Impact of CDK4/6 Inhibitors on Aromatase Inhibitor-Associated Musculoskeletal Syndrome (AIMSS) in the Adjuvant Setting

Background Third-generation aromatase inhibitors (AIs) are the mainstay of treatment in hormone receptor (HR)-positive breast cancer. Even though it is considered to be a well-tolerated therapy, AI-induced musculoskeletal symptoms are common and may be accused for treatment discontinuation. Recently, selective cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors changed the therapeutic setting, and currently, ribociclib, palbociclib, and abemaciclib are all approved in combination with nonsteroidal AIs in patients with ER-positive, HER2-negative advanced or metastatic breast cancer. This systematic review aims to identify the frequency of aromatase inhibitor-associated musculoskeletal syndrome (AIMSS) in the adjuvant setting in patients under AI monotherapy compared to patients under combination therapy with AIs and CDK4/6 inhibitors and demonstrate the underlying mechanism of action. Methods This study was performed in accordance with PRISMA guidelines. The literature search and data extraction from all randomized clinical trials (RCTs) were done by two independent investigators. Eligible articles were identified by a search of MEDLINE and ClinicalTrial.gov database concerning the period 2000/01/01–2021/05/01. Results Arthralgia was reported in 13.2 to 68.7% of patients receiving AIs for early-stage breast cancer, while arthralgia induced by CDK4/6 inhibitors occurred in a much lower rate [20.5–41.2%]. Bone pain (5–28.7% vs. 2.2–17.2%), back pain (2–13.4% vs. 8–11.2%), and arthritis (3.6–33.6% vs. 0.32%) were reported less frequently in patients receiving the combination of CDK4/6 inhibitors with ET. Conclusions CDK4/6 inhibitors might have a protective effect against joint inflammation and arthralgia occurrence. Further studies are warranted to investigate arthralgia incidence in this population.


Introduction
Breast cancer is the most common malignancy in females afecting more than 2,000,000 women each year [1]. Clinical practice has changed substantially in recent decades leading to a more prolonged survival. At the time of diagnosis, approximately 64% of patients have localized disease, while 6% of patients have already developed metastases. Te 5-year survival rate of metastatic disease is estimated to be around 27% [2]. Around 75% of all breast tumors express the estrogen receptor (ER) and/or the progesterone receptor (PgR) and are considered as hormone receptor-positive (HR) tumors [3]. HR-positive breast cancer demonstrates the best survival rates (90.3%-92.5% 4-year survival rate) compared with epidermal growth factor receptor 2 (HER2)-positive tumors (82.7% 4-year survival rate) and triple-negative subtype (survival rate 77.0%) [4].
Adjuvant endocrine therapy remains the mainstay of HR-positive early breast cancer along with chemotherapy and adjuvant radiation therapy. For postmenopausal women, next generation aromatase inhibitors (AIs) nonsteroidal (anastrozole and letrozole) or steroidal (exemestane) are widely used in the adjuvant setting [5][6][7]. However, administration of AIs has been associated with joint pain and musculoskeletal symptoms that can even lead to treatment discontinuation. Aromatase inhibitor-associated musculoskeletal syndrome (AIMSS) is the most common adverse event encountered by breast cancer patients. In early-stage breast cancer, an incidence of 47% of arthralgias was reported in postmenopausal women treated with AI, including a 23.5% rate of new-onset arthralgia and a 23.5% rate of exacerbation of pre-existing arthralgias [8]. Another recent meta-analysis showed a 17.9% rate of AI-induced arthralgia in postmenopausal women which was lower in early-stage breast cancer compared with advanced disease [9]. Overall, approximately 13-25% of patients discontinue AI therapy mainly because of arthralgias [10,11].
Te development of selective cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors which regulate cell cycle progression has altered daily clinical practice. Currently, ribociblib (LEE011), palbociclib (PD0332991), and abemaciclib (LY2835219) are all approved in combination with nonsteroidal AIs as frst-line therapy of ER-positive, HER2negative advanced or metastatic breast cancer. Results from Phase III studies showed a 40-45% improvement in progression-free survival (PFS) from the addition of CDK4/6 inhibitors in the metastatic setting [12]. CDK4/6 inhibitors are currently under investigation in the adjuvant setting as well. Adverse events commonly reported after CDK4/6 inhibitor administration includes neutropenia (75-80%) upon ribociclib or palbociclib treatment, elevated amino-transferases (15%), and prolongation of corrected QT interval by Fredericia (QTcF) (3.3%) with ribociclib and diarrhea (81.3%) upon abemaciclib treatment. Remarkably, musculoskeletal symptoms have not been linked with CDK4/6 treatment despite the coadministration with AIs. Whether CDK4/6 inhibitors afect the arthralgia rate reported in Phase III trials of AIs remains unknown. We conducted this systematic review to identify the rate of musculoskeletal manifestations induced by AIs and CDK4/ 6 inhibitors in the adjuvant setting and the potential efect of CDK4/6 inhibitors on joint symptoms. We have previously published the musculoskeletal toxicity of CDK4/6 and aromatase inhibitors in the metastatic setting, so we are now presenting the existing data in the adjuvant setting [13].

