Understanding and targeting resistance mechanisms in cancer

1Cancer remains a global health burden, ranking first or second in the leading causes of death before the age of 70 years in more than 60% of the countries worldwide. In 2020, there were an estimated 19.3 million newly diagnosed cases and 10 million cancer deaths globally. Currently, the major therapies for cancer include surgery, radiation therapy, chemotherapy, hormone therapy, targeted therapy, and immunotherapy, either as monotherapy or in combination.,,, Despite the great development and improvement of cancer treatment in recent decades, resistance to cancer therapies has been a commonly observed phenomenon in clinical practice., Moreover, cancer cells with resistant characteristics often exhibited cross‐resistance to a variety of anticancer drugs that can be structurally irrelevant, namely, the multidrug resistance (MDR) phenomenon. MDR has been a major obstacle impeding therapeutic success and a dominating cause of cancer relapse and cancer‐related death.

2Cancer therapeutic resistance can be categorized into intrinsic and acquired resistance based on the timeline of resistance occurrence. The intrinsic resistance, also known as primary resistance, is mediated by the endogenous factors that are present in tumor cells or tissues before therapeutic applications, which provide cancer cells with survival advantages and adaptability to primary therapeutic stress.,, However, acquired drug resistance is developed after receiving cancer treatment, which is generally mediated by the adaptive alterations against the given therapy in initially sensitive tumors, resulting in compromised treatment effectiveness.,, The reduced responsiveness of cancer cells can be associated with various mechanisms, which usually involve the coactions of genetic factors and nongenetic contributors. Genetic factors in tumor cells have been considered critical contributors to therapeutic resistance, such as genetic diversity, acquired mutations of drug targets, amplification of oncogenes in compensatory or bypass pathways, and epigenetic modifications, which can further affect intratumor heterogeneity, tumor cell plasticity, DNA repair, and the susceptibility of tumor cells to cell death pathways, leading to multifactor‐mediated resistance.,, However, resistant cases in the absence of genetic alteration have been extensively recognized in different types of cancers. Phenotype changes may be independent of genotype alteration in resistance mediated by metabolic inactivation of drugs,, reduced intracellular drug concentration by transporters,,, drug compartmentation, and drug‐induced reversible transcriptional or posttranslational regulations on adaptive pathways. In addition to the factors within the tumor cells, the tumor microenvironment (TME) is also considered to be involved in the development of resistance in some cancers (Figure 1).

3Notably, the resistance mechanisms are not mutually exclusive and can act jointly to induce therapeutic irresponsiveness in cancer. Therefore, to develop effective strategies to obviate cancer resistance, it is an urgent need to gain a better understanding of the mechanisms collectively. In this review, we will discuss the discoveries of the diverse mechanisms of cancer resistance, the approaches to improve anticancer efficacy via targeting specific mechanisms, and the progress in developing polytherapy fighting against multiple resistance mechanisms.

4Drug resistance in cancer can arise from genetic, epigenetic, and microenvironmental factors. A better understanding of the molecular mechanisms underlying cancer resistance is necessary to develop effective strategies to overcome it. In the following context, the identified resistance mechanisms will be discussed with reviews of recent research updates, followed by the development of corresponding combating approaches, such as novel drug designs, combination therapies with specific targets, tackling alternative signaling pathways, and the modulation of the TME.

5The activation and deactivation processes of many chemotherapeutic agents are regulated by drug‐metabolizing enzymes (DMEs). Dysregulation of DMEs and metabolic signaling pathways, which can lead to the detoxification of drugs or failure in the conversion of drugs into active metabolites, is one of the major mechanisms for chemoresistance in cancer.

6Some anticancer drugs require activation by metabolic enzymes. For example, the biotransformation of irinotecan into the active metabolite SN‐38 is catalyzed by carboxylesterase, and the thymidine phosphorylase is responsible for the activating metabolism of 5‐fluorouracil (5‐FU) into fluorodeoxyuridine monophosphate. Cytarabine (AraC), a nucleoside drug used in patients with acute myeloid leukemia (AML) and non‐Hodgkin's lymphoma, relies on the phosphorylation catalyzed by deoxycytidine kinase (DCK) to become the cytotoxic form cytarabine triphosphate, and the deficiency of DCK has been considered to be associated with AraC resistance in AML. Wu et al. reported that DCK mutations were found in 4 of 10 patients with AML relapse after complete remission and high‐dose AraC post‐remission treatment. In vitro studies using drug‐selected AraC‐resistant AML and lymphoma cell models indicated that mutation or deficient expression and function of DCK could be acquired after receiving AraC therapy, resulting in the reduced cellular response to AraC and possible cross‐resistance to other nucleoside drugs like gemcitabine.,, A potential strategy to overcome DCK deficiency‐mediated nucleoside drug resistance is using nucleoside analog phosphate prodrugs with anticancer efficacy without the demand for DCK phosphorylation, among which NUC‐1031, a gemcitabine phosphoramidate prodrug, has entered clinical trials for the treatment of gemcitabine‐resistant cancers., Other reported attempts in preclinical evaluations include enhancing DCK activity by etoposide in combination with AraC for the treatment of AML, and bypassing the drug resistance mechanism associated with downregulated DCK expression using second‐generation deoxyadenosine analog clofarabine in acute lymphoblastic leukemia cells.

7The other aspect of metabolism‐associated anticancer drug inactivation is the detoxification of drugs by metabolizing enzymes. For instance, certain isoforms of aldehyde dehydrogenases (ALDH), including ALDH1A1 and ALDH3A1, have been reported to specifically cause resistance against chemotherapeutic drugs of nitrogen mustard type, such as cyclophosphamide, mafosfamide, and ifosfamide. In contrast, cytochrome P450 (CYP450) in phase I metabolism, and glutathione‐S‐transferase (GST) as well as uridine diphosphoglucuronosyltransferase (UGT) in phase II conjugating biotransformation, are involved in the inactivation of a broader spectrum of anticancer drugs.,

8The CYP450 enzyme superfamily consists of 57 members that are responsible for the phase I oxidation of most clinically used drugs. Of the CYP450 enzymes, CYP1B1, 2C8, 3A4, and 3A5 have been reported to be correlated to cancer resistance,, with CYP1B1 being mostly studied, which is found to be exclusively overexpressed in various types of cancers but has relatively low expression in normal tissues. Intratumoral CYP1B1 overexpression may contribute to the diminished effectiveness of a diversity of chemotherapeutic drugs, such as paclitaxel and docetaxel, mitoxantrone, flutamide, and gemcitabine.,,, However, whether the resistance against these drugs is contributed by CYP1B1‐catalyzed drug inactivation remains controversial. McFadyen et al. suggested that docetaxel was metabolized by CYP1B1, and the docetaxel resistance in CYP1B1 overexpressing Chinese hamster ovary cancer cells could be reversed by a CYP1 inhibitor., However, Martinez et al. later demonstrated that CYP1B1 did not directly inactivate docetaxel because the cytotoxicity of docetaxel in MCF‐7 breast cancer cells was not affected by silencing CYP1B1 or adding recombinant CYP1B1. Although it is possible that CYP1B1 may play different roles in regulating drug resistance in different cancer types, the inhibition of CYP1B1 activity has been considered to be a therapeutic target for improving chemotherapy. Although lacking high selectivity, phytochemicals are the most common source of CYP1B1 inhibitors, including stilbene, flavonoids, coumarins, anthraquinones, and alkaloids.,, By modifying the structures of phytochemicals, highly potent and selective inhibitors of CYP1B1 have been developed and under preclinical evaluation, such as TMS ((E)‐2,3′,4,5′‐tetramethoxystilbene) and α‐naphthoflavone derivatives.

9GSTs are involved in drug detoxification in phase II metabolism and catalyze the glutathione (GSH) conjugation to drugs. Overexpression of GSTs in cancer cells may enhance the detoxification of anticancer drugs. Moreover, after GSH conjugation, the conjugated drugs may become substrates of ATP‐binding cassette (ABC) transporters, particularly the MDR‐associated proteins (MRPs, belongs to ABCC subfamily), and got actively pumped out of the cancer cells.,, Among the GST superfamily, GST alpha 1 (GSTA1), GST Mu 2 (GSTM2), and GST Pi 1 (GSTP1) have been found to be associated with cisplatin resistance in ovarian, lung, and gastric cancers. Notably, the overexpression of GSTP1 has been considered a factor impairing the efficacy of platinum drugs like cisplatin, carboplatin, and oxaliplatin by promoting the formation of platinum‐GSH conjugates., The I150V polymorphism of GSTP1 with a phenotype of reduced enzymatic capacity has been correlated with better therapeutic outcomes in gastric cancer patients receiving oxaliplatin‐based chemotherapy., Besides platinum drugs, the sensitivity to doxorubicin and various alkylating agents in cancer may also be affected by the detoxification process mediated by GSTA subclass, GSTP1, and GSTM1 enzymes., However, additional mechanisms other than GSH conjugation may be involved in GST‐related chemoresistance to doxorubicin and alkylators, such as free radical scavenging and apoptosis suppression via the inhibition of the mitogen‐activated protein kinase (MAPK) pathway.,, Thus, there has been an increasing focus on targeting the GSTs to overcome chemoresistance. Inhibitors of GSTs have been shown effective in treating resistant cancer cells. Ethacrynic acid and analogs are among the earliest investigated GST inhibitors that exhibited resensitizing effects on tumor cells to alkylating agents via covalent binding to GSTs and reducing the enzyme activity., NBDHEX (6‐(7‐nitro‐2,1,3‐benzoxadiazol‐4‐ylthio)hexanol) is another potent inhibitor against GSTP1, GSTM2, and other GST isoenzymes, which has been found beneficial in combating cisplatin‐resistant osteosarcoma when used in combination with cisplatin. Apart from inhibiting GST activity, NBDHEX can also induce the dissociation of GSTP1 from its complex with c‐Jun N‐terminal kinase or tumor necrosis factor receptor‐associated factor 2 thereby promoting the activation of apoptosis pathways in cancer cells. However, one of the significant roadblocks that GST inhibitors encounter in clinical trials is their insufficient specificity, prompting the development of novel NBDHEX analogs with improved selectivity, among which MC3181 was screened out as a specific GSTP1‐1 inhibitor and tested highly efficient in inhibiting vemurafenib‐resistant melanoma in vitro and in vivo., Although the GST inhibitors showed promising effects on inducing cell apoptosis via inhibiting GST activities, investigations on combinations with chemotherapeutic drugs are lacking; thus, whether the inhibitors can reduce drug inactivation remains to be determined. On the other hand, utilizing prodrugs like canfosfamide (TLK286) and brostallicin, which are activated by GSTs to become cytotoxic, has entered clinical evaluations as a practical strategy to circumvent the GST‐mediated resistance mechanism.,

