Incidence of Hypercalcemia and Vertebral Fractures Following Denosumab Withdrawal in Lung Cancer Patients: A Longitudinal Cohort Study
Article information
Abstract
Background
Bone-target agents (BTAs), including denosumab (DMAb), are one of the bone metastasis treatments that should continue indefinitely. However, BTAs may be interrupted in some cases. In osteoporosis, DMAb withdrawal causes a rebound effect characterized by an increased bone turnover with spine fractures and hypercalcemia; evidence of the DMAb withdrawal effect in oncology is lacking.
Methods
This study aimed to identify the DMAb withdrawal effect amongst lung cancer patients treated with DMAb for bone metastases between January 2020 and December 2021. Patients who discontinued DMAb were included. Encounter notes, radiological and laboratory findings were comprehensively reviewed.
Results
Thirty patients were included with a median follow-up of 21 months (interquartile range [IQR], 10–30) after DMAb discontinuation. Bisphosphonates were administered before starting DMAb in 7 patients (23.3%) and after DMAb withdrawal in 4 cases (13.3%). Three cases of DMAb withdrawal-related hypercalcemia and 3 cases of spine fractures following DMAb cessation were identified in 5 patients (16.7%), all of them were females and the median age was 65 years old (IQR, 65–70). No statistical difference in DMAb duration or number of injections was found in patients developing DMAb withdrawal-related spine fractures or hypercalcemia compared with others (binary logistic regression, P=0.688 and P=0.938, respectively).
Conclusions
Patients with bony-metastatic lung cancer, especially post-menopausal women, are at risk of fractures and calcium abnormalities after DMAb discontinuation, suggesting that DMAb withdrawal effect may also be present in the oncological setting. A close follow-up and careful monitoring during and after discontinuation of DMAb is necessary.
INTRODUCTION
Lung cancer, which includes patients with both non-small cell lung cancer (NSCLC) and small cell lung cancer, is one of the most commonly diagnosed cancer types worldwide and is the leading cause of all cancer-related deaths.[1,2] Much progress has been recently made, leading to a significant drop in death rates both in men and women.[3] The skeletal system is one of the most common sites for metastases; indeed approximately two-thirds of patients with lung cancer have bone metastases at diagnosis, and 30% to 40% of patients with NSCLC will develop bone metastases.[4,5]
The formation of bone metastasis is a complex process, which affects the bone remodeling that occurs continuously in the skeleton. Although partially understood, this process includes the production of cytokines and growth factors by the proliferating tumor cells, which increase the osteoblast production of receptor activator of nuclear factor-κB ligand (RANKL).[6,7] It results in bone destruction and the occurrence of skeletal-related events, including pathological fractures, the need for radiotherapy to bone, the need for surgery to bone, spinal cord compression, and hypercalcemia.[7–9] Treatments include bone-target agents (BTAs), such as bisphosphonates (BPs) and denosumab (DMAb), which is a monoclonal neutralizing antibody to RANKL that inhibits the osteoclast differentiation, survival, and activity.[6,8] As the cure is not achievable in the majority of cases, the bone metastasis treatment aims to prevent disease progression and palliate symptoms.[6] Regarding lung cancer, a recent meta-analysis demonstrated that for overall survival, DMAb was more effective than zoledronic acid, which is a BP, and estimated to confer a mean of 3.3 months (95% confidence interval [CI], 0.3–6.3) of increased overall survival compared with untreated patients.[9] According to the European Society for Medical Oncology (ESMO) guidelines, BTAs should continue indefinitely, including the hospice setting. Treatment may be interrupted in selected patients with good prognostic features such as oligometastatic disease, a perceived low risk of bone complications and durable response to systemic treatment.[6]
Due to its anti-resorptive mechanism, DMAb is also used as an anti-osteoporotic therapy. To note, the dose in osteoporosis is 60 mg every 6 months [10] and in bone metastasis is 120 mg every 4 weeks.[6] In osteoporotic patients, the discontinuation of DMAb results in a rebound phenomenon characterized by increased bone turnover marker (BTM), increased bone resorption leading to declining bone mineral density (BMD), and increased risk of major osteoporotic fractures, especially vertebral fractures.[11] The increased risk tends to be revealed within 1-year post-discontinuation and to be greater after a longer treatment duration.[11–14] Moreover, DMAb cessation has been rarely reported to be followed by the onset of hypercalcemia.[15,16] Therefore, osteoporotic patients transitioning off DMAb, especially after 3 years of treatment, should initiate zoledronic acid or another BP to reduce bone turnover and BMD loss and prevent vertebral fractures.[14] In oncologic patients, the evidence of rebound phenomenon is rare,[16] likely because DMAb is commonly continued life-long and because of the difficulty to distinguish the rebound phenomenon from the risk of bone complications related to progression of bone metastases. This study aimed to detect the DMAb-withdrawal rebound phenomenon, including spine fractures and hypercalcemia, in lung cancer patients with bone metastatic disease.
