Cytotoxicity of targeted PLGA nanoparticles: a systematic review

Recent advances in nanotechnology have contributed tremendously to the development and revolutionizing of drug delivery systems in the field of nanomedicine. In particular, targeting nanoparticles based on biodegradable poly(lactic-co-glycolic acid) (PLGA) polymers have gained much interest. However, PLGA nanoparticles remain of concern for their effectiveness against cancer cells and their toxicity to normal cells. The aim of this systematic review is to identify a promising targeting PLGA nanoformulation based on the comparison study of their cytotoxicity potency in different cell lines. A literature search was conducted through the databases of Google Scholar, PubMed, ScienceDirect, Scopus and SpringerLink. The sources studied were published between 2009 and 2019, and a variety of keywords were utilized. In total, 81 manuscripts that met the inclusion and exclusion criteria were selected for analysis based on their cytotoxicity, size, zeta potential, year of publication, type of ligand, active compounds and cell line used. The half maximal inhibitory concentration (IC50) for cytotoxicity was the main measurement in this data extraction, and the SI units were standardized to μg mL−1 for a better view of comparison. This systematic review also identified that cytotoxicity potency was inversely proportional to nanoparticle size. The PLGA nanoparticles predominantly exhibited a size of less than 300 nm and absolute zeta potential ∼20 mV. In conclusion, more comprehensive and critical appraisals of pharmacokinetic, pharmacokinetic, toxicokinetic, in vivo and in vitro tests are required for the investigation of the full value of targeting PLGA nanoparticles for cancer treatment.


Introduction
Recent advances in nanotechnology have contributed tremendously to the development and revolutionizing of the drug delivery system in the nanomedicine eld. The application of nanoparticles has long been recognized as a controlled release formulation for delivering a therapeutic agent to a specic targeted site. Nanoparticles provide a high therapeutic effect against cancers, which has earned them remarkable research interest among researchers. Nanoparticles offer a highly efficient targeted therapy compared to traditional cancer therapies. This targeted therapy can be done easily on nanoparticles using an ideal targeting ligand. Thus, targeting nanoparticles based on biodegradable polymers have gained much interest for treating cancer cells with minimal systemic side effects. 1,2 Poly(lactic-co-glycolic acid) (PLGA) is a biodegradable polymer that has attractive properties as a nanocarrier for cancer therapy. PLGA is a hydrophobic copolymer and mainly composed of two monomers: lactic acid and glycolic acid (Fig. 1). PLGA is approved by the European Medicine Agency and the US Food and Drug Administration (FDA) as an ideal material for designing a drug delivery system due to its biocompatibility and biodegradability. PLGA is widely adapted for preparing nanoparticles encapsulating hydrophilic and hydrophobic anti-therapeutic agents. 3,4 PLGA offers an enhanced permeability and retention (EPR) effect, sustained and controlled drug delivery for cancer therapy, enhanced accumulation of drugs in tumor vasculature and targeted delivery by surface conjugation with targeting ligands. 5 Active and passive mechanisms are oen practiced when targeting cancer cells using nanoparticles. Passive targeting is the application of polymeric nanoparticles owing to their size, shape and surface charge enabling them to be accumulated the selected articles based on the types of cells, IC 50 for cytotoxicity, year of publication, treatment duration, types of active compounds used, types of targeting ligands, types of studies (in vitro) and the size and zeta potential of the nanoparticles. The data were described and presented in a table. The IC 50 for cytotoxicity was the main measurement in this data extraction. The SI units were standardized to mg mL À1 for a better view of comparison. The IC 50 is the dose required to inhibit 50% of the cell viability. Based on the availability of the data of IC 50 , size and zeta potential of the nanoparticles, PLGA nanoparticles with the active targeting properties were selected for the purpose of studying the correlation between the particle size/ zeta potential of PLGA nanoparticles and cytotoxicity potency.

Cytotoxicity of PLGA nanoformulations
The databases of Google Scholar, PubMed, ScienceDirect, Scopus and SpringerLink were searched. These produced 113, 32, 113, 20 and 5 articles, respectively. From all the databases, a total of 266 articles were retrieved aer 17 duplicates were removed. Following this, 170 articles that were not compliant with the inclusion criteria were identied and excluded from the study. The 96 articles le were thoroughly assessed according to the exclusion criteria dened in Introduction. Aer critical assessment, 15 articles were omitted due to the methodology and cytotoxicity data being insufficiently described.