10Similar to GSTs, the UGT enzymes are involved in phase II metabolism, and they catalyze the glucuronidation process. Among the three subfamilies (UGT1A, 2A, and 2B) of the UGTs, UGT1A enzymes, particularly UGT1A1, have been shown to overexpress in tumor tissues and play a role in resistance to anticancer drugs. UGT1A1 is physiologically responsible for facilitating bilirubin elimination by catalyzing the glucuronic acid conjugation of bilirubin. Elevated tumoral UGT1A1 levels can increase the glucuronidation of SN‐38, the active metabolite of anticancer drug irinotecan, resulting in the inactive SN‐38 glucuronide and reduced therapeutic efficacy. UGT1A enzymes have also been reported to be associated with primary resistance to some targeted drugs, such as heat shock protein HSP90 inhibitors ganetespib and luminespib and epidermal growth factor receptor (EGFR) inhibitor erlotinib. Drug‐induced increase of UGT1A expression has been reported in AML patients treated with ribavirin, which resulted in acquired resistance. Apart from substrate drugs, the efficacy of monoclonal antibody anticancer drugs may be affected by UGT1As. For instance, UGT1A6 has been found to be correlated with resistance to programmed cell death 1 (PD‐1) antibody nivolumab in patients with advanced renal clear‐cell cancer. As UGTs are not known to conjugate proteins like antibodies, UGT‐mediated resistance to nivolumab may not be related to directed drug inactivation, but to other mechanisms, such as indirect metabolism regulations by UGTs and other resistance‐correlated signaling regulations involving UGTs. Because of their association with hampered drug response, the expression of UGT1As might be a useful indicator for patient stratifying so as to avoid the application of substrate drugs to patients who are likely to have low response due to high UGT1A levels. Meanwhile, UGTs can be potential pharmacological targets to overcome drug resistance in cancer. However, the use of UGT inhibitors has been limited by toxic side effects, and selectivity remains the major challenge in developing novel inhibitors of UGTs.,

11Notably, direct detoxification is not the only mechanism of DME‐associated cancer resistance, other involving factors like drug–drug interactions and signaling molecules inherent in cancer cells should be taken into consideration in developing strategies targeting metabolic enzymes in cancer. Regulating the expression and activity of DMEs through regulatory signaling transductions can be beneficial to overcoming cancer resistance not only by means of retaining drug bioavailability but also in other aspects like inhibiting cancer cell proliferation or invasion.

12Drugs can cross the cell membrane through diffusion, endocytosis, or transporters, which can be affected by the permeability and lipid composition of the plasma membrane, and the functions or expression levels of membrane transporters. Alteration of cell membrane structure can impair drug diffusion across the plasma membrane and the endocytosis process. The plasma membranes from drug‐resistant cancer cells have a different lipid composition compared to those from the parental drug‐sensitive cells: The cholesterol and phospholipid levels are elevated, and the protein/lipid ratio is increased up to 60% in MDR cells compared to sensitive cells. Besides, the slight alkalic pH of the cytoplasm in MDR cancer cells could attenuate the repulsions between the polar groups from membrane lipids by shielding the negative charges, thereby increasing lipid packing and membrane rigidity. Therefore, MDR cells have plasma membranes with a relatively lower fluidity and reduced permeability leading to decreased drug absorption. Chemotherapeutic drugs, including vinblastine, doxorubicin, and cisplatin, are vulnerable to this resistance mechanism.,, Additionally, using model membranes by molecular dynamic simulations, Rivel et al. recently demonstrated that, cancer cell membranes are common with the loss of lipid asymmetry compared to normal cell membranes, which may be a contributor to the slower diffusion of cisplatin into cancer cells. These findings have suggested that lipid composition assessment may be useful for cancer prognosis, and modulating cell membranes can be a potential strategy to ameliorate drug resistance in cancer.

13It has been shown that treating cancer cells with short‐chain ceramides can increase cell membrane permeability and fluidity, resulting in an increased uptake of amphiphilic anticancer drugs such as doxorubicin, either in free form or encapsulated form with lipid‐based nanoparticles., The other way to modulate the lipid composition of the cell membrane is by enhancing sphingomyelinase activity using an agonist, such as daunorubicin, etoposide, and ara‐C, thereby decreasing sphingomyelin levels and increasing ceramide levels, leading to increased membrane fluidity., Moreover, biomimetic cell membrane‐coated nanoparticle delivery systems are gaining increased recognition. Cancer cell membrane‐based nanoparticles contain surface proteins from cancer cells; therefore, they can reduce side effects by specifically targeting cancer cells through homotypic binding. Fang et al. reported that the uptake of membrane‐coated nanoparticles in human breast cancer MDA‐MB‐435 cells was increased by 20‐fold compared to naked nanoparticles. Many studies have demonstrated the selectivity and biosafety of cancer cell membrane‐based biomimetic nanoparticles using in vitro and in vivo MDR cancer models,,, suggesting a novel drug delivery strategy to improve therapeutic efficacy for MDR cancer.

14Intracellular drug concentrations largely depend on the activity of uptake transporters. The main transporters involved in drug uptake are the drug solute carriers (SLCs). The SLC superfamily consists of more than 400 members categorized into 52 families. Chemoresistance‐relevant SLCs include the organic anion transporting proteins (OATPs), organic cation transporters, concentrative nucleoside transporters, equilibrative nucleoside transporters, and copper transporters., Reduced uptake of chemotherapeutic drugs can be caused by either genetic variants resulting in truncated uptake transporter proteins with reduced or absent function or acquired expression changes in uptake transporters mediated by pharmacological selection pressures. Among them, OATPs (SLCO family) are mostly studied. In particular, amplified expression of OATP1A2, OATP1B1, and OATP1B3 has been found in various cancer tissues, which is considered associated with chemosensitivity because of its role in the uptake of several classes of anticancer drugs, such as taxanes, platinum‐based drugs, camptothecin analogs, methotrexate, and some tyrosine kinase inhibitors (TKIs)., Downregulation of OATP1B3 was reported in a patient‐derived docetaxel‐resistant prostate tumor xenograft model, where the intratumoral concentrations of docetaxel and cabazitaxel were both lower than chemotherapy‐naive tumors. In contrast, prostate tumors with OATP1B3 expression exhibited an increased uptake of both taxanes and higher chemotherapeutic sensitivity. Overexpression of OATP1B3 in prostate cancer may be beneficial to chemotherapy but can be detrimental to hormone therapy because OATP1B3 drives testosterone uptake in cancer cells, leading to the development of resistance to androgen deprivation therapy (ADT). Increased OATP1B3 expression in prostate cancer can be induced by ADT, which provides a clue for sequential therapy design: Upregulated tumor SLCO expression following ADT could potentially enhance the uptake of chemotherapeutic drugs like taxanes and improve treatment response. Interestingly, the genetic variant of OATP transporters may confer resistance to TKIs via the uptake activity instead of enhancing sensitivity.

15Haberkorn et al. recently discovered that cancer‐type OATP1B3 protein, a splice variant of liver‐type OATP1B3, is localized in the lysosomal membrane of colorectal cancer (CRC) cells and contributes to the transport of encorafenib and vemurafenib into lysosomes, resulting in decreased drug concentrations in the cytoplasm and reduced drug efficacy. Therefore, it is necessary to identify the genotype of intratumorally expressed SLCs in the process of confirming their roles in drug resistance.

16Although enhancing drug uptakes by modulating SLC transporters may improve chemosensitivity, it may not be a practical approach in cancer treatment. Due to the ubiquitous expression of SLCs in many tissues, modulating the function of SLCs could interfere with the transport activities of nutrients and xenobiotics in normal tissues, which can lead to unfavorable side effects and drug–drug interactions. On the other hand, besides the uptake of anticancer drugs, modulating SLCs may also influence the nutrient uptake in cancer cells, risking promoting cancer progression and metastasis. The best strategy to overcome the uptake transporter‐mediated resistance in cancer may be to use drugs or delivery systems that can circumvent the aberrant transporter, either with improved diffusion efficacy or the capability to utilize alternative transporters.

17In addition to reduced drug uptake, drug efflux mediated by the overexpression of ABC transporters is the major mechanism contributing to insufficient intracellular drug concentration. In humans, the ABC transporter superfamily consists of 49 members categorized into 7 subfamilies (ABCA–ABCG) functioning to transport various substrates, including lipids, ions, peptides, and xenobiotics. The involvement of ABC transporters in MDR was first discovered in 1976 when ABCB1, which is also known as P‐glycoprotein (P‐gp) or multidrug resisitance 1 (MDR1) protein, was discovered in mouse MDR cell lines. To date, at least 13 ABC transporters have been found to directly mediate chemoresistance via the efflux of anticancer drugs. These ABC transporters mainly distribute on the plasma membrane and are triggered upon binding to substrate drugs, resulting in an ATP hydrolysis–driven conformational change of the transporter and the extrusion of the substrate drugs. Consequently, the overexpression of these transporters has been correlated with poor chemotherapy response and unfavorable patient prognosis in many different types of cancers. Most studies have focused on the ABC transporters with wide spectrums of substrates, such as ABCB1, ABCG2 (also known as breast cancer resistance protein), and the ABCC subfamily (also known as multidrug resistance‐associated proteins, MRPs) with ABCC1 and ABCC10 as representative members. Many cytotoxic chemotherapeutic drugs can be pumped out of cancer cells by ABCB1 and ABCCs, such as taxanes, vinca alkaloids, and anthracyclines. The spectrum of anticancer drugs ABCG2 confers resistance to largely overlap with that of ABCB1 and ABCCs but with some specificities; anthracenedione (such as mitoxantrone) and camptothecins (such as topotecan and SN‐38) are especially vulnerable to the efflux activity of ABCG2, whereas taxanes and vinca alkaloids are not much affected by ABCG2. Certain clinically used kinase inhibitors, such as imatinib, nilotinib, dasatinib, palbociclib, and some newly developed targeted drugs, such as OTS964 and ARS‐1620, have also been reported as substrates of ABC transporters.,,, Moreover, ABCCs generally have high affinities to GSH‐conjugated or glucuronate‐conjugated forms of drugs; therefore, the efflux activity of ABCC transporters may be regulated by phase II metabolic enzymes. Additionally, for brain tumors and metastatic cancer in the brain, chemosensitivity is also affected by the o expression of ABCB1 and ABCG2 transporters at the luminal membrane of endothelial cells of brain microvessels, which are known to impede anticancer drug delivery across the blood–brain barrier.,

18Ongoing efforts have been taken to develop modulators of ABC transporters in order to overcome this resistance mechanism. Numerous small molecule compounds have been developed and tested as inhibitors against ABC transporters, particularly ABCB1. There have been three generations of ABCB1 inhibitors; however, these inhibitors have failed to enter clinical use as combination therapy with substrate anticancer drugs due to different drawbacks. The first generation of ABCB1 inhibitors, such as verapamil, erythromycin, and cyclosporine A, mostly lacks sufficient therapeutic efficacy, which results in considerable side effects by the required high dosages. Although the second‐generation inhibitors, such as dexverapamil and valspodar, have exhibited improved potency, they failed to proceed to clinical application because of their unwanted interactions with the CYP450 enzymes leading to unfavorable pharmacokinetic profiles. The third generation, exampled by tariquidar and zosuquidar, is able to overcome the selectivity problem; however, the performance in clinical settings turns out to be unsuccessful due to interpatient variability. In the recent decade, some repurposing targeted anticancer drugs, such as selonsertib, tepotinib, and poziotinib, have been discovered as dual inhibitors for ABCB1 and ABCG2.,, Inhibitors that can antagonize both ABCB1‐ and ABCC1‐mediated resistance have also been reported, exampled by cediranib and CBT‐1 (tetrandrine)., Targeting multiple ABC transporters may be more prospective given that co‐expression of different ABC transporters is common in tumor tissues and multitargeting drugs may be beneficial to avoid compensatory upregulation of another transporter induced by selective inhibition of a single transporter.

