Abstract
Background/Aim: This study investigated in vivo synergism between eribulin and palbociclib in a breast cancer patient-derived xenograft (PDX) model, with expanded scope to include fulvestrant as a third drug. Materials and Methods: Eribulin plus palbociclib combinations were tested in vitro in six cell lines each of estrogen receptor positive and triple-negative breast cancer, and in vivo in the OD-BRE-0192 PDX model using weekly eribulin plus 5×/week or 7×/week palbociclib (holiday or no-holiday schedules, respectively). When included as a third drug, fulvestrant was dosed weekly. Results: In vitro, combining palbociclib with eribulin led to increased eribulin IC50s in 11 of 12 cell lines, suggesting that the drugs antagonized each other due to mutual exclusion of the mitotic and G1/S cell cycle block points for eribulin and palbociclib. An in vivo study in the OD-BRE-0192 PDX model compared weekly eribulin plus either palbociclib holiday or no-holiday schedules to gauge the importance of post-palbociclib cell cycle synchronization. Results showed no advantage of holiday over no-holiday schedules, arguing that differing pharmacokinetics of the drugs were sufficient to overcome cell cycle-based mechanistic antagonism. In vivo comparisons of doublet and triplet combinations of eribulin, palbociclib, and fulvestrant showed that all three doublets were superior to individual monotherapies, and that the triplet combination was markedly superior to all three doublets, being the only group to show tumor regression in 100% of the mice. Conclusion: Results show complex synergistic interactions between eribulin, fulvestrant, and palbociclib, and point to a particularly robust synergy when combining all three drugs.
Eribulin (ER-086526, E7389, and NSC-707389), a macrocyclic ketone analog of the marine sponge natural product halichondrin B (Figure 1) (1), has been approved by various agencies for the treatment of patients with advanced breast cancer, advanced liposarcoma, or soft tissue sarcoma (2-5). In addition to the cytotoxic antimitotic effects that typify most microtubule-targeting agents (1, 6, 7), eribulin also affects tumor biology in ways unrelated to mitosis, including tumor vasculature remodeling leading to increased tumor perfusion and mitigation of tumor hypoxia, reversal of epithelial-mesenchymal transition (EMT) in breast cancer and cell differentiation in sarcoma, and therapeutically beneficial alterations of the tumor immune microenvironment including increased post-treatment CD8 T cells in patient tumor samples and activation of innate immune signaling (4, 5, 8-14). These non-mitotic effects of eribulin on tumor biology prompted us to combine eribulin with other anticancer drugs acting through different mechanisms to explore the breadth of therapeutic possibilities under combination conditions (15). In one combination study using patient-derived xenografts (PDX) of luminal B breast cancer, we found synergistic anticancer activities between eribulin and the cyclin-dependent kinase 4/6 (cdk 4/6) inhibitor palbociclib (Figure 1) (15). Here, we further explored this synergy in both hormone receptor-positive (HR+) and triple-negative breast cancer (TNBC) cell lines, while also testing drug holiday versus no-holiday dosing schedules in vivo. As palbociclib is approved in the context of HR+ breast cancers (16, 17), we further explored both doublet and triplet combinations of eribulin, palbociclib, and the selective estrogen receptor (ER) degrader (SERD) fulvestrant (Figure 1). Results indicate that while any of the three possible doublet combinations can lead to tumor stasis, only the triplet combination of eribulin + fulvestrant + palbociclib can induce marked and long-lasting tumor regression.
Chemical structures of eribulin, palbociclib, and fulvestrant.
Materials and Methods
Test compounds. Eribulin mesylate (salt form used in Halaven® clinical formulation; hereinafter referred to as eribulin) was synthesized by Eisai Co., Ltd. at the Kashima, Japan manufacturing facility. Palbociclib was purchased from Selleck Chemicals LLC (Houston, TX, USA) for cell-based studies and from Euromedex (Strasbourg, France) for in vivo studies. Fulvestrant was purchased from Albrecht (Grevenbroich, Germany).
