Remineralization Rate of Lytic Lesions of the Spine in Multiple Myeloma Patients Undergoing Radiation Therapy.

Introduction

Multiple myeloma (MM) is the second most diagnosed hematologic malignancy and is rising in incidence in developed countries. 1 , 2 It is a malignant disease characterized by monoclonal plasma cell proliferation in the bone marrow. 1 In 60%–90% of the MM cases, these malignant plasma cell proliferations cause lytic bone lesions and/or diffuse osteopenia due to increased activity of osteoclast and decreased osteoblastic activity. 3 , 4 This leads to lytic lesions and therefore to a loss of structural integrity of the spine, increasing the risk of pathologic fractures and severe pain with a negative impact on quality of life. 3 , 5

In general, MM patients are treated with systemic therapy including high-dose chemotherapy, in some cases followed by stem cell transplantation. 6 Although systemic MM treatment has improved greatly over the years with the emergence of novel therapies such as monoclonal antibodies and cellular therapies, the disease remains incurable. 7 Radiation therapy can also have an important supportive role in the palliative management of MM related osteolytic lesions. 4 It has been demonstrated that even low doses of radiation offer significant symptom relief for tumor deposits in the bone or soft tissue. 4 , 8 However, radiation can result in myelosuppression and irreversible bone marrow fibrosis. This is detrimental, because a sufficient bone marrow reserve is necessary for eligibility for systemic therapy and future stem cell transplantation. 9 , 10 Hence, the selection of patients for radiation therapy should be based on careful consideration of the potential risks and benefits.

Previous clinical trials have shown the potential of radiotherapy to induce radiation-associated remineralization in patients with solid tumor metastases (renal cell, breast, prostate, lung, urothelial cell, stomach, etc.), i.e. increasing bone mineral density (BMD) and thus (partially) restoring the structural integrity of the vertebrae. 11 - 15 In patients with MM, however, the success rate was lower; one study measured the size of lytic lesions before and after radiation and found that 48% of lesions showed some remineralization. 16 Although, the studies including MM patients did not include quantification of the effect on BMD using DEXA or Hounsfield Unit measurements. Thus, in patients with MM, the impact of radiation on the trabecular and cortical bone surrounding the lytic lesions, as well as the potential for bone reforming within these lytic areas, remains largely unknown. MM often leads to widespread bone disease rather than isolated lesions, adding further complexity to the understanding of the effect of radiation. Therefore, our study aims to investigate the degree of radiation-associated remineralization in MM patients to gain a better understanding of its potential impact on BMD and, consequently, fracture prevention.

Methods

Study Design

This retrospective cohort study was conducted at two affiliated tertiary care hospitals in the United States. Institutional review board approval was obtained prior to initiation of this study (registration number 2018P000688).

Patient Selection

Eligibility for inclusion was assessed based on the following criteria: (1) a known diagnosis of MM, (2) received conventional external beam radiotherapy for spinal MM lesions, (3) a CT scan of the radiated area within 12 weeks before radiation, and (4) at least one follow-up CT scan with the same modality between 3-24 months after radiation. ICD-9 and ICD-10 codes for Multiple Myeloma (203.0 and C90) and CPT codes for Radiation Treatment Delivery were utilized to identify patients. In total, 569 potentially eligible patients were identified. Patients or specific vertebrae were excluded if they (1) underwent surgery in the treated area prior to radiation or before the first follow-up CT-scan (as this could influence the Hounsfield Unit (HU) measurement) (n = 25), or (2) had previously undergone radiation to the same area (n = 9). We manually screened medical records of those potentially eligible patients and 119 patients ultimately met the selection criteria (See appendix, Figure A1 ). General baseline characteristics were manually extracted from patients’ electronic medical records. We specifically documented vertebral level, chemotherapy regimen, bisphosphonate use, and radiation scheme to use in further analyses and all variables mentioned in Table 1 for descriptive purposes. Table 1. Baseline and Treatment Characteristics. Overall (n = 119) Male 79 (66.4%) Age at radiation 64 [11.7] BMI (kg/m 2 ) 28.3 [24.7 – 32.0] ISS disease stage - Plasmacytoma 17 (14.3%) ISS 1 47 (39.5%) ISS 2 33 (27.7%) ISS 3 22 (18.5%) Back pain (NRS) 8.0 [5.0 – 9.0] No pain (NRS 0) 5 (4.2%) Mild (NRS 1-3) 16 (13.4%) Moderate (NRS 4-6) 25 (21.0%) Severe (NRS 7-10) 73 (61.3%) Neurologic symptoms - No neurology 75 (63.0%) Only motor 15 (12.6%) Only sensory 16 (13.4%) Combined 13 (10.9%) ECOG performance score - 0 26 (22.0%) 1 53 (44.9%) 2 27 (22.9%) 3 12 (10.2%) ASIA impairment score - E 96 (80.7%) D 20 (16.8%) C 3 (2.5%) Charlson comorbidity index 4.3 [2.8] Baseline VCFs - No VCF 74 (63.8%) One VCF 34 (29.3%) Two VCFs 5 (4.3%) Three or more VCFs 3 (2.6%) Chemotherapy regimen - RVd (lite), Rd or RV 56 (47.1%) DRd, DVd or DPd 14 (11.8%) KRd, Kd or KCd 11 (9.2%) Other regimens 24 (20.2%) No chemotherapy 14 (11.8%) Start chemotherapy - Before radiation 69 (65.7%) After radiation (before end of follow-up) 36 (34.3%) Bisphosphonates 78 (65.5%) Once 7 (6.8%) 2-5 times 7 (6.8%) Regularly/quarterly 64 (62.1%) Pre-radiation steroids 48 (40.3%) Former chronic prednisone use 15 (12.6%) Site treated - Cervical spine 7 (5.9%) Cervicothoracic 11 (9.2%) Thoracic spine 48 (40.3%) Thoracolumbar 10 (8.4%) Lumbar spine 22 (18.5%) Lumbosacral 12 (10.1%) Sacral 9 (7.6%) Number of levels radiated - 1-2 29 (24.4%) 3-5 72 (60.5%) 6-10 16 (13.4%) 10+ 2 (1.7%) Radiotherapy technique - AP/PA/3D 106 (91.4%) VMAT/IMRT 8 (6.9%) SRS 2 (1.7%) Dose schedule 1 × 8 Gy 8 (6.7%) 1 × 15 Gy 1 (.8%) 5-6 x 4 Gy 36 (30.3%) 10-12 × 2.5 Gy 11 (9.2%) 10 × 3 Gy 45 (37.8%) 15-17 × 2 Gy 4 (3.4%) 20 x 2-2.5 Gy 5 (4.2%) 25 × 1.8 Gy 9 (7.6%) Biologically effective dose α/β = 10 39 [28 – 39] Values are presented as number (%) and mean [standard deviation] for normally distributed data or median [interquartile range] for non-normally distributed data. Abbreviations: BMI, Body Mass Index; ISS, International Staging System; NRS, Numeric Rating Scale; ECOG, Eastern Cooperative Oncology Group; ASIA, American Spinal Cord Injury Association, VCF, Vertebral Compression Fracture; R, Revlimid; V, Velcade; d, Dexamethasone; D, Daratumumab, P, Pomalidomide; K, Carfilzomib; C, Cyclophosphamide; AP/PA, Anteroposterior-posteroanterior; 3D, three-dimensional; VMAT, Volumetric modulated arc therapy; IMRT, Intensity modified radiation therapy, SRS, Stereotactic radiosurgery; Gy, Gray.