Materials and Methods
Tis systematic review was performed in accordance with PRISMA guidelines [14]. Te protocol of this systematic review has been submitted to the Institutional Review Board of Alexandra Hospital and Medical School of Athens and is available upon request. All eligible studies were identifed using MEDLINE and ClinicalTrial.gov database for search concerning the period 2000/01/01-2021/05/01. Te search algorithm applied consisted of the following words: (breast AND (cancer OR neoplasm) and (aromatase inhibitors OR letrozole OR anastrozole OR exemestane OR CDK4/6 OR palbociclib OR ribociclib OR abemaciclib) AND (phase III) AND (adjuvant). Language restrictions were not applied. In order to maximize the amount of synthesized information, we systematically examined the reference lists of the articles retrieved for potentially eligible studies.
Eligible studies included all randomized controlled Phase III trials exploring third-generation AI monotherapy treatment in early-stage breast cancer in postmenopausal women and all randomized controlled Phase III trials evaluating AIs in combination with CDK4/6 inhibitors in the adjuvant setting. Phase I-II trials, case reports, and reviews were excluded. Trials of AIs in combination with other antineoplastic drugs, e.g., trastuzumab, or in combination with bone-protective therapies [15][16][17][18][19] were also excluded. In case of overlapping publications emerging from the same study, the larger sample size study was evaluated, e.g., TEAM trial and N-SAS BC 04 trial [6,20]. In case of additional information provided from multiple papers from the same trial, each article was evaluated separately [5,21].
From each of the eligible studies, the following data were extracted: frst author, year of publication, trial number, treatment arms, sample size, median age, median follow-up time, disease-free survival, overall survival, arthralgia rate, myalgia rate, backpain rate, bone pain rate, arthritis rate, osteoporosis, and osteoporotic fracture rates. Two investigators (ES and AA), working independently, searched the literature, and extracted data from each eligible study. Only data when AIs were administered as monotherapy treatment were considered eligible, and thus, data emerging from the sequential administration of tamoxifen and AIs were not included in our fnal results. In addition, whenever study results were available on Clinical.Trials.gov, results extracted from eligible articles were updated according to Clinical.Trials.gov data. Any diferences in extracted data were resolved via within-pair consensus. Odds Ratio (OR) with 95% confdence interval (CI) was applied as an efect estimate of arthralgia occurrence when CDK4/6 inhibitors were added to ET therapy. Pooled OR with 95% CI was calculated using the random-efects model due to the underlying variations across the included trials. Heterogeneity across included trials was assessed using I 2 and Q statistics, and signifcant heterogeneity was defned as I 2 > 50.0% or P < 0.10. Statistical analysis was conducted using Review Manager (RevMan) v.5.4.1 software.

Arthralgia.
In the adjuvant setting, the reported percentage of arthralgias in patients under monotherapy with AIs ranged from 13.2 to 68.7%, while in patients receiving therapy with tamoxifen, the reported arthralgia rates ranged from 11.8 to 31.8% and in patients receiving placebo from 15 to 50% (Table 1). Te highest rates of arthralgias were observed in studies when 4-6 years of AI treatment had preceded before enrolment, such as MA.17R study and SOLE study where arthralgia was reported in 53% and 68.7% of patients, respectively [32,36]. Anastrozole and letrozole were associated with approximately the same arthralgia rate (48.2% vs. 47.9%, respectively) according to the FACE trial [28]. More specifcally, letrozole monotherapy induced arthralgia with an incidence of 13.2-68.7%, while the incidence of arthralgia in anastrozole and exemestane monotherapy was 24.5-57.7% and 24-55.4% respectively. Overall, the occurrence of exemestane-induced arthralgia is lower than those of letrozole and anastrozole.