19Silencing the efflux transporters using gene editing technologies such as siRNA and CRISPR/Cas9 system has been increasingly investigated as a novel approach to reducing drug resistance, which can potentially benefit genetic variants with intrinsic ABC transporter‐related resistance. However, many studies conducted gene editing in cell line‐based settings before the step of tumor xenograft model establishment, which is unlikely to represent gene therapy for patients. Cancer‐targeted delivery remains a major challenge in developing gene therapies against ABC transporters. Another trending strategy against ABC transporter‐dependent MDR is to target the regulatory factors of the ABC transporter expression, including transcriptional regulation and epigenetic modifications on the gene, and posttranslational modifications on the protein. Inhibiting the cancer‐specific regulators of ABC transporters may render an advantage to targeting tumor tissues selectively and increasing safety to normal tissues. Dysregulated noncoding RNAs, including microRNAs (miRNAs) and long noncoding RNAs (lncRNAs), have been found to be associated with the overexpression of ABCB1, ABCG2, and ABCCs in chemoresistant cancers. The effect on suppressing ABC transporters and reversing cancer MDR by nanoparticle‐delivered ncRNAs mimics or inhibitors has been verified from in vitro and in vivo studies., Another epigenetic factor‐linked overexpression of ABC transporters in cancer is the hypomethylation of the transporter gene promoter, and evidence has shown that agonists of DNA methyl transferases can induce the hypermethylation of the ABCG2 promoter resulting in the lower ABCG2 expression and increased intracellular concentration of substrate drugs., Alternatively, the expression of ABC transporters can be modulated by targeting the upstream regulatory pathways. For instance, in breast cancer cells, inhibiting the overexpressed receptor tyrosine kinase‐like orphan receptor 1 (ROR1), which is an upstream regulator of ABCB1 transcription via MAPK/extracellular signal‐regulated kinase (ERK) and p53 pathways, can reduce ABCB1‐mediated drug efflux and resensitize breast cancer cells to doxorubicin. Another recent study showed that suppressing mitochondrial respiration using methylation‐controlled J protein mimetics can decrease ATP production thereby decreasing the energy for ABCB1 and ABCG2 efflux activity and overcoming chemoresistance in vitro and in vivo.

20Indeed, the direct approach to circumvent this resistance mechanism is to avoid using substrate drugs or develop novel anticancer drugs that can bypass the efflux transporters. This may rely on advanced computer‐aided drug designs to modify drug structures of current known and actively reported substrates of ABC transporters.

21The leading limitation of targeted cancer therapies is that cancer cells can be intrinsically irresponsive or acquire resistance after a period of treatment because of either the mutation of target molecules (“on‐target” mechanism) or tumor cells gaining survival advantages from a new mechanism independent of the target (“off‐target” mechanism), such as the activation of downstream signaling pathways or compensatory pathways, (Figure 2).

22The primary approach to overcome “on‐target” mechanism is to develop new generations of targeted drugs against drug‐resistant tumors. For instance, the first‐ and second‐generation EGFR–TKIs, such as gefitinib, erlotinib, and dacomitinib, are less effective in treating patients with additional EGFR mutation T790M. This has led to structural modification and the development of the third‐generation EGFR–TKIs therapy. Osimertinib (AZD9291), a third‐generation EGFR–TKI, presented a superior clinical response and outcome in EGFR‐mutated non‐small cell lung cancer (NSCLC). However, the rapidly acquired resistance to osimertinib conferred by EGFR C797S mutation has been observed. Besides, some patients who exhibit primary resistance remain unresponsive to third‐generation and other newly developed EGFR–TKIs. Complete remission is rare, and all patients eventually develop resistance, suggesting that both primary and acquired resistance mechanisms reduce the efficacy of the drug., To reverse these types of resistance, more recently, the fourth‐generation EGFR inhibitors, which can inhibit both T790M and C797S signaling, have been introduced into clinical evaluation. So far, EAI045 is the first allosteric TKI developed for this purpose. The C797S mutation is unlikely to impair the efficacy of EAI045 because its allosteric binding pocket is not affected by this cysteine residue. EGFR receptor dimerization invalidates drug‐mediated inhibition alone. The activity against T790M and C797S can be restored by a combination regimen with cetuximab, an antibody against EGFR dimerization. Other fourth‐generation EGFR‐TKIs, such as JND3229 and JBJ‐04‐125‐02, were recently found to be active in EGFR C797S‐T790M‐L858R signal transduction in vitro and in vivo.,

23For the first‐generation tropomyosin receptor kinase (TRK) kinase inhibitors, such as entrectinib and larotrectinib, the secondary mutations occurring at the ATP binding pocket of the TRK kinase domain, including G667C and G595R mutations in NTRK1 gene, and G696A and G623R mutations in NTRK3 gene, are common acquired‐resistance mechanisms., In addition, second‐generation TRK inhibitors like selitrectinib have been developed to overcome this acquired resistance.

24The multitargeted TKI crizotinib was approved by the Food and Drug Administration (FDA) in 2011 to treat patients with advanced NSCLC harboring ALK rearrangements. However, despite a high response rate of 60% in ALK‐rearranged NSCLC, most patients develop resistance to crizotinib, typically within 1–2 years. Studies of ALK‐rearranged lung cancers with acquired resistance to crizotinib have identified ALK fusion gene amplification and secondary ALK tyrosine kinase domain mutations in about one third of cases., To date, seven different acquired resistance mutations have been identified among crizotinib‐resistant patients. The most frequently identified secondary mutations are L1196M and G1269A. In addition to these mutations, the 1151T‐ins, L1152R, C1156Y, G1202R, and S1206Y mutations have also been detected in crizotinib‐resistant cancers.,, In approximately one third of crizotinib‐resistant tumors, there is evidence of activation of bypass signaling tracts such as EGFR. In the remaining one third of crizotinib‐resistant tumors, the resistance mechanisms remain to be identified. Next‐generation ALK inhibitors with improved potency and selectivity compared with crizotinib have been developed to overcome crizotinib resistance in the clinic. The ability of several ALK TKIs (TAE684, AP26113, ASP3026, and CH5424802) to inhibit ALK activity was evaluated in models harboring different ALK secondary mutations., These studies revealed variable sensitivity to these ALK inhibitors depending on the specific resistance mutation present. For example, the gatekeeper L1196M mutation was sensitive to TAE684, AP26113, and ASP3026, whereas 1151T‐ins conferred resistance to all next‐generation ALK TKIs. Ceritinib, an ATP‐competitive, potent, and selective next‐generation ALK inhibitor, has shown favorable selectivity to ALK. In clinical studies involving ALK‐positive NSCLC patients, ceritinib has exhibited marked antitumor activity has been observed in both crizotinib‐relapsed and crizotinib‐naive patients., On the basis of this impressive clinical activity, ceritinib received FDA approval on April 29, 2014.

25Despite increasing successes in efforts to target oncogenic driver amplifications or mutations, several of the most formidable oncogenes and tumor suppressor genes remain undruggable, including RAS (KRAS, NRAS, and HRAS) and RAF (ARAF, BRAF, and CRAF). Effective targeting of KRAS signaling has been tough to realize in patients., Blocking the localization of KRAS at the plasma membrane, a vital element for its activation, has been ineffective because multiple compensatory pathways modulate this process. Similarly, targeting effector signaling downstream of KRAS has not achieved remarkable clinical benefits on the account of paradoxical signaling activation generated by the inhibitor or because of on‐target toxicity limiting the maximum tolerated dose in patients., Tumor cells with KRAS or BRAFV600E mutations are addicted to a downstream ERK signaling cascade for their growth, viability, and other malignant properties. The adaption of selective inhibitor target KRASG12C only briefly inhibits KRAS within 1–3 days of treatment, and initial signaling suppression is accompanied by the re‐storage of active KRAS and reactivation of ERK1/2 signaling pathway., Accumulation of active KRAS illustrates that compensatory activation of upstream receptor tyrosine kinases (RTKs), including EGFR, is to a large extent responsible for the adaptive alterations noted during KRASG12C inhibitor treatment. In RASMut tumors, RAF inhibitors (RAFis) are ineffective because they drive paradoxical ERK1/2 pathway activation and adventitious tumor progression. However, MAPK kinase (MEK) inhibitors (MEKis) do not present the same limitation as RAFis, but the relief of feedback inhibition and pathway reactivation restricts MEKi monotherapy in RASMut‐driven tumors., By the amplification of KRASG13D or BRAFV600E, CRC cells acquire resistance to the MEKi. Furthermore, MEKi resistance driven by BRAFV600E amplification is thoroughly reversible upon prolonged drug withdrawal as BRAFV600E amplification brings about a selective disadvantage in the absence of MEKi, and MEKi withdrawal drives ERK1/2 activation beyond a key point that is optimal for cell proliferation and viability. Remarkably, MEKi resistance driven by KRASG13D amplification is not reversible. Upon MEKi withdrawal, CRC cells experience an ERK1/2‐dependent epithelial‐to‐mesenchymal transition (EMT) and emerge resistance to frequently used chemotherapeutics instead of exhibiting growth defects. Consequently, the appearance of MEKi resistance, drug‐addiction, and the possible options of intermittent dosing schedules, may depend on the nature of the amplified oncogenes (e.g., KRASG13D BRAFV600E), further highlighting the hardships of targeting RAS and RAF mutant tumors. Mutations of tumor suppressor can also result in resistance to targeted therapies, for example, mutations in phosphatase and tensin homolog deleted on chromosome 10 (PTEN) can activate phosphatidylinositol‐3‐kinase (PI3K) β signaling in response to PI3Kα inhibitors, as well as reverse mutations in breast cancer 1 (BRCA1) or BRCA2 in response to poly(ADP‐ribose) polymerase (PARP) inhibitors.