Outcome Measures

Due to the difference in absolute HU values across vertebral levels and across different patients, our primary outcome measure was percent change in bone density (Δ% HU) between pre- and post-radiation measurements, compared to non-targeted vertebral levels that functioned as reference. The Δ% HU was calculated for each vertebral level using the following equation: Δ % HU = H U p o s t − H U p r e H U p r e x 100

All outcome measurements were pooled into five consecutive follow-up intervals of 3-6 months, 6-9 months, 9-12 months, 12-18 months, and 18-24 months.

HU Measurements

The hospital eUnity PACS software system was used to access imaging and to measure the bone density values in HU. Mean HU score for each vertebra was calculated by placing a circular Region of Interest (ROI) on the mid-axial plane of CT-scans. The center of the ROI was placed on the estimated 3D centroid of the vertebral body, maximizing the diameter while still excluding the vertebral cortices to prevent volume averaging due to the normally high-attenuating cortex (See Appendix, Figure A2 ). Careful attention was taken to ensure that the ROIs drawn on the axial slices were in the mid-axial portion of the vertebral body by cross-checking on the sagittal slices. The baseline measurement was performed on CT scans obtained within 12 weeks prior to radiotherapy. The post-radiation measurements were done in the same way and as close to the original measured location and diameter as possible, on CT scans performed between 3-24 months after radiation. These measurements were only done if a comparable follow-up CT scan was available (e.g. CT Thorax without contrast agent compared with a CT Thorax without contrast agent). If multiple comparable CT scans were acquired within this timeframe, a maximum of three follow-up measurements were done, ensuring the measurements were at least three months apart. Vertebral levels C1, C2 and S2-S5 were excluded because of their different anatomy. For each patient, two reference measurements were performed on one cranial and one caudal vertebra outside the radiated area (if visible on CT-images) to detect changes in bone density of non-targeted vertebrae. These control measurements were taken from the nearest non-adjacent vertebra to allow the adjacent vertebra to serve as a buffer for some potential imprecision of radiation delivery and the scattered effect of radiation. In the presence of any implants (e.g. cement or metallic implants), the first non-adjacent implant-free vertebra was chosen.

Statistical Analysis

All HU measurements were performed by a single rater (JPV). In order to ensure validity of these measurements, a random sample of 20% of the entire patient cohort was also measured by a second rater (HZ). Subsequently, inter-rater reliability was assessed with an intraclass correlation coefficient (ICC) which was calculated using a single-measurement, absolute agreement, two-way random effects model. Measurement of Δ% HU was considered reliable, with an ICC of .96 (95% CI .94-.97; P < .001), indicating excellent reliability. 17

To compare HU values of radiated vertebrae to non-targeted vertebrae per timepoint, unpaired t-tests were used. In order to assess changes across time within the same vertebrae, paired t-tests were used to compare mean HU values from baseline to five consecutive follow-up intervals. To test if there is a difference in remineralization rate between radiated and control vertebrae, a linear mixed effect model using the restricted maximum likelihood estimator (REML) approach was fitted, since multiple vertebrae per patient were included, and these ‘clusters’ of vertebrae cannot be treated as uncorrelated measurements. In mixed models, a random subject and cluster effect can be added to account for the correlation of the data. 18 Fixed effects are variables that are constant across individuals whereas random effects vary. To estimate the effect of radiation on remineralization rate over time, variables radiation (yes/no) and time from radiation to HU measurement (in months) were fitted as fixed effects and together in an interaction term to compute the remineralization rate. Δ% HU was fitted as dependent variable. To account for multiple measurements (vertebrae) per patient, a random intercept was fitted per patient with vertebral level as a nested effect within patient. The effect of three covariates (chemotherapy, bisphosphonates, and radiation fractionation scheme) were evaluated because of their presumed effect on bone mineral density, by adding them as a fixed variable, and in an interaction term with time and a three-way interaction term with time and radiation .

