Antibody–drug conjugates (ADC) represent a fast-growing drug class in oncology. However, ADCs are associated with resistance, and therapies able to overcome it are of utmost importance. Recently, enfortumab vedotin-ejfv (EV) was approved in nectin-4+ metastatic urothelial cancer. We previously described PVRL4/nectin-4 as a new therapeutic target in breast cancer and produced an efficient EV-like ADC comprising a human anti–nectin-4 mAb conjugated to monomethyl auristatin-E (MMAE) named N41mab-vcMMAE. To study the consequence of the long-term treatment with this ADC, we developed a preclinical breast cancer model in mice, and report a mechanism of resistance to N41mab-vcMMAE after 9-month treatment and a way to reverse it. RNA-sequencing pointed to an upregulation in resistant tumors of ABCB1 expression, encoding the multidrug resistance protein MDR-1/P-glycoprotein (P-gp), associated with focal gene amplification and high protein expression. Sensitivity to N41mab-vcMMAE of the resistant model was restored in vitro by P-gp pharmacologic inhibitors, like tariquidar. P-gp is expressed in a variety of normal tissues. By delivering the drug to the tumor more specifically than classical chemotherapy, we hypothesized that the combined use of ADC with P-gp inhibitors might reverse resistance in vivo without toxicity. Indeed, we showed that the tariquidar/N41mab-vcMMAE combination was well tolerated and induced a rapid regression of ADC-resistant tumors in mice. In contrast, the tariquidar/docetaxel combination was toxic and poorly efficient. These results show that ABC transporter inhibitors can be safely used with ADC to reverse ADC-induced resistance and open new opportunities in the fight against multidrug resistance.

Designed to deliver anticancer drugs in the most targeted way, antibody–drug conjugates (ADC) represent a fast-growing drug class in oncology (1). Twelve ADCs are currently approved by the FDA, including KADCYLA [ado-trastuzumab emtansine (T-DM1)] in refractory HER2+ metastatic breast cancer, and PADCEV (enfortumab vedotin-ejfv) in nectin-4+ metastatic urothelial cancer (2). However, primary or secondary resistance to ADCs inevitably occurs (3). It is thus essential to define the mechanisms and develop alternative strategies to overcome ADC-based resistance.

We previously described PVRL4/nectin-4, a type I transmembrane cell adhesion molecule, as a new marker, tumor antigen and target in breast cancer (4–6). Nectin-4 is expressed during fetal development, with expression declining in adult life, and as a tumor-associated antigen with prooncogenic properties in various carcinomas including breast cancer (4, 7–11). We developed an ADC (N41mab-vcMMAE) comprising a human anti–nectin-4 mAb conjugated to the monomethyl auristatin-E (MMAE) drug, which showed strong antitumor activity in preclinical models of triple-negative breast cancer (5). In the meantime, a phase II study assessing an anti–nectin-4 ADC linked to MMAE (enfortumab vedotin-ejfv) in patients with metastatic urothelial cancer reported a very encouraging rate of objective responses in heavily pretreated cases (12). This ADC was recently approved after the publication of the phase III study (2).

In many circumstances, ADCs have improved the outcome of patients in therapeutic failure, but so far have not achieved cures because of resistance issue (1). Suggested resistance mechanisms are diverse for ADCs and may overlap with those described for chemotherapy like enhanced efflux of drugs (3). However, they remain largely underexplored and there is a crucial need to develop both preclinical models and overcoming strategies. These observations prompted us to develop an in vivo breast cancer preclinical model of resistance through long-term nectin-4 targeting with the N41mab-vcMMAE ADC. Using this model, we describe a mechanism of resistance to N41mab-vcMMAE ADC involving overexpression of P-gp and a way to efficiently and safely reverse it in vivo with P-gp–specific inhibitors. Provided that these findings are extended to patients’ tumors, our results suggest that P-gp inhibition may extend the effective treatment window in patients treated with anti–nectin-4 ADC and probably other FDA-approved ADCs.

Reagents and antibodies

The ADC N41mab-vcMMAE and N41mab-DM1 were produced by Concortis. Conjugates were produced from purified N41mab mAb, established in our laboratory (5). For comparison, enfortumab and N41mab naked antibodies both bind to the distal IgV domain of nectin-4 but do not compete and thus recognize different epitopes. For N41mab-vcMMAE, the linker used was the maleimidocaproyl-l-valine-l-citrulline-p-aminobenzyl alcohol p-nitrophenyl carbonate covalently conjugated to monomethyl auristatin-E (MMAE) (MC-vs.-PAB-MMAE). This cleavable linker was selected because it induced potent bystander killing. The drug-to-antibody ratio (DAR) was 4.73. For the N41mab-DM1, the noncleavable linker used was the succinimidyl-trans-4-(N-maleimidomethyl) cyclohexane-1-carboxylate covalently conjugated to emtansine (DM1; Lys-SMCC-DM1). The DAR was 3.96. Structures of linker-payload are presented in Supplementary Fig. S1. The ADC brentuximab vedotin (CD30-MMAE) and trastuzumab emtansine (T-DM1) aliquots were obtained from hospital. Tariquidar (XR9576), zosuquidar (LY335979), and elacridar (GF120918) were purchased from Selleckchem.

Cell line

The SUM190 breast cancer cell line was kindly provided by Dr. Ethier (https://sumlineknowledgebase.com/) and was cultured in Ham F12 medium with 5% FBS, nonessential amino acids, 10 μg/mL insulin, 1 μg/mL streptomycin, and 2 mmol/L glutamine. Absence of Mycoplasma contamination was regularly controlled by PCR (Eurofins Genomics), and MycoAlert PLUS test (Lonza). This cell line expresses high level of nectin-4 and represents a suitable model for targeting with anti–nectin-4 ADC. Integrity and authentication of cell lines after cell culture were verified by comparing their array-CGH profiles with that of the reference original SUM190 line.