Hence, 81 studies have been integrated into the qualitative synthesis involving assessment of the data of in vitro studies.

In vitro studies
The data of the cytotoxicity of PLGA nanoparticles conjugated with particular targeting ligands that deliver specic active compounds against different types of cellssuch as brain, breast, lung, colon, stomach, gastric, liver, ovary, cervix, prostate, uterus, pancreas, skin, umbilical vein endothelial, esophagus, bladder, head, neck and kidney cellsare shown in Table  1. From the table, it can be seen that PLGA nanoparticles are time/dose-dependent on cytotoxicity.
2.1.1 Brain. Magnetic silica PLGA nanoparticles conjugated with transferrin showed the most potent cytotoxic effect against brain cancer cells (U-87) with an IC 50 of 0.13 mg mL À1 . 10 This could be due to the dual drug delivery of doxorubicin and paclitaxel designed for the PLGA nanoparticles, compared to the single drug carriers listed in Table 1. The application of targeting ligands can improve the bioavailability of drug-loaded nanoparticles. In this study, transferrin was actively targeted to the overexpressed transferrin receptors in brain capillary endothelium and glioma cells. The IC 50 was much lower when the treatment duration was prolonged from 48 h (1.03 mg mL À1 ) to 96 h (0.13 mg mL À1 ), showing the time-dependent effect of the treatment.
2.1.3 Lung. Arginine-glycine-aspartic acid (RGD) peptide-modied and paclitaxel-loaded PLGA-chitosan nanoparticles (PTX-PLGA-CSNP-RGD) had the most potent cytotoxic effect against H1975 lung cancer cells with an IC 50 of 0.0017 mg mL À1 . 12 The PTX-PLGA-CSNP-RGD nanoparticles showed enhanced uptake due to the nature of RGD peptide, which is highly targeted to the overexpressing integrin a v b 3 receptor specically found in lung cancer cells. In addition, less toxicity was received by the normal lung cells due to the weak expression of integrin a v b 3 in normal lung cells.
2.1.4 Colon. PLGA nanoparticles loaded with paclitaxel and conjugated with WGA wheat germ agglutinin (WNP) showed the most promising cytotoxic potency against colon cancer cells (HT-29) with an IC 50 of 0.028 mg mL À1 . 13 WGA actively binds to the highly expressed N-acetyl-D-glucosamine-containing glycoprotein found in the membrane of colon cancer cells, thus increasing the cellular uptake of WNP in colon cancer cells. Since WGA tends to bind to the glycoprotein in colon cancer cells, WNP is more effective in delivering paclitaxel to colon cancer cells to enhance the bioavailability of paclitaxel compared to non-targeted nanoparticles.
2.1.5 Stomach. Only one study about nanoparticles targeting stomach cancer cells was included for review aer the databases were screened. PLGA nanoparticles modied with polyethylene glycol and conjugated with an engineered antihuman CD44v6 Fab (AbD15179) were developed to specically target human CD44 isoforms containing exon v6 (CD44v6) present in stomach cancer cells. The PLGA nanoparticles were reported to exhibit anti-proliferative potency against GP202, MKN74-CD44v6+ and MKN74 stomach cancer cells. The cytotoxicity of the PLGA nanoparticles was 50 mg mL À1 and highly stable against uid-mimicking gastrointestinal conditions. 14 No IC 50 data were reported in this study.
2.1.6 Gastric. Dual-targeting hybrid nanoparticles made of PLGA and a lipoid shell prepared by conjugating the anti-HER2/ neu peptides (AHNP) and n-hexadecylamine (HDA) to the carboxyl groups of hyaluronic acid (HA) were reported to deliver 7-ethyl-10-hydroxycamptothecin (SN38 agent) specically to gastric cancer cells (HGC27 cells) with overexpression of (1) CD44 cluster determinant 44 and (2) HER2 (human epidermal growth factor receptor 2). An IC 50 of 0.05 mg mL À1 was reported for the dual-targeting nanoparticles. 15 Studies on the cytotoxicity mechanism have indicated that the enhanced cellular uptake of dual-targeting nanoparticles and suppression of CD44 and HER2 expression by HA and AHNP inhibit the growth of HGC27 cells.