A two-tailed P -value of less than .05 was considered statistically significant. All statistical analyses were performed using Python programming language, version 3.9.7 (Python Software Foundation, https://www.python.org/ ) and the MixedLM procedure in the Statsmodels package. 19

Results

Patient Cohort

The mean age of patients at the time of radiation was 64 years (standard deviation (SD) 11.7) and 66% were male ( Table 1 ). ISS 1 was the most common disease stage (n = 47, 39.5%), most patients were capable of self-care (ECOG scores 0-2; n = 106, 89.8%) and fully neurologically intact (ASIA E, n = 96, 80.7%). Forty-two patients (36.2%) presented with at least one vertebral compression fracture (VCF) in the radiated area, of which 20 were classified as severe according to the Genant criteria. 20 A combination of Revlimid, Velcade and Dexamethasone (RVD) was the most common chemotherapy regimen, used in 56 patients (47.1%, Table 1 ). Sixty-nine patients (65.7%) had already started chemotherapy before initiation of radiotherapy. Bisphosphonates were administered in 78 patients (65.5%), most of which received them in regular intervals (n = 64, 62.1%). The median survival time was 33 months (95%CI 23-46).

Remineralization

Within the radiated vertebrae, there was an initial reduction in bone density in the first 3-6 months following radiation (mean difference (MD) −22.0; P = .004). Between 9 to 18 months, there was an increase in bone density at both timepoints, relative to the baseline (9-12: MD = 8.3; P = .017, and 12-18: MD = 19.9; P = .015, Table 2 ). For the non-targeted vertebrae, there was no initial reduction in bone density in the first 3-6 months compared to baseline (MD = −2.0, P = .910). There was, however, an increase in bone density of both timepoints between 9 to 18 months as well (9-12: MD = 26.0; P = .002, and 12-18: MD = 40.7; P = .006). Comparison of the bone density of the targeted vs non-targeted vertebrae demonstrated no statistically significant difference at any time point ( Table 2 , Figure 1 ). Table 2. Measurements at Baseline and Five Consecutive Timepoints for the targeted (n = 366) and Non-targeted (n = 199) Vertebrae. Targeted (n = 366) Non-targeted (n = 199) Timepoint Mean SD MD P -value* Mean SD MD P -value* MD P -value** Baseline (n = 565, 100%) 196.5 92.8 - - 187.9 92.3 - - 8.5 .296 3-6 months (n = 351, 62.1%) 174.4 88.5 −22.0 .004 186.0 89.5 −2.0 .910 −11.5 .245 6-9 months (n = 174, 30.8%) 211.0 119.1 14.5 .184 177.8 93.1 −10.1 .951 33.2 .056 9-12 months (n = 144, 25.5%) 204.7 115.0 8.3 .017 213.9 126.7 26.0 .002 −9.2 .660 12-18 months (n = 135, 23.9%) 216.4 99.8 19.9 .015 228.7 98.4 40.7 .006 −12.3 .485 18-24 months (n = 80, 14.2%) 197.7 128.9 1.2 .372 184.5 88.4 −3.5 .990 13.3 .629 Abbreviations: SD, standard deviation; MD, mean Difference. Bold P -values indicate statistical significance at P < .05 * P -values for paired t test, evaluating the differences between different timepoints compared to the baseline for both targeted and non-targeted vertebrae. ** P -values for unpaired t-tests, comparing the differences between targeted and non-targeted vertebrae for each timepoint. Figure 1. Change in bone density with 95% confidence intervals after radiation until 24 months.

Linear mixed modeling, assessing the effect of radiation rate over time, demonstrated that radiation was not associated with a greater increase in bone density per month compared to non-targeted (β = .67; 95%CI -.17 to 1.52; P = .118). The β-coefficient for months follow-up shows that the HU of all vertebrae increased over time (β = 1.037; 95%CI .38 to 1.69; P = .002; Table 3 ) regardless of whether the vertebrae were targeted or not. Table 3. Results of Multivariable (Adjusted) linear Mixed Models. Parameter β 95% CI P -value Radiation versus no radiation Difference in △% HU at starting point (3-6 months) −11.98 −21.41 to −2.54 .013 Deviation of targeted vertebrae from non-targeted vertebrae Mean remineralization rate (per month) 1.04 .38 to 1.69 .002 Both targeted and non-targeted vertebrae Difference in remineralization rate .67 −.17 to 1.52 .118 Deviation of targeted vertebrae from non-targeted vertebrae Covariate analysis β 95% CI P -value Chemotherapy pre-RT versus post-RT or no chemotherapy ( Figure 2A ) Difference in △% HU at starting point (3-6 months) −.05 −16.37 to 16.27 .995 When chemotherapy was started pre-RT - all vertebrae Difference in mean remineralization rate (per month) 1.53 −.04 to 3.10 .056 When chemotherapy was started pre-RT - all vertebrae Difference in remineralization rate (per month) −.28 −2.13 to 1.58 .772 When chemotherapy was started pre-RT - targeted vertebrae Bisphosphonates versus no bisphosphonates ( Figure 2B ) Difference in △% HU at starting point (3-6 months) 11.19 −4.43 to 26.81 .160 When bisphosphonates are administered - all vertebrae Difference in mean remineralization rate (per month) .47 −1.11 to 2.05 .560 When bisphosphonates are administered - all vertebrae Difference in remineralization rate (per month) .66 −1.20 to 2.52 .486 When bisphosphonates are administered - targeted vertebrae Multi-fraction versus single fraction ( Figure 2 C) Difference in △% HU at starting point (3-6 months) −4.51 −28.56 to 19.55 .713 When multi-fraction is used - all vertebrae Difference in mean remineralization rate (per month) .63 1.49 to 2.76 .559 When multi-fraction is used - all vertebrae Difference in remineralization rate (per month) 1.21 1.52 to 3.93 .867 When multi-fraction is used - targeted vertebrae Abbreviations: CI, Confidence Interval; HU, Hounsfield Units; RT, radiotherapy. Bold P -values indicate statistical significance at P < .05.