Cytotoxicity/cell growth measurement

Experiments were done by incubating 3,000 cells per well in duplicate with serial dilutions of ADC at day 0 in 96-well plates. To evaluate the effect of ADC, cell growth/cytotoxicity was measured using the CellTiter-Blue Reagent staining procedure as recommended (Promega). The spectral properties of CellTiter-Blue Reagent change upon reduction of resazurin to resorufin. Resazurin is dark blue in color and has little intrinsic fluorescence until it is reduced to resorufin by mitochondrial oxidation, which is pink and highly fluorescent (579 nm Excitation/584 nm Emission). Hence, the intensity of the fluorescent signal reflects cell viability. Fluorescence was analyzed on the BMG LABTECH microplate reader FLUOstar Optima at 560Ex/590Em.

RNA sequencing

Tumor RNAs were extracted from sensitive and resistant preclinical models and controlled on Agilent Bioanalyzer (Agilent Technologies). RNA-Seq libraries were created using Truseq RNA-Seq Library kit (Illumina). Briefly, 1 μg of total RNA was used to isolate mRNA poly(A). cDNA was synthetized by random hexamer primers and reverse transcriptase (Superscript II, Life Technologies). Next, end repair and 3′ ends adenylation of the fragments were performed. Bar-coded adapters were ligated to the cDNA fragments and a PCR reaction was performed to produce the sequencing libraries. The quality check of libraries and quantification were performed using Tape station High sensitivity D1000. The libraries were then sequenced on an Illumina NextSeq500 platform, and 150 bp paired-end reads were generated. Sequencing reads were aligned to the human genome (hg19) using STAR 2.4.2a. The reads were summarized at the gene-level using htseq-count 0.11.1. Counts were then normalized and transformed to Log(cpm) with the bioconductor package EdgeR (13). To compare the expression profiles of sensitive (ADCS-TS SUM190) versus resistant (ADCR-PR and ADCR-DR SUM190) preclinical models, we applied a supervised analysis by using a moderated t test and the following significance thresholds (P < 5%, q<25%, and IFold Change FCI>2).

Array-comparative genomic hybridization

Tumor DNAs were extracted from sensitive and resistant preclinical models (14) and controlled on Agilent Bioanalyzer. Genomic profiles of samples were established by using array-CGH onto 4×180K CGH microarrays (SurePrint G3 Human CGH Microarray Kit, Agilent Technologies). Human female DNA was used as reference (G152A, Promega). Both approaches and analysis methods were previously described (15). All probes for array-CGH were mapped according to the hg19/NCBI human genome mapping database. The copy number was estimated for each gene by taking the value of the segment with the highest amplitude, then categorized into “Amp” (log2 ratio >1), “Gain” (0.5< log2 ratio ≤1), “Loss” (−1≤ log2 ratio <−0.3), and “Del” (log2 ratio ←1). Focal events were defined as genomic alterations with a size less than 5 Mb and a copy number higher than the surrounding segments.

Flow cytometry

FACS analysis of SUM190 cells was done using 2 μg/mL N41mab antibody in PBS supplemented with 5% FCS. Cells were then stained with phycoerythrin-conjugated (PE) goat anti-mouse antibody (Beckman Coulter). Anti-CD243 conjugated to PE (e-Bioscience) was also used at 5 μg/mL. Cells were analyzed with the Fortessa cytofluorimeter (Becton Dickinson).

Internalization studies

Thiol-conjugation of N41mab antibody was performed as recommended (Promega, Ref G9835). Briefly, mAb was reduced with 2.5 mmol/L DTT for 1 hour at 25°C. DTT was removed with Zeba spin desalting column. Thiol-conjugation was done with the pHdye reagent dissolved in 50% DMSO, for 3 minutes at 25°C under gentle agitation. Free dye was removed with Zeba spin desalting column. Concentration and dye-to-antibody ratio were calculated at A280 and A532 as recommended. The N41mab-pHdye dye-to-antibody ratio was 4. Internalization experiments were done in 96-well plates using 2 μg/mL of N41-pH dye for 24 hours at 37°C under 5% CO2. Cells were washed three times with 150 μL PBS and fluorescence quantified on the Clariostar Plus spectrofluorimeter using Cy3 setting (530–20 nm/580–30 nm).

In vivo studies

All experiments were done in agreement with the French Guidelines for animal handling and approved by the local ethics committee (agreement no. 01152–01). NOD/SCID/γc null mice (NSG) were obtained from Charles Rivers. Mice were housed under sterile conditions with sterilized food and water provided ad libitum and maintained on a 12-hour light and 12-hour dark cycle. ADCR-DR-SUM190 cells were inoculated in both flanks in the mammary fat pads with 0.5 × 106 cells suspended in 50% phenol red-free Matrigel (BD Biosciences). Mice were treated when tumor reached average volume of 700 mm3. Mice were either treated by N41mab-MMAE alone, docetaxel alone, N41mab-MMAE in association with tariquidar, or docetaxel in association with tariquidar. The “tariquidar groups” were treated once daily during 7 days with tariquidar (15 mg/kg, i.v.). ADC (2×10 mg/kg, i.v). and docetaxel (10 mg/kg, i.v.) treatments started after the first treatment with tariquidar. The tariquidar regimen we applied was adapted from two previous studies (16, 17). Tumor growth was monitored by measuring with a digital caliper and by calculating the tumor volume (length × width2 × π/6). All animals were randomly assigned into treatment groups, such that the mean tumor volume for each group was 100 to 200 mm3. Pain and distress were evaluated for each animal. Animal weight was monitored every 3 days to evaluate the toxicity of the different treatments.

