Open Access

Pharmacodynamic Modeling Identifies Synergistic Interaction Between Chloroquine and Trastuzumab in Refractory HER2- positive Breast Cancer Cells


11Center for Pharmacometrics and Systems Pharmacology, Department of Pharmaceutics, University of Florida, College of Pharmacy, Orlando, FL, U.S.A.

22Quantitative Pharmacology and Pharmacometrics (QP2), Merck & Co., Inc, Kenilworth, NJ, U.S.A.

Cancer Diagnosis & Prognosis Mar-Apr; 3(2): 175-182 DOI: 10.21873/cdp.10198
Received 24 November 2022 | Revised 07 December 2023 | Accepted 09 December 2022
Corresponding author
Sihem Ait-Oudhia, MS, PharmD, Ph.D., Quantitative Pharmacology and Pharmacometrics (QP2), Merck & Co., Inc, 2000 Galloping Hill Rd, Kenilworth, NJ 07033, U.S.A. Tel: +1 7162509828, email:


Background/Aim: Despite improvements in HER2-positive breast cancer (BC) patients’ outcomes with trastuzumab, the occurrence of intrinsic or acquired resistance presents a clinical challenge. Here, we quantitatively assess the combinatorial effects of chloroquine, an autophagy inhibitor, with trastuzumab on JIMT-1 cells, a HER2-positive BC cell-line primarily resistant to trastuzumab. Materials and Methods: The temporal changes in JIMT-1 cellular viability were assessed using the CCK-8 kit, where JIMT-1 cells were exposed for 72-h to trastuzumab (0.007-17.19 μM) or chloroquine (5-50 μM) as single-agents, in combination (trastuzumab: 0.007-0.688 μM; chloroquine: 5-15 μM), or control (no drug). Concentration-response relationships were built for each treatment arm to determine drugs’ concentrations inducing 50% of cell-killing (IC50). Cellular pharmacodynamic models were built to characterize the time-trajectory of JIMT-1 cellular viability under each treatment arm. The nature of trastuzumab and chloroquine interaction was quantified by estimating the interaction parameter (Ψ). Results: The IC50 were estimated at 19.7 and 24.4 μM for trastuzumab and chloroquine. The maximum killing effect was about thrice higher for chloroquine than trastuzumab (0.0405 vs. 0.0125 h-1), validating chloroquine’s superior anti-cancer effect on JIMT-1 cells compared to trastuzumab. The time-delay for chloroquine cell-killing was twice longer than that for trastuzumab (17.7 vs. 7 h), suggesting a chloroquine time-dependent anti-cancer effect. The Ψ was determined at 0.529 (Ψ<1), indicating a synergistic interaction. Conclusion: This proof-of-concept study on JIMT-1 cells identified chloroquine and trastuzumab synergistic interaction, warranting further in vivo investigations.
Keywords: Mathematical modeling, chemotherapy resistance, drugdrug interaction, cytotoxicity, synergy

The human epidermal growth factor receptor-2 (HER2) is overexpressed in approximately 20-30% of breast cancer (BC) cases. HER2 positive BC cells have a tendency to proliferate and spread aggressively leading to higher recurrence and mortality rates (1,2). The HER2 receptor belongs to the epidermal growth factor (EGF) family of receptor tyrosine kinases; although this receptor does not bind to ligands, it dimerizes with HER1, HER3 and HER4 to activate downstream signaling cascades that control biological processes such as cell growth and survival (3).

Trastuzumab (TZB) is a recombinant humanized monoclonal antibody that targets the HER2 receptor. Once TZM binds to the extracellular domain of the HER2 receptor, multiple mechanisms are triggered including the prevention of HER2-receptor dimerization, shedding of the extracellular domain, immune activation, as well as inhibition of its downstream intracellular tyrosine kinases (4). Despite the success of HER2 targeted therapies, such as TZB, the occurrence of intrinsic and acquired resistance has been reported. Some mechanisms of TZM resistance include the truncation of the HER2 receptor, increased signaling from other HER receptors such as HER3 or EGFR, increased signaling by the insulin-like growth factor receptor (IGF-1), loss of the AKT regulator phosphatase and TENsin homolog (PTEN), and/or lack of immune response (5).