Covariates

When patients were started on chemotherapy before radiation compared to when patients received no chemotherapy or this was started after radiation, there was an increase in bone density around 9-12 months for both radiated and non-targeted vertebrae ( Figure 2A ). Receiving bisphosphonates also seemed to increase bone density in all vertebrae, regardless of radiation ( Figure 2B ). A multi-fraction radiation scheme resulted in a higher BMD around 9-18 months for targeted vertebrae, but it did not seem to affect BMD for non-targeted vertebrae ( Figure 2C ). When calculating remineralization slopes for these three covariates, there was no difference in the remineralization rate over time between radiated or non-targeted vertebrae or within radiated vertebrae ( Table 3 ). Figure 2. Bone density over time, stratified for different covariates.

Retreatments

Within 2 years following radiation therapy, a subset of our study cohort underwent subsequent treatments, including additional radiation therapy or surgical interventions. Of the analyzed patients, 15 (12.6%) received further treatment after the follow-up CT scan. Specifically, eight patients underwent kyphoplasty due to fracture or persistent pain, four patients underwent laminectomy (two with fusion and two without fusion), and three patients received a second course of radiotherapy.

Role of the Funding Source

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Discussion

Multiple myeloma often causes osteolytic lesions in the spine. 3 These destructive lesions change the biomechanical nature of the spine and have been known to result in painful compression fractures as well as unstable patterns of injury. 5 Treatment for these spinal lesions is multidisciplinary in nature often involving both systemic therapies as well as radiotherapy. 21 There is little knowledge regarding the effect of available treatments on the structural stability of the spine. Data regarding the structural integrity of the spinal column in the setting of MM lesions could potentially be helpful for both prognostication and treatment decision purposes. Therefore, in this study, we set out to measure the influence of radiation on remineralization of radiated MM vertebrae as measured by bone density in HU on CT scans before and after irradiation. Our results demonstrate that following radiation, BMD increased over time for both targeted and non-targeted vertebrae. However, no conclusive evidence was found that targeted vertebrae have a higher remineralization rate than non-targeted vertebrae in patients with MM.

The process of bone modeling and remodeling is primarily regulated by the RANKL/RANK/OPG pathway ( Figure 3A ). 22 - 24 The primary roles of receptor activator of nuclear factor-kappa B (RANK) and its ligand, RANKL, (two proteins of the tumor necrosis factor [TNF] superfamily) are regulating bone remodeling by promoting osteoclast differentiation and activation of T cells. 23 Osteoprotegerin (OPG) is released from osteoblasts and regulates the bone remodeling process by inhibiting osteoclast activation and development through blocking the binding of RANK and RANKL. 23 , 24 MM patients have decreased OPG concentrations in the bone marrow micro-environment, caused by intracellular interactions with osteoblasts and downregulating OPG release by bone marrow stem cells. 24 In addition, myeloma cells show upregulated RANKL expression. 24 , 25 Furthermore, B-cell maturation antigen (BCMA) is a transmembrane glycoprotein (also part of the TNF superfamily) which is highly overexpressed on the surface of plasma cells in MM patients. 26 , 27 Along with other proteins, BCMA normally supports B-cell proliferation, survival and maturation and differentiation into long-lived plasma cells. However, in MM it promotes proliferation and survival of malignant myeloma cells. Additionally, BCMA interacts with osteoclasts in the bone marrow microenvironment, leading to their growth and activation. 27 , 28 The BCMA overexpression and the imbalance of the RANKL/OPG axis promotes pathological activation of osteoclasts and therefore underlies the development of the disease-defining lytic lesions in myeloma patients. 23 Figure 3. A. (1) Binding of RANK-ligand to RANK as its receptor which promotes the transition of macrophage cells to pre-fusion osteoclasts (“osteoclast precursor maturation”). (2) Osteoprotegerin (OPG) provides an alternative binding site for RANKL and acts as a decoy receptor by blocking the RANK/RANKL binding. 3) Fusion of multiple mononucleated pre-fusion osteoclasts to form a multinucleated osteoclast is blocked in the absence of enough stimulation with RANK. Consequently, there are less activated osteoclasts to resorb bone. 4) In MM, there are decreased OPG concentrations in the bone marrow and upregulated RANKL, which causes increased formation of activated mononucleated pre-fusion osteoclasts. 5) Fusion of multiple mononucleated pre-fusion osteoclasts to form a multinucleated osteoclast. 6) Multinucleated osteoclasts are activated and can start resorbing bone. a) BCMA expression on plasma cells: normally supports B-cell proliferation, survival and maturation and differentiation into long-lived plasma cells. In MM it promotes proliferation and survival of malignant myeloma cells. Figure created with https://www.biorender.com/ . B. Illustration of how radiation affects bone (remodeling) . 1a) Radiation induces higher concentrations of multiple inflammatory cytokines (e.g. IL-1,6,7,11) and TNF-α shortly after radiation exposure, leading to systemic activation of osteoclastogenesis. 1b) Cell cycle arrest (apoptosis) and reduced proliferation rate of osteoblasts after radiation, leading to decreased OPG which subsequently prevents inhibition of RANKL. 2a) After initial reaction to radiation, healthy bone reinitiates osteoblast activity, which restores normal OPG concentrations, and consequently normalizes osteoclast activity. 2b) Osteoblasts fill up Howship Lacunae with unmineralized matrix which hardens when minerals (i.e. calcium) are deposited in it. 3) Normal bone homeostasis restores after several cycles. 4) In MM, the pre-existing imbalance of the RANKL/OPG axis and overexpression of BCMA cause prolonged osteoclastic activity and thereby delayed neutralization to normal bone hemostasis. Figure created with https://www.biorender.com/ .