26In NSCLC, the majority of all EGFR mutations are EGFRL858R in exon 21 and activating EGFR exon 19 deletions, and tumor cells harboring the specific activating mutations exhibit high sensitivity to EGFR–TKIs., These constitutively activate mutant EGFR oncoproteins signaling by regulating MAPK and PI3K/AKT/mTOR signaling cascades to promote tumorigenesis. In a series of EGFR–TKI‐resistant tumor samples, 1% of patients exhibited resistance to early‐generation EGFR–TKIs via the acquisition of BRAFV600E or BRAFG469A. Increased mTOR level was related to EGFR–TKI resistance in clinical samples, and in mouse models, the application of rapamycin (mTOR inhibitor) presented the positive progression of EGFRmut lung tumors. EGFR–TKIs in combination with PI3K/AKT/mTOR signaling pathway inhibitors have proved improved tolerability and efficacy. Another pattern of acquired resistance to kinase inhibition is the amplification of upstream genes, which can exemplify by mesenchymal–epithelial transition (MET) amplification leading to resistance to EGFR–TKIs.

27The clinical value of proto‐oncogene 1 (ROS1)‐directed TKIs was first prospectively explored in NSCLC patients, and in 2016, the FDA and EMA approved crizotinib for the treatment of advanced‐stage ROS1‐rearranged NSCLC. Afterward, new ROS1–TKIs got into clinical trial, leading to the FDA and Ministry of Health, Labor and Welfare of Japan approvals of entrectinib for the treatment of ROS1‐aberrant NSCLC. So far, all developed ROS1–TKIs are multikinase inhibitors, which can also target MET (e.g., crizotinib and cabozantinib), ALK (e.g., ceritinib and crizotinib), TRK (e.g., entrectinib and repotrectinib), and several other kinases (e.g., SRC, EGFR, JAK2, and FLT3) with equivalent or lower potency.,, Various compensatory pathways involved in TKI therapy of ROS1‐rearranged tumors progression, such as KRAS, NRAS, BRAF, HER2, EGFR, MEK, and MET. For example, mutations of KRASG12D and BRAFV600E have emerged with crizotinib treatment in the clinical testing, and NRASQ61K has occurred while entrectinib treatment. Upregulated HER2 phosphorylation has been identified in patient‐derived, crizotinib‐resistant CD74–ROS1‐positive CUTO23 (NSCLC cell line), and afatinib combined applied recovered crizotinib sensitivity in CUTO23. Usually, ROS1 fusions de novo barely co‐occur with EGFR mutations in clinic, but ROS1 fusions come into sight as resistance mechanisms to EGFR–TKI therapy in EGFR‐mutant NSCLC, suggesting feedback or potential signaling reciprocity among ROS1 and EGFR. In patients treated with ROS1–TKIs, activating PIK3CA mutations have also been reported., In ROS1 fusion‐positive NSCLCs patients treated with ROS1–TKIs, 11% were found to have concurrent MEK/MAPK alterations. Subsequently, it was proved that the deletion of MEK1 or MEKK1 in cells conferred resistance to ROS1–TKIs and combined ROS1 and MEKis inhibited cell growth. Until 2020, MET amplification was found to act as an important participator of ROS1–TKI (e.g., lorlatinib) resistance. ROS1 kinase domain mutations that trigger the resistance of ROS1 fusion‐positive tumors to ROS1–TKIs have been identified by various preclinical and clinical studies. Some of these mutations are paralogous to ALK resistance mutations that emerge with targeted therapy for ALK fusions, among which ROS1G2032R is the most common resistance substitution observed in patients treated with crizotinib. Considering its clinical benefits but its lack of activity against ROS1G2032R, lorlatinib can rebuild enduring disease control in ROS1‐rearranged tumors with the limited spectrum of resistance mutations. However, FDA has conferred repotrectinib as the Fast Track Designation agent in patients who previously experienced one ROS1–TKI and platinum‐doublet chemotherapy. Up to now, next‐generation ROS1–TKIs have not been approved for the treatment of ROS1 rearrangement‐positive tumors.

28The DNA damage response (DDR) is a mechanism for repairing drug‐induced DNA damage. Mutation or dysregulation of certain genes and DDR mechanisms are common in many types of cancer and can influence the sensitivity to DNA‐damaging chemotherapy drugs, such as cisplatin and 5‐FU., The nucleotide excision repair (NER) responsible for repairing DNA damage repairing and mismatch repair (MMR) responsible for maintaining genomic integrity are known to be related to 5‐FU resistance. It is demonstrated by Liu et al. that the acquired upregulation of the excision repair cross‐complementing 1 (ERCC1), a component of the NER system, was induced by 5‐FU treatment in gastric cancer cells, which can subsequently weaken the anticancer activity of 5‐FU. Further mechanism studies revealed that 5‐FU‐induced ERCC1 overexpression may be regulated by ERK 1/2 and p38 signaling‐mediated activation of the transcription factor c‐jun/activator protein‐1, which provides a clue to overcome this resistance mechanism using ERK inhibitors or p38 kinase inhibitors. In contrast, the role of MMR in 5‐FU resistance remains uncertain. It was recently reported by Oliver et al. that both MMR‐proficient and MMR‐deficient CRC cells exhibited 5‐FU resistance after a period of treatment, suggesting that 5‐FU chemoresistance in CRC cells may be independent of MMR status. Nevertheless, MMR genes human homolog 1 (hMLH1), a component of MMR, was upregulated in an MMR‐deficient cell line, which indicated a potential involvement in 5‐FU sensitivity.

29MMR deficient–associated microsatellite instability has been linked to cisplatin resistance in clinical germ cell tumor sample, whereas in ovarian cancer, cisplatin resistance is generally related to homologous recombination repair (HRR) for DNA damage. The BRCA1 and BRCA2 tumor suppressor genes are critical for the HRR of DNA double‐strand breaks by the HRR pathway; therefore, BRCA1/BRCA2‐mutated ovarian cancers are usually more sensitive to cisplatin. However, acquired secondary mutations of BRCA1/2 induced by long‐term cisplatin exposure can reconstruct the function of BRCA1/2 and enhance DNA repair, resulting in resistance to cisplatin as well as inhibitors of DNA repair poly (ADP‐ribose) polymerase (PARP). Recently, another DNA repair–related protein, actin‐like protein ACTL6A, has been reported to promote DNA lesion induced by cisplatin damage through the SWI/SNF chromatin remodeling complex in several types of cancers, suggesting a novel cisplatin‐resistance mechanism. As ACTL6A is also a component of NuA4/TIP60 histone acetylase, treatment with a histone deacetylase inhibitor can be useful to attenuate this resistance mechanism and resensitize tumor cells to cisplatin.

30In addition to gene mutations that activate DNA repair genes, epigenetic regulations, including DNA methylation, histone modifications, and miRNAs, can all affect the expression of DNA repair genes, contributing to drug resistance.

31DNA methylation can suppress gene expression. In the context of DNA repair–associated drug resistance, DNA methylation has been implicated in the regulation of key DNA repair genes, such as O 6‐methylguanine‐DNA methyltransferase (MGMT), BRCA1/2, and MLH1. MGMT promoter methylation is associated with a loss of MGMT protein expression and activity in the tumor and has been shown to correlate with a better outcome of therapy in several studies., Consequently, the methylation status of the MGMT promoter has emerged as a useful biomarker for predicting the treatment response of glioblastoma (GBM) patients to methylating chemotherapeutic agents like temozolomide. In breast cancer, the hypermethylation of the BRCA1 promoter has been associated with decreased BRCA1 expression and sensitivity to DNA‐damaging drugs and PARP inhibitors. In contrast, the hypermethylation of the MLH1 promoter has been associated with decreased MLH1 expression and DNA repair capacity in cancer cells, which can increase the accumulation of somatic mutation in cancer cells and induce drug resistance. These findings highlight the complex interplay among DNA methylation, DNA repair, and cancer resistance and suggest that targeting DNA methylation may be a promising strategy for overcoming drug resistance in cancer.

32Histone modifications also play an important role in DNA repair and cancer resistance. A recent study demonstrated that, in high‐grade serous ovarian carcinoma cell lines and patient‐derived xenograft models, the upregulation of histone methyltransferases EHMT1 and EHMT2 are responsible for the trimethylation of histone H3 lysine 9 (H3K9me3) and contribute to PARP inhibitor resistance through inhibiting the expression of the tumor suppressor gene CDKN1A, which is involved in the regulation of the cell cycle and DNA repair. However, another study showed that the loss of the histone methyltransferase, enhancer of zeste homolog 2 (EZH2), led to decrease in global H3K27 methylation, resulting in the upregulation of genes associated with cisplatin resistance and the downregulation of genes involved in DDR. This study also showed that treatment with EZH2 inhibitors could reverse chemotherapy resistance in TGCTs, indicating the potential for epigenetic therapies in treating drug‐resistant tumors.

33MiRNAs and LncRNAs have been implicated in the regulation of multiple pathways, including DNA repair. For example, miR‐15a and miR‐16 are found to downregulate the expression of B‐lymphoma Moloney murine leukemia virus insertion region‐1 protein (BMI1), a protein involved in the ubiquitin‐mediated DNA repair pathway, thereby decreasing DNA repair capacity and increased sensitivity of breast cancer cells to doxorubicin treatment. Upregulation of LncRNA HOTAIR has been observed in breast cancer cells following radiation therapy, which promotes DNA repair and radioresistance by interacting with the EZH2 protein. These findings have implicated that the specific miRNAs and LncRNAs could serve as biomarkers for predicting therapeutic response in cancer patients and could be potential target to overcome drug resistance.