Statistical analyses

Data are presented as mean + SEM. The comparison of tumor volumes and mouse weights between two treatments (ADC alone vs. ADC + tariquidar, and docetaxel alone vs. docetaxel + tariquidar) were done using the Mann–Whitney test. To evaluate whether treatments are significantly different from each other, Bonferroni method was used for pairwise comparison of the means. The GraphPad Prism software was used for statistical analyses. A P < 0.05 was considered statistically significant. Data are representative of at least three experiments.

Data availability

The data generated in this study were deposited in the ArrayExpress database at EMBL-EBI under the accession numbers E-MTAB-11681 and E-MTAB-11682 for array-CGH and RNA-seq, respectively.

Long-term treatment with anti–nectin-4 ADC and induction of resistance

A preclinical model of nectin-4+ BC was developed in NSG mice xenografted orthotopically on both flanks with SUM190 breast cancer cells (5). The treatment of 5 mice with N41mab-vcMMAE induced a regression of tumors that lasted up to 40 days. One mouse was studied over time (Fig. 1A). Relapses occurred, but tumors on both flanks remained sensitive to new injections. After the fifth cycle (day 264), a primary resistant tumor (PR) occurred in the left fat pad (LFP) and was removed. At day 438, a local relapse (LR) in the LFP and a metastasis in an inguinal lymph node (distant relapse; DR) were observed. In the right fat pad, the tumor kept sensitivity to ADC (TS). After euthanasia at day 438, the two tumors (DR, TS) were removed and processed for comparative studies with PR tumor as indicated in Fig. 1A. First, ADCR-PR and ADCR-DR tumors were grafted in NSG mice and confirmed as resistant (Fig. 1B) to a new cycle of ADC treatment. Second, three ex vivo cell lines were established in vitro and called PR, DR, and TS. The cell line derived from TS was sensitive to ADC and called ADCS-TS SUM190 (IC50 = 4 ng/mL; 27 pmol/L) and the two cell lines derived from PR and DR were resistant and called ADCR-PR and ADCR-DR SUM190 (>1,000 ng/mL; >6,750 pmol/L; Fig. 1C). ADCS-TS SUM190 and ADCR-DR SUM190 cells displayed similar expression of nectin-4 and internalization capacities (Supplementary Figs. S2 and S3). Third, because this acquired resistance was kept after multiple passages in mice, as well as in ex vivo–derived tumor cells, we hypothesized that resistance might be due to an intrinsic molecular modification of the tumor cells, which we profiled.

Figure 1.

Development of a pre-clinical model resistant to anti-nectin-4 ADC (N41mab-MMAE). A, NSG mice were xenografted in the right and left mammary fat pads with SUM190 breast cancer cells. NSG mice were repetitively treated with ADC at 10 mg/kg/injection with 2 intravenous injections at 4-days interval (double arrows). Measurement of tumor volume was processed in the left (white dot) and right (black dot) fat pad. PR, primary resistant tumor; LR, local relapse; DR, distant relapse; TS, sensitive tumor. B, Passages of ADCR-PR and ADCR-DR resistant tumors in NSG were compared for ADC sensitivity. Treatment with N41mab-MMAE (arrows) did not reduce tumor growth. C, The cell lines derived from the sensitive right fat pad (TS – left figure) and from distant relapse (DR – right figure) were treated with CD30-MMAE (negative control (white circle) and N41mab-MMAE (black circle). The SUM190 cell line derived from TS (left) was sensitive to N41mab-MMAE and called ADCS-TS SUM190. The SUM190 cell line derived from DR (right) were resistant to N41mab-MMAE and called ADCR-DR SUM190.

Figure 1.

Development of a pre-clinical model resistant to anti-nectin-4 ADC (N41mab-MMAE). A, NSG mice were xenografted in the right and left mammary fat pads with SUM190 breast cancer cells. NSG mice were repetitively treated with ADC at 10 mg/kg/injection with 2 intravenous injections at 4-days interval (double arrows). Measurement of tumor volume was processed in the left (white dot) and right (black dot) fat pad. PR, primary resistant tumor; LR, local relapse; DR, distant relapse; TS, sensitive tumor. B, Passages of ADCR-PR and ADCR-DR resistant tumors in NSG were compared for ADC sensitivity. Treatment with N41mab-MMAE (arrows) did not reduce tumor growth. C, The cell lines derived from the sensitive right fat pad (TS – left figure) and from distant relapse (DR – right figure) were treated with CD30-MMAE (negative control (white circle) and N41mab-MMAE (black circle). The SUM190 cell line derived from TS (left) was sensitive to N41mab-MMAE and called ADCS-TS SUM190. The SUM190 cell line derived from DR (right) were resistant to N41mab-MMAE and called ADCR-DR SUM190.

Close modal

Characterization of resistance to anti–nectin-4 ADC

Using RNA-seq we found that ABCB1, encoding an ATP-binding cassette transporter (MDR-1, P-gp, CD243), was the most differentially expressed gene, overexpressed in both ADCR-PR and -DR as compared with ADCS-TS tumors (Fig. 2A; Supplementary Table S1). This was confirmed at the protein level by using flow cytometry on cell lines (Fig. 2B) and IHC on tumors (Fig. 2C). Moreover, we found that resistant cells displayed a focal amplification of the ABCB1 gene in chromosome region 7q21.1 (Fig. 2D).

Figure 2.

Molecular characterization of resistance to anti-nectin-4 ADC: role of P-gp. A, Volcano plot representation of differentially expressed genes between ADCS and ADCR-PR and ADCR-DR SUM190 tumors. The red dots represent the 342 transcripts upregulated in ADCR versus ADCS cells (positive x-values), and the green dots represent the 279 transcripts downregulated in ADCR versus ADCS cells (negative x-values). ABCB1 is the most overexpressed gene and presents the highest fold-change of expression. B, FACS analysis of P-gp cell surface expression on ADCS-TS SUM190 cells (bold black) and on ADCR-DR SUM190 cells (dashed black; fill gray: negative control histogram). P-gp expression is higher in ADCR than in ADCS cells. C, P-gp expression by IHC on FFPE sections of ADCS-TS, ADCR-PR and ADCR-DR SUM190 tumors. D, DNA copy-number profiles of ADCS-TS (red lines), ADCR-PR (green), and ADCR-DR (blue) tumors established by using array-CGH (top). Zoom of profiles of chromosome 7 for ADCS-TS, ADCR-PR, and ADCR-DR tumors (bottom left). Both ADCR-PR and ADCR-DR present an amplification of the ABCB1 gene (bottom right), whereas no amplification was detected in ADCS.