Human breast cancer cell lines and in vitro antiproliferative assays. In vitro antiproliferative cell-based studies were performed by Charles River Discovery Services (Morrisville, NC, USA) under contract from Eisai Inc. The 12 established human breast cancer cell lines utilized included six ER+ lines (MCF-7, MDA-MB-134VI, MDA-MB-175VII, MDA-MB-415, T-47D, and ZR-75-1) and six TNBC lines (BT-549, HCC1806, HCC70, Hs578t, MDA-MB-231, and MDA-MB-436). Cell lines were grown and maintained under standard tissue culture conditions optimized for use in the Charles River commercial testing program.
In vitro antiproliferative activities of eribulin and palbociclib, alone and in combination, were determined by standard ATP-based luminescence assay techniques (Cell Titre-Glo®, Promega, Madison, WI, USA). Compounds were added after overnight incubation following the initial cell seeding. Compound exposure times were 72 h for all cell lines except MDA-MB-134VI, MDA-MB-175VII, MDA-MB-415, and MDA-MB-436. Due to extremely slow growth rates of these four cell lines, cell growth and drug exposure times were doubled to 144 h to provide sufficient dynamic range between initial seeding densities and final cell numbers achieved in control wells. Eribulin and palbociclib IC50s determined for the eight 72 h lines versus the four 144 h lines showed no statistically significant differences associated with exposure time length, arguing that the IC50s measured represent true time-independent biological sensitivities to the drugs.
In vivo PDX studies. All studies using laboratory animals were approved by the Institutional Animal Care and Use Committees of Oncodesign Services (Dijon, France) and adhered to applicable laws and guidelines for the humane care and use of laboratory animals. The OD-BRE-0192 luminal B breast cancer PDX model of Oncodesign Services has been previously described (15). Briefly, this PDX was generated using a clinical sample from a 45-year-old female with luminal B invasive lobular breast carcinoma with lymph node metastases. The patient had responded poorly to prior therapies, including epirubicin, 5-fluorouracil, cyclophosphamide, taxotere, paclitaxel, bevacizumab, and gemcitabine. The resulting PDX tumor was passaged by serial subcutaneous transplantation into immunocompromised Balb/c nu/nu mice (Charles River Laboratories, Lyon, France). When needed for PDX line preservation, tumors were cut into small fragments (30-50 mg) and snap-frozen in liquid nitrogen. As needed for regeneration, frozen samples were retrieved, thawed, and re-engrafted into nu/nu mice for amplification. Since OD-BRE-0192 PDX tumors require continuous estrogen stimulation for growth, mice carrying this tumor received drinking water continuously supplemented with 2.5 μg/ml estradiol.
In vivo efficacy studies of drugs in the OD-BRE-0192 model were performed in 6-7 week old female Swiss nu/nu mice (Charles River Laboratories) using subcutaneous engraftment of 30-40 mg tumor fragments. Engraftment was performed 24-48 h after whole body irradiation with a γ-irradiation source (2 Gy [60Co]; BioMEP, Bretenières, France). When engrafted tumors reached volumes of 200-300 mm3, mice were randomized into different treatment groups consisting of 8-10 mice per group.
Eribulin was administered by intravenous (iv) tail vein injection in 0.9% (v/v) NaCl at previously determined suboptimal doses of 0.125 or 0.25 mg/kg using weekly schedules of one injection every 7 days (Q7D) (15), with the dose used and number of weekly cycles depending on the experiment. Suboptimal doses of palbociclib were administered in 0.9% (v/v) NaCl by oral gavage at 150 or 107 mg/kg on Q1Dx5 or Q1Dx7 weekly schedules for palbociclib holiday and no-holiday schedules, respectively, with the number of weekly cycles depending on the experiment. Fulvestrant was administered subcutaneously (sc) at a suboptimal dose of 0.625 mg/kg on a Q7D schedule for four weekly cycles.