2.1.7 Liver. LFC131 peptide-conjugated PLGA nanoparticles composed of D-a-tocopheryl polyethylene glycol succinate (TPGS) moieties were prepared to deliver epirubicin and specically bind with CXCR4-overexpressing human hepatic carcinoma cells (HepG2). TPGS is a vitamin E derivative and was used to stabilize the PLGA nanoparticles. It was also used as an inhibitor of P-glycoprotein in overcoming multi-drug resistance. 16 LFC131 peptide-conjugated nanoparticles exhibited a threefold higher cellular uptake in HepG2 cells than nontargeted nanoparticles. CX-EPNP showed a promising antiproliferative effect against HepG2 cells with an IC 50 of 0.78 and 0.38 mg mL À1 for 24 and 48 h of treatment, respectively. 17 Herein, LFC131 peptide-conjugated nanoparticles showed a time-dependent effect on the cytotoxicity studies.
2.1.8 Ovarian. The potential use of PLGA nanoparticles as a paclitaxel carrier for ovarian cancer stem cells (OCSCs) was reported. PLGA nanoparticles loaded with paclitaxel were developed by an emulsion solvent evaporation method and then conjugated with a targeting ligand: folic acid (FA). Cytotoxicity results reveal that FA-conjugated nanoparticles had an IC 50 of 0.00075 mg mL À1 . 18 Folate receptors (FR) are biomarkers that are over-expressed in human cancer cells, such as ovarian cancer cells. Therefore, FA was applied as the targeting ligand in the study to target paclitaxel to FR-positive OCSCs over the normal cells.
2.1.9 Cervical. Magnetic PLGA nanoparticles with surface modied with folate-chitosan conjugate, which served as an anti-cancer and magnetic resonance imaging (MRI) contrast agent, were reported in one study. Docetaxel and super paramagnetic iron oxide nanoparticles (SPIONs) were loaded into the PLGA nanoparticles for delivery to folate-positive KB cancer cells. The folate-chitosan conjugate was prepared using the carbodiimide method and then used as a shell for the loaded PLGA nanoparticles to target the FR in KB cells. This specic targeting of FA in PLGA nanoparticles improved the cellular uptake by FR-positive KB cancer cells with an IC 50 of 0.0057 mg mL À1 . 19 2.1.10 Prostate. Theragnostic PLGA nanoparticles loaded with superparamagnetic iron oxide (SPIO) nanocrystals and docetaxel were prepared for both ultrasensitive MRI and cancer Table 1 The cytotoxicity of PLGA nanoparticles on different types of cells treatment. PLGA nanoparticles were formed by using a single emulsion evaporation method. The active targeting ligand single-chain prostate stem cell antigen antibodies (scAb PSCA ) were conjugated to PLGA via a poly(ethylene glycol) linker.
Overexpression of prostate stem cell antigen (PSCA), a prostate-specic glycosyl phosphatidylinositol-anchored glycoprotein found in prostate cancer PC3 cells, was the binding site for the scAb PSCA -conjugated nanoparticles. 20 Targeted PLGA nanoparticles demonstrated improved cellular uptake and cytotoxicity in PC3 prostate cancer cells exhibited an IC 50 of 0.01403 mg mL À1 (24 h), 0.00579 mg mL À1 (48 h) and 0.00322 mg mL À1 (72 h). 21 Herein, the scAb PSCA -conjugated nanoparticles showed a time-dependent cytotoxicity against prostate cancer PC3 cells. 2.1.11 Uterine. One study reported a comparison of the cytotoxicity and cellular uptake activity of targeted and nontargeted PLGA nanoparticles for delivering doxorubicin against multi-drug resistance in uterine (MES-SA/Dx5) cancer cells. HER2 antibody-conjugated nanoparticles and nontargeted nanoparticles showed higher cellular uptake of doxorubicin than free doxorubicin in MES-SA/Dx5 cancer cells. No signicant difference was found regarding cytotoxicity in MES-SA/Dx5 cells for targeted and non-targeted PLGA nanoparticles. This was due to no HER2 receptor overexpression being observed in MES-SA/Dx5 cells. Higher cytotoxicity was observed for both targeted and non-targeted PLGA nanoparticles compared to free doxorubicin, showing suppression of the overexpression of P-glycoprotein in Dx5 cells. HER2 antibodyconjugated nanoparticles were able to overcome the multidrug resistance (MDR) effect in Dx5 cells since cytotoxicity and cellular uptake results at 10 mM extracellular doxorubicin concentration were comparable. 22 2.1.12 Pancreas. Bortezomib, a proteasome inhibitor, was loaded into PLGA nanoparticles with poloxamer 407 as an emulsier against S2-013 pancreatic cancer cells. Surface-modied of the PLGA nanoparticles with transferrin was done to achieve pancreatic cancer cell targeting. Cellular uptake studies have shown high uptake of the targeted PLGA nanoparticles by cancer cells for a sustained release of bortezomib from targeted PLGA nanoparticles. Targeted PLGA nanoparticles showed cytotoxic effects against pancreatic cancer cells with a GI 50 of 0.0028 mg mL À1 . 23 Low toxicity to normal pancreatic cells demonstrated that the targeted PLGA nanoparticles enhanced the delivery of bortezomib to S2-013 pancreatic cancer cells.