METHODS
1. Study design and participants
This study was performed at Brigham and Women’s Hospital (Boston, MA, USA) and was approved by the Institutional Review Board of Partners HealthCare. Patient records were searched using the Research Patient Database Registry (RPDR), which is an electronic registry of patient clinical records in the Partners HealthCare system.
The RPDR contains data from more than 7 million patients. Outpatient clinical notes starting from January 1st, 2020 through December 31st, 2021, were searched for the following keyword combination: “lung cancer” AND “denosumab” (Fig. 1). The query yielded 326 patients. Medical records from these patients, including encounter notes and laboratory results, were comprehensively reviewed. Patients who discontinued DMAb for more than 4 weeks were included. Exclusion criteria were DMAb administered as antiosteoporotic treatment (60 mg every 6 months), DMAb treatment for cancers other than lung, ongoing DMAb, patients lost to follow-up, follow-up <30 days after DMAb withdrawal, DMAb administered for hypercalcemia of malignancy (HCM), DMAb withdrawal for cancer progression/poor clinical condition according to medical clinical notes, missing baseline laboratory evaluation, missing laboratory and radiological exams after DMAb withdrawal, kidney and liver failure.
Flowchart of identification, selection and inclusion of the patients. DMAb, denosumab; HCM, hypercalcemia of malignancy.
2. Data collection and definitions
Data collected included demographic information, past medical history, family history of osteoporosis or fragility fractures (at the spine, hip, or wrist), and antiresorptive treatment before and after DMAb withdrawal. Clinical notes, radiological reports and laboratory exams performed before, during and after DMAb withdrawal were also collected and analyzed. As DMAb for bone metastasis is administered monthly,[6] the 4 weeks following the last DMAb injection were included in the DMAb treatment period (Fig. 2).
Timeline of the study definitions. DMAb, denosumab.
Hypercalcemia was defined as a serum calcium level above the upper limit of our institutional laboratory’s normal range (8.7–10.4 mg/dL). Diseases and drugs potentially causing hypercalcemia (primary hyperthyroidism, thyrotoxicosis, adrenal insufficiency, pheochromocytoma, acute renal failure, thiazide diuretics, lithium, vitamin A, and parathyroid hormone)[17] were evaluated and excluded by analyzing the clinical and laboratory notes. We further defined hypercalcemia related to cancer in the case of radiologically documented cancer progression; otherwise, in the absence of radiological evidence of metastasis and other possible etiologies, hypercalcemia was classified as related to the DMAb withdrawal. Hypocalcemia was defined as a serum calcium level below the lower limit of our institutional laboratory’s normal range. In the presence of hypoalbuminemia, hypercalcemia and hypocalcemia were defined based on ionized calcium levels or albumin-adjusted calcium concentration using the equation: Corrected calcium (mg/dL)=(0.8×[normal albumin–albumin])+ calcium.[18,19]
The occurrence of fractures was identified through clinical notes and radiological reports for each patient during oncological follow-up. Radiological exams included thoracic and abdominal X-rays, computerized tomography, or magnetic resonance imaging (MRI). Pathological bone fracture is defined as a bone fracture that occurs without adequate trauma and is caused by a preexistent pathological bone lesion.[20] In the present study, cancer-related fracture was defined as the occurrence of pathological bone fracture in the presence of a pre-existing or a new bone metastasis detected by MRI and bone scan. DMAb withdrawal- related fracture was defined as bone fractures occurring in a bone without radiological evidence of metastasis, as a result of a low or non-traumatic force based on medical records. As the discontinuation of DMAb in osteoporosis results in an increased risk of vertebral fractures [12] and the oncological follow-up was based on thoracic and abdominal imaging, we focused the analysis on the occurrence of vertebral fractures.