For the palbociclib holiday versus no-holiday comparison, the two palbociclib dose levels on their corresponding schedules yield the same dosing intensity per week (750 mg/kg/week), allowing direct comparison of the two schedules. Timing of the holiday was such that the last dose of Q5D palbociclib occurred two days before eribulin, with the first day of the next palbociclib cycle occurring one day after eribulin. This timing asymmetry was intended to provide sufficient time for cells to recover from palbociclib-induced G1/S blocks, whereas cells blocked in mitosis by eribulin had no opportunity to reenter the cell cycle due to eribulin’s known antimitotic irreversibility (18).
For all in vivo studies, tumor sizes and body weights were measured twice a week. Tumor volumes (TVs) were measured by caliper and calculated using the following formula: TV=width2 × [length/2], with TV in mm3 and width and length measurements in mm.
Statistical analyses. Statistical analyses, including unpaired t-tests, one-way ANOVA, and Kaplan–Meier analyses were performed using GraphPad Prism version 9.4.1 (GraphPad Prism Software, San Diego, CA, USA). Where needed, multiple comparison tests were performed using Tukey’s multiple comparison tests. Values of p<0.05 were considered statistically significant.
Results
Antagonism between eribulin and palbociclib under continuous exposure conditions in vitro. Previous in vivo studies examining combinations of eribulin with other anticancer agents showed potential synergy between eribulin and the cdk 4/6 inhibitor palbociclib in PDX models of luminal B breast cancer (15). We first tested the in vitro combinations of the two drugs in six ER+ and six TNBC human breast cancer cell lines. Although palbociclib is used clinically for HR+ breast cancers, TNBC cell lines were included since all proliferating cells regardless of HR status must pass through the G1/S cell cycle boundary, the site of action of cdk 4/6. Indeed, some recent reports suggest possible roles for cdk 4/6 inhibitors in TNBC, particularly retinoblastoma protein (Rb) wild-type TNBC (19-21).
Initial in vitro proliferation assays showed that single agent potencies of the two drugs differed by at least 3-4 orders of magnitude, with eribulin IC50s being <1 nM in all 12 cell lines whereas palbociclib IC50s were ≥1 μM for 11 lines and ≥10 μM for 4 lines (Table I). Since such large differences in inherent potency could confound interpretation of Combination Index strategies to evaluate synergy versus antagonism (22), we asked instead if the continuous presence of 1 μM or 10 μM palbociclib could affect measured eribulin IC50 values. Unexpectedly, despite the previously observed in vivo synergy between the two drugs (15), the presence of either 1 μM or 10 μM palbociclib increased eribulin IC50s in 11 of 12 cell lines (Table I), indicating that the continuous presence of palbociclib actually antagonized, not enhanced, responses to eribulin. Since palbociclib is a known G1/S cell cycle blocker (23-25), it seems likely that G1/S blocks under continuous exposure conditions set up cell cycle-based antagonism by preventing cells from reaching mitosis where eribulin’s antimitotic effects occur (1, 6, 7). In this scenario, under intermittent in vivo dosing conditions, differing pharmacokinetics of the two drugs would overcome cell cycle-based antagonism that might manifest itself only under the continuous exposure conditions of in vitro incubations. If so, we hypothesized that cell cycle-based antagonism could be leveraged in vivo for intentional tumor cell synchronization by judicious selection of intermittent drug administration schedules, leading to enhanced combination responses.
In vitro antagonism of eribulin by palbociclib under continuous simultaneous exposure conditions.