Figure 2.

Molecular characterization of resistance to anti-nectin-4 ADC: role of P-gp. A, Volcano plot representation of differentially expressed genes between ADCS and ADCR-PR and ADCR-DR SUM190 tumors. The red dots represent the 342 transcripts upregulated in ADCR versus ADCS cells (positive x-values), and the green dots represent the 279 transcripts downregulated in ADCR versus ADCS cells (negative x-values). ABCB1 is the most overexpressed gene and presents the highest fold-change of expression. B, FACS analysis of P-gp cell surface expression on ADCS-TS SUM190 cells (bold black) and on ADCR-DR SUM190 cells (dashed black; fill gray: negative control histogram). P-gp expression is higher in ADCR than in ADCS cells. C, P-gp expression by IHC on FFPE sections of ADCS-TS, ADCR-PR and ADCR-DR SUM190 tumors. D, DNA copy-number profiles of ADCS-TS (red lines), ADCR-PR (green), and ADCR-DR (blue) tumors established by using array-CGH (top). Zoom of profiles of chromosome 7 for ADCS-TS, ADCR-PR, and ADCR-DR tumors (bottom left). Both ADCR-PR and ADCR-DR present an amplification of the ABCB1 gene (bottom right), whereas no amplification was detected in ADCS.

Close modal

Overcoming the resistance to anti–nectin-4 ADC in vitro

We wondered whether pharmacologic inhibition of P-gp could restore the sensitivity to our ADC. ADCR cells were treated with third-generation inhibitors (tariquidar, elacridar, zosuquidar) combined with either brentuximab vedotin (BV; CD30-MMAE ADC as a negative control) or N41mab-vcMMAE. The three P-gp inhibitors used at up to 200 nmol/L concentration were not toxic as shown by the CD30-MMAE curves. They equally restored the sensitivity of ADCR cells to N41mab-vcMMAE, with IC50 similar for the three inhibitors (20, 10, and 10 nmol/L, respectively; Fig. 3A). Interestingly, P-gp inhibitors (here shown for 100 nmol/L zosuquidar; Fig. 3B) restored the sensitivity of ADCR cells to N41mab-vcMMAE to the level of ADCS cells, confirming the major role of P-gp–mediated drug efflux in the resistance induced by the long-term ADC treatment (Fig. 3B). ADCR cells also showed resistance to free auristatin-E (MMAE), free maytansine (DM1), N41-DM1 ADC, and to a lesser extent to T-DM1. This resistance was similarly restored by zosuquidar (Fig. 3C). As expected, ADCS cells were sensitive to all drugs, without or with zosuquidar.

Figure 3.

Reversion of the resistance to ADC in vitro using P-gp inhibitors. A, ADCR-DR cell line was treated with 1 μg/mL of N41mab-MMAE (black circle) or CD30-MMAE negative control (white circle) in the presence of indicated concentrations of tariquidar, elacridar or zosuquidar. The three P-gp inhibitors were not toxic (CD30-MMAE curves), and restored the sensitivity to N41mab-MMAE. B, Treatment of ADCS-TS (left) and ADCR-TR (right) cell lines with indicated concentrations of N41mab-MMAE with or without 100 nmol/L zosuquidar (ZOSU). Zosuquidar had no effect on the sensitivity of ADCS to N41mab-vcMMAE (left, black vs. white triangles), but restored the sensitivity of ADCR SUM190 cells to the level of ADCS SUM190 cells (IC50: 5 ng/mL; right, black vs. white triangles). Zosuquidar was not toxic (CD30-MMAE curves). C, Resistance to monomethyl auristatin-E and maytansine (DM1). ADCS-TS (triangles) and ADCR-DR (squares) cell lines were treated with indicated concentrations of free MMAE (top left), free DM1 (top right), N41mab-DM1 ADC (bottom left), and with T-DM1 ADC (bottom right). ADCR cells were resistant to all drugs (white squares), but treatment with 100 nmol/L zosuquidar restored sensitivity (black squares).

Figure 3.

Reversion of the resistance to ADC in vitro using P-gp inhibitors. A, ADCR-DR cell line was treated with 1 μg/mL of N41mab-MMAE (black circle) or CD30-MMAE negative control (white circle) in the presence of indicated concentrations of tariquidar, elacridar or zosuquidar. The three P-gp inhibitors were not toxic (CD30-MMAE curves), and restored the sensitivity to N41mab-MMAE. B, Treatment of ADCS-TS (left) and ADCR-TR (right) cell lines with indicated concentrations of N41mab-MMAE with or without 100 nmol/L zosuquidar (ZOSU). Zosuquidar had no effect on the sensitivity of ADCS to N41mab-vcMMAE (left, black vs. white triangles), but restored the sensitivity of ADCR SUM190 cells to the level of ADCS SUM190 cells (IC50: 5 ng/mL; right, black vs. white triangles). Zosuquidar was not toxic (CD30-MMAE curves). C, Resistance to monomethyl auristatin-E and maytansine (DM1). ADCS-TS (triangles) and ADCR-DR (squares) cell lines were treated with indicated concentrations of free MMAE (top left), free DM1 (top right), N41mab-DM1 ADC (bottom left), and with T-DM1 ADC (bottom right). ADCR cells were resistant to all drugs (white squares), but treatment with 100 nmol/L zosuquidar restored sensitivity (black squares).