3. Statistical analysis
Frequencies and percentages reported for categorical variables. The Shapiro-Wilk test was utilized to determine the normality of data distribution. Continuous variables were summarized as mean±standard deviation if normally distributed; otherwise, they were presented as median and interquartile range (IQR). Binary logistic regression models were fitted to identify the association between DMAb-related fractures or hypercalcemia and predictor variables. Statistical tests were 2-tailed with the 2-sided statistically significant level set at P value equal to 0.05. Analyses were conducted using SPSS version 26 (SPSS Inc., Chicago, IL, USA).
RESULTS
Thirty patients treated with DMAb for lung cancer bone metastasis stopped the treatment. The reasons for DMAb withdrawal were: good prognosis/stable bone metastasis (N=13), concern about dental issues (N=4), osteonecrosis of the jaw (N=2), intolerance (N=2), patients’ choice (N=2), hypocalcemia (N=1), impossibility to attend the clinic (N=1), incompatibility to protocol treatment (N=1), concern about side effect (N=1), or reason not further specified (N=3). Seven patients (23.3%) received BPs before starting DMAb as anti-osteoporosis or bone metastasis treatment. In this sub-group, BPs were switched to DMAb for oncological reasons (N=2), patients’ preference (N=2), and reasons not further specified (N=3). Patient characteristics and laboratory evaluation at DMAb withdrawal are presented in Table 1 and Supplementary Table 1.
Patient characteristics at denosumab withdrawal
The median follow-up after the last DMAb injection was 21 months (IQR, 10–30). At the last follow-up, 18 patients (60%) were alive. Incidence of spine metastasis, spine fractures and occurrence of hypercalcemia before, during and after DMAb administration are summarized in Table 2 and Figure 3.
Incidence of spine metastasis, spine fractures, and occurrence of hypercalcemia before, during, and after denosumab treatment
Incidence of denosumab (DMAb) treatment-related hypocalcemia and DMAb withdrawal-related hypercalcemia and fractures (N=30). Blue bar represents percentage of each condition.
Hypocalcemia occurred in 11 cases (36.7%) after DMAb withdrawal and required calcium gluconate in 3 cases. No PTH or vitamin D levels were available at hypocalcemia occurrence. The median time from the last DMAb to hypocalcemia was 4 months (IQR, 3–11). At the time of hypocalcemia, tumor progression was radiologically evident in 3 cases (10%). Moreover, hypocalcemia was detected during hospitalization in 3 cases (10%). Data are summarized in Table 2 and Supplementary Table 2.
Three cases of DMAb withdrawal-related new-onset hypercalcemia and three cases of DMAb withdrawal-related spine fracture following DMAb cessation occurred in 5 patients (16.7%), all of them were females and the median age was 65 years old (IQR, 65–70) (Table 2, Fig. 4). The median time from the last DMAb to the occurrence of DMAb withdrawal-related new-onset hypercalcemia was 20 months (IQR, 11–20). Hospitalization and additional treatments were not required (Supplementary Table 3). DMAb withdrawal-related spine fractures (N=3) occurred after a median time from the last DMAb of 16 months (IQR, 15–16) and were at multiple levels in 2 cases (Supplementary Table 4).
Kaplan-Meier analysis (solid line) for (A) hypercalcemia and (B) spine fracture post- denosumab (DMAb) withdrawal among lung cancer patients (N=30). The dashed lines represent the 95% confidence interval.
In particular, in patient no. 19, the fractures in L2 occurred after a low-grade trauma; according to clinical notes, the fracture was attributed to osteoporosis because of the absence of bone metastasis on radiological exams.
In patient no. 29, the fractures in T3, T10, and L1 occurred after a low-grade trauma. The radiological report stated that the fractures were likely related to osteoporosis, not cancer. Indeed, the disease was overall stable and there was no evidence of bone metastasis progression.
In patient no. 30, the fractures in T12, L1, and L2 occurred in absence of any trauma. There were no evidence of underlying lesions and overall the disease was stable. Therefore, the clinical report stated that the compression fractures were likely benign, due to osteoporosis.
One patient received BPs because of the occurrence of a DMAb-withdrawal related spine fracture and one patient because of cancer-related hypercalcemia, which occurred 16 and 18 months after the last DMAb injection, respectively (Supplementary Table 5). Additionally, BPs were administered in 2 patients within 6 months after the last DMAb, without evidence of hypercalcemia or fracture. No episode of hypercalcemia or fractures were reported during BPs treatment.
There was no difference in DMAb treatment duration or the number of injections in patients developing hypercalcemia or DMAb-related fracture compared with others in the entire cohort (N=30, binary logistic regression, P=0.688 and P=0.938, respectively) or excluding patients treated with BPs before or after DMAb (N=21, binary logistic regression, P=0.840 and P=0.616, respectively).