34Cancerous cells are known to highly regulate the apoptotic pathways, and apoptosis plays different roles in tumor: eliminates infected cells from the human body, assists the functioning of the immune system, as well as contributes to the maintenance of homeostasis. Generally, those signals induce the activation of effector caspases that mediate intracellular signaling, resulting in DNA fragmentation. Caspase mutations occurring in tumor therapy chemoresistance have been widely reported. For example, caspase‐8 mutations are mainly detected in gastric cancers, and the procaspase‐8Q482H mutation abrogates apoptosis by leading to the dimerization attenuation of the procaspase‐8 protein monomers, resulting in resistance to chemotherapy. The intrinsic apoptosis pathway by mitochondrial damage is activated through numerous exogenous or endogenous stimuli that induce DNA damage. The activation of this pathway involves the activation or inhibition of proteins from the B‐cell lymphoma 2 (Bcl‐2) family, and multiple drugs are known to induce apoptosis of cancerous cells through this mechanism. For example, small molecule RAF kinase inhibitor sorafenib induces apoptosis of AML through the downregulation of Mcl‐1 (an anti‐apoptotic protein) and the upregulation of BIM (an activator of the intrinsic pathway). Previous studies (in vitro and in vivo) have suggested that upregulating of Bcl‐2 in breast cancer cells promotes metastasis to the lung in mice by the EMT process and triggers resistance to chemotherapy (PD168393, a specific inhibitor of EGFR; AG490, an inhibitor of JAK2)., In the treatment of hematological malignancies, the aim to integrate the application of Bcl‐2 inhibitors has led to further research and development of Bcl‐2 selective and dual Bcl‐2 and Bcl‐xL inhibitors. Similar to ABT‐199 (the first‐ever Bcl‐2 inhibitor known as Venetoclax), Navitoclax (ABT‐263) is one of the most studied dual inhibitors; it acts as a BH3‐mimetic, permitting the release of BH3‐only proteins and further MOMP by Bax/Bak activation. One clinical trial detected the effect of ABT‐263 in 118 chronic lymphocytic leukemia (CLL) patients and reported partial responses in 34.6% of patients in phase I trials. Rituximab, a monoclonal antibody targeting the protein CD‐20 was combined with ABT‐263 and showed a statistically significant overall response rate of 70% in untreated patients with CLL compared to rituximab alone. However, because of the character of Bcl‐xL in platelet survival, a main concern with ABT‐263 is the negative effect: inducing thrombocytopenia in a dose‐dependent manner. Furthermore, the additional concern is regarding the emergence of tumor cell resistance to this agent, through the upregulation of Mcl‐1 phosphorylation and further sequestration of the proapoptotic protein Bim. Overall, the clinical use of ABT‐263 is under restriction, and further studies are necessary before its approval.

35Autophagy is a conserved and tightly regulated catabolic self‐degrading mechanism whereby cells keep homeostasis and respond to stress through recycling damaged cellular organelles, proteins, and other cellular components. In tumors, autophagy is a double‐edged sword, and its anti‐ or pro‐tumorigenic character relies on the oncogenic context and stage of tumorigenesis. In the early stages of tumorigenesis, autophagy exhibits tumor suppression function by removing aggregated and misfolded proteins, damaged organelles, limiting cell growth, necrosis, and chronic inflammation. Upon advanced periods of tumor progression, autophagy helps to keep tumor cell survival and increases cellular resistance by conferring stress tolerance by providing the recycling substrate to promote cancer cell survival and contributes to cancer progression and drug resistance. For example, in ovarian cancer, leukemia, hepatocellular carcinoma (HCC), and CRC, the increased levels of basal autophagy contribute to tumor cell growth and increase aggressiveness. Another study showed that autophagy was activated in mice with Myc‐induced lymphomas when treated by chemotherapy; autophagy inhibition sensitizes tumor cells to cell death induced by chemotherapeutic agents, suggesting that autophagy induction leads to a self‐defense effect in this case. Several barriers result from TKI resistances that emerge in a later period of tumor attribute to the mutation in the kinase domain or function of basal autophagy as housekeeping as well as the regulator of RTK activity, which challenges effective treatment against cancer targeted therapy. Therefore, the combination of autophagy modulators with TKIs has been considered in cancer therapy. For example, in NSCLC, erlotinib and gefitinib can induce a high level of autophagy as a drug resistance and cytoprotective mechanism that was accompanied by the suppression of the PI3K/AKT/mTOR signaling cascade. In addition, autophagy inhibition by the pharmacological inhibitor, chloroquine, and siRNAs targeting autophagy‐related gene 5 (ATG5) and ATG7, enhances the erlotinib and gefitinib cytotoxicity. Gastrointestinal stromal tumors (GISTs), in which the activation of CD117, stem cell growth factor receptor, or platelet‐derived growth factor receptor‐α mutations comes up, rarely answer to only‐imatinib mesylate (IM) treatment. IM triggers autophagy as a survival mechanism in quiescent GIST cells that result in acquired resistance and ineffective treatment. Nevertheless, the combination therapy by which IM was used with siRNAs targeting ATGs potentiates GIST cytotoxicity. Another study revealed that IM induces autophagy in chronic myeloid leukemia by induction of ER stress in a different way from apoptosis induction and the combination therapy with autophagy inhibition using either siRNA targeting ATGs or pharmacological inhibitors promotes cell death. Moreover, other TKIs like dasatinib or nilotinib, combined with autophagy suppression treatment, exhibited similar effects.

36Ferroptosis is an intracellular iron‐dependent mechanism of cell death that is distinct from apoptosis, necrosis, and autophagy. Extensive preclinical and clinical studies have concentrated on conquering drug resistance, and inducing ferroptosis has been proven to reverse drug resistance. According to the theoretical framework for initiating the process of ferroptosis, there are three diverse pathways to reverse chemotherapy resistance: the canonical glutathione peroxidase 4 (GPX4)‐regulated pathway, iron metabolism pathway, and lipid metabolism pathway. Genetic inhibition of GPX4 can induce tumor cell ferroptosis and suppress tumor growth in vivo. In addition, the inactivation of GPX4 results in the accumulation of phospholipid hydroperoxides to induce cell membrane damage and ferroptotic death. In GBM, androgen receptor ubiquitination induced by the curcumin analog was found to inhibit GPX4, generating ferroptosis and reversing temozolomide resistance. In CRC, elevated kinesin family member 20A (KIF20A) expression was associated with oxaliplatin resistance, cellular ferroptosis may be induced by disturbing the KIF20A/NUAK1/GPX4 signaling pathway to reverse the resistance of CRC to oxaliplatin. Cystine/glutamate antiporter (xCT) system, as an essential component of the GPX4‐regulated pathway, has also been revealed to be capable of triggering ferroptosis. A study reported that erastin and sulfasalazine (xCT system inhibitors) could induce head‐and‐neck cancer cell ferroptosis and overcome cisplatin resistance. Another study showed that in gastric cancer, inducing ferroptosis by restraining Nrf2/Keap1/xCT signaling, can sensitize cisplatin‐resistant cells to cisplatin treatment. Ferroptosis is defined as an iron‐catalyzed form of regulated necrosis. Upregulation of cellular labile iron pool (LIP) leads to increase vulnerability to ferroptosis. Du et al. reported that tumor cell ferroptosis contributed to catastrophic LIP accumulation after dihydroartemisinin treatment in pancreatic ductal adenocarcinoma, and dihydroartemisinin treatment could overcome cisplatin resistance by triggering ferroptosis. Divalent metal transporter 1 (DMT1) and secreted glycoprotein lipocalin‐2 are both key proteins in regulating iron homeostasis. Turcu et al. found that elevated cellular LIP caused by DMT1 inhibition induced ferroptosis, thereby wiping out breast cancer stem cells (CSCs) and reversing MDR. A recent report by Chaudhary et al. indicated that targeting lipocalin‐2 overcame 5‐FU resistance in CRC by elevating intracellular iron levels, which in turn resulted in tumor cell ferroptosis. The lipid metabolism pathway is also tightly associated with cell vulnerability to ferroptosis. Acyl–CoA synthetase long‐chain family member 4 (ACSL4), an important participant in ferroptosis execution, is involved in producing phospholipid hydroperoxides in an enzymatic manner. In pancreatic cancer, the activation of ACSL4 conquered gemcitabine resistance by inducing ferroptosis.

37Recently, Tsvetkov et al. revealed a previously uncharacterized cell death mechanism termed cuproptosis. Cuproptosis a mechanism distinct from all other known manners of regulated cell death, such as apoptosis, pyroptosis, necroptosis, and ferroptosis. Cuproptosis is regulated by protein lipoylation in the citric acid (TCA) cycle, where lipoylation is necessary for enzymatic function., Their research illuminates the connection between the sensitivity to copper‐mediated cell death and mitochondrial metabolism: respiring. TCA‐cycle active cells have upregulated levels of lipoylated TCA enzymes (particularly, the pyruvate dehydrogenase complex), and the lipoyl moiety acts as a direct copper binder, inducing the aggregation of lipoylated proteins and destabilization of Fe–S cluster proteins, ultimately leading to proteotoxic stress and cell death. Genetic variation in copper homeostasis leads to serious disease, and copper chelators and copper ionophores have been proposed as anticancer agents.,,, For genetic disorders of copper homeostasis like Menke's disease and Wilson's disease, copper chelation is an effective treatment. However, in tumors, copper ionophores like elesclomol have been tested in clinical trials, but neither the benefit of a biomarker of the appropriate patient population nor an understanding of the drug's mechanism of action is considered in such testing. For example, a phase III combination clinical trial of elesclomol in melanoma patients indicated the absence of efficacy in this unselected population, yet a post hoc analysis of samples with low plasma lactate dehydrogenase (LDH) levels displayed evidence of antitumor activity. Low LDH represents a higher cellular dependency on mitochondrial metabolism, in accordance with Tsvetkov's finding that cells more reliant on mitochondrial respiration are nearly 1000‐fold more sensitive to copper ionophores than cells going through glycolysis. Considering the distinct mechanism of cuproptosis from other forms of cell death, deeply understanding how cuproptosis is initiated, progressed, and ultimately executed may exhibit great significance for specific therapeutic interventions and workable combination therapies.