The primary goal of radiation in patients with MM is to induce cell apoptosis of malignant plasma cells or deprive them of their multiplication potential to slow progression and achieve local control. 29 Histomorphometric analyses of bone after radiation have revealed higher concentrations of multiple inflammatory cytokines (e.g. IL-1,6,7,11) and TNF-α in targeted and non-targeted contralateral bone shortly after radiation exposure, leading to systemic activation of osteoclastogenesis ( Figure 3B ). 30 - 33 Additionally, some studies have observed cell cycle arrest and reduced proliferation rate of osteoblasts after radiation, leading to decreased OPG which subsequently prevents inhibition of RANKL. 30 , 33 - 35 These factors lead to an initial decrease in BMD, which is in line with our results and other studies observing the same effect in both murine models 30 , 34 and in humans. 12 , 36 After this initial dip, bone homeostasis should be restored back to normal after several cycles and osteoblast production is reinitiated. When tumor infiltration in vertebrae is reduced and osteoclasts have excavated an area of bone ( Howship lacunae ), osteoblasts fill up this bone resorption pit with unmineralized matrix which hardens when minerals (i.e. calcium) are deposited in it. 37 Return to full density may take up to 6 months. 38

A relative increase in bone density compared with the baseline can occur when this remineralization process happens in a tumor cavity. 33 Many authors investigating bone density after radiation in patients with other spinal metastatic diseases have demonstrated this positive effect on BMD and found a greater increase in bone density in targeted vertebrae compared with non-targeted vertebrae. 12 - 14 , 39 However, due to the overexpression of BCMA and imbalance of the RANKL/OPG axis, neutralization of bone homeostasis is delayed in MM 40 and therefore presumably less effective compared to bone metastases originating from non-B-cell derived malignancies. We presume that this is the reason we did not find a difference in remineralization between targeted and non-targeted vertebrae for patients with MM. Notably, within our initial study cohort, a relatively high proportion of patients (n = 25, 17.6%) required surgery within the first 3-6 months post-radiation (and were therefore excluded from further analyses). While the exact reasons for these surgeries were not uniformly documented, the observed dip in BMD during the initial post-radiation period could potentially contribute to the need for surgical interventions.

Our findings show that both targeted and non-targeted vertebrae show remineralization after radiation and a pronounced increase in absolute bone density compared with the baseline. Targeted vertebrae had a slightly higher correlation coefficient (slope) for remineralization effect per month compared with non-targeted vertebrae, although this difference was not statistically significant. These findings are in line with the aforementioned systemic effect of radiation in MM patients. Previous researchers have investigated this systemic effect by measuring targeted and contralateral-shielded bone in mice. 30 , 34 , 41 , 42 In these models, it is evident that radiation causes increased osteoclast numbers in both targeted and contralateral-shielded (non-targeted) bone sites due to the systemic activation of osteoclastogenesis. 30 - 33 Additionally, while we did take into account the possible imprecision of radiation delivery and the scattered effect of radiation by taking the first non-adjacent vertebrae (skipping one vertebral level from the targeted site), it cannot be ruled out that these ‘control-vertebrae’ did receive some radiation.

Systemic treatment of MM has improved greatly within the recent years and the emergence of novel therapies have changed the therapeutic landscape for MM. While the disease remains incurable, BCMA-targeted (immune) therapies like Chimeric Antigen Receptor (CAR) T-cell treatments have achieved remarkable clinical success in MM, especially for patients who are refractory to standard treatment. 7 Research has shown a rapid decrease of malignant plasma cells in the bone marrow and a remarkable decrease in serum BCMA expression following CAR T-cell treatment. 43 Bisphosphonate treatment (e.g., zoledronate) has been demonstrated to decrease the occurrence of skeletal-related events (such as spinal cord compression, pathological fractures, need for radiation and need for surgery) in patients with myeloma bone disease. 6 Bisphosphonates protect the bone from bone resorption by promoting apoptosis of osteoclasts. 23 In addition, denosumab is becoming increasingly common in the treatment of myeloma bone disease. Denosumab is a monoclonal antibody that had already proven to reduce the frequency in skeletal-related events in patients with advanced solid tumors, and more recent clinical trials have suggested its value in myeloma treatment as well by blocking the RANK/RANKL interaction. 23 , 44 Additionally, it is not known to be nephrotoxic. 44 In short, BCMA-targeted therapies, denosumab and bisphosphonates inhibit bone resorption by osteoclasts and therefore support bone remineralization after radiation. In the current study, our data demonstrate that both radiated and non-targeted vertebrae show remineralization after radiation, which could be the result of systemic treatment. Although the study results suggest a positive effect of radiation on BMD locally and systemically, it is possible that systemic treatment contributes more to this effect. Additionally, the emergence of novel systemic therapies has opened new perspectives for the concomitant administration of radiation and systemic drugs, as possible synergy has been observed in pre-clinical studies in MM. 45 In terms of future research, it would be useful to extend the current findings by evaluating bone density of patients who received novel BCMA-targeted therapies without radiation to any location.