DISCUSSION
In this study, we observed the occurrence of DMAb withdrawal-related hypercalcemia and spine fractures in 10% and 10%, respectively, of bone metastatic lung cancer patients.
DMAb could be used in the oncological setting for two main reasons. The first one is to prevent bone loss in breast cancer and prostate cancer undergoing endocrine therapy. In this setting the DMAb dose is the anti-osteoporotic one (60 mg every 6 months) and the effect of DMAb withdrawal seems similar to the one observed in osteoporosis.[11,21,22] The second use of DMAb is for treating bone metastasis, indication in which the dose is 12 times the standard osteoporosis dose and in which the evidence of the rebound effect following DMAb discontinuation are limited to two case reports [23,24] and one study.[25] Our results suggested that the DMAb-withdrawal effect observed in osteoporosis, which results in new-onset fragility fractures [13,14] and acute hypercalcemia,[15] could be present also in patients receiving the oncology doses of DMAb.
1. Vitamin D and calcium levels evaluation
Adequate vitamin D and calcium supplements are critical in calcium-level homeostasis.[26] Although 70% of the present cohort patients were taking vitamin D supplements, vitamin D levels at the time of DMAb withdrawal were not available in most patients. Our findings underscore the importance of monitoring serum vitamin D and calcium levels during and after discontinuing DMAb treatment. Given hypercalcemia and hypocalcemia can both occur with DMAb withdrawal, it is important to reassess calcium and vitamin D intake if DMAb is discontinued.
2. Hypocalcemia
We analyzed the laboratory data performed before, during and after DMAb treatment and we defined the events occurring within the 4 weeks after the last DMAb injection as part of DMAb treatment. This method permitted us to minimize the well-known DMAb hypocalcemia, that is caused by the neutralization of the RANKL receptor and the reduced signal for osteoclast formation, function, and survival.[27] However, a great percentage of our patients (36.7%) had hypocalcemia after 4 weeks following DMAb withdrawal. It was mild in the majority of the patients as calcium gluconate was administrated in just 3 cases.
On one hand, it is unknown when stops the anti-resorptive effect of DMAb and when starts the rebound phenomenon during the oncological administration of DMAb.[27] Therefore, we cannot fully exclude that the observed hypocalcemia could be partially caused by the anti-resorptive action of DMAb.
On the other hand, hypocalcemia could be caused by the rebound phenomenon itself, in which there is a transient increase in bone resorption markers that can cause calcium abnormalities, particularly in patients with risk factors such as chronic kidney disease or vitamin D deficiency. [28] In addition, hospitalization and tumor progression contribute to hypocalcemia.
3. Hypercalcemia and fragility fractures
The key finding of the present study was the detection of post-DMAb hypercalcemia or spine fractures in 5 lung cancer patients (16.7%), which were all post-menopausal females. Moreover, we highlighted some points to discuss.
First, hypercalcemia following DMAb discontinuation has been rarely described in oncological patients, mainly in children with giant cell tumors of bone.[15] Hypercalcemia is likely caused by the increase of bone resorption markers during the post DMAb-rebound effect.[28] In the present study, hypercalcemia after DMAb withdrawal was detected in 3 (10%) cases. A recent study of 21 bony-meta-static breast cancer patients reported 6 (28.6%) cases of rebound hypercalcemia at a median time of 7.5 months post-DMAb. No episode of hypocalcemia was reported. As in the present study, prior BPs use was not an exclusion criterion, and it was not significantly different in hypercalcemic patients compared to others. In the breast cancer study, DMAb treatment duration and total treatment doses were higher in patients who developed post-DMAb hypercalcemia (median 41 months, 40 doses) compared to those who remained normocalcemic (median 10 months, 5 doses; P=0.009).[25] In our study, such a difference was not found. Further studies to clarify risk factors of post-DMAb hypercalcemia are needed.