2.1.13 Skin. PLGA nanoparticles containing dioleoylphosphatidic acid (DOPA) coated cisplatin and rapamycin induced potent cytotoxic effects on A375-luc human melanoma cells with an IC 50 of 0.09 mg mL À1 . 24 This was due to the synergistic effects of rapamycin and cisplatin towards A375-luc human melanoma cells. Rapamycin acts as a mammalian target of rapamycin inhibitors and a sensitizer. DOPA was coated onto cisplatin to achieve compatibility between PLGA and the dual drugs cisplatin and rapamycin. A high anti-proliferative effect was observed in PLGA nanoparticles conjugated with targeting ligand anisamide, which has a high affinity towards sigma receptor membrane-bound proteins that are overexpressed in A375-luc human melanoma cells.
2.1.14 Umbilical vein endothelial. One study reported in vitro synergistic effects of paclitaxel or combretastatin A4loaded charge reversible pullulan-conjugated PLGA nanoparticles formulated with poly(b-amino ester) for the treatment of human umbilical vein endothelial cells (HUVECs). This work revealed IC 50 s of less than 0.0396 and 0.0118 mg mL À1 for the respective paclitaxel and combretastatin A4-loaded nanoparticles. 25 Pullulan-conjugated PLGA nanoparticles had high cytotoxicity activity in HUVECs due to the polysaccharide backbone of pullulan having a high affinity towards the asialoglycoprotein receptor (ASGPR) in HUVECs. 26 The pH sensitivity of the PLGA nanoparticles was attributed to pullulan's cleavage of b-carboxylic amide bond towards changes in pH in the microenvironment of cells.
2.1.15 Esophageal. Surface-modied PLGA nanoparticles with DTPA-hEGF allowed PLGA nanoparticles to attain radiolabeling and targeting towards the EGF receptor (EGFR). DTPA represents diethylenetriaminepentaacetic acid, while hEGF refers to human epidermal growth factor. Surface-modied PLGA nanoparticles were radiolabeled with 111 In to achieve high affinity to EGFR-overexpressing esophageal cancer cells. The 111 In radiolabeled PLGA nanoparticles induced radiotoxicity via cellular DNA damage. Ru1 or Ru(phen) 2 (tpphz) 2+ (phen represents 1,10-phenanthroline, while tpphz represents tetrapyridophenazine) is a ruthenium-based DNA replication inhibitor and radiosensitizer. Ru1 was loaded into 111 In radiolabeled PLGA nanoparticles for DNA damage enhancement. hEGF-PLGA-Ru1 nanoparticles showed remarkable cytotoxicity in EGFR-overexpressing OE21 cells due to the active targeting of the hEGF ligand. However, compared to OE21 cells, lower cytotoxicity was observed in EGFR-normal OE33 cells, with >70% proliferation. 27 No IC 50 data were reported for the in vitro cytotoxicity results of hEGF-PLGA-Ru1 nanoparticles.