Another explanation for the increased bone density in both targeted and non-targeted vertebrae, is that patients who received radiotherapy experience less pain after radiation, 4 and therefore could become more physically active. Physical activity induces more mechanical stress on the bone, which is important for regulation of the bone metabolism and promotion of bone growth, according to Wolff’s law. 37 , 38 In this study, level of activity was not measured, so no comparative analyses can be done to determine the effect of physical activity on BMD. It would be interesting to include activity level in any further studies investigating BMD after radiation.

Limitations and Recommendations

There are several limitations to this study. First, due to the retrospective nature of this study, CT-scans were not acquired at specific, predetermined timepoints or according to set protocols. In some patients, CT simulation scans acquired just prior to radiation were not available for baseline measurements. In those cases, we had to use a CT scan from an earlier moment. This increases the risk of missing any bone breakdown between that CT scan and the start of radiation. We did ensure that the pre-radiation CT scan was only compared with a follow-up CT using the same protocol. Second, it is still unclear what the effect of systemic therapy is on remineralization. Prospective studies with larger sample sizes and including a group of patients receiving only systemic therapy and a group receiving only radiation and no systemic therapy would be needed to confirm the exact relationship between radiation and bone strength. Unfortunately, data regarding the use of Denosumab was not available so an analysis of the effect of Denosumab was impossible. Nevertheless, despite these limitations, this study is the first to investigate bone density after radiation in patients with MM, using a relatively large patient population.

Lastly, our data revealed that a notable proportion of patients required subsequent surgeries or reirradiation within 2 years post-radiation, underscoring the complex and evolving nature of clinical care in MM-related spinal lesions. These findings highlight the ongoing need for a multidisciplinary approach, including spine surgeons, in the management of MM patients following radiation therapy. Radiation therapy should only be used after careful patient selection, considering its potential benefits (pain reduction) and limitations in bone remineralization. Close follow-up post-radiation is essential to monitor for any emerging complications, including fractures, allowing for timely intervention if needed.

Conclusion

The current study investigated vertebral remineralization rate in patients with MM undergoing radiation therapy. Our data showed that in the early period after radiation, targeted vertebrae demonstrated a decline in bone density compared with non-targeted vertebrae. However, after nine months, both targeted and non-targeted vertebrae showed increases in bone density, with no additional effect of radiation. This could be explained by a systemic effect of local radiation, a higher activity level of patients after radiation or the effect of (newer) systemic therapeutic approaches. The results can be useful for guiding patient selection for the use of radiation in the treatment of MM and highlights the need for future research for evaluating bone remineralization in patients who received novel targeted therapies without radiation to any location in a prospective study with controlled treatment protocols.

Supplemental Material

Supplemental Material - Remineralization Rate of Lytic Lesions of the Spine in Multiple Myeloma Patients Undergoing Radiation Therapy

Supplemental Material for Remineralization Rate of Lytic Lesions of the Spine in Multiple Myeloma Patients Undergoing Radiation Therapy by Hester Zijlstra, Jens P. te Velde, Brendan M. Striano, Olivier Q. Groot, Tom M. de Groot, Noopur Raje, Chirayu Patel, Jad Husseini, Diyar Delawi, Diederik H. R. Kempen, Jorrit-Jan Verlaan, and Joseph H. Schwab in Global Spine Journal

Acknowledgments

We thank Mr. D.L. Hayden, PhD and Mr. H. Lee, PhD, from the MGH Biostatistics Center for their support and advice on design and statistical analysis.

CRediT Roles: Conceptualization: HZ, JH, DD, DHRK, JJV and JHS; Data curation: HZ and JPV; Formal analysis: HZ; Investigation: HZ; Methodology: HZ, JH and JHS; Supervision: BMS, OQG, NR, CP, JH, DD, DHK, JJV and JHS; Visualization: HZ and TG; Writing – original draft: HZ; Writing - review & editing: BMS, OQd, TdG, NR, CP, JH, DD, DHK, JJV and JHS.

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material: Supplemental material for this article is available online.

Ethical Statement

Ethical Approval

Investigation performed at Massachusetts General Hospital, Boston USA. Local Institutional Review Board (IRB) approval was obtained for this study (registration number 2018P000688).

ORCID iDs

Hester Zijlstra https://orcid.org/0000-0003-3962-6871

Olivier Q. Groot https://orcid.org/0000-0002-5091-3110

References

1 Mitsiades CS Mitsiades N Munshi NC Anderson KC . Focus on multiple myeloma . Cancer Cell . 2004 ; 6 ( 5 ): 439 - 444 . doi: 10.1016/j.ccr.2004.10.020 15542427

2 Padala SA Barsouk A Barsouk A , et al. Epidemiology, staging, and management of multiple myeloma . Med Sci . 2021 ; 9 ( 1 ): 3 . doi: 10.3390/MEDSCI9010003 PMC7838784 33498356

3 Kim C Bhatta S Cyprien L Fonseca R Hernandez RK . Incidence of skeletal-related events among multiple myeloma patients in the United States at oncology clinics: Observations from real-world data . J Bone Oncol . 2019 ; 14 : 100215 . doi: 10.1016/j.jbo.2018.100215 30666288 PMC6329702

4 Tsang RW Campbell BA Goda JS , et al. Radiation therapy for Solitary Plasmacytoma and multiple myeloma: Guidelines from the international Lymphoma radiation oncology group . Int J Radiat Oncol Biol Phys . 2018 ; 101 ( 4 ): 794 - 808 . doi: 10.1016/j.ijrobp.2018.05.009 29976492

5 Sonmez M Akagun T Topbas M , et al. Effect of pathologic fractures on survival in multiple myeloma patients: a case control study . J Exp Clin Cancer Res . 2008 ; 27 ( 1 ): 11 . doi: 10.1186/1756-9966-27-11 18577267 PMC2438338