Myalgias.
In the adjuvant setting, the incidence of myalgia reported in 9 of RCT was 7.5-37.1% in patients receiving AIs as monotherapy, 7% in patients receiving tamoxifen, and 6-25% in patients receiving placebo. Te highest myalgia rates were reported in MA.17R trial (28%) and SOLE trial (35.9%/37.1%) where patients enrolled had received 4 to 5 years of previous endocrine therapy with AIs [32,36]. In the rest of RCTs, the myalgia rate reported was lower than 20% (7.5-17.7%). Myalgia incidence was

Bone Pain. Data concerning bone pain emerge from 9
RCTs and afect 5-28.7% of patients receiving AIs in the adjuvant setting, while 6-14% and 16% of patients sufered from bone pain in the placebo and tamoxifen groups, respectively. Te incidence of bone pain in exemestane group was 5.4-10.1%, whereas it was 5-28.7% and 5.9-10.9% in the letrozole and anastrozole groups.

Backpain.
Te percentage of backpain observed in patients treated with AIs in the adjuvant setting was 2-13.4% according to reported data emerging from 5 RCTs. Te incidence of backpain for letrozole was 2-10.3% compared with 9.4-13.4% and 5.7-12.7% reported in anastrozole and exemestane arms.

3.5.
Arthritis. Arthritis is of major concern in patients treated with AIs, and data concerning arthritis were reported in 6 trials. In the adjuvant setting, AIs induced arthritis in 3.6-33.6% of women, which was higher than the 12% and 5-30% rates, reported in tamoxifen monotherapy and placebo groups, respectively. However, the rate of AI-induced arthritis was 6-7% in the majority of RCTs [6,[22][23][24] with the exception of MA.17R trial where AI treatment for 4.5-6 years preceded the enrolment (33%) [36]. Te arthritis rate caused by letrozole was 6-33%, while arthritis was reported in 6.4% of women in anastrozole group and 3.6-7% in exemestane group.
Only one trial has reported the incidence of arthritis induced by CDK4/6 inhibitors in the adjuvant setting [38]. Arthritis was identifed in 0.32% of patients receiving abemaciclib plus ET in the adjuvant setting.

Osteoporosis-Osteoporotic Fractures.
Data concerning osteoporosis and osteoporotic fractures were eligible in 13 and 19 of the 25 RCTs, respectively. In breast cancer patients receiving letrozole, osteoporosis was confronted in 4.8-46.8% of patients. Osteoporosis incidence was 10.9-36.4% and 10-33% in patients receiving anastrozole and exemestane, respectively. Te highest rates of osteoporosis were reported in SOLE trial where about half of patients treated with letrozole for fve years after prior endocrine therapy for 4-6 years developed osteoporosis. Osteoporosis incidence increased consistently with duration of AI treatment according to results of DATA, IDEAL, and GIM4 LEAD trials [29,31,47]. CDK4/6 inhibitors (abemaciclib) plus ET induced osteoporosis less commonly (1.61%) than ET alone (4.8-46.8%) in MonarchE trial [38].
Fractures represent one of the serious adverse events in AI treatment. In the adjuvant setting, incidence of fractures observed in patients treated with AIs ranged from 0.5% to 14%. Te highest fracture rate (14%) was reported in MA.17R trial where patients received AIs for a total of 10 years while the lowest rate (0.5%) was reported in GIM4 LEAD trial where patients received AI treatment for a maximum of 5 years after treatment with tamoxifen for 2 to 3 years [36,47]. In contrast, incidence of fracture was 1-7.7% and 1.1-9% in patients treated with tamoxifen or placebo, respectively. Te incidence of fracture was greater in patients treated with AIs compared with tamoxifen treatment as shown in BIG 1-98, ATAC, TEAM, IES, ABCSG-6a, and ABCSG-8/ARNO-95 trials [5-7, 21, 22, 40, 42, 45, 46]. Te duration of AI treatment seemed to correlate with fracture rate. DATA trial reported an 7.5% and 10% fracture rate in patients treated with anastrozole for 3 and 6 years, respectively [29]. Consistently, patients administered with letrozole for 2.5 and 5 years presented with fractures in a rate of 2.8% and 5% respectively in IDEAL trial [31]. SALSA trial demonstrated an increase from 4% to 6% of fracture rate in patients receiving anastrozole for 2 versus 5 years,  Figure 2: Forest plot of the arthralgia incidence in patients receiving CDK4/6 inhibitors plus ET versus ET alone. 8 Te Breast Journal respectively [48,49]. Overall, letrozole induced fractures in 0.5-14% of patients, whereas the same rates were 0.8-11% and 4-5% for anastrozole and exemestane population.