Second, the risk of rebound fractures in cancer patients following DMAb cessation has not been specifically studied. In the present study, 10% of patients had at least one spine fracture related to DMAB withdrawal and were at multiple levels in 2 cases. This incidence is in line to the vertebral fracture rate of 7.1 per 100 participant-years after DMAb discontinuation for osteoporosis treatment.[12] To note, the DMAb-related spine fractures occurred in three post-menopausal women, one of them with osteoporosis and 1 with osteopenia. We tried our best to exclude pathologic vertebral fractures by recording presence of spinal metastases before, during, and after DMAb treatment. Tyan et al. [24] reported a case of a 55-year-old woman with metastatic lung cancer, without other relevant risk factors, presenting multiple atraumatic spinal fractures that required kyphoplasty 15 months after receiving the last oncology dose of DMAb. Her bone densitometry, performed after the fractures, revealed osteopenia at the femoral level, and lumbar spine densitometry was uninterpretable due to kyphoplasty.[24] Therefore, fracture after DMAb cessation seems to be not uncommon and may be related to several factors, such as underlying osteoporosis.
Third, in osteoporotic patients BPs have been showed to limit the post-DMAb rebound factor.[14] In oncology, ESMO guidelines recommend the administration of BPs if DMAb is discontinued for more than 6 months to prevent the possible rebound effect, based on DMAb mechanism of action and outcomes in osteoporotic patients.[6] Interestingly, in the present study, post-DMAb BPs (zoledronic acid, 4 mg) were administered in 4 patients (13.3%): the two patients who received BPs within 6 months had no evidence of hypercalcemia or fractures; the other 2 patients received BPs after sustaining post-DMAb hypercalcemia or fractures. Similar to our study, Wang et al. [29] reported only one out of 18 patients transitioned directly to BPs therapy after stopping DMAb. Overall, it is likely that the post-DMAb transition to BPs is not an established clinical practice. However, as cancer mortality has declined in the last decade,[3] it is essential to provide evidence and updated guidelines regarding the management of discontinuation of DMAb.
Overall, oncologists prescribing BTAs for bone metastasis should be aware that DMAb is associated with the withdrawal rebound effect, especially in patients at high risk of osteoporosis.
4. Limitation
This is a single-center retrospective study with a small sample size and results are not generalizable. The distinction of hypercalcemia and fractures related to cancer itself or to DMAb withdrawal is challenging. Indeed, the nature of the study may introduce biases during the assessment of causative relationships between DMAb withdrawal, hypercalcemia and vertebral fractures. Moreover, 1,25-dihydroxyvitamin D and PTH-related peptide dosage at the time of hypercalcemia are missing, causing the inability to rule out HCM. However, we tried our best to define hypercalcemia and spine fractures related to cancer or DMAb withdrawal based on the laboratory exams, MRI, bone scans, and clinical reports in a real-world homogeneous population referring to our hospital. Moreover, to limit the misdiagnosed paraneoplastic syndromes, which are associated with progression,[30] we defined the fracture related to cancer in presence of a pre-existing or a new bone metastasis. On the contrary, DMAb withdrawal-related fracture was defined as bone fractures occurring in a bone without radiological evidence of metastasis.
As the discontinuation of DMAb in osteoporosis results in an increased risk of vertebral fractures [4] and the oncological follow-up was based on thoracic and abdominal imaging, we focused the analysis on the occurrence of vertebral fractures. A minor limitation of this analysis was that follow-up radiographs taken after DMAb discontinuation were obtained at routine scheduled appointments, and not for intercurrent back pain or any other clinical vertebral fractures symptomatology. However, the routine radiographic surveillance allowed for the capture of all morphometric vertebral fractures, which are usually not clinically recognized. Another limitation is the missing BTM and BMD measurements, which might help to diagnose the bone reabsorption effect and the bone metabolism changes following DMAb withdrawal, which could explain the occurrence of fragility fractures.
CONCLUSIONS
Patients with bony-metastatic lung cancer are at risk of fractures and serum calcium abnormalities after DMAb discontinuation. An adequate evaluation of vitamin D and calcium levels is fundamental to limit the occurrence of hypocalcemia during and after DMAb treatment. A baseline evaluation of the bone density at the time of DMAb withdrawal may help in patient management, as patients with osteoporosis may consider transition to osteoporotic-dosing of DMAb or other osteoporosis treatment. Emphasis needs to avoid sudden and unprepared DMAb withdrawal, including recommendations for maintaining good oral hygiene. Given the large number of lung cancer patients with bone metastasis treated with DMAb and their better overall survival, prospective studies are urgently needed to analyze and prevent risk of DMAb withdrawal-associated vertebral fractures.
Supplementary Information
Notes
Funding
The authors received no financial support for this article.
Ethics approval and consent to participate
The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Institutional Review Board of Partners HealthCare.
Conflict of interest
No potential conflict of interest relevant to this article was reported.