2.1.16 Bladder. PLGA nanoparticles were surface-modied with a novel cell-penetrating polymerpoly(guanidinium oxanorbornene) (PGON)to improve tissue penetration tenfold in mouse bladder and human ureter. PGON is a synthetic polymer that mimics cell-penetrating peptides and possesses low toxicity to normal/cancer cells. PGON PLGA nanoparticles showed a signicant enhancement in intracellular uptake of nanoparticles compared to unmodied nanoparticles. Belinostat, a histone deacetylase (HDAC) inhibitor, was loaded into the PGON PLGA nanoparticles to assess their biological activity. In comparison to uncapsulated belinostat, belinostat-loaded nanoparticles exhibited a signicantly low IC 50 (0.509 mg mL À1 ) in cultured bladder cancer cells (UM-UC-3 and RT-4) and sustained hyperacetylation. 28 2.1.17 Head and neck. A PLGA/polydopamine core/shell nanoparticle system was designed for light induced cancer thermochemotherapy. Overexpression of the epidermal growth factor receptor (EGFR) drove the high binding of anti-EGFR antibody-conjugated nanoparticles towards head and neck cancer cells. This enhanced cellular uptake of nanoparticles by head and neck cancer cells and induced the conversion of nearinfrared light to heat, triggering drug release from the nanoparticles and cancer cell ablation due to the increased temperature. The study revealed that PLGA/polydopamine nanoparticles were effective in inhibiting cancer cells at 10 mM doxorubicin concentration when coupled with near-infrared (NIR) irradiation. A doxorubicin concentration of 5 mM or higher was required to achieve the NIR irradiation effect on PLGA/polydopamine nanoparticles in order to produce heat needed for cancer cell ablation and trigger drug release. 29 2.1.18 Kidney. Folate-targeted and reduction-triggered PLGA nanoparticles were prepared for targeted delivery of doxorubicin to the COS-7 kidney broblast-like cell line. Folatetargeted PLGA nanoparticles were prepared from a PLGA core containing a monolayer soybean lecithin and a reducible outer layer monomethoxy-poly(ethylene glycol)-S-S-hexadecyl (mPEG-S-S-C 16 ). Disulde bonds (-S-S-) are highly degradable in the reducing environment (high concentration of glutathione) of cancer cells and thus release drugs at the targeted site. [30][31][32] FA was conjugated to the mPEG-S-S-C 16 outer layer to achieve tumor targeting. No signicant difference was observed in the cytotoxicity in FR-negative COS-7 cells for targeted and nontargeted PLGA nanoparticles. This shows that the cellular uptake of folate-targeted PLGA nanoparticles involved a FRmediated endocytosis. 33 The study showed a dose-dependent effect where the cytotoxic effect of the nanoparticles was directly proportional to the concentration of doxorubicin, which proves the non-cytotoxicity of the nanoparticles.

Size and cytotoxicity
The cytotoxicity results show that PLGA nanoparticles induced an anti-proliferative effect as early as 3 h, taking 240 h at most. The particle size of PLGA nanoparticles ranged from 58 to 407 nm, with an average size of 189 nm. It was evident that particles smaller than 100 nm and larger than 300 nm were less commonly prepared, as a certain size is required to carry the load of the drugs and nor bigger size to promote the EPR effect in cancer cells. The IC 50 of cytotoxicity of PLGA nanoparticles was plotted against particle size of the PLGA nanoparticles to assess the correlation between particle size and cytotoxicity (Fig. 3). It can be seen that as the particle size increased, the IC 50 of cytotoxicity also increased, which indicates that smaller nanoparticles induced lower IC 50 values and higher cytotoxicity potency. Although the application of corona materials such as hyaluronic acid increase the particle size, the efficacy of the encapsulated drugs to the desired site was enhanced. This is due to the cancer targeting properties provided by the binding of hyaluronic acid to the PLGA nanoparticles. 53,58,60,70,72 Selective targeting enhances the cellular uptake of drugs in cancer cells and reduces the cytotoxic effect of drugs in normal cells. This further justies that the ideal size and ability of PLGA nanoparticles to selectively deliver drugs to desired site are crucial in developing efficient and safe drug delivery systems. The IC 50 values were mostly lower than 10 mg mL À1 , but reached up to 80 mg mL À1 for FA-conjugated chitosan-functionalized PLGA nanoparticles (CPN) with particle size of 207 nm. This was due to the extremely low entrapment efficiency of bicalutamide in CPN (only about 1%). 95

Zeta potential and cytotoxicity
A plot of the zeta potential of the PLGA nanoparticles against their cytotoxicity IC 50 values can be seen in Fig. 4. It is noticeable that PLGA nanoparticles were predominantly in negative charged because of its nature negatively charge of the carboxyl group end chain in PLGA. PLGA nanoparticles ranging from À13.2 to À19.3 mV showed the lowest IC 50 value of 0.00031 mg mL À1 at À18.9 mV, with an average IC 50 value of 1.15 mg mL À1 (n ¼ 20) (ESI data †). Surface charge of the particle is dened by the absolute value of zeta potential. Zeta potential is a critical factor in designing a drug delivery system because it denes the stability of the nanoparticles. Interestingly, a former study reported that the cytotoxicity potency of PLGA nanoparticles is directly proportional to the absolute zeta potential. 102 This is in agreement with a previous report that an absolute value of zeta potential 20 mV or much lower results in nanoparticles with adequate stability. 103 The stability of nanoparticles is directly proportional to the absolute value of zeta potential. Since stronger repulsive forces were formed between the nanoparticles with high absolute values of zeta potential, stable nanoparticles with uniform size distributions were produced. 104 PLGA is a hydrophobic polymer and can be stabilized by the hydrophilic corona materials such as chitosan. 19,47,84 This is because of the absolute zeta potential of PLGA nanoparticles is increasing with the concentration of chitosan and indirectly resulted in increased stability of the PLGA nanoparticles. High Fig. 3 The correlation between particle size and cytotoxicity. Fig. 4 The correlation between zeta potential and cytotoxicity.