Synergy between eribulin and palbociclib under intermittent dosing conditions in vivo. To test whether a palbociclib treatment holiday strategy might increase the in vivo antitumor responses of the eribulin + palbociclib combination, a study was performed in OD-BRE-0192 luminal B breast cancer PDX (HR+/HER2−) comparing weekly iv eribulin with or without 5 day per week oral palbociclib (palbociclib 48 h holiday schedule) versus weekly iv eribulin with or without 7 day per week oral palbociclib (palbociclib no-holiday schedule) for a total of six weekly cycles (Figure 2A). Two partially effective (suboptimal) doses of eribulin were selected (0.125 mg/kg and 0.25 mg/kg) based on previous studies in this model (15). Palbociclib doses in the two arms were adjusted for equal total dosing intensity during the treatment period (150 mg/kg/dose and 107 mg/kg/dose palbociclib for holiday and no-holiday arms, respectively). All dosing regimens were well tolerated and substantially below maximum tolerated dose (MTD) as shown by <10% reversible mean body weight losses for all monotherapy and combination groups (Figure 2B). Although synergy of eribulin with palbociclib was observed for all combinations regardless of dose or schedule (compare Figure 2D with 2E), unexpectedly, no differences in antitumor efficacy were observed between palbociclib holiday and no-holiday schedules at either 0.125 mg/kg or 0.25 mg/kg eribulin dosing levels (Figure 2C-E). We hypothesize that intermittent in vivo dosing together with different pharmacokinetics properties of the two drugs is sufficient to sidestep any antagonism that can be seen with continuous exposure to both drugs in cell culture (Table I). Thus, the results of this experiment provide no evidence for additional benefit of a palbociclib dosing holiday in vivo, beyond what is already provided by the once daily dosing of the drug.
In vivo antitumor activity of eribulin plus palbociclib combinations in the OD-BRE-0192 breast cancer PDX model. (A) Dosing schedule showing 0.125 mg/kg or 0.25 mg/kg eribulin given intravenously (iv) on Q7D weekly cycles, and 150 mg/kg or 107 mg/kg palbociclib given orally (po) on Q5D or Q7D weekly cycles for holiday or no-holiday schedules, respectively. Since body weight losses were found to be both minor (<10%) and reversible after the first 3 weekly cycles, the study was extended for another 3 weekly cycles starting on day 83. (B) Percent mean body weight changes relative to day 58 (first day of dosing); the mean day 58 body weight of all 8 groups was 27.5±0.2 g (SD), with individual group mean weights on this day ranging from 27.2-27.8 g. (C) TVs, all groups. (D) TVs, vehicle, and monotherapy groups only. (E) TVs, vehicle, and combination groups only. In B-E, data are presented as means±SEM, with the day 58-103 treatment period (detailed in panel A) indicated by solid bars on the x-axes.
Comparison of doublet and triplet combinations of eribulin, palbociclib and fulvestrant. Palbociclib is approved in the USA for advanced or metastatic HR+/HER2− breast cancer in combination either with an aromatase inhibitor, as initial endocrine therapy, or with fulvestrant after progression on a previous endocrine therapy. Hormone response pathways are thus invoked by both approved indications. Accordingly, we next tested suboptimal doses of eribulin, palbociclib and fulvestrant as monotherapies and all possible doublet or triplet combinations in the OD-BRE-0192 PDX model (Figure 3). All dosing regimens were well tolerated and were substantially below MTD as shown by <10% reversible mean body weight losses for all groups (Figure 3B). Relative to monotherapies (Figure 3C and D), all three doublet combinations showed synergistic interactions defined initially by tumor stasis beginning about one week after dosing initiation and persisting until about a week after dosing cessation (Figure 3C and E). Based on regrowth rates after dosing cessation, the two fulvestrant-containing doublets led to longer tumor regrowth suppression compared to the palbociclib plus eribulin doublet. Importantly, the triplet combination of eribulin plus palbociclib plus fulvestrant was the only combination group to show mean tumor regression, which started about one week after dosing initiation and persisted for five weeks post-dosing to the end of the experiment (Figure 3C and E). Thus, despite each doublet only eliciting tumor stasis, the triplet combination led to marked and long-lasting tumor regression.