6 Suvannasankha A Grethlein SJ . Multiple myeloma treatment protocols. Medscape . Published May 11 , 2021. Accessed October 13, 2021 . https://emedicine.medscape.com/article/2006866

7 Rodríguez-Lobato LG Ganzetti M Fernández de Larrea C Hudecek M Einsele H Danhof S . CAR T-cells in multiple myeloma: State of the art and future directions . Front Oncol . 2020 ; 10 : 1243 . doi: 10.3389/fonc.2020.01243 32850376 PMC7399644

8 Nehlsen AD Sindhu KK Moshier E , et al. The safety and efficacy of radiation therapy with concurrent dexamethasone, cyclophosphamide, etoposide, and cisplatin-based systemic therapy for multiple myeloma . Clin Lymphoma Myeloma Leuk . 2022 ; 22 : 192 - 197 . doi: 10.1016/j.clml.2021.09.015 34736880 PMC9040190

9 Momm F Greil C Schäfer H . Towards individualized radiation therapy in multiple myeloma . Haematologica . 2020 ; 105 ( 7 ): 1763 - 1764 . doi: 10.3324/haematol.2019.243451 32611577 PMC7327648

10 Elhammali A Amini B Ludmir EB , et al. New paradigm for radiation in multiple myeloma: lower yet effective dose to avoid radiation toxicity . Haematologica . 2020 ; 105 ( 7 ): e355 - e357 . doi: 10.3324/haematol.2019.235804 31919088 PMC7327656

11 Pielkenrood BJ Visser TF van Tol FR , et al. Remineralization of lytic spinal metastases after radiotherapy . Spine J . 2023 ; 23 : 571 - 578 . doi: 10.1016/j.spinee.2022.12.018 36623735

12 Koswig S Budach V . [Remineralization and pain relief in bone metastases after after different radiotherapy fractions (10 times 3 Gy vs. 1 time 8 Gy). A prospective study] . Strahlenther Onkol . 1999 ; 175 ( 10 ): 500 - 508 . doi: 10.1007/s000660050061 10554645

13 Wang Q Sun B Meng X Liu C Cong Y Wu S . Density of bone metastatic lesions increases after radiotherapy in patients with breast cancer . J Radiat Res . 2019 ; 60 ( 3 ): 394 - 400 . doi: 10.1093/jrr/rry098 30785994 PMC6530615

14 Foerster R Eisele C Bruckner T , et al. Bone density as a marker for local response to radiotherapy of spinal bone metastases in women with breast cancer: a retrospective analysis . Radiat Oncol . 2015 ; 10 ( 1 ): 62 - 69 . doi: 10.1186/s13014-015-0368-x 25884703 PMC4355128

15 Foerster R Hees K Bruckner T , et al. Survival and stability of patients with urothelial cancer and spinal bone metastases after palliative radiotherapy . Radiol Oncol . 2018 ; 52 ( 2 ): 189 - 194 . doi: 10.1515/raon-2017-0038 30018523 PMC6043884

16 Matuschek C Ochtrop TA Bölke E , et al. Effects of radiotherapy in the treatment of multiple myeloma: a retrospective analysis of a single institution . Radiat Oncol . 2015 ; 10 ( 1 ): 71 - 79 . doi: 10.1186/s13014-015-0374-z 25889851 PMC4391087

17 Koo TK Li MY . A guideline of selecting and reporting intraclass correlation coefficients for reliability research . J Chiropr Med . 2016 ; 15 ( 2 ): 155 - 163 . doi: 10.1016/j.jcm.2016.02.012 27330520 PMC4913118

18 Introduction to Linear Mixed Models . UCLA: Statistical Consulting Group . Accessed November 8, 2022 . https://stats.oarc.ucla.edu/other/mult-pkg/introduction-to-linear-mixed-models/

19 Seabold S Perktold J . Statsmodels: Econometric and statistical modeling with python . In: 9th Python in Science Conference ; 2010 .

20 Genant HK Wu CY Van Kuijk C Nevitt MC . Vertebral fracture assessment using a semiquantitative technique . J Bone Miner Res . 1993 ; 8 ( 9 ): 1137 - 1148 8237484 10.1002/jbmr.5650080915

21 Terpos E Zamagni E Lentzsch S , et al. Treatment of multiple myeloma-related bone disease: recommendations from the bone working group of the international myeloma working group . Lancet Oncol . 2021 ; 22 ( 3 ): e119 - e130 . doi: 10.1016/S1470-2045(20)30559-3 33545067

22 Tobeiha M Moghadasian MH Amin N Jafarnejad S . RANKL/RANK/OPG pathway: a mechanism involved in Exercise-induced bone remodeling . BioMed Res Int . 2020 ; 2020 : 6910312 . doi: 10.1155/2020/6910312 32149122 PMC7053481

23 Raje NS Bhatta S Terpos E . Role of the RANK/RANKL pathway in multiple myeloma . Clin Cancer Res . 2019 ; 25 ( 1 ): 12 - 20 . doi: 10.1158/1078-0432.CCR-18-1537 30093448

24 Fili S Karalaki M Schaller B . Therapeutic implications of osteoprotegerin . Cancer Cell Int . 2009 ; 9 : 26 . doi: 10.1186/1475-2867-9-26 19747396 PMC2754428

25 Spanoudakis E Papoutselis M Terpos E , et al. Overexpression of RANKL by invariant NKT cells enriched in the bone marrow of patients with multiple myeloma . Blood Cancer J . 2016 ; 6 ( 11 ): e500 . doi: 10.1038/bcj.2016.108 27834938 PMC5148055

26 Guo R Lu W Zhang Y Cao X Jin X Zhao M . Targeting BCMA to Treat multiple myeloma: Updates from the 2021 ASH Annual Meeting . Front Immunol . 2022 ; 13 : 839097 . doi: 10.3389/fimmu.2022.839097 35320942 PMC8936073

