Cilengitide Synthesis Essay

Open Access This article is
  • freely available
  • re-usable


Reduced Cytokine Release in Ex Vivo Response to Cilengitide and Cetuximab Is a Marker for Improved Survival of Head and Neck Cancer Patients

Susan Cedra, Susanne Wiegand, Marlen Kolb, Andreas Dietz and Gunnar Wichmann *

Academic Editor: Helen M. Sheldrake

Received: 30 June 2017 / Accepted: 2 September 2017 / Published: 5 September 2017



Targeting of αVβ3 and αVβ5 integrins by cilengitide may reduce growth of solid tumors including head and neck squamous cell carcinoma (HNSCC). Preclinical investigations suggest increased activity of cilengitide in combination with other treatment modalities. The only published trial in HNSCC (ADVANTAGE) investigated cisplatin, 5-fluorouracil, and cetuximab (PFE) without or with once (PFE+CIL1W) or twice weekly cilengitide (PFE+CIL2W) in recurrent/metastatic HNSCC. ADVANTAGE showed good tolerability of the cilengitide arms and even lower adverse events (AEs) compared to PFE but not the benefit in overall survival expected based on preclinical data. As we found in the FLAVINO assay, a short-time ex vivo assay for prediction of chemosensitivity, only a subgroup of HNSCC had an increased suppressive effect of cilengitide containing combination therapies on colony formation of epithelial cells (CFec) and release of pro-angiogenetic and pro-inflammatory cytokines, whereas other HNSCC failed to respond. Response to αVβ3 and αVβ5 integrin targeting by cilengitide classifies HNSCC regarding outcome. We present FLAVINO data arguing for further development of cilengitide plus cetuximab in treatment of a subgroup of HNSCC potentially identified by the FLAVINO assay using a set of biomarkers for response evaluation.

Department of Otolaryngology, Head and Neck Surgery, University of Leipzig, 04103 Leipzig, Germany


Correspondence: Tel.: +49-341-972-1926


head and neck cancer; head and neck squamous cell carcinoma (HNSCC); predictive assay; chemoresponse ex vivo; cilengitide; integrin; αVβ3; targeted therapy; biomarker; interleukin 6; monocyte chemotactic protein-1

1. Introduction

As reviewed by Ahmedah and colleagues earlier this year, αV integrins play a crucial role in the development, progression, and metastatic spread of head and neck squamous cell carcinoma (HNSCC), thus supporting the therapeutic potential of integrin targeting [1].

In our translational study, we focused on the efficacy of cilengitide applied either alone or as part of multi-component chemotherapy in HNSCC ex vivo. Here we present our findings including the involvement of pro-inflammatory and pro-angiogenic cytokines as potential prognostic markers.

Cilengitide, an N-methylated cyclic pentapeptide (cyclo-Arg-Gly-Asp-D-Phe-(N-methyl)-Val; EMD 121974; Merck KGaA, Darmstadt, Germany) containing the characteristic RGD-recognition sequence, specifically inhibits integrins αVβ3 and αVβ5 [2,3,4]. In the ex vivo short-time chemoresponse assay FLAVINO, we used alterations in colony formation and cytokine release of interleukin 6 (IL-6) and monocyte chemotactic protein-1 (MCP-1) as read-out for response of HNSCC to cilengitide plus cetuximab.

IL-6 is a pleiotropic pro-inflammatory cytokine involved in numerous biological processes such as cell differentiation, wound healing, and apoptosis [5]. Dysregulated or excessive expressed IL-6 is associated with a variety of chronic inflammatory diseases and the development and maintenance of malignant tumors including HNSCC [6]. Elevated IL-6 levels in patients with HNSCC correlate with higher tumor stage, lymph node metastasis, increased proliferative tumor-activity, decreased immunologic response, and distinctive cachexia [7]. Furthermore, IL-6 supports vascular endothelial growth factor A production, resulting in enhanced tumor-angiogenesis [8]. There is evidence that IL-6 concentration decreases during successful anti-tumoral therapy, suggesting predictive qualities of the course of IL-6 concentration alteration [7].

As shown in previous studies by our laboratory, cilengitide modulates effects of MCP-1 production in HNSCC ex vivo [9,10], thus indicating MCP-1 as a valid molecule for testing. MCP-1 is a pleiotropic chemotactic cytokine that is essential for the recruitment of monocytes, macrophages, and natural killer cells during tissue injury and inflammation [11]. It is known that high concentrations of this CC-chemokine have impact on tumor environment and that it is intertwined with tumor invasiveness, progress of disease, metastatic spread, and tumor-angiogenesis [12,13]. MCP-1 demonstrates both tumor-supporting and anti-tumor effects due to its diverse influences on different cell types, but it is still not clear how it decides to act either way—probably concentration dependency plays a leading role in this question [10]. Many studies have linked elevated MCP-1 serum levels in patients with the presence, development, or recurrence of solid tumors, including advanced HNSCC, with significant lower overall survival (OS) and tumor-specific survival (TSS), indicating MCP-1 as a prognostic biomarker for solid tumors including HNSCC and clinical outcome [14].