In vivo antitumor activity of eribulin, palbociclib and fulvestrant combinations in the OD-BRE-0192 breast cancer PDX model. (A) Dosing schedule showing 0.25 mg/kg eribulin given intravenously (iv) on Q7D weekly cycles, 150 mg/kg palbociclib given orally (po) on Q5D weekly cycles, and 0.625 mg/kg fulvestrant given subcutaneously (sc) on Q7D weekly cycles. (B) Percent mean body weight changes relative to day 47 (first day of dosing); the mean day 47 body weight of all 8 groups was 27.1±0.5 g (SD), with individual group mean weights on this day ranging from 26.2-27.8 g. (C) TVs, all groups. (D) TVs, vehicle, and monotherapy groups only. (E) TVs, vehicle, and combination groups only. In B-E, data are presented as means±SEM, with the day 47-73 treatment period (detailed in panel A) indicated by solid bars on the x-axes. Three groups (vehicle, palbociclib, eribulin) lost >20% of mice due to large tumor volumes before the end of the experiment (day 111); these three groups were humanely euthanized in their entireties as soon as >20% of mice were lost, accounting for the early cessation of data in those groups relative to the experiment end day.
Closer analysis of the combination groups provides additional insights. As shown in Figure 4A, mean TVs from the three doublet combinations measured five days after cessation of dosing were statistically indistinguishable from each other, whereas the triplet combination group was statistically different from all three doublets. Increased efficacy of the triplet combination relative to the doublets was also observed by Kaplan–Meier analysis of time to reach >500 mm3 TV (Figure 4B). As shown, median days to >500 mm3 TV ranged from 61-70 days for the vehicle and monotherapy groups, while the three doublet groups showed medians of 97-111 days. In marked contrast, not a single mouse in the triplet combination group reached 500 mm3 TV during the course of the experiment (Figure 4B).
Detailed analyses of combination groups from the OD-BRE-0192 PDX study shown in Figure 3. (A) Comparison of TVs in combination groups on experiment day 78 (5 days after last dose). Numbers above brackets represent p-values of pairwise comparisons determined using Tukey multiple comparison tests; comparisons that were not statistically significant at p<0.05 are not shown. (B) Kaplan–Meier plot of time to reach TVs >500 mm3. Median values are shown in parentheses in the legend lines for each group. (C) Best response categorization as progressive disease (PD), stable disease (SD), partial response (PR) and complete response (CR) using mRECIST criteria (26) adapted for murine tumor studies (27). Response evaluation window started on experiment day 58, corresponding to dosing day 11 (see Materials and Methods). One mouse in the palbociclib group was found dead on dosing day 5, accounting for the group size of 9; based on the day 4 body weight this death did not appear to be drug related. (D) Summary table of best response categories shown in 4C.
Evaluation of best overall responses of individual mice by mRECIST criteria (26), as adapted for murine PDX studies (27), again shows the superiority of both the doublet combinations over the monotherapies and the triplet combination over the doublets (Figure 4C with tabulated results in 4D). As shown, with the exception of one mouse in the fulvestrant group, all mice in the vehicle and three monotherapy groups showed progressive disease (PD). In contrast, 60-80% of the mice in the three doublet groups showed stable disease (SD), with one mouse each in the two fulvestrant doublets achieving partial response (PR). Compared to the three doublet groups, the triple combination group showed markedly better responses, with 9 of 10 mice achieving PR and the tenth mouse maintaining SD. Thus, whether analyzed by mean TV over the course of the experiment, TV at 2 days after dosing cessation, days to >500 mm3 TV, or best response by mRECIST criteria, the results of this study clearly establish superiority of (i) any of the three doublet combinations over the monotherapies, and (ii) the triplet combination over any of the three doublets.