stability in nanoparticles ensures that there is no early drug release along the route to the target site, which translates to a higher cytotoxic effect on the targeted site in cancer cells.

Discussion
This systematic review provides an overview of the application of different targeting ligands and active compounds/drugs used in PLGA nanoparticles to achieve active targeting for drug delivery to a particular cell line. This can give insight to researchers in regard to the designing of a potential drug delivery nanoparticle system for different types of cell lines. The potential of treating multiple cell lines using a single formulation makes the designation of a drug delivery system become more exible in terms of its usage. For instance, a drug delivery nanoparticle system involving the encapsulation of paclitaxel and conjugation of transferrin as the targeting ligand can be used to treat several types of cancer cells, including breast and brain cancer cells. This can be explained by the fact that FR is a well-known biomarker that has a high affinity for FA in cancer cells due to its overexpression on several types of cancer cell lines, while paclitaxel has long been recognized as a mitotic inhibitor for treating various types of cancer cells. Both the drug and targeting ligand are suitable for treating breast and cervical cancers within a single formulation; this provides cost efficiency and compatibility to patients. The systematic review gives a summary of the IC 50 of cytotoxicity of different types of cancer cells treated with various formulated PLGA nanoparticles, in which the overview of the cytotoxicity of PLGA nanoparticles is better understood. For instance, the exposure time and IC 50 concentration for selected types of cells could serve as references for other researchers to use in planning and identifying their research interests and protocols, which would make these processes easier and save time.
The cellular uptake of nanoparticles with the presence of targeting ligands provides a specic binding to the target or cancer site without causing any or less harms to the healthy cells. This phenomenon is attracting more interest from researchers due to its safeness and less toxicity being imposed as a result of its specic controlled release of anti-cancer drugs. These statements are in line with the ndings of previous studies that reported that the surface conjugation of a targeting ligand in nanoparticles more greatly enhances the cellular uptake of drugs and cytotoxicity potency compared to nonconjugated nanoparticles. 12,21,27 For instance, the cell viability for RGD antibody-conjugated nanoparticles (PTX-PLGA-CSNP-RGD, 35.2%) was signicantly lower (p < 0.01) than for nonconjugated nanoparticles (PTX-PLGA-CSNP, 45.7%). 12 On top of this, 24 h treatment of A549 cells with FluTax-PLGA-CSNP-RGD, on the basis of the uorescence intensity of uorescent paclitaxel (FluTax), revealed higher uptake of 28 pg/1 Â 10 5 cells, compared to PTX-PLGA-CSNP by 28 pg/1 Â 10 5 cells. The reduced cell viability and higher uorescence intensity of Flu-Tax in RGD antibody-conjugated nanoparticles demonstrates the selective absorption of paclitaxel through integrin receptormediated endocytosis compared to non-targeted delivery at the same dose.
Identifying the zeta potential and particle size of PLGA nanoparticles is crucial because an ideal nanoparticle size can provide an EPR effect for cancer cells' uptake of drugs, while proper zeta potential can provide nanoparticles with high stability. From this systematic review, it is evident that nanoparticles <300 nm in size and $20 mV in absolute zeta potential are favorable. This is because a decent size ($200 nm) is required for nanoparticles to load the drugs, while high zeta potential provides a uniform and narrow size distribution, as well as high stability. 105 Since nanoparticles are highly stable, the leakage of drugs is negligible during transportation along the target site, which means greater cytotoxicity potency and cellular uptake of nanoparticles.