In addition, induction of autophagy, a cellular survival mechanism activated in response to metabolic stress, hypoxia, or chemotherapy-induced cell death, has also been linked to TZB resistance. Briefly, during autophagy, autophagosomes are formed and engulf cytoplasm/cytoplasmic organelles. Then the autophagosomes fuse with lysosomes to degrade the contents of the autophagosome and make amino acids and other by-products available for metabolism and building of macromolecules (6). In a publication by Vazquez-Martin et al., it was reported that TZB resistant BC cells displayed increased cellular levels of the LC3-II protein, which is correlated with increased autophagosome numbers (7,8). The anti-malarial chloroquine (CQ) has been found to inhibit autophagy by impairing autophagosome fusion with lysosomes and increasing the pH inside lysosomes to inhibit enzymatic function. CQ also induces disorganization of the golgi and endo-lysosomal systems, which may further contribute to fusion impairment. As a result, adding CQ to TZB therapy may decrease resistance to HER2 therapy in TZM-treated breast cancer (7,9).

In this work, we sought to test in vitro the hypothesis that inhibition of autophagy with a common autophagy inhibitor such as CQ may be a valuable mechanism to reverse the resistance to TZM therapy in refractory HER2-positive BC. We examined the anti-cancer effects of TZM and CQ as single agents and in combination on JIMT-1 cells, a HER2 positive BC cell-line primarily resistant to TZM, and quantitatively assessed the nature of their drug-drug interaction. Our findings of synergistic combinatorial killing effects of TZM and CQ on JIMT-1 cells support the aforementioned hypothesis; however, an in vivo confirmatory study is warranted.

Materials and Methods

Drugs and reagents. CQ was purchased from Selleck chemical (Houston, TX, USA) and TZB was acquired from the pharmacy of the University of Florida (Gainesville, FL, USA). Dulbecco’s Modified Eagle’s Medium (DMEM), Penicillin/Streptomycin, and Phosphate Buffered Saline (PBS) were purchased from Hyclone, GE Healthcare Biosciences (Chicago, IL, USA). MEM non-essential amino acids, molecular biology grade water and 0.25% trypsin/2.21 mM EDTA were acquired from Corning (Corning, NY, USA). Fetal bovine serum (FBS) and Cell Counting kit-8 (CCK-8) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

CQ was dissolved in molecular biology grade water to make a 100 mM stock. CQ stocks were stored at –80˚C and TZM was stored at 4˚C, per manufacturer instructions. Fresh serial dilutions of drugs were prepared prior to experiments.

Cell culture. JIMT-1 cells, a TZM-resistant HER2+ cell line, were acquired from AddexBio (San Diego, CA, USA) and maintained in DMEM supplemented with 10% FBS, 1% sodium bicarbonate, 1% MEM non-essential amino acids, and 1% Penicillin/Streptomycin. Cells were incubated at 37˚C in a humidified atmosphere with 5% CO2 and passaged once confluent with 0.25% trypsin/2.21 nM EDTA.

In vitro cytotoxicity. JIMT-1 cells were seeded at a density of 3,000 cells/100 μl/well in a 96-well plate and incubated for at least 24 h to ensure adhesion. Concentration-response relationships were generated by exposing JIMT-1 cells to either single agent TZM (0.007-51.54 μM) or CQ (0.5-1000 μM) over 72 h. Experiments were performed in quadruplicates and compared against vehicle control (cell culture media). Cell viability was determined by incubating the cells in CCK-8 solution (10 μl/well of a 96-well plate) for 1.5 h and measuring absorbance at 450 nm using a microplate spectrophotometer (Biotek, Winooski, VT, USA). In subsequent experiments, JIMT-1 cells were exposed to TZM (0.007-17.19 μM) or CQ (5-50 μM) single agents or combinations of these (TZB: 0.007-0.688 μM; CQ: 5-15 μM) over a 72-h time course. Cellular viability was measured using the CCK-8 assay as described above.

Mathematical modeling.

Concentration-response relationships. The maximal inhibitory effects (Imax) for CQ and TZB and their corresponding concentrations leading to 50% of Imax (IC50) were estimated at 72 h by modeling their respective concentration-response curves with an inhibitory Hill function (10) such as:

where, R is the response to treatment (% cell viability), R0 is the baseline response (% viability under control conditions), Imax is the maximal effect, C is the concentration of drugs, IC50 is the drug concentration corresponding to 50% of Imax, and γ is the Hill coefficient. All mathematical modeling was performed with Monolix version 2016R1 (Antony, France: Lixoft SAS, 2016).

Time course cellular response pharmacodynamic models. JIMT-1 cellular response following exposure to single agent CQ, TZB, or combinations of CQ and TZB were measured for 0, 24, 48 and 72 h.

Cell growth for the control arm was described with an exponential growth function:

where kg represents the first order JIMT-1 growth rate constant, and R represents cellular response.