2. Results

In analyses including specimens of 39 HNSCC patients (Table 1) we found a tremendous heterogeneity in the release of the cytokines IL-6 and MCP-1 after short-time culturing (72 h; see Material and Methods Section) of HNSCC.

This is consistent with earlier findings in investigations applying cilengitide either alone or in combination with other treatments to solid tumors in vitro and in vivo [15,16,17,18,19,20,21,22,23]. We also detected varying efficacy regarding reduction of colony formation. The heterogeneity in response of HNSCC to cilengitide and in particular to cilengitide in binary combination with cetuximab (Cil+E) is substantial (Figure 1).

We analyzed the outcome of patients under study regarding OS using receiver operator characteristic (ROC) curves (Supplementary Figures S1–S3) and found significant areas under the curve (AUC) regarding predicted survival for MCP-1 ≤ 75%, and statistical trends for CF ≤ 45%, and IL-6 ≤ 90% of controls. The grouping of HNSCC patients according to these cutoff values revealed significantly different TSS and OS of patients (Figure 2). Suppression of MCP-1 concentration below 75% of controls and IL-6 concentration below 90% of controls under treatment of cetuximab plus cilengitide were associated with significantly improved OS (MCP-1 p < 0.006; IL-6 p < 0.007) and TSS (MCP-1 p < 0.08; IL-6 p < 0.004). This trend for improved survival was also seen in patients whose HNSCC responded to Cil+E with suppression of CFec below 45% of controls (OS p < 0.066; TSS p < 0.061). Since readouts of cytokine release in triplicate measurements achieved more consistent and objective results than manual microscopic cell counting for CFec [9], measurement of IL-6 and MCP-1 might be a suitable new strategy for the prediction of patients’ likely outcomes. Independent from the chosen readout, the statistical analyses using Fisher’s exact test revealed no significant difference in distribution of values above and below the cutoff regarding response of HNSCC samples taken from early stages (UICC I and II) vs. locoregional advanced HNSCC stages (UICC III and IV) or T categories T1 and T2 vs. T3 and T4 to either cilengitide, cetuximab, or Cil+E and the detected prognostic value (all p > 0.387). Only slight and insignificant differences in the patients’ OS associated with UICC stage and T category were found in Kaplan Meier curves comparing binary classified patients according to the cutoff for the readouts CFec (p = 0.951 and p = 0.465), or the release of IL-6 (p = 0.955 and p = 0.501) or MCP-1 (p = 0.883 and p = 0.771).

3. Discussion

Our set of data demonstrates a huge heterogeneity of head and neck cancer cells in response to Cil+E. A strongly reduced release of IL-6 and MCP-1 by Cil+E treated HNSCC is demonstrated as being a potential classifier for the patients’ OS and TSS. Our results also imply the existence of a subgroup of patients potentially benefitting from combined Cil+E treatment.

A more detailed look at results obtained in ADVANTAGE [24,25] reveals a 10% higher objective response rate (ORR) in PFE + CIL1W vs. PFE of 42% (95% confidence interval, CI, 30–55%) vs. 32% (95% CI 21–45%) corresponding to an odds ratio (OR) of 1.516 (95% CI 0.732–3.141) in the independent read (sensitivity analysis). This higher ORR in PFE+CIL1W vs. PFE was accompanied by a prolonged median time to treatment failure (5.6 vs. 4.2 months) but led to a slightly increased hazard ratio (HR) for progression-free survival (PFS; HR 1.15, 95% CI 0.74–1.79; p = 0.528), whereas the OS of patients treated with PFE+CIL1W was not inferior (HR 0.94, 95% CI 0.61–1.47; p = 0.800).

As the primary objective of the trial (i.e., reasonably improved OS) was not met, the investigators concluded that further development of PFE+CIL1W or PFE+CIL2W in R/M HNSCC cannot be recommended. Does this mean that also further development of the binary combination of cilengitide plus cetuximab for treatment of R/M HNSCC should not be recommended?