Discussion
Breast cancer is the most common cancer worldwide and the leading cause of cancer-related deaths among women (28). Approximately 70-80% of breast cancers are classified as HR+/HER2−, so-called luminal A breast cancers (29). Initially, HR+/HER2− patients are typically treated with hormone therapy (HT) including antiestrogens that interfere with the ER pathway (e.g., tamoxifen or fulvestrant) or estrogen deprivation strategies including aromatase inhibitors or ovariectomy. Unfortunately, most patients develop HT resistance, with chemotherapy previously being the only remaining option. However, the recent development of cdk 4/6 inhibitors, such as palbociclib, ribociclib and abemaciclib significantly changed treatment options for HR+/HER2− patients with HT resistance (30-32), pushing chemotherapy back into later lines. Ultimately, however, most such patients eventually develop resistance to cdk 4/6 inhibitor-containing regimens (33, 34), raising the question of what the next best treatment option is, pure chemotherapy or adding chemotherapy to existing cdk 4/6 inhibitor-containing regimens.
As a microtubule dynamics inhibitor, eribulin exerts cytotoxic antimitotic activities that are typical of most tubulin-targeted agents (1, 6, 7). In addition, however, eribulin’s non-cytotoxic tumor biology-modifying effects (4, 5, 8-14) support the hypothesis that it interacts with additional pathways beyond those directly involved with mitosis. Such additional activities suggest that eribulin may be a good candidate to combine with other drugs acting via different mechanisms, especially those exerting effects at the G1/S cell cycle transition point where proliferation and cellular differentiation signaling pathways converge.
We previously observed synergistic antitumor effects between eribulin and the cdk 4/6 inhibitor palbociclib in two PDX models of HR+/HER2− luminal breast cancer (15). As an initial approach to exploring the basis of this synergy, we investigated in vitro cell-based combinations between eribulin and palbociclib in 12 established human breast cancer cell lines, including 6 HR+ and 6 TNBC lines. Unexpectedly based on the prior in vivo results, evidence for antagonism, not synergy, was observed in 11 of the 12 cell lines tested (Table I), leading to the hypothesis that continuous exposure to the two drugs under cell culture conditions was causing cell cycle-based antagonism, wherein the palbociclib-induced block at G1/S prevented cells from reaching mitosis where eribulin’s antimitotic effects occur, while the eribulin-induced mitotic block prevented cells from reaching G1/S where palbociclib effects occur. If so, it seems possible that such cell cycle-based antagonism could have a therapeutic advantage in vivo with judicious selection of dosing schedules. Accordingly, we tested the possibility that a 48 h palbociclib dosing holiday before administration of weekly eribulin might cause synchronous release of cells from the palbociclib-induced G1/S block, resulting in enhanced cell capture at the eribulin-induced mitotic block. Surprisingly, however, direct in vivo comparison of palbociclib holiday versus no-holiday dosing schedules failed to show superiority of one regimen over the other (Figure 2), although both schedules still yielded the previously observed in vivo synergy between the two drugs. While one explanation could be that the cell cycle-based antagonism hypothesis is simply incorrect, it seems equally possible that with intermittent in vivo dosing, the differing pharmacokinetic properties of the two drugs are sufficient to overcome any cell cycle-based antagonism that might only be observable under continuous exposure conditions in cell culture.
We next extended exploration of eribulin plus palbociclib in vivo synergy by introducing the SERD fulvestrant to directly inhibit ER pathway activity. Using the OD-BRE-0192 PDX model, double and triplet combinations of eribulin, palbociclib and fulvestrant were tested (Figure 3, Figure 4). As expected, monotherapy with intentionally suboptimal doses of the three drugs led to only modest inhibition of tumor growth rates. In contrast, all three doublet combinations led to tumor stasis during the administration period, followed by a three week delay before tumor growth resumed. Based on tumor regrowth rates, the two fulvestrant-containing doublets were the most effective. In contrast to the stasis achieved with the three doublets, the triplet combination of eribulin + palbociclib + fulvestrant led to marked tumor regression during the treatment period, followed by post-treatment persistence of stasis at very small TVs for five weeks until the end of the study. Superiority of the triplet combination over the three doublets was confirmed by statistical comparison of TVs five days after the last dose, Kaplan–Meier analysis of times to reach 500 mm3 TVs, and best response of all groups by mRECIST criteria. Notably, nine of ten mice in the triplet combination group achieved PR status, with the tenth mouse achieving SD; thus, 100% of the mice in the triplet group benefitted and showed some degree of disease control.