Close modal

Overcoming the resistance to anti–nectin-4 ADC in vivo without toxicity

Because normal tissues physiologically express P-gp, a major problem of chemotherapy is its increased toxicity when associated with P-gp inhibitors (18). In tumors with P-gp–mediated resistance, the targeted delivery of ADC treatment should leave intact normal tissues and reduce the toxicity. To address this issue in vivo, ADCR cells were xenografted back in NSG mice. Tariquidar is known to inhibit murine P-gp in vitro and in vivo (19). We compared the efficacy and toxicity of docetaxel versus anti–nectin-4 ADC treatment in association or not with repetitive doses of tariquidar in the resistant models (Fig. 4A). As expected, N41mab-MMAE alone had no effect on tumor growth. Docetaxel alone led to slight reduction of tumor growth. Association of tariquidar with docetaxel led to a slightly more durable growth reduction than docetaxel alone, likely attributable to tariquidar (P < 0.05 at day 41, D41; Supplementary Table S2). In contrast, association of tariquidar with N41mab-MMAE led to a much more marked reduction of tumor growth than N41mab-MMAE alone (P < 0.001 at D31 and D41), indicating that tariquidar resensitized the resistant tumor to ADC in vivo.

Figure 4.

Reversion of the resistance to ADC in vivo using P-gp inhibitors. A, Association of tariquidar with ADC or chemotherapy to restore sensitivity of P-gp–resistant tumors in mice. ADCR-DR cells were orthotopically xenografted in NSG mice. Mice were treated when tumor reached average volume of 700 mm3. Mice were either treated by N41mab-MMAE (black circles), or docetaxel (black squares) alone, or in association with tariquidar (gray). The “tariquidar groups” were treated once daily during 7 days with tariquidar (15 mg/kg, i.v.). ADC (2 × 10 mg/kg, i.v at 4-day interval) and docetaxel (10 mg/kg, i.v.) treatments started after the first treatment with tariquidar. The comparison of tumor volumes between two treatments (ADC alone vs. ADC + tariquidar, and docetaxel alone vs. docetaxel + tariquidar) were done using the Mann–Whitney test. To evaluate whether treatments are significantly different from each other, Bonferroni method was used for pairwise comparison of the means. *, P < 0.05; ***, P < 0.001. B, Body weight measurement of mice was done over time in the different groups. The legend is similar to A.

Figure 4.

Reversion of the resistance to ADC in vivo using P-gp inhibitors. A, Association of tariquidar with ADC or chemotherapy to restore sensitivity of P-gp–resistant tumors in mice. ADCR-DR cells were orthotopically xenografted in NSG mice. Mice were treated when tumor reached average volume of 700 mm3. Mice were either treated by N41mab-MMAE (black circles), or docetaxel (black squares) alone, or in association with tariquidar (gray). The “tariquidar groups” were treated once daily during 7 days with tariquidar (15 mg/kg, i.v.). ADC (2 × 10 mg/kg, i.v at 4-day interval) and docetaxel (10 mg/kg, i.v.) treatments started after the first treatment with tariquidar. The comparison of tumor volumes between two treatments (ADC alone vs. ADC + tariquidar, and docetaxel alone vs. docetaxel + tariquidar) were done using the Mann–Whitney test. To evaluate whether treatments are significantly different from each other, Bonferroni method was used for pairwise comparison of the means. *, P < 0.05; ***, P < 0.001. B, Body weight measurement of mice was done over time in the different groups. The legend is similar to A.

Close modal

Assessment of toxicity was a crucial point and was based upon measurement of animal weight over time. Association of tariquidar with docetaxel induced a marked weight decrease when compared to docetaxel alone, close to the ethical limit point (20%), suggesting high toxicity (P < 0.001 at D34, D36, and D39, Fig. 4B). By contrast, association of tariquidar with N41mab-MMAE induced only a slight weight decrease when compared to N41mab-MMAE alone (P < 0.05, only at D31, Fig. 4B). These preclinical data demonstrated that the P-gp inhibitor combined with our ADC could efficiently kill P-gp–resistant tumor cells while remaining much less toxic than when combined with chemotherapy.

ADC treatments are currently undergoing an unprecedented expansion in oncology. Several clinical trials are ongoing (generally after multi-prior regimens) demonstrating encouraging efficacy. But tumor relapse occurs almost irremediably and biological mechanisms of resistance remain poorly elucidated in patients (1). In some rare cases, biological mechanisms have been reported from small clinical series implicating modifications in antigen expression and/or in internalization process, and/or resistance to the payload such as drug efflux pumps or target mutation (20–23). For example, in a recent analysis of pre- and post-sacituzumab-govitecan tumor samples from three treated patients with metastatic triple-negative breast cancer (TNBC), TROP2 target expression was a pretreatment determinant of response, whereas acquired mutations involving the direct targets of antibody (TROP2) and of drug payload (TOP1) were identified in one patient in the postprogression samples (23). However, most resistance mechanisms have been described from in vitro and in vivo studies (24) and ways to overcome them have been poorly studied.

In this report, we were interested in the resistance to anti–nectin-4 ADC in breast cancer and the way to overcome it. Nectin-4 is a new therapeutic target in various carcinomas. Two anti–nectin-4 ADC conjugated to MMAE via a cleavable linker have shown promising efficacy, including ours in TNBC (5). Our study provides the first in vivo preclinical model of resistance induced by an anti–nectin-4 ADC in breast cancer and the first model developed directly in mouse implanted with human breast cancer cells and then treated with an ADC. It also provides insights into the molecular mechanism and, importantly, proposes a therapeutic alternative both efficient and safe to overcome such resistance in vitro and in vivo, whereas no difference was observed between the resistant versus sensitive tumors regarding nectin-4 expression and internalization capacities (Supplementary Figs. S2 and S3), we found an upregulation of P-gp expression in resistant tumors and restored sensitivity to our ADC both in vitro and in vivo using third-generation P-gp inhibitors.

