convention.Denominators in Table 4 are slightly lower than the full DXA cohort (n = 7496) and than in Table 3 (n = 5570) because a subset of scanned reports could not be fully parsed into discrete site-level fields during structured data extraction; patients with at least one evaluable site-level measurement are included in Table 4. Table 3 does not report BMD or Z-score, as the composite classification is determined by whichever site yielded the lowest T-score for each patient and does not correspond to a single, clinically meaningful absolute BMD value (BMD reference ranges differ by skeletal site)

In a multivariable logistic regression analysis, female sex was independently associated with higher odds of undergoing DXA assessment compared with male sex (adjusted OR 2.13, 95% CI 1.79–2.50; p < 0.001). Increasing age was also associated with greater odds of DXA assessment (aOR per year 1.03, 95% CI 1.03–1.03; p < 0.001). Patients with breast cancer and prostate cancer had markedly higher odds of DXA assessment compared with patients with other cancer types (aOR 5.82 and 5.29, respectively; both p < 0.001). Collinearity between covariates was assessed and no clinically meaningful multicollinearity was identified (Table 5).

Table 5.

Multivariable logistic regression analysis of factors associated with DXA performance

Predictor	Adjusted OR (aOR)	95% CI	p value
Female sex (vs. male)	2.13	1.79–2.50	< 0.001
Age (per year)	1.03	1.03–1.03	< 0.001
Breast cancer (vs. other cancers)	5.82	5.15–6.58	< 0.001
Prostate cancer (vs. other cancers)	5.29	4.42–6.33	< 0.001

Discussion

In this large real-world oncology cohort, bone mineral density assessment following a new cancer diagnosis was markedly underutilized and frequently delayed. Fewer than 40% of patients underwent DXA at any time after diagnosis, and only a small minority were evaluated within the first 6 months. Despite low screening rates, nearly two-thirds of patients who underwent DXA had abnormal bone density, and more than one in six met criteria for osteoporosis. These findings highlight a substantial gap between guideline recommendations and routine clinical practice, underscoring the need for improved integration of bone health assessment into oncology care pathways.

Importantly, the DXA-performing cohort was heavily enriched with breast cancer patients, who accounted for approximately 42% of all bone density assessments. Given that breast cancer represents a population with well-established guideline recommendations for skeletal monitoring, one might expect high rates of early DXA integration. Nevertheless, overall DXA utilization in breast cancer remained below 50%, and early assessment within six months of diagnosis was uncommon. These findings suggest that gaps in bone health evaluation persist even within guideline-targeted populations and are not confined to nonhormone-sensitive malignancies. Although breast cancer comprised the largest single cancer type, more than half of the cohort (51.4%) consisted of nonbreast and nonprostate malignancies, providing a broad representation of solid and hematologic cancers in routine practice. In this group, DXA use was substantially lower, despite well-established skeletal risks related to systemic therapy, corticosteroids, and transplantation. This gap is particularly relevant in the modern therapeutic era. The skeletal effects of newer systemic therapies—including immune checkpoint inhibitors and targeted agents—remain incompletely characterized. Emerging mechanistic and observational data suggest that immune modulation may influence bone remodeling pathways; however, longitudinal bone mineral density and fracture outcomes remain insufficiently defined [32, 33]. In a recent meta-analysis of randomized trials involving over 19,000 patients, immune checkpoint inhibitor use was associated with a nonsignificant 18% increase in clinical fracture incidence compared with controls (OR 1.18, 95% CI 0.82–1.70), although fracture events were infrequently reported and estimates were imprecise [34]. Together, these findings highlight ongoing uncertainty regarding long-term skeletal risk in the era of immune modulation. In this evolving therapeutic landscape, reliance on traditional risk frameworks may underestimate bone vulnerability across contemporary cancer care.

Sex-based disparities were also evident. Women were more likely to undergo DXA assessment. While this may partly reflect the predominance of breast cancer and greater familiarity with osteoporosis screening in women, it raises concern that men with cancer—particularly those receiving bone-depleting therapies—remain underserved with respect to bone health evaluation.

The delayed timing of DXA observed in this cohort further emphasizes missed opportunities for early intervention. Bone loss associated with cancer therapies often occurs early in the treatment course, and timely identification of low bone mass allows for preventive strategies that may reduce fracture risk and preserve functional independence. Several practical strategies could help narrow the gaps identified in this study. Embedding an automated DXA referral prompt within the electronic oncology record at the time of diagnosis or initiation of bone toxic therapy, analogous to existing decision support alerts for chemotherapy dosing, could reduce reliance on ad hoc clinician recall. Incorporation of bone health assessment into standardized survivorship care plans and multidisciplinary tumor boards, particularly for nonhormone-sensitive malignancies for which guideline awareness is lower, may help extend screening beyond breast and prostate cancer. Given the pronounced sex disparity observed, targeted education for clinicians managing male oncology patients, and closer collaboration between oncology and endocrinology or bone-health services, may help address the underuse of DXA in men. Finally, defining a dedicated referral pathway for patients receiving chemotherapy, radiotherapy, or glucocorticoids—groups in whom skeletal risk is well-documented but not covered by disease-specific guidelines—may facilitate more systematic, guideline-concordant bone health evaluation across the broader oncology population.