38How tumor cells react to therapy not merely relies on the genomic aberrations they harbor but also is mediated by the TME. TME is a complicated element of tumors and is quite heterogeneous. The significant character of the TME in changing tumor activity has been stated by plentiful studies uncovering how the TME can influence the malignant behavior and therapeutic resistance of the tumor cells (Figure 3). The TME contains a multitudinous cellular and acellular milieu; diverse stromal and immune cells are recruited to develop and maintain such self‐sustained circumstances. The extracellular matrix (ECM), including proteins like laminins, fibronectin, proteoglycans, vitronectin, tenascin‐C, and collagen, devotes to the majority component of the TME and is vital for the maintenance of TME and the induction of cellular adhesion, migration, invasion, and metastasis. Moreover, the composition and organization of ECM can also influence the sensitivity to drug therapy. For example, it has been reported that ECM proteins (laminin, fibronectin, and vitronectin) mediated cell adhesion‐mediated drug resistance (cilengitide, an integrin inhibitor, and/or carmustine, an alkylating chemotherapy) in glioblastoma (GBM). Enrichment of the above proteins in the TME facilitates GBM cell proliferation via integrin αv‐mediated FAK/paxillin/AKT signaling cascade and suppresses p53‐involved tumor apoptosis. In breast cancer, cancer‐associated fibroblasts (CAFs) promote drug resistance by increasing hyaluronan production. Another study showed that the inhibition of the β1 integrin activity by monoclonal antibody AIIB2 markedly promotes radiotherapy efficacy and elevates sensitivity to HER2‐targeting agents of breast cancer cells., In melanoma, CAFs promote cell metastasis and drug resistance by upregulating the level of matrix metalloproteinase 1 (MMP1) and MMP2. As a significant aspect of the TME, the cell‐to‐ECM interaction and cellular crosstalk induce the release of soluble factors (like, angiogenic factors and chemokines) in charge of ECM remodeling and immune evasion, which further bring about therapy resistance. For example, the interaction between α4β1 integrin on cancer cells and fibronectin induce drug resistance in AML and CLL through the PI3K/AKT/BCL2 signaling pathway., Stromal cell–derived factor 1 (SDF1) can interplay with CXC motif chemokine receptor type 4 (CXCR4) and activate AKT and ERK1/2 signaling pathways, resulting in antiapoptotic effects and contributing to tumor cell survival in CLL cells. In AML cells, CXCR4 activation induced by SDF1‐mediated leads to resistance to cytarabine through reducing miRNA let‐7a and facilitating transcriptional activation of MYC and Bcl‐xl. In CRC, CXCR4 is upregulated in chemoresistant tumor cells, and lymph node–derived stromal cells enhance resistance to oxaliplatin and 5‐FU via an SDF1/CXCR4 dependent manner.

39Exosomes, intraluminal vesicles of multivesicular bodies with a diameter of 30–100 nm, are ubiquitously present in most body fluids like blood, cerebrospinal fluids, urine, saliva, and lymphatic fluid., Exosome contents not only reflect the composition of the donor cell but also mirror a regulated sorting mechanism. A complicated of multifarious proteins, including ECM proteins, enzymes, transcription factors, receptors, lipids, and nucleic acids (DNA, mRNA, and miRNA) inside and on the surface of the exosomes constitute their content., Extensive evidence has shown that exosome‐mediated factors can promote tumorigenesis, metastasis, and therapeutic resistance of cancer cells via intercellular communication within TME.,, Tumor‐derived exosomes (TDEs) involve in the initiation, development, and progression of diverse tumor courses, including angiogenesis, TME remodeling, metastasis, and therapy resistance.,, TDEs induce the differentiation of various kinds of TME cells to CAFs that are the main cell population of TME in the majority of tumors; thus, exosomes play a vital role in ECM remodeling and TME reprogramming., Exosomes derived from CAFs contain different molecules such as miRNAs and growth factors that have distinct influences on the target cells of TME. For example, the gemcitabine therapy of CAFs in pancreatic cancer stimulates the expression of miR‐146a and snail and prolonged exosome secretion, thereby promoting epithelial cell proliferation. Drug‐resistant tumor cells can pack the chemotherapeutic drugs in exosomes and shuttle therapeutic agents out of cancer cells. Besides, the delivery of exosomal cargo containing proteins, mRNA, and miRNA to cancerous cells is related to therapeutic resistance., For instance, GBM cell–derived exosomes, which comprise MET‐protein and tyrosine phosphatase receptor type Z (PTPRZ1) fusion proteins, obtain temozolomide resistance via EMT. TDEs from gemcitabine‐resistant pancreatic cancer cells induce chemoresistance by trapping MRP5 and P‐gp or letting gemcitabine to flow back to TME. Exosomes derived from MCF7WT breast cancer cells contain P‐gp and ubiquitin C‐terminal hydrolase‐L1 proteins that are able to induce doxorubicin resistance by elevating the level of P‐gp. TDEs promote platinum resistance by upregulating EMT markers and changing TGF‐β/SMAD signaling pathway in ovarian cancer cells, and exosomes from epithelial ovarian cancer A2780 platinum‐resistant cells attain resistance via promoting EMT. Exosome‐transmitted miR‐567 reverses trastuzumab resistance in breast tumor cells by inhibiting ATG5. In MDA‐MB‐231 and MCF‐7 breast cancer cells, exosome‐mediated miR‐155 induces chemoresistance via upregulating EMT markers and targeting CCAAT/enhancer‐binding protein β (C/EBP‐β), TGF‐β, and forkhead box O 3α (FOXO‐3α) mRNA. In HER2‐positive breast cancer, lncRNA‐SNHG14 induces exosome‐mediated trastuzumab resistance by targeting the apoptosis regulator Bcl‐2/BAX signaling. In NSCLC, exosome‐mediated miR‐21 delivery induces the upregulation of p‐AKT level, thereby resulting in increased gefitinib resistance. Exosomes containing miR‐32‐5p generate MDR in HCC cells by inhibiting PTEN, activating PI3K/AKT signaling cascade and promoting angiogenesis and EMT process. HCC‐derived exosomes containing miR‐221 induce sorafenib resistance through modulating apoptosis inhibition and caspase‐3 activity. Considering that exosomes are involved in various pathophysiological conditions, understanding the molecular mechanisms underlying exosome biogenesis and chemoresistance will benefit in developing novel therapeutics targeting exosome‐mediated tumorigenesis, progression, and chemoresistance.

40One of the important functions of TDEs is to induce tumor vascular development. Angiogenesis, a multistep course by which cancers form new vasculature, is essential for tumor progression. Since Judah Folkman revealed the significant character of vascular networks for the proliferation and progression of solid tumors, establishing angiogenesis as a therapeutic target has become a focal aim. Up to now, antiangiogenic therapy has emerged as significant targeted therapeutic and diverse antiangiogenic agents are currently used in combination with other chemotherapeutic drugs for the treatment of various malignant tumors. Nowadays, over 11 antiangiogenic drugs have been approved by the U.S. FDA, including bevacizumab, aflibercept, sorafenib, ramucirumab, sunitinib, pazopanib, axitinib, vandetanib, Lenvatinib, and regorafenib. However, the treatment is in doubt on the grounds of sustainable efficacy, side effects, off‐target toxicities, and therapy resistance., Clinical and experimental studies have illustrated that cancers employ compensatory/alternative/bypass angiogenic pathways and other adaptive mechanisms for their sustained growth, proliferation, and metastasis, after undergoing a therapy episode(s) with antiangiogenic drugs. That is, various compensatory signaling pathways driving tumor growth and metastasis invariably become the underlying cause of tumor refractoriness. Several preclinical and clinical studies have linked poor performance and resistance to antiangiogenic agents to the activation of a range of compensatory angiogenic signaling pathways and various angiogenic factors that support the angiogenic bypass mechanism. The vascular endothelial growth factor (VEGF)/VEGF receptor signaling pathway is the most promising angiogenic target due to its pivotal role in angiogenesis and tumor growth., Previous studies have suggested that revascularization appears even after blocking VEGF signaling pathways because of the activation of redundant angiogenic pathways.

41Serial evidence accumulated in recent years linked the function of various compensatory/alternative/bypass angiogenic canonical mechanisms sustaining the progression of cancers while exposed to antiangiogenic drugs.,, The current experimental, clinical, and epidemiological data has definitely outlined at least four potential different mechanisms, which can be considered for an explanation of evasive resistance to antiangiogenic therapies. The first mechanism refers to the activation and/or upregulation of compensatory pro‐angiogenic signaling pathways within the cancer. The second mechanism is mainly driven by myeloid/stromal cells, which compensate for the requirement of the VEGF‐mediated pathway, thereby promoting tumor angiogenesis. The third manner is attributed to the dual role of pericytes; first, in establishing the increased pericyte coverage of the tumor vasculature, and second, their potential angiogenic attributes, both serving as escape mechanisms from VEGF‐mediated angiogenesis. The fourth way is associated with remodeling and accessing normal vasculature for the invasion and metastasis of tumors in lieu of obligate neovascularization.

42As an important component of TME, the immune system plays a critical role in regulating the response of cancer cells to therapies, particularly immunotherapies. Immune‐targeted therapies approved for cancer include monoclonal antibodies against immune checkpoints like cytotoxic T‐lymphocyte‐associated protein 4 (CTLA4), PD‐1, and programmed cell death ligand 1 (PD‐L1), which aim to modulate the antitumor T‐cell immune response. The resistance to immune‐checkpoint inhibitors (ICIs) has been attributed to genomic and nongenomic mechanisms that are influenced by the tumor–host–microenvironment relationship. One mechanism is the upregulation of alternative immune checkpoint pathways. Koyama et al. revealed that, in mouse xenograft models with lung adenocarcinoma, other immune checkpoint proteins, including T‐cell immunoglobulin mucin 3, lymphocyte activation gene 3, and CTLA‐4, were upregulated after PD‐1 blockade therapy. The increased expression of alternative immune checkpoints in PD‐1 antibody‐bound T cells can inhibit T cell activity and confer resistance to PD‐1 blockade. Other mechanisms include the presence of immunosuppressive cells, such as regulatory T cells (Tregs) or myeloid‐derived suppressor cells (MDSCs), which can inhibit T cell function. Besides, the presence of stromal cells in TME, such as CAFs, has also been found to be associated with suppressive CD8+ T cell infiltration and insensitivity to αPD‐L1 antibody plus αCTLA‐4 antibody dual immune checkpoint blockade.