Efflux of drugs by ABC transporters is a well-known mechanism of resistance to chemotherapy (25), and ABC transporter inhibitors showed activity in preclinical studies. However, their toxicity in normal cells, particularly in combination with chemotherapy, prevented their clinical use. First-generation P-gp inhibitors, such as verapamil and cyclosporine-A (CsA), were micromolar range inhibitors with poor specificity for P-gp. However, CsA was recently associated with brentuximab vedotin (BV) in a phase I trial to treat 14 patients with Hodgkin lymphoma resistant to this ADC (26); efficacy results were encouraging in term of reversion of BV resistance (26, 27), but toxicity was high and increased compared with BV alone, probably due to high dose of CsA used. Associated with the immunosuppressive effect of CsA, such toxicity led the authors to suggest association with new-generation P-gp inhibitors. Second-generation inhibitors such as valspodar, more potent and more selective, showed unfavorable pharmacokinetic interactions with chemotherapy often requiring dose reduction and limited benefit (28). This led us to use the third-generation P-gp inhibitors such as tariquidar with nanomolar range efficacy, higher specificity, and improved pharmacokinetic properties. A few years ago, the use of tariquidar had been tested in clinical trials in association with chemotherapy but was stopped because of an efficacy limited by a high toxicity (29), notably in breast cancer (30). We hypothesized that the targeted drug delivery by ADC to tumor cells and not to normal cells, by contrast to chemotherapy, could favor tariquidar efficacy while reducing systemic toxicity.

Therefore, we used tariquidar in association with our anti–nectin-4 ADC. In vitro, tariquidar, as well as elacridar and zosuquidar, were successful in restoring sensitivity in cell lines resistant to our ADC. Two other studies previously showed the capacity of third-generation P-gp inhibitors to restore the sensitivity to ADCs in vitro: fequidar with gemtuzumab ozogamicin in AML cell lines (31) and elacridar with brentuximab vedotin in Hodgkin lymphoma cell lines (32). Importantly in vivo, whereas our anti–nectin-4 ADC alone had no effect on the tumor growth of our resistant model, combination with tariquidar not only stopped the growth but also led to major significant tumor regression. Of note, the three third-generation P-gp inhibitors when tested alone showed no effect on the growth of our models in vitro (Fig. 3B). Docetaxel alone was slightly more efficient on the resistant model than our anti–nectin-4 ADC alone, but addition of tariquidar to docetaxel did not significantly improve its efficacy in vivo. This may reflect the limited tumor bioavailability of docetaxel in vivo. Clearly, the most efficient treatment on our resistant model was anti–nectin-4 ADC combined with tariquidar. Regarding toxicity, as expected our ADC was less toxic than docetaxel, which was less toxic than the docetaxel/tariquidar combination. Addition of tariquidar to our ADC increased its toxicity, but such combination displayed toxicity similar to that of docetaxel alone and significantly inferior to that of the docetaxel/tariquidar combination. We acknowledge two limitations for our in vivo analysis. First, because we tested one drug regimen only, which was based on data from two previous studies (16, 17), we cannot warrant that this regimen is optimal in terms of P-gp inhibition (complete or partial) and duration of inhibition; of note, Chen and colleagues (26) had tested also one drug regimen in vivo, whereas they had tested several different regimens in the subsequent phase I study. Second, we did not assess the effect of tariquidar alone in vivo: thus, it is unclear if the body weight loss seen with N41mab-MMAE+tariquidar is due to tariquidar alone or to an interaction effect between the two treatments. But yet and to our knowledge, our study is the first in vivo study to show efficacy and toxicity of combining third-generation P-gp inhibitors and ADC. In a recent study, the combination of BV with tariquidar showed a modest survival improvement of mice xenografted with BV-resistant epithelioid inflammatory myofibroblastic sarcoma (17), despite P-gp overexpression and sensitivity to tariquidar in vitro; however, the mice received only one injection of 12 mg/kg tariquidar one day before each BV injection, whereas we used in our model seven repetitive daily doses of tariquidar (15 mg/kg) one day before, during and two days after N41mab-vcMMAE treatment; this may account for the marked efficacy observed.

Drug efflux has been reported as a resistance mechanism for other non–nectin-4 ADCs including FDA-approved ADCs, and involves P-gp or other ABC transporters (33). In this context, the use of inhibitors against these transporters represents one new means to fight resistance to ADCs that provide a strong rationale to revisit their therapeutic potential. These ABC transporter inhibitors will probably contribute to expand the percentage of patients eligible for the ADC/anti-ABC combined therapy. However, because our results are based on the analysis of a single resistant breast cancer in mice, in the current state, we cannot extrapolate our results to ADC-resistant patients, nor to all cancer types or all ADCs. Tools for selection of patients candidate to receive a given ABC transporter inhibitor associated to ADC will have to be developed, perhaps based on the presence of one prominent transporter. In this context, analysis of large series of pre- and post-ADC clinical samples is urgently warranted.

In conclusion, we developed a mouse model of breast cancer resistant to our nectin-4 ADC, identified ABCB1/P-gp upregulation as a major mechanism of resistance, and showed that the tariquidar third-generation P-gp inhibitor restored sensitivity to ADC both in vitro and in vivo with limited toxicity.

A. Gonçalves reports non-financial support from Novartis outside the submitted work. No disclosures were reported by the other authors.