This study has several limitations. Its retrospective design and reliance on electronic health record data introduce the potential for incomplete or inconsistent documentation, particularly for treatment exposures and lifestyle factors. Information on cancer histology, disease stage, and treatment dose or duration—factors that may influence skeletal outcomes—was not uniformly available and could not be fully incorporated into the analysis. Hematologic malignancies and stem cell transplantation could not be systematically identified, precluding detailed subgroup analyses in these high-risk populations. DXA examinations performed outside Assuta Medical Centers could not be captured in the present dataset. However, given that bone density testing is routinely reimbursed and widely available within the network, performance of DXA at external facilities is considered unlikely for patients whose oncologic care and routine imaging are both based at Assuta. This is supported by internal review of Assuta’s DXA registry, which confirms very high volume, routine use of on-site DXA among Maccabi Healthcare Services members—who make up the majority of Maccabi-affiliated patients treated at this center—making substantial ascertainment gaps unlikely for this subgroup. Additionally, DXA performance was used as a surrogate for bone health evaluation, and fracture outcomes were not assessed, so it is possible that some patients without a DXA in our cohort nonetheless had osteoporosis by other clinical criteria. The site-specific bone mineral density analysis was based on structured extraction of individual DXA report fields; a subset of reports could not be fully parsed into discrete site-level values, so site-specific denominators (n = 5457–5536) are modestly smaller than the full DXA cohort (n = 7496). These restrictions are unlikely to have introduced systematic bias in the reported means, as the excluded records reflected a data-parsing limitation rather than a clinical selection criterion, but the exact proportions in Table 4 should be interpreted as representative of, rather than identical to, the full DXA-performing cohort. Ethnicity and population subgroup were not consistently captured in the structured electronic medical record fields used for this analysis and could therefore not be reported or examined as a potential determinant of DXA utilization. Prior work using the Maccabi Healthcare Services database has documented that its enrollees include all major population groups composing Israeli society—Arab, nonimmigrant Jewish, immigrant Jewish, and ultra-Orthodox Jewish members—although the exact proportions of these groups within Maccabi’s membership do not necessarily mirror the national population [35]. We therefore cannot exclude ethnicity as a potential residual confounder, but we have no specific reason to expect it to be a dominant driver of the low and delayed DXA utilization observed in this cohort. Assuta Medical Centers is a privately owned hospital network, but it functions as a preferred, in-network provider for Maccabi members, who are referred there for oncologic care and whose treatment is reimbursed by Maccabi under Israel’s National Health Insurance Law; care delivered through this pathway is therefore standard, HMO-covered oncologic care rather than exclusive or income-restricted private pay treatment, and DXA is included in the nationally mandated health basket for recognized indications. It is also possible that patients with the most complex or severe oncologic disease—for example, those requiring highly specialized multidisciplinary management, transplantation, or intensive inpatient supportive care—are preferentially treated at Israel’s large public academic medical centers, which could result in a study population somewhat enriched for less complex disease trajectories relative to the broader national cancer population. We cannot evaluate this possibility with the data available, but if present, such case-mix selection could affect the generalizability of our findings. Our data are limited to a single national network within Israel’s specific healthcare system, and findings may not generalize to countries with different insurance structures or DXA access. Nonetheless, the large cohort size and real-world setting provide valuable insight into current practice patterns and highlight clinically relevant gaps in care. Importantly, more than half of the cohort consisted of nonbreast and nonprostate malignancies, ensuring broad representation beyond traditionally studied hormone-sensitive cancers.

In summary, our real-world experience reveals that bone density assessment following cancer diagnosis is often delayed and selectively applied, despite a high prevalence of low bone mass at the time of evaluation. These observations highlight missed opportunities for early skeletal risk identification and support more proactive incorporation of bone health assessment into routine oncology and survivorship care.

Conclusions

In conclusion, in our real-world experience, bone density assessment following cancer diagnosis was infrequently performed and commonly delayed, with substantial disparities according to sex and cancer type. These observations should be interpreted in the context of current guideline recommendations, which primarily target patients in hormone-sensitive cancers receiving endocrine therapies or those with additional fracture risk factors. Nevertheless, the high prevalence of low bone mass among patients who underwent DXA suggests that skeletal risk is often recognized only after clinically relevant bone loss has already occurred. Collectively, these findings support the need for more systematic and proactive integration of bone health evaluation into oncology and survivorship care, including populations beyond traditionally recognized high-risk malignancies.

Author contribution

V.R. conceived and designed the study. Z.V. and M.S. contributed to data acquisition and data curation. V.R. and Z.V. performed data analysis and interpretation. V.R. drafted the initial manuscript. Y.G. and A.M. contributed to study supervision, critical revision of the manuscript for important intellectual content, and provided senior clinical oversight. All authors reviewed, edited, and approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Funding

This research received no external funding.

Data availability

Data Availability Statement The data that support the findings of this study were extracted from the MDClone platform using fully anonymized electronic medical record data from Assuta Medical Centers. The dataset generated and analyzed during the current study is not publicly available due to institutional data governance policies and patient privacy regulations. Access to de-identified data may be considered upon reasonable request to the corresponding author and subject to approval by Assuta Medical Centers and the relevant Institutional Review Board, in accordance with national data protection regulations.

Declarations

Ethics approval and consent to participate

The study was reviewed by the Assuta Medical Centers Institutional Review Board and was granted exemption from formal ethical approval due to the use of fully anonymized retrospective data. Informed consent was waived in accordance with institutional and national guidelines.

Competing interests

The authors declare no competing interests.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work, the author (Vanessa Rouach) used ChatGPT (OpenAI) to assist with language editing and refinement of the text. After using this tool, the author reviewed and edited the content as needed and takes full responsibility for the content of the published article.