Carpal Tunnel Syndrome.
Carpal tunnel syndrome was reported in 0-3% of patients treated with AIs in the adjuvant setting in contrast with 0.2-1% incidence reported in tamoxifen arms according to data reported in 4 RCTs. Tere are no data of the incidence of carpal tunnel syndrome in patients receiving CDK4/6 inhibitors in the adjuvant setting.

Discussion
Aromatase inhibitor-induced musculoskeletal syndrome (AIMSS) has emerged as a major cause of treatment discontinuation in hormone receptor-positive patients treated with AIs. In our systematic review of phase III RCTs, arthralgias were reported in 13.  [8]. Consistently, 45% of women with early-stage breast cancers experienced joint symptoms according to results from COBRA trial [50]. Median time to onset of symptoms was 1.6 months although joint symptoms could appear as early as a few days after initiation of treatment [50]. Laroche et al. reported an initial appearance of joint pain after 6 weeks of treatment and a consequent more difuse pain after 12 months of treatment [51]. Whatever the time course, AI-induced musculoskeletal syndrome remains of major clinical signifcance in women with breast cancer. AIs may lead to joint symptoms via three diferent mechanisms. Estrogen deprivation seems to be the most likely mechanism of AIMSS [52]. Treatment with AIs results in a decrease in serum estrogen levels below postmenopausal levels. Te expression of the two estrogen receptors alpha and beta (ERα and ERβ) in human articular cartilage provides evidence that estrogens modulate the metabolism of chondrocytes [53]. Concentrations of ERα and ERβ receptors are increased in men, and thus, male joints may be more protected against the development of arthritis [53]. Animal studies demonstrated that estrogen defciency may accelerate cartilage turnover and increase cartilage surface erosion [54]. Exogenous estrogen or SERM administration suppresses the progression of cartilage erosion exerting a chondroprotective efect on articular cartilage. Moreover, hormone replacement therapy results in a reduction up to three-fold of osteoarthritis incidence [55,56]. In addition, estrogen has direct efects on opioid pain fbers in the central nervous system [57]. Estrogen receptors are localized in opioid-containing neurons in the spinal cord and brain. Apart from their role in central nervous system, estrogens enhance neural transmission of peripheral nociceptive input by inducing an infammatory environment within the joint. As a result, transmission of stimuli produced from articular structures is facilitated via estrogen mediation.
Te second mechanism of AI-induced arthralgia might implicate an autoimmune process [52]. Morel et al. described a case of rheumatoid arthritis in a woman treated with exemestane [58]. Moreover, AI treatment may lead to autoimmune diseases in patients with breast cancer, including rheumatoid arthritis, subacute cutaneous lupus erythematosus (SCLE), and vasculitis [59]. Tis could be due to AI-induced inhibition of diferentiation of naïve T-cells to regulatory T-cells and increased interferon-c (IFN-c) and interleukin-12 (IL-12) cytokine levels. In some cases, autoimmune disorders including arthralgias reversed upon AI cessation, along with a decrease in autoantibody levels, such as antinuclear antibodies (ANA) and rheumatoid factor (RF) [60]. Te last and less likely mechanism of AI-induced musculoskeletal syndrome is through a direct of target efect of the AI medication or one of its metabolites [52].
CDK4/6 inhibitors have been widely applied in everyday clinical practice in patients with HR-positive disease. Cyclindependent kinase 4 or 6 (CDK4/6) modulates cell cycle progression to the S phase via catalyzing retinoblastoma (RB) protein hyperphosphorylation [12]. In the hypophosphorylated state, RB exerts a repressive efect on the E2F family of transcription factors blocking cell cycle progression through the G1-S checkpoint. In response to mitogenic signals, including ER pathway, cyclin D associates with the protein kinases CDK4 and CDK6, and the cyclin D-CDK4/6 complex subsequently phosphorylates the RB protein.
Phosphorylation induces structural changes in RB protein rendering it unable to interact with E2F transcription factors. CDK4/6 inhibitors restore the onco-suppressive efect of RB by preventing its hyperphosphorylation by CDK4/6 kinases.
Te mechanism through which CDK4/6 inhibitors might reduce the rate of AIMSS remains unclear. Handschick et al. demonstrated that CDK6 physically and functionally interact with NF-kB subunit p65 [61]. Aberrant CDK6 expression contributes to NF-kB p65 phosphorylation and NF-kB-induced gene expression. NF-kB is activated upon stimulation from proinfammatory cytokines, such as interleukin-1 (IL-1) or tumor necrosis factor alpha (TNF-a). It was shown that 34-45% of IL-1-induced genes required CDK4 or CDK6, while RNAi-mediated knockdown of CDK6 suppressed several IL-1-induced genes such as IL-8, IL-6, and NFKB1A [61]. Moreover, CDK6 colocalized with NF-kB p65 subunit at several binding sites modulating the expression of many NF-kB target genes. CDK4/6 inhibitors might attenuate CDK6-dependent infammatory gene expression and the development of an intraarticular infammatory microenvironment. Te inhibition of AI-Te Breast Journal 9 induced infammation by CDK4/6 inhibitors could contribute to the decreased arthralgia rate observed in this population. CDK4/6 inhibitors reinstate the suppressive efect of RB on the E2F family of transcription factors. Wang et al. demonstrated that E2F2 regulates the infammation produced via STAT1 and PI3K/AKT/NF-kB pathways in a rheumatoid arthritis animal model [62]. Indeed, NF-kB associates with the promoter region of E2F2, and activated E2F2 subsequently binds to the promoter of IL-6, which in turn leads to the development of arthritis. SiRNAmediated knockdown of E2F2 attenuates the expression of infammatory cytokines in rheumatoid arthritis synovial fbroblasts. E2F2 regulates nuclear translocation of STAT1 and activation of PI3K/AKT/NF-kB pathway, inhibiting the expression of infammatory cytokines such as IL-1 and TNF-a. Tis efect of E2F2 is further supported by the fact that E2F2 is overexpressed in RA synovial tissues [62]. CDK4/6 inhibitors might suppress E2F2 through an RB-dependent way and thus inhibit the hyperplasia of RA synovial fbroblasts and joint damage caused by E2F2-induced infammatory factors. Tomita et al. reported a decrease in cartilage erosion by infltrating synovium upon transfection with E2F decoy oligodeoxynucleotides (ODN) [63]. Te production of IL-1, IL-6, and MMP-1 proinfammatory factors was also impaired by E2F decoy ODN. CDK4/6 inhibitors might have a similar protective function against articular cartilage invasion.