Single agent pharmacodynamic models. Pharmacodynamic (PD) models to characterize the cellular response profiles of CQ and TZB required two and four transit compartments (11) to adequately capture the observed delays between drug exposure and declines in cell viability as a result of the stimulation of cell death. The respective equations are described below:

Where τ represents mean transit time, K1-K4 represent the transit compartments, C represents drug concentrations, Smax represents the maximal value of the stimulatory cell death constant, SC50 represents the concentration corresponding to half of Smax, and γ represents the hill coefficient.

Chloroquine and trastuzumab pharmacodynamic model. Since TZB and CQ both exhibited cytotoxic effects over JIMT-1 cells, an interaction parameter, psi (Ψ), was applied to equation 3, to determine the nature of the CQ-TZM drug-drug interaction. Psi values below 1 indicate a synergistic interaction, values above 1 indicate an antagonistic interaction, while values equivalent to 1 indicate an additive interaction. The equations are described as follows:

Statistical analysis. To determine if the reductions in cellular viability following exposure to CQ+TZB combinations were significantly different from those observed post-CQ single agent exposure, statistical analyses of relative cell viabilities at 24, 48 and 72 h were performed via one-way ANOVA followed by Dunnett’s multiple comparison test, with an α of 0.05. Statistical analyses were performed with GraphPad Prism ver. 5 (GraphPad Software, La Jolla, CA, USA).


Concentration-response relationships. The model fits of the measured percent cell viability for JIMT-1 cells vs. a range of concentrations of TZB and CQ are depicted in Figure 1 and the corresponding model fitted parameters are summarized in Table I. The observed data were characterized well with inhibitory Hill functions, with all model parameters estimated with reasonable precision. The IC50, or concentration required for 50% of maximal drug effect, for CQ and TZB were 24.4 and 19.7 μM, respectively. For both drugs the highest tested concentrations reduced cell viability to approximately 0%, thus Imax (maximal effect) was fixed to 1.

Cellular response time course. Cellular response pharmacodynamic (PD) models were built to characterize JIMT-1 cellular viability following exposure to CQ and TZM single agents and combinations over a 72-h time course. The schematic summarizing the PD models for the single agents and combinations is depicted in Figure 2, graphs of cellular viability for all treatment arms are depicted in Figure 3, and the model parameters are summarized in Table II.

Both CQ and TZB stimulated JIMT-1 cell death, with 2 transit compartments required to capture the observed delays in cytotoxicity for CQ and 4 compartments required for TZB. The estimated maximal value of the stimulatory death constant (Smax) for CQ was 0.0405 h-1, with an SC50 (concentration required to achieve half of Smax) of 26.3 μM, while the estimated Smax and SC50 for TZB were 0.0124 h-1 and 12.6 μM. As both CQ and TZB exerted cytotoxic effects over JIMT-1 cells, the interaction parameter (Ψ) was estimated as 0.529±0.014, indicating a synergistic interaction.

Statistical analysis. Finally, to determine if the observed cellular viabilities between CQ single agent treated cells were significantly different from CQ+TZB treated cells, relative cell viabilities for the 24-, 48- and 72-h time points were analyzed by one-way Analysis of Variance (ANOVA), followed by Dunnett’s multiple comparison test. Figure 4 depicts the average relative cell viabilities±standard deviation (SD) and Table III summarizes the ANOVA results. Generally, this analysis demonstrated that cell viability of the combination treatment arms was significantly different from that of single agent arm for the 24 and 72-h time points.


HER2-positive BC represents 25% of all BC subtypes and is an aggressive BC that is associated with poor treatment outcomes and low patients’ survival rates (1,2). In this work, we designed an in vitro proof-of-concept study aiming to test the hypothesis that CQ, an autophagy inhibitor, may alleviate resistance to TZM in HER2-positive BC that is refractory to TZM. To this end, we performed in vitro experiments on JIMT-1 cells, a HER2-positive BC cell-line that is primarily resistant to TZM. We quantified JIMT-1 cells viability under single and combinatorial effects of CQ with TZM at increasing concentrations of each drug and exposure times. We also characterized mathematically these concentration- and time-effect relationships.