Challenges are always present in designing a drug delivery system. Although active targeting of nanoparticles reduces toxicity to normal cells, the conjugation of targeting ligands to the nanoparticles might increase the particle size. This is timeconsuming, as optimization of the preparation step in making nanoparticles of an ideal size is required. In addition, targeting ligand conjugation might also reduce the release of drugs from the nanoparticles due to the probability that the ligand may serve as an external barrier when releasing the drug from the nanoparticles. The process of conjugating targeting ligands onto the surface of nanoparticles oen involves a two-step reaction. This two-step reaction may cause leakage or early release of drugs from the nanoparticles due to the involvement of a reaction incubation period as well as sonication to minimize the agglomeration of nanoparticles.
The studies analyzed in this systematic review involved different factors, including length of exposure time, active compounds and targeting ligands used in assessing cytotoxicity assays of PLGA nanoparticles in a single cell line. Comparisons among the available studies for a particular cell line are difficult to make due to the variations in the factors involved. Hence, there is still a lack of ideally designed PLGA nanoparticle drug delivery systems. Moreover, some in vitro studies lack proper descriptions for data, such as IC 50 data for cytotoxicity assessment. Additionally, different in vitro cytotoxicity assays have been used (e.g. MTT, MTS, SRB), which may have resulted in variations in cytotoxicity data. 106 Hence, it is difficult to obtain completely accurate results from comparison studies of cytotoxicity for PLGA nanoparticles.
Currently, only 19 drug formulations based on PLGA have been approved by the FDA. 107 This negligible number displays that the development of PLGA formulations is very challenging. Thus far, none of the 19 FDA approved PLGA formulations are based on PLGA nanoparticles as they are mainly composed of PLGA microparticles, solid implant and in situ gel. This lacking availability of PLGA nanoformulations further indicates that more evaluations on its efficacy and safety requirements are needed before getting approved for clinical use. Poor drug entrapment efficiency and drug release kinetics from PLGA nanoformulations are the main challenges faced in order to deliver drugs effectively to the target site. For instances, initial burst drug release is the typical issue of low efficacy in designing PLGA nanoformulations. Although PLGA nanoformulations were reported to be safe and selective targeting to the cancer cells in in vitro and animal studies, they are insufficient to prove that the same outcomes will be observed in human trials. Safety concerns of a drug delivery system are the top priority when it comes to establish their application.
Nevertheless, challenges still arise along with opportunities. Despite the great challenges, PLGA nanoparticles are still of interest due to its biocompatibility and particularly, an FDA approved material for drug delivery. Over the past few decades, a signicant advancement in the development of PLGA nanoparticles for the application of drug delivery has led to the revolution of pharmaceutical industry. The uniqueness of PLGA nanoformulations, specically their sustained and controlled drug release as well as cancer targeting properties provide the assurance of great potentials as promising nanocarriers. There are still great potentials in PLGA nanoformulations by modifying their physicochemical properties with better understanding of the physiology of the cancer cells and pharmacokinetics of the drug delivery system. In terms of biocompatibility and toxicity in biological systems, the physicochemical properties of PLGA nanoformulations play an important role. 108 Particle size, surface charge and selective targeting properties of PLGA nanoformulations are critical benchmarks to consider when it comes to efficacy. The potential of these state-of-the-art innovations and strategies to develop PLGA nanoformulations are contributing to the advancement of cancer treatment. Nevertheless, intensive evaluations for pharmacokinetics, biodistribution and toxicity are essential before progressing PLGA nanoformulations into clinical studies.

Conclusions
In conclusion, PLGA nanoparticles with active targeting properties have higher cytotoxicity and cellular uptake of drugs compared to non-targeting nanoparticles, regardless of the types of cells studied. The size and zeta potential of PLGA nanoparticles play a crucial role in dening the resulting cytotoxicity. Therefore, in future studies, greater focus should be placed on assessing the pharmacodynamic, pharmacokinetic and toxicokinetic proles of the drug delivery system using cancer-targeting PLGA nanoparticles. In addition, although the listed PLGA nanoparticles exhibited the potency of pharmacological actions based on the in vitro data, there was a lack of in vivo data for many of them. This raises concerns regarding efficacy and safety of usage in the application of PLGA nanoparticles in human trials. Therefore, more relevant in vivo data on the efficacy and toxicity of PLGA nanoparticles are desired before human clinical trials should be commenced, as a pharmaceutical formulation can only be considered successful when both safety of usage and efficacy are guaranteed.

Conflicts of interest
There are no conicts to declare.