43Combination treatments appear to be the promising strategy to overcome resistance to immune checkpoint inhibitors. Combinations of ICIs with radiation, chemotherapy, targeted drug, or tumor vaccines have been proposed and tested. Radiation can trigger an adaptive mechanism to increase the expression of PD‐L1 on tumor cells. Additionally, radiation may increase the depth and duration of immune responses by promoting a more diverse adaptive antitumor immune response. As reported by Victor et al., the combination of radiation with dual checkpoint blockade (anti‐CTLA‐4 and anti‐PD‐1) can enhance tumor control and survival in a preclinical model of melanoma, where radiation prime immune response by increasing T cell receptor repertoire of intratumoral T cells and expanding peripheral clones. Combination with chemotherapy aims to assist ICI treatment by releasing neoantigens or modulating the TME via depleting Tregs and MDSCs. However, it has been reported that the efficacy of this combination therapy is model (cancer type)‐dependent and regimen‐type‐dependent. Combination with chemotherapeutic drugs is not always able to tune the TME or enhanced antitumor efficacy. Even with the effect of depleting MDSCs, the combination may not be beneficial to enhance treatment responsiveness. Additional studies combining single agents with ICIs are still required to better understand potential interactions between drugs and the heterogeneity of TME. Compared to the less predictable ICI–chemotherapy combination, the combination of TME‐targeted drugs with immune checkpoint blockade is considered a relatively logical approach to enhance the stimulation of antitumor immune response. TME‐targeted therapies, such as angiogenesis inhibitors, stromal cell depletion agents, and immunomodulatory agents, can modulate the TME and enhance the immune response against the tumor. Combining these agents with ICIs can lead to synergistic effects of blocking inhibitory signals in T cells and activating the immune response, thereby improving treatment outcomes. For example, the VEGF inhibitor bevacizumab can augment intratumoral CD8+ T cell infiltration through vascular normalization and endothelial cell activation, thereby potentiating PD‐L1 checkpoint inhibition with PD‐L1 inhibitor atezolizumab. This combination has shown promising anticancer efficacy and has been granted FDA approval for the treatment of patients with advanced unresectable or metastatic HCC.

44Combining oncolytic vaccines with ICIs is an emerging approach to overcome ICI resistance. Oncolytic vaccines are a type of oncolytic virus that are engineered to express tumor‐specific antigens, which can stimulate an immune response against the cancer cells. As ICI‐resistant tumors are with a suppressed immune microenvironment, the application of an oncolytic virus can reshape TME and assist ICIs to boost the immune system leading to a stronger antitumor response. This combination therapy has shown promising results in preclinical and clinical studies in a variety of cancers, including melanoma, breast cancer, and bladder cancer.,,

45Recent studies have shown that bacteria can reside within tumors and play a role in controlling cancer response to therapy by modulating the TME. For example, Pseudomonas aeruginosa and Escherichia coli are two types of bacteria found to colonize tumors in various types of cancer, including lung, breast, and pancreatic cancer.,, These bacteria can metabolize chemotherapy drugs such as gemcitabine, reducing their concentration within the tumor and rendering them less effective., Moreover, certain commensal bacteria or pathogens within the TME may interact with immune cells and alter the production of cytokines and chemokines, which are involved in the immune response to cancer., The influence of gut microbiota on cancer resistance has also gained extensive attention. A recent study found that in mice and patients with castration‐resistant prostate cancer, an adaptive change in the gut commensal microbiota was observed after ADT: There was an expansion of commensal microbiota species that can convert androgen precursors to produce androgens, which were absorbed into the systemic circulation thus inducing castration or endocrine therapy resistance. This type of resistance could be delayed by applying antibiotics to reduce gut microbiota or could be reversed by fecal microbiota transplantation hormone‐sensitive prostate cancer patients, implicating two possible ways to overcome bacterial‐mediated resistance. Although using antibiotics can directly eradicate the bacteria within the TME, this approach may have limitations due to the risk of antibiotic resistance. Another approach is to develop drug delivery systems specifically targeting the tumor cells while avoiding bacterial‐mediated metabolism., For example, nanotechnology‐based drug delivery systems can be engineered to release drugs only in the tumor cells, minimizing the exposure of the drug to the bacteria., The presence of bacteria within the TME and possible influences from gut microbiota are thought to be significant factors contributing to treatment resistance in cancer. Understanding the complex interactions between bacteria and tumors is essential for developing effective cancer treatments to overcome this resistance.

46CSCs are a small population of cells within a tumor that possess stem cell–like properties and are thought to be responsible for tumor initiation, growth, and metastasis., These cells have been identified in various cancers, including breast, colon, and GBM. CSCs are less sensitive to chemotherapy and radiation therapy due to their ability to activate DNA damage repair and antiapoptotic signaling pathways, which may lead to incomplete tumor eradication and recurrence after treatment. Based on these findings, the researchers put forward the “CSC theory” and found evidence of the existence of CSCs in some cancer tissues., CSCs can resist therapy mainly because of the overexpression of MDR transporters that mediate drug efflux, the more active DNA repair capacity, and the tendency to form new microvascular for the tumor. These characteristics enable CSCs to tolerate treatment, maintain the tumor with nutrients and oxygen, and rapidly repopulate the tumor. This mechanism is similar to that of normal tissue stem cells in response to trauma, which may explain that the bladder CSCs actively contribute to the cause of chemotherapy resistance after several cycles of chemotherapy., Targeting this trauma response of CSCs can become a new therapeutic intervention. CSCs between different treatment cycles actively regenerate and respond to chemotherapy‐induced injury or apoptosis, much as normal tissue stem cells respond to trauma‐induced damage. Dying cells release a metabolite called prostaglandin E2 (PGE2) that stimulates proliferation and causes CSC to repopulate cancers shrunk by chemotherapy., In normal cells, this is an integral part of the wound repair process, where PGE2 induces the regeneration of tissue stem cells. Ironically, PGE2 induces more CSC regeneration between chemotherapy cycles in cancers. An essential characteristic of CSC is maintaining self‐renewal ability, and the mechanism may be one key factor in promoting cancer development and metastasis., CSC theory helps people understand the occurrence and development of cancers. In addition, CSCs can activate signaling pathways that promote cell survival and proliferation, such as the Wnt/β‐catenin and Notch pathways.,, Therefore, understanding the mechanisms underlying CSC‐mediated resistance to therapy is essential for developing effective cancer treatments. It provides a new perspective on cancer development, metastasis, drug resistance, and recurrence. Leukemia therapy targeting LSC (leukemia stem cells) has achieved good clinical efficacy, and beneficial research has also been carried out in solid cancers.

47EMT is one of the essential mechanisms of cancer metastasis. The latest research suggests that cancer cells with dual characteristics of EMT and stem cells are the key to cancer metastasis and drug resistance. There is a direct link between EMT and CSCs, but the molecular mechanism is still unclear. EMT phenomenon refers to a reversible process in which relatively stable epithelial cells lose cell polarity and intercellular adhesion and transform into spindle‐shaped mesenchymal cells with migration ability. It is ubiquitous in epithelial cancer cells. Its essential feature is the loss of the expression of epithelial markers, such as E‐cadherin, β‐catenin, tight junction protein, and epithelial cell adhesion molecules on the cell membrane of cancer cells,,,, and meanwhile obtaining a mesenchymal phenotype with the increased expression of vimentin, N‐cadherin, fibronectin, and β1 and β3 integrins.,, The mechanism of EMT is mainly due to changes in the epithelial cells themselves or the surrounding microenvironment, leading to the activation of a series of signal transduction pathways, and related transcription factors in the nucleus play a regulatory role. Precise intracellular signaling mechanisms regulate different degrees of epithelial cell transformation. Various extracellular signals activate different nuclear transcription factors by binding to specific receptors on the cell surface. The common feature of these transcription factors is that they can recognize the DNA binding sequence of the E‐box motif on the target gene promoter, thereby regulating the expression of the target gene and initiating EMT. Loss of E‐cadherin expression is currently considered the most prominent feature of EMT. Decreased E‐cadherin levels can lead to decreased cell adhesion and make cells easily invade and metastasize. Cancer cells lose some of the characteristics of epithelial cells and acquire the characteristics of mesenchymal cells through EMT, so cancer cells can obtain stronger invasion and migration capabilities.,

48Snail is involved in triggering EMT during the progression of epithelial cancers and is a crucial point in the occurrence of EMT. Snail expression correlates with the reduced loss of E‐cadherin and acts as a direct repressor of E‐cadherin transcription. Twist is another essential transcription factor that regulates EMT. Its mechanism of inducing EMT is to directly or indirectly bind the E‐cadherin promoter through the E‐box motif, thereby inhibiting the expression of E‐cadherin. Snail/Twist1 knockout breast cancer model demonstrates that chemotherapy resistance is associated with EMT. Studies have shown that EMT may be involved in drug resistance of cancer cells, such as lung, pancreatic, and breast cancer.,, Numerous studies have shown that cancer cells develop resistance to carboplatin or paclitaxel by acquiring mesenchymal cell phenotype, indicating that EMT may be the instigator of chemotherapy resistance., The latest research shows that cancers may have a type of partial EMT, which means the cancer cells have both epithelial and mesenchymal cell phenotypes, and these cancer cells are more aggressive than the others. EMT has been described as a significant cause of EGFRi failure in EGFR‐mutated NSCLC. Cellular transdifferentiation, a phenotypic shift from adenocarcinoma to squamous cell carcinoma or neuroendocrine carcinoma, occurs in 3%–14% of EGFR‐mutant NSCLC patients, and approximately 17% of prostate cancer failed with abiraterone/enzalutamide therapy.

49From the discussion above, it is evident that targeting EMT processes or cellular plasticity has great potential to circumvent drug resistance. However, only a few compounds adesigned to inhibit the EMT process are currently in clinical trials. Inhibitors targeting Notch, TGF‐β, and Wnt signaling pathways are also promising candidates for inhibiting EMT and cellular plasticity. TGF‐β heterogeneity in the cancer microenvironment creates rapidly dividing CSCs that accelerate cancer growth and others that invade surrounding healthy tissue and evade treatment., For example, PF‐03446962 and galunisertib are antagonists designed to inhibit TGF‐β receptors, currently in phase I clinical studies in solid cancers (NCT00557856, NCT02423343). Both PF‐03446962 and galunisertib inhibit the EMT program, thereby preventing cancer development. In addition, Wnt inhibitors such as ETC‐1922159 and OMP‐54F28 have been reported to inhibit the EMT program and are currently in phase I clinical trials in cancer (NCT02521844, NCT01608867).

50Cancer chemotherapy is still a required treatment method for most cancer patients. However, cancer cells are prone to MDR, a significant problem limiting the efficacy of current chemotherapeutic drugs lacking selectivity. Many cancer patients have a remarkable curative effect in the early stage of chemotherapy. However, the drug resistance of cancer cells increases with the prolongation of treatment time, eventually leading to treatment failure. Anticancer strategies are constantly evolving, and targeted therapy drugs have encountered increasing resistance after the initial excitement, including primary and acquired resistance. Strategies to address drug resistance through combination therapy have made some progress, although considerable challenges have been faced, and the full potential of combination therapy has not yet been realized.