O. Cabaud: Conceptualization, supervision, visualization, methodology. L. Berger: Validation, investigation, visualization, methodology. E. Crompot: Conceptualization, formal analysis, validation, investigation, visualization, methodology, writing–review and editing. J. Adélaide: Visualization, methodology. P. Finetti: Software, validation. S. Garnier: Visualization, methodology. A. Guille: Software, validation. N. Carbuccia: Visualization, methodology. A. Farina: Validation, visualization, methodology. E. Agavnian: Validation, visualization, methodology. M. Chaffanet: Project administration. A. Gonçalves: Writing–review and editing. E. Charafe-Jauffret: Resources. E. Mamessier: Funding acquisition, writing–review and editing. D. Birmbaum: Funding acquisition, writing–review and editing. F. Bertucci: Resources, supervision, funding acquisition, investigation, writing–original draft, project administration, writing–review and editing. M. Lopez: Conceptualization, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft.

This work has been supported by Inserm, Institut Paoli-Calmettes and grants from the Ligue Nationale Contre Le Cancer (EL2016.LNCC/DaB, EL2019.LNCC/FB), Association Ruban Rose (to F. Bertucci) and Foundation Groupe EDF (to D. Birmbaum). We thank Jean-Charles Graziano and Jean-Christophe Orsoni for their helpful assistance in animal facilities, and Manon Richaud and Françoise Mallet for their contribution on the flow cytometry platform.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Tarantino
P
,
Carmagnani Pestana
R
,
Corti
C
,
Modi
S
,
Bardia
A
,
Tolaney
SM
, et al
.
Antibody-drug conjugates: Smart chemotherapy delivery across tumor histologies
.
CA Cancer J Clin
2022
;
72
:
165
82
.
2.
Powles
T
,
Rosenberg
JE
,
Sonpavde
GP
,
Loriot
Y
,
Duran
I
,
Lee
JL
, et al
.
Enfortumab vedotin in previously treated advanced urothelial carcinoma
.
N Engl J Med
2021
;
384
:
1125
35
.
3.
Drago
JZ
,
Modi
S
,
Chandarlapaty
S
.
Unlocking the potential of antibody-drug conjugates for cancer therapy
.
Nat Rev Clin Oncol
2021
;
18
:
327
44
.
4.
Fabre-Lafay
S
,
Monville
F
,
Garrido-Urbani
S
,
Berruyer-Pouyet
C
,
Ginestier
C
,
Reymond
N
, et al
.
Nectin-4 is a new histological and serological tumor associated marker for breast cancer
.
BMC Cancer
2007
;
7
:
73
.
5.
M-Rabet
M
,
Cabaud
O
,
Josselin
E
,
Finetti
P
,
Castellano
R
,
Farina
A
, et al
.
Nectin-4: a new prognostic biomarker for efficient therapeutic targeting of primary and metastatic triple-negative breast cancer
.
Ann Oncol
2017
;
28
:
769
76
.
6.
Reymond
N
,
Fabre
S
,
Lecocq
E
,
Adelaide
J
,
Dubreuil
P
,
Lopez
M
.
Nectin4/PRR4, a new afadin-associated member of the nectin family that trans-interacts with nectin1/PRR1 through V domain interaction
.
J Biol Chem
2001
;
276
:
43205
15
.
7.
Challita-Eid
PM
,
Satpayev
D
,
Yang
P
,
An
Z
,
Morrison
K
,
Shostak
Y
, et al
.
Enfortumab vedotin antibody-drug conjugate targeting nectin-4 is a highly potent therapeutic agent in multiple preclinical cancer models
.
Cancer Res
2016
;
76
:
3003
13
.
8.
Derycke
MS
,
Pambuccian
SE
,
Gilks
CB
,
Kalloger
SE
,
Ghidouche
A
,
Lopez
M
, et al
.
Nectin 4 overexpression in ovarian cancer tissues and serum: potential role as a serum biomarker
.
Am J Clin Pathol
2010
;
134
:
835
45
.
9.
Nishiwada
S
,
Sho
M
,
Yasuda
S
,
Shimada
K
,
Yamato
I
,
Akahori
T
, et al
.
Nectin-4 expression contributes to tumor proliferation, angiogenesis and patient prognosis in human pancreatic cancer
.
J Exp Clin Cancer Res
2015
;
34
:
30
.
10.
Pavlova
NN
,
Pallasch
C
,
Elia
AE
,
Braun
CJ
,
Westbrook
TF
,
Hemann
M
, et al
.
A role for PVRL4-driven cell-cell interactions in tumorigenesis
.
Elife
2013
;
2
:
e00358
.
11.
Takano
A
,
Ishikawa
N
,
Nishino
R
,
Masuda
K
,
Yasui
W
,
Inai
K
, et al
.
Identification of nectin-4 oncoprotein as a diagnostic and therapeutic target for lung cancer
.
Cancer Res
2009
;
69
:
6694
703
.
12.
Rosenberg
JE
,
O'Donnell
PH
,
Balar
AV
,
McGregor
BA
,
Heath
EI
,
Yu
EY
, et al
.
Pivotal trial of enfortumab vedotin in urothelial carcinoma after platinum and anti-programmed death 1/programmed death ligand 1 therapy
.
J Clin Oncol
2019
:
JCO1901140
.
13.
Robinson
MD
,
McCarthy
DJ
,
Smyth
GK
.
edgeR: a Bioconductor package for differential expression analysis of digital gene expression data
.
Bioinformatics
2010
;
26
:
139
40
.
14.
Adelaide
J
,
Finetti
P
,
Bekhouche
I
,
Repellini