The mechanisms behind the regression-inducing triple synergy effect are currently unknown, but close inspection of the data reveals several aspects that may point in certain directions. First, the suboptimal doses of the three drugs administered alone led to initially equivalent, albeit only modest, decreases in tumor growth rates during the first three weeks of the four week dosing period (Figure 3D). During the last week of dosing, tumors in the fulvestrant group entered a short stasis period that was not observed in the eribulin or palbociclib groups, suggesting a slow onset mechanism by which fulvestrant can trigger stasis. Second, despite widely different mechanisms, all three doublets led to virtually identical stasis during the last three weeks of dosing, indicating that regardless of mechanism, combining any two of the three drugs is sufficient to fully halt tumor proliferation, yet without tumor regression (Figure 3E). Interestingly, 2-3 weeks after dosing cessation, tumor regrowth in the two fulvestrant-containing doublets resumed slowly while regrowth in the eribulin + palbociclib group accelerated rapidly (Figure 3E), suggesting that fulvestrant exerts long-lasting post-treatment effects that can alter tumor phenotype towards slower growth. In this regard, it is notable that none of the three doublets induced tumor regression, only stasis. In contrast, clear and persistent tumor regression was observed in the triple combination group starting only one week after initiation of dosing, ultimately resulting in very small tumors that remained static for the remainder of the experiment. Mechanistically, inhibition of proliferation resulting in tumor stasis and cytotoxicity resulting in tumor regression are different: cytotoxicity is not simply “more stasis”. Thus, considering that only stasis is achieved by any of the three doublets, the rapid appearance of tumor regression with the triple combination was unexpected. The two aspects already mentioned point to possible phenotype-altering effects of fulvestrant, and it is known that eribulin itself can modify tumor phenotype by several mechanisms (4, 5). However, even such phenotypic effects by both drugs seem insufficient to trigger tumor regression without cdk 4/6 inhibitors, since none of the three doublets led to tumor regression. It thus seems inescapable that some unique aspect of the triple combination triggers tumor regression, which is not achieved by combining only two of the three drugs. This then leads to the question of whether regression with the triple combination reflects direct cytotoxicity (e.g., tumor cell apoptosis) or initiation of an antitumor immune response, which could account for both TV reduction and prolonged maintenance of small tumor stasis through the end of the experiment. Since the study was performed in nude mice lacking T cells (35, 36), immunological explanations for the observed regression would require myeloid cells, NK cells, B cells or some combination thereof. In this context, eribulin is known to induce innate immune signaling (13, 14), whereas fulvestrant has been reported to inhibit both innate immune pathways but also immunosuppressive myeloid-derived suppressor cells (37, 38). It is possible that in the fulvestrant + eribulin doublet, opposing effects balance each other out, but in the context of the triple combination including cdk 4/6 inhibitor the balance tips toward antitumor immune responses. Such scenarios are currently speculative, and further studies will be required to understand the mechanisms behind the unexpected tumor regression seen only with the triple combination of eribulin + palbociclib + fulvestrant.
Footnotes
Authors’ Contributions
IB contributed to data analysis and manuscript development. MHdB contributed to study design, data analysis, and manuscript development. EMdC contributed to performing experiments. BAL contributed to study design, data analysis, and manuscript development.
Conflicts of Interest
IB, MHdB, and EMDC are full-time employees of Oncodesign Services. BAL is a full-time employee of Eisai Inc., which discovered, developed and currently manufactures and markets eribulin mesylate (as Halaven®) according to FDA-approved indications in the USA and indications in other countries approved by the appropriate regulatory authorities.
- Received October 18, 2023.
- Revision received December 11, 2023.
- Accepted December 18, 2023.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