27 Cho SF Anderson KC Tai YT . Targeting B cell maturation antigen (BCMA) in multiple myeloma: potential uses of BCMA-based Immunotherapy . Front Immunol . 2018 ; 9 : 1821 . doi: 10.3389/fimmu.2018.01821 30147690 PMC6095983

28 Kleber M Ntanasis-Stathopoulos I Terpos E . BCMA in multiple myeloma—a Promising Key to therapy . J Clin Med . 2021 ; 10 ( 18 ): 4088 . doi: 10.3390/jcm10184088 34575199 PMC8472544

29 Talamo G Dimaio C Abbi KKS , et al. Current role of radiation therapy for multiple myeloma . Front Oncol . 2015 ; 5 : 40 . doi: 10.3389/fonc.2015.00040 25741475 PMC4332323

30 Wright LE Buijs JT Kim HS , et al. Single-limb Irradiation induces local and systemic bone loss in a murine model . J Bone Miner Res . 2015 ; 30 ( 7 ): 1268 - 1279 . doi: 10.1002/jbmr.2458 25588731 PMC4478128

31 D’Amico L Roato I . Cross-talk between T cells and osteoclasts in bone resorption . BoneKEy Rep . 2012 ; 1 : 82 . doi: 10.1038/bonekey.2012.82 23951473 PMC3727728

32 Costa S Reagan MR . Therapeutic irradiation: Consequences for bone and bone marrow adipose tissue . Front Endocrinol . 2019 ; 10 : 587 . doi: 10.3389/fendo.2019.00587 PMC6727661 31555210

33 Donaubauer AJ Deloch L Becker I Fietkau R Frey B Gaipl US . The influence of radiation on bone and bone cells—Differential effects on osteoclasts and osteoblasts . Int J Mol Sci . 2020 ; 21 ( 17 ): 6377 . doi: 10.3390/ijms21176377 32887421 PMC7504528

34 Zou Q Hong W Zhou Y , et al. Bone marrow stem cell dysfunction in radiation-induced abscopal bone loss . J Orthop Surg Res . 2016 ; 11 ( 1 ): 3 . doi: 10.1186/s13018-015-0339-9 26739584 PMC4704383

35 Chandra A Lin T Tribble MB , et al. PTH1-34 Alleviates radiotherapy-induced local bone loss by improving osteoblast and Osteocyte survival . Bone . 2014 ; 67 : 33 - 40 . doi: 10.1016/j.bone.2014.06.030 24998454 PMC4154509

36 Reinbold WD Wannenmacher M Hodapp N Adler CP . Osteodensitometry of vertebral metastases after radiotherapy using quantitative computed tomography . Skeletal Radiol . 1989 ; 18 ( 7 ): 517 - 521 . doi: 10.1007/BF00351751 2588030

37 Rowe P Koller A Sharma S . Physiology, bone remodeling . In: StatPearls . StatPearls Publishing ; 2022 . https://www.ncbi.nlm.nih.gov/books/NBK499863/ Accessed 18 November 2022. 29763038

38 Croarkin E . Osteopenia in the patient with cancer . Phys Ther . 1999 ; 79 ( 2 ): 196 - 201 . doi: 10.1093/ptj/79.2.196 10029059

39 Lang K König L Bruckner T , et al. Stability of spinal bone lesions in patients with multiple myeloma after radiotherapy—a retrospective analysis of 130 cases . Clin Lymphoma Myeloma Leuk . 2017 ; 17 ( 12 ): e99 - e107 . doi: 10.1016/j.clml.2017.09.005 28993162

40 Schmiedel BJ Scheible CA Nuebling T , et al. RANKL expression, function, and therapeutic targeting in multiple myeloma and chronic lymphocytic leukemia . Cancer Res . 2013 ; 73 ( 2 ): 683 - 694 . doi: 10.1158/0008-5472.CAN-12-2280 23139212

41 Oest ME Policastro CG Mann KA Zimmerman ND Damron TA . Longitudinal effects of single Hindlimb radiation therapy on bone strength and morphology at local and contralateral sites . J Bone Miner Res . 2018 ; 33 ( 1 ): 99 - 112 . doi: 10.1002/jbmr.3289 28902435 PMC5776033

42 Zhai J He F Wang J Chen J Tong L Zhu G . Influence of radiation exposure pattern on the bone injury and osteoclastogenesis in a rat model . Int J Mol Med . 2019 ; 44 ( 6 ): 2265 - 2275 . doi: 10.3892/ijmm.2019.4369 31638191 PMC6844641

43 Raje N Berdeja J Lin Y , et al. Anti-BCMA CAR T-cell therapy bb2121 in Relapsed or refractory multiple myeloma . N Engl J Med . 2019 ; 380 ( 18 ): 1726 - 1737 . doi: 10.1056/NEJMoa1817226 31042825 PMC8202968

44 Raje N Terpos E Willenbacher W , et al. Denosumab versus zoledronic acid in bone disease treatment of newly diagnosed multiple myeloma: an international, double-blind, double-dummy, randomised, controlled, phase 3 study . Lancet Oncol . 2018 ; 19 ( 3 ): 370 - 381 . doi: 10.1016/S1470-2045(18)30072-X 29429912

45 Lee YS Heo W Nam J Jeung YH Bae J . The combination of ionizing radiation and proteasomal inhibition by bortezomib enhances the expression of NKG2D ligands in multiple myeloma cells . J Radiat Res . 2018 ; 59 ( 3 ): 245 - 252 . doi: 10.1093/jrr/rry005 29518205 PMC5967576