L
,
Geneix
J
,
Sircoulomb
F
, et al
.
Integrated profiling of basal and luminal breast cancers
.
Cancer Res
2007
;
67
:
11565
75
.
15.
Bertucci
F
,
Finetti
P
,
Guille
A
,
Adelaide
J
,
Garnier
S
,
Carbuccia
N
, et al
.
Comparative genomic analysis of primary tumors and metastases in breast cancer
.
Oncotarget
2016
;
7
:
27208
19
.
16.
Mistry
P
,
Stewart
AJ
,
Dangerfield
W
,
Okiji
S
,
Liddle
C
,
Bootle
D
, et al
.
In vitro and in vivo reversal of P-glycoprotein-mediated multidrug resistance by a novel potent modulator, XR9576
.
Cancer Res
2001
;
61
:
749
58
.
17.
Fordham
AM
,
Xie
J
,
Gifford
AJ
,
Wadham
C
,
Morgan
LT
,
Mould
EVA
, et al
.
CD30 and ALK combination therapy has high therapeutic potency in RANBP2-ALK-rearranged epithelioid inflammatory myofibroblastic sarcoma
.
Br J Cancer
2020
;
123
:
1101
13
.
18.
Callaghan
R
,
Luk
F
,
Bebawy
M
.
Inhibition of the multidrug resistance P-glycoprotein: time for a change of strategy?
Drug Metab Dispos
2014
;
42
:
623
31
.
19.
Weidner
LD
,
Fung
KL
,
Kannan
P
,
Moen
JK
,
Kumar
JS
,
Mulder
J
, et al
.
Tariquidar Is an Inhibitor and Not a Substrate of Human and Mouse P-glycoprotein
.
Drug Metab Dispos
2016
;
44
:
275
82
.
20.
Chen
R
,
Hou
J
,
Newman
E
,
Kim
Y
,
Donohue
C
,
Liu
X
, et al
.
CD30 downregulation, MMAE resistance, and MDR1 upregulation are all associated with resistance to brentuximab vedotin
.
Mol Cancer Ther
2015
;
14
:
1376
84
.
21.
Mittendorf
EA
,
Wu
Y
,
Scaltriti
M
,
Meric-Bernstam
F
,
Hunt
KK
,
Dawood
S
, et al
.
Loss of HER2 amplification following trastuzumab-based neoadjuvant systemic therapy and survival outcomes
.
Clin Cancer Res
2009
;
15
:
7381
8
.
22.
Sakai
H
,
Tsurutani
J
,
Iwasa
T
,
Komoike
Y
,
Sakai
K
,
Nishio
K
, et al
.
HER2 genomic amplification in circulating tumor DNA and estrogen receptor positivity predict primary resistance to trastuzumab emtansine (T-DM1) in patients with HER2-positive metastatic breast cancer
.
Breast Cancer
2018
;
25
:
605
13
.
23.
Coates
JT
,
Sun
S
,
Leshchiner
I
,
Thimmiah
N
,
Martin
EE
,
McLoughlin
D
, et al
.
Parallel genomic alterations of antigen and payload targets mediate polyclonal acquired clinical resistance to sacituzumab govitecan in triple-negative breast cancer
.
Cancer Discov
2021
;
11
:
2436
45
.
24.
Gandullo-Sanchez
L
,
Ocana
A
,
Pandiella
A
.
Generation of antibody-drug conjugate resistant models
.
Cancers (Basel)
2021
;
13
:
4631
.
25.
Bukowski
K
,
Kciuk
M
,
Kontek
R
.
Mechanisms of multidrug resistance in cancer chemotherapy
.
Int J Mol Sci
2020
;
21
:
3233
.
26.
Chen
R
,
Herrera
AF
,
Hou
J
,
Chen
L
,
Wu
J
,
Guo
Y
, et al
.
Inhibition of MDR1 overcomes resistance to brentuximab vedotin in hodgkin lymphoma
.
Clin Cancer Res
2020
;
26
:
1034
44
.
27.
Wei
W
,
Lin
Y
,
Song
Z
,
Xiao
W
,
Chen
L
,
Yin
J
, et al
.
A20 and RBX1 regulate brentuximab vedotin sensitivity in hodgkin lymphoma models
.
Clin Cancer Res
2020
;
26
:
4093
106
.
28.
Boesch
D
,
Gaveriaux
C
,
Jachez
B
,
Pourtier-Manzanedo
A
,
Bollinger
P
,
Loor
F
.
In vivo circumvention of P-glycoprotein-mediated multidrug resistance of tumor cells with SDZ PSC 833
.
Cancer Res
1991
;
51
:
4226
33
.
29.
Binkhathlan
Z
,
Lavasanifar
A
.
P-glycoprotein inhibition as a therapeutic approach for overcoming multidrug resistance in cancer: current status and future perspectives
.
Curr Cancer Drug Targets
2013
;
13
:
326
46
.
30.
Pusztai
L
,
Wagner
P
,
Ibrahim
N
,
Rivera
E
,
Theriault
R
,
Booser
D
, et al
.
Phase II study of tariquidar, a selective P-glycoprotein inhibitor, in patients with chemotherapy-resistant, advanced breast carcinoma
.
Cancer
2005
;
104
:
682
91
.
31.
Matsui
H
,
Takeshita
A
,
Naito
K
,
Shinjo
K
,
Shigeno
K
,
Maekawa
M
, et al
.
Reduced effect of gemtuzumab ozogamicin (CMA-676) on P-glycoprotein and/or CD34-positive leukemia cells and its restoration by multidrug resistance modifiers
.
Leukemia
2002
;
16
:
813
9
.
32.
Liu-Kreyche
P
,
Shen
H
,
Marino
AM
,
Iyer
RA
,
Humphreys
WG
,
Lai
Y
.
Lysosomal P-gp-MDR1 confers drug resistance of brentuximab vedotin and its cytotoxic payload monomethyl auristatin E in tumor cells
.
Front Pharmacol
2019
;
10
:
749
.
33.
Chang
CH
,
Wang
Y
,
Zalath
M
,
Liu
D
,
Cardillo
TM
,
Goldenberg
DM
.
Combining ABCG2 inhibitors with IMMU-132, an anti-trop-2 antibody conjugate of SN-38, overcomes resistance to sn-38 in breast and gastric cancers
.
Mol Cancer Ther
2016
;
15
:
1910
9
.

Supplementary data