Conclusion
Overall, we report a decreased incidence of AI-induced arthralgia in postmenopausal breast cancer patients treated with AI and CDK4/6 inhibitor combination in the adjuvant setting. CDK4/6 inhibitors might have a protective efect against joint infammation and arthralgia occurrence although the exact mechanism remains unknown. Ongoing Phase III trials of CDK4/6 inhibitors in the adjuvant setting remain to address this issue.

Data Availability
Te data supporting our fndings can be found in PubMed MEDLINE bibliographical and ClinicalTrial.gov database.

Conflicts of Interest
MAD has received honoraria from participation in advisory boards from Amgen, Bristol-Myers-Squibb, Celgene, Janssen, Takeda. FZ has received honoraria for lectures and has served in an advisory role for Astra-Zeneca, Daiichi, Eli-Lilly, Merck, Novartis, Pfzer, and Roche. Evanngelos Terpos received honoraria from Amgen, Astra/Zeneca, Bristol Myers Squibb, Eusa Pharma, GSK, Integris Pharma, Janssen, Pfzer, Sanof, and Takeda, research grants from Amgen, GSK, Janssen, Sanof and Takeda, and travel grants from Amgen, Eusa Pharma and Takeda. Te remaining authors declare that they have no conficts of interest.

Authors' Contributions
Angeliki Andrikopoulou, Efthymia Skafda, Christos Markellos, and Moustafa S contributed data curation. Angeliki Andrikopoulou, Efthymia Skafda, Christos Markellos, and Moustafa S contributed formal analysis. Angeliki Andrikopoulou, Efthymia Skafda, Christos Markellos, and Moustafa S performed investigation. Meletios-Athanasios Dimopoulos and Flora Zagouri performed project administration; Pectasides D, Dimopoulos MA, Zagouri F, Vassilopoulos D, Terpos E. Visualization, Flora Zagouri, Dimopoulos MA, and Vassilopoulos D did supervision. Angeliki Andrikopoulou and Efthymia Skafda wrote the original draft. Angeliki Andrikopoulou, Efthymia Skafda, and Terpos E wrote the article, and reviewed and edited the article. All the authors have read and approved the fnal manuscript. All authors agree with the content, and all give consent to submit the manuscript in the current form. FZ is the corresponding author and guarantor of the review. Efthymia Skafda and Angeliki Andrikopoulou contributed equally to this work.