The concentration-response relationships in JIMT-1 cells from single agents CQ and TZM were quantified pharmacologically and the IC50 of each drug was determined. Our results confirm that JIMT-1 cells are indeed resistant to TZM, since the IC50 was estimated at 19.7 μM, which is equivalent to 2,867 μg/ml. This concentration is approximately 23-fold higher than the reported clinical serum concentration at steady-state of 123 μg/ml (12). Besides the fact that JIMT-1 cells are inherently resistant to TZM, one other possible interpretation to the lower concentration of TZM in vivo compared to our estimated in vitro IC50 is that in the tumor micro-environment, the TZM main mechanism of action of TZM is known to be through antibody-dependent cell-mediated cytotoxicity (ADCC). The ADCC is an adaptive immune response mediated mainly by the natural killer (NK) cells that bind the Fc portion of TZM, triggering the lysis of tumor cells. The absence of NK cells in the in vitro cell culture setting and, hence, lack of ADCC effect by TZM led to the need for a larger TZM concentration to achieve a desired level of JIMT-1 cell killing.

Despite the unclear clinical utility of CQ in cancer therapy, our findings agree with other published studies conducted in vitro on various cancer cell lines including lung (13), breast (14), bladder (15) and colon cancer (16). In these studies, CQ has been reported to induce apoptosis at concentrations ranging from 25 to 128 μM. Notwithstanding that in adults, CQ toxicity is dose dependent. It was shown that the ingestion of CQ at doses between 2 to 4 g may cause neurological symptoms corresponding to blood concentrations between 2.5 and 5 mg/l (4.84-9.69 μM), while CQ blood concentrations above 5 mg/l have been associated with serious toxicity (17). While the CQ estimated IC50 from our experimental data on JIMT-1 cells is close to the lower bound of its concentration interval in various in vitro cancer cells (24.4 vs. 25 μM), it is calculated to be ~2.5-fold the in vivo serum concentration causing serious safety concerns (24.4 vs. 9.69 μM). Hence, based on the supra-therapeutic estimated in vitro IC50 for CQ and TZM in JIMT-1 cells, we conclude that as single-agents CQ and TZM require above therapeutic blood concentrations to be efficacious at killing HER2-positive BC cells refractory to TZM.

In subsequent experiments, we exposed JIMT-1 cells to a range of CQ and TZB concentrations, as single-agents and in combination over a 72-h time course. Cell-based PD models were built to characterize the time course data and to determine the nature of their drug-drug interaction. Both CQ and TZB stimulated JIMT-1 cell death, with an estimated interaction parameter Ψ of 0.529±0.014, indicating synergism (Ψ<1) between these agents. Upon visual inspection of the cellular response plots, it appeared that JIMT-1 cellular viability under the combination treatment was slightly lower than the JIMT-1 cellular viability under the treatment with single agent CQ. To determine if these differences were statistically significant, the relative cell viability at 24-, 48-, and 72-h time points was analyzed by one-way ANOVA test followed by Dunnett’s multiple comparison test. These analyses revealed that generally for the 24- and 72-h time points, the mean relative cell viability of single agent CQ at concentrations 5, 10 and 15 μM was significantly different from the mean relative cell viability of combinations containing these same concentrations of CQ. Overall, these results agree well with the synergistic interaction between CQ and TZM, as estimated via the interaction parameter Ψ. Additionally, the mean-transit-time parameter τ, which represents the time for each drug to achieve apoptosis of JIMT-1 cells, was estimated to be twice longer for CQ than for TZM (17.7 vs. 7 h), suggesting that CQ anti-tumoral activity is time-dependent.


In summary, this work demonstrated that as single agents CQ and TZM are inefficacious on JIMT-1 cells, however their combinatorial effects are synergistic. This finding may support the primary hypothesis of CQ lifting TZM resistance in HER2-positive BC refractory to TZM therapy. However, our findings also suggest that despite the fact that CQ and TZB act synergistically in vitro for the killing of JIMT-1 cells, the clinical utility of this combination may be limited due to CQ induced toxicity. Further in vivo studies could be considered to investigate whether CQ cytotoxic effects can be achieved while keeping drug levels within therapeutically acceptable ranges.

Conflicts of Interest

The Authors declare that there are no conflicts of interest.

Authors’ Contributions

YLF designed the study, conducted the experiments, and wrote the manuscript. CK helped conducting the experiments, SAO designed the study, provided scientific mentoring to all co-authors, and wrote and edited the manuscript.