51According to statistics, in the early 1990s, the main reasons for the failure of new drug development were concentrated in poor pharmacokinetics and limited biological activity., With the introduction of absorption, distribution, metabolism, and excretion for predictive analysis and research applications, the clinical failure rate dropped from 40% to 10% in 2000. With the accumulation of basic research on targets and signaling pathways and the advent of the era of genome sequences, the success rate of clinical development of targeted drugs is also increasing. For example, the familiar TKIs have become one of the most popular research directions. In recent years, with a series of significant breakthroughs in technological applications, such as the application of high‐throughput genomics to discover new targets,, the identification of molecular biomarkers,, and the progress of statistical methods in biological and chemoinformatics,, the progress and the success rate of research and development have been greatly improved. However, new problems are still emerging, with lacking clinical efficacy and increased toxicity still the main directions of failure in the second research and development phase.

52Nanodrugs represent a promising approach to overcoming resistance to cancer. These drugs are engineered to be nanoscale size and have unique physicochemical properties that enable them to target cancer cells while minimizing off‐target effects. Nanodrugs can be designed to overcome resistance by delivering drugs to cancer cells in more targeted and controlled manner and bypassing resistance mechanisms such as efflux pumps and DNA repair pathways.,, One key advantage of nanodrugs is their ability to improve the pharmacokinetics and pharmacodynamics of drugs. By encapsulating drugs within nanocarriers, the drug concentration within the tumor can be increased, improving drug efficacy. Additionally, nanocarriers can protect the drug from degradation and clearance, leading to a longer half‐life and a sustained drug release., Nanodrugs can also be engineered to overcome specific resistance mechanisms. For example, nanodrugs can target CSCs by incorporating specific ligands that bind to stem cell markers. Nanodrugs can also be engineered to overcome bacterial‐mediated resistance by releasing drugs only within the TME and avoiding bacterial exposure. By improving drug delivery, targeting, and overcoming specific resistance mechanisms, nanodrugs can potentially improve the efficacy of cancer treatment and patient outcomes. However, further research is needed to optimize the design and delivery of nanodrugs and to ensure their safety and efficacy in clinical use.

53Polytherapy can make cancer cells less likely to develop compensatory resistance mechanisms than a single or sequential drug regimen. There are many problems in the reversal of MDR with combination drugs, including clinical toxicity and pharmacokinetic interactions. The therapeutic range of each drug may be narrow when administered in combination, and there may be an overlap in toxicity. In some cases, in addition to the expected overlapping toxicities, combination therapy has some characteristic toxicities., We cannot accurately predict toxicity from preclinical models, which adds to the challenge of optimizing the toxicity–potency balance of drug combinations. Meanwhile, the pharmacokinetic interaction is the question of how to combine two or more drugs., For example, lapatinib is a substrate and moderate inhibitor of CYP450‐3A4 and substantially reduces the clearance of drugs such as other CYP450‐3A4 substrates. In the phase I clinical trial of lapatinib and pazopanib, investigators compared the historical pharmacokinetic parameters of the two drugs and concluded that the drug combination did not alter drug exposure. However, in a more detailed phase II pharmacokinetic analysis, the combination indicated a significant drug–drug interaction contributing to the reduced efficacy of lapatinib in glioma patients., In addition, the frequent use of antiepileptic drugs in this patient population also reduced pazopanib exposure, resulting in poorer patient outcomes. This highlights the importance of detailed pharmacokinetic assessment for the combined evaluation of anticancer drugs. Combination therapies approved or in clinical trials in the recent 5 years are summarized in Table 1.

54Antibodies, cytotoxic drugs, and conjugates bridge ADC (antibody–drug–conjugates) drugs. The antibody will specifically recognize and guide the drug to the lesion. The conjugate can generally be broken in the pH environment of the lesion site, thereby releasing the cytotoxin with a therapeutic effect. Cytotoxins mainly target the DNA and tubulin of cancer cells, block the proliferation of tumor cells, and induce apoptosis., After the second new ADC drug Adcetris (brentuximab vedotin) was approved by the FDA in 2011, 10 ADC drugs have been approved so far, 7 of which have been approved in the last 3–4 years. ADC drugs can improve the selectivity of tumor treatment and can better deal with the drug resistance of targeted monoclonal antibodies., It is considered one of the essential directions for developing monoclonal antibodies (especially in the field of tumor‐targeted therapy) in the next decade. The tissue specificity and cytotoxicity of the new generation of ADCs have been improved compared with the previous generation products, allowing them to show fantastic activity in treating refractory cancers., The currently available evidence shows that the potency of ADCs is based on complex and delicate interactions among antibodies, conjugates, various components of the cytotoxin, and the tumor and its microenvironment. Combining ADCs with other antibody therapies, such as ADCs with the anti‐VEGF monoclonal antibody bevacizumab, has also shown activity in preclinical models. This may be due to bevacizumab's enhanced drug delivery efficiency by altering tumor vascular innervation.

55Some researchers proposed to reverse chemotherapy resistance by targeting inhibitor of apoptosis (IAP) inhibitors. The IAP protein family plays an essential role in controlling programmed cell death, and the expression level of IAP in cancer cells is significantly increased. Therefore, by directly or indirectly regulating the expression of apoptotic proteins on the extrinsic apoptotic pathway (transmembrane apoptotic pathway) and the intrinsic apoptotic pathway (mitochondrial apoptotic pathway), the chemosensitivity of cell apoptosis can be improved and obtained a potential response. However, the current combination of this method and chemotherapy only targets a small number of tumor cells, and there are cases where tumor cells are resistant to IAP inhibitors.

56The technologies for reversing tumor MDR at the gene level mainly include antisense nucleic acid technology, ribozyme technology, and RNA interference (RNAi) technology., Drugs that regulate miRNA expression (such as miR‐125, −20, −24) may have specific clinical application prospects in reversing tumor MDR., Although many constructed‐cell line banks exist, the genetic characterization of these cells and human tumor samples varies enormously. Researchers are attempting to create patient‐derived cell culture models and then test these models on pharmacogenomics platforms to rapidly discover multiple effective drug combinations., These models can better reflect the biological complexity of human drug‐resistant tumor cells than constructed cell models. However, the combination therapy results still need to be verified by randomized clinical trials.

57The study of functional drug delivery systems that can reverse tumor MDR will have broad application prospects in improving the efficacy of chemotherapy drugs and reducing side effects. Due to tumors’ heterogeneity and drug resistance, it is usually challenging to achieve an excellent therapeutic effect using one drug alone. Therefore, people have been devoting themselves to designing nanocarriers that can be loaded with multiple anticancer drugs and improve the therapeutic effect through synergistic therapeutic effects., Combining chemotherapeutics and MDR‐reversing agents using a drug delivery system has been a promising strategy for reversing MDR in recent years.,, Common nano‐drug carriers that have been reported to be used for the co‐delivery of drugs include liposomes, nanoparticles, micelles, nano‐emulsions, and nanogels. Combining MDR drug reversal agents, RNAi/DNA, targeted drugs with nanocarrier drug delivery systems, through nonspecific internalization, reduce the efflux of drugs by ABC transporters in tumor cells, increases the uptake of drugs, and through RNAi/DNA Delivery, active targeting, and increased responsiveness to physiological stimuli can reverse tumor cell MDR.,, Combination types of drugs delivered in combination include chemotherapeutics and chemotherapeutics,, chemotherapeutics and MDR reversal agents,,, chemotherapy drugs and siRNA,,, chemotherapy drugs and monoclonal antibodies., Among them, combining chemotherapy and other drugs is the most common type of combination administration. The rapid development of nano‐drug delivery carriers provides a good carrier platform for improving the therapeutic effect of drugs and overcoming the MDR problem of tumors. Combining drug administration with nano‐drug carriers can enhance the effect of reversing MDR through various forms and achieve a substantial combined effect by co‐delivering different types of drugs. Nano‐drug carrier‐mediated co‐administration is a promising strategy for reversing tumor MDR. Currently, there is more and more research on reversing MDR of tumors by combined administration of nano‐drug carriers., With the clinical application of nanocarrier drug delivery system, this technology shows great potential in reversing tumor MDR.

58Tumor immunotherapy can improve tumor patients’ immune function, kill or inhibit tumor cells, and reverse tumor MDR with little adverse reactions. There are many studies on tumor immunohistology, enzymology, and cytokines,, but most are still experimental research. Among them, the systems biology approach involves the analysis of large amounts of data, including genetics, transcriptomics, proteomics, or factors affecting posttranslational regulation. Some systems’ biology approaches study the importance of proteins in intracellular signaling networks and explore the possibility of reversing tumor drug resistance based on the differential effects of drug targets on intracellular associations.

59The new technologies, for example, the novel computational method, can be used for predicting polytherapy switching strategies to overcome tumor heterogeneity and evolution. In addition, the widely used gene silencing tools, including shRNA and CRISPR, could help researchers propose and predict effective drug effects both in vitro and in vivo. Finally, the clinical trial designs that reflect drug toxicity and utilize intermittent doses and adaptive trial designs that give dynamic combinations after the emergence of resistance will significantly accelerate the development of more effective therapeutic combinations to reverse the MDR in cancer cells (Figure 4).

60As summarized in Table 2, MDR in cancer is a multifactorial phenomenon that results in drug inactivation, efflux, target alteration, cancer cell death inhibition, DNA damage repair, cellular heterogeneity, and more. Tumor drug resistance has become a significant problem in oncology, affecting the treatment effect and prognosis of cancer patients, and may lead to tumor progression or even recurrence. Therefore, it becomes crucial to understand the causes and underlying mechanisms of cancer drug resistance, which will facilitate the development of various therapies or combinations for treating different cancers. Combination therapy is considered the most important treatment option for personalized medicine and overcoming MDR. Simultaneous use of two or more treatment methods can effectively overcome MDR, avoid clinical toxicity, and improve patients’ survival rate and quality of life. The specificity of cancer cells is a promising research and therapeutic area, which will help to develop tumor‐targeted therapies with low toxicity to normal cells and higher tumor specificity.

61Zhe‐Sheng Chen, Yihang Pan, and Leli Zeng designed the review. Zi‐Ning Lei, Qin Tian, and Qiu‐Xu Teng did the literature search and wrote the manuscript. Zi‐Ning Lei and Qiu‐Xu Teng prepared the table and figures. Zhe‐Sheng Chen, Yihang Pan, and John N. D. Wurpel reviewed and revised the manuscript. All authors listed have made a substantial contribution to the work. All authors have read and approved the article.

62The authors declare no conflict of interest. Author Zhe‐Sheng Chen is the Editorial Board Member of MedComm. Author Zhe‐Sheng Chen was not involved in the journal's review of, or decisions related to, this manuscript.

63Not applicable.