1 Loibl S & Gianni L HER2-positive breast cancer. Lancet. 389(10087) 2415 - 2429 2017. PMID: 27939064. DOI: 10.1016/S0140-6736(16)32417-5
2 Park JW Neve RM Szollosi J & Benz CC Unraveling the biologic and clinical complexities of HER2. Clin Breast Cancer. 8(5) 392 - 401 2008. PMID: 18952552. DOI: 10.3816/CBC.2008.n.047
3 Hurvitz SA Hu Y O’Brien N & Finn RS Current approaches and future directions in the treatment of HER2-positive breast cancer. Cancer Treat Rev. 39(3) 219 - 229 2013. PMID: 22658319. DOI: 10.1016/j.ctrv.2012.04.008
4 Hudis CA Trastuzumab—mechanism of action and use in clinical practice. N Engl J Med. 357(1) 39 - 51 2007. PMID: 17611206. DOI: 10.1056/NEJMra043186
5 Lavaud P & Andre F Strategies to overcome trastuzumab resistance in HER2-overexpressing breast cancers: focus on new data from clinical trials. BMC Med. 12 132 2014. PMID: 25285786. DOI: 10.1186/s12916-014-0132-3
6 Glick D Barth S & Macleod KF Autophagy: cellular and molecular mechanisms. J Pathol. 221(1) 3 - 12 2010. PMID: 20225336. DOI: 10.1002/path.2697
7 Vazquez-Martin A Oliveras-Ferraros C & Menendez JA Autophagy facilitates the development of breast cancer resistance to the anti-HER2 monoclonal antibody trastuzumab. PLoS One. 4(7) e6251 2009. PMID: 19606230. DOI: 10.1371/journal.pone.0006251
8 Luque-Cabal M García-Teijido P Fernández-Pérez Y Sánchez-Lorenzo L & Palacio-Vázquez I Mechanisms behind the resistance to trastuzumab in HER2-amplified breast cancer and strategies to overcome it. Clin Med Insights Oncol 10(Suppl. 1) 21 - 30 2016. PMID: 27042153. DOI: 10.4137/CMO.S34537
9 Cufí S Vazquez-Martin A Oliveras-Ferraros C Corominas-Faja B Cuyàs E López-Bonet E Martin-Castillo B Joven J & Menendez JA The anti-malarial chloroquine overcomes primary resistance and restores sensitivity to trastuzumab in HER2-positive breast cancer. Sci Rep. 3 2469 2013. PMID: 23965851. DOI: 10.1038/srep02469
10 Holford NH & Sheiner LB Kinetics of pharmacologic response. Pharmacol Ther. 16(2) 143 - 166 1982. PMID: 6752972. DOI: 10.1016/0163-7258(82)90051-1
11 Lobo ED & Balthasar JP Pharmacodynamic modeling of chemotherapeutic effects: application of a transit compartment model to characterize methotrexate effects in vitro. AAPS PharmSci. 4(4) E42 2002. PMID: 12646013. DOI: 10.1208/ps040442
12 Goldenberg MM Trastuzumab, a recombinant DNA-derived humanized monoclonal antibody, a novel agent for the treatment of metastatic breast cancer. Clin Ther. 21(2) 309 - 318 1999. PMID: 10211534. DOI: 10.1016/S0149-2918(00)88288-0
13 Fan C Wang W Zhao B Zhang S & Miao J Chloroquine inhibits cell growth and induces cell death in A549 lung cancer cells. Bioorg Med Chem. 14(9) 3218 - 3222 2006. PMID: 16413786. DOI: 10.1016/j.bmc.2005.12.035
14 Jiang PD Zhao YL Shi W Deng XQ Xie G Mao YQ Li ZG Zheng YZ Yang SY & Wei YQ Cell growth inhibition, G2/M cell cycle arrest, and apoptosis induced by chloroquine in human breast cancer cell line Bcap-37. Cell Physiol Biochem. 22(5-6) 431 - 440 2008. PMID: 19088425. DOI: 10.1159/000185488
15 Lin YC Lin JF Wen SI Yang SC Tsai TF Chen HE Chou KY & Hwang TI Chloroquine and hydroxychloroquine inhibit bladder cancer cell growth by targeting basal autophagy and enhancing apoptosis. Kaohsiung J Med Sci. 33(5) 215 - 223 2017. PMID: 28433067. DOI: 10.1016/j.kjms.2017.01.004
16 Zheng Y Zhao YL Deng X Yang S Mao Y Li Z Jiang P Zhao X & Wei Y Chloroquine inhibits colon cancer cell growth in vitro and tumor growth in vivo via induction of apoptosis. Cancer Invest. 27(3) 286 - 292 2009. PMID: 19194831. DOI: 10.1080/07357900802427927
17 Mégarbane B Bloch V Hirt D Debray M Résiére D Deye N & Baud FJ Blood concentrations are better predictors of chioroquine poisoning severity than plasma concentrations: a prospective study with modeling of the concentration/effect relationships. Clin Toxicol (Phila). 48(9) 904 - 915 2010. PMID: 21080867. DOI: 10.3109/15563650.2010.518969