Our findings in the FLAVINO study [9] and the here shown suitability of these data to classify HNSCC regarding their prognosis together with the data published for ADVANTAGE strongly suggest considerable therapeutic potential of cilengitide in combination with cetuximab in preselected HNSCC patients potentially identified using the outcome measured in the FLAVINO assay. As cilengitide even in combination with multiple other chemotherapeutics was well tolerated [21,24], there might be a place for cilengitide in binary combination with cetuximab (Cil+E) instead of cetuximab monotherapy in R/M HNSCC without adequate general health to receive PFE or after failure of platinum-based regimens. Future research may reveal if cilengitide deserves another chance in R/M HNSCC.

4. Materials and Methods

4.1. Patient Characteristics

HNSCC samples of 43 patients were included in this study. With patients’ informed consent, biopsies of tumor tissue were obtained during surgery or panendoscopy. All patients received therapy according to the consented decision made by the Head and Neck Cancer Tumor Board of the University Hospital Leipzig according to German therapy guidelines; none of the patients were treated with cilengitide and/or cetuximab. Thirty-nine histopathologically confirmed HNSCC of these 43 patients, 34 male and 5 female patients (mean age of 60.3 years; Table 1), could be analyzed regarding cytokine production and colony formation after treating the specimens with either Cil or E alone or combined (Cil+E).

4.2. FLAVINO-Assay

The same procedures as the protocol of the FLAVINO assay were used, as previously described [9]. Freshly obtained tumor specimens were put into phenol red- and riboflavin-free medium supplemented with 10% fetal calf serum and antibiotics (TM). After mechanical disintegration and digestion by 230 mU/mL collagenase IV (Sigma–Aldrich, Deisenhofen, Germany), 10,000 viable HNSCC cells were added to triplicate wells coated with human laminin (Roche, Germany) containing either 66.7 µg/mL E, 10 µM Cil, Cil+E, or (for reference) TM alone, adjusting the total volume to 300 µL. Supernatants harvested after 3 days were analyzed by ELISA and adherent cells ethanol-fixated and underwent pan-cytokeratin staining by FITC-labeled antibodies and counting of green fluorescent colonies of epithelial cells. Thirty-nine HNSCC had adherent growth (mean CFec ≥ 4/well in triplicate controls).

4.3. ELISA

The cytokines IL-6 and MCP-1 released into supernatants were measured using indirect sandwich ELISAs (OptEIA Kits; BD Biosciences, Heidelberg, Germany) following the manufacturer’s instructions and using tetra-methyl benzidine as substrate. The optical density of each well was determined measuring the optical density at λ1 = 450 nm and λ2 = 620 nm on the Synergy2TM multi-mode microplate reader (BioTek Instruments, Inc., Winooski, VT, USA) according to 4-parameter calibration curves calculated using Gen5 software (BioTek). The lower limit of detection (LLD) was ≤4 pg/mL, the lower limit of quantification (LLQ) was ≤7 pg/mL for both cytokines, two orders of magnitude below median cytokine concentrations in untreated controls [9].

4.4. Statistical Analysis

All data shown are based on triplicate measurements. Differences were compared by Student’s t-test for paired samples. The maximum Youden scores (sensitivity plus specificity) were used for finding in ROC curves with significant areas under the curve the optimum cutoff for binary classification of HNSCC patients for survival data analyses according to the Kaplan-Meier method by applying the log-rank test. Statistics were done using SPSS Statistics for Windows, version 20.0.0 (SPSS Inc., Chicago, IL, USA). P ≤ 0.05 was regarded as significant.

4.5. Ethical Approval

The study was approved by the Ethics Committee of the Medical Faculty of the University Leipzig (study codes 180-2007, 201-10-12072010, and 341-15-05102015) and performed in accordance with the ethical standards as laid down in the 1964 declaration of Helsinki and its later amendments.

5. Conclusions

In a subgroup of HNSCC, the ex vivo targeting αVβ3 and αVβ5 integrins by cilengitide and EGFR by cetuximab leads to reduced CFec and release of IL-6 and MCP-1, thus proving that cilengitide in binary combination with cetuximab does have potential in the treatment of this subgroup. As response in the FLAVINO assay was associated with response in vivo [9,26], the FLAVINO assay may allow for the detection of HNSCC patients that could benefit from the combined targeted therapy. In this respect, IL-6 and MCP-1 gain significance as prognostic biomarkers for improved overall survival when cetuximab plus cilengitide achieve a decrease of IL-6 production <90% and of MCP-1 production <75% relative to control.

Future research may focus on the development of cilengitide plus cetuximab in preselected R/M HNSCC.

Supplementary Materials

The following are available online at Supplementary Figures S1–S3.


The study was supported by Merck Serono GmbH. The authors received funds for covering the costs to publish in open access from the Medical Faculty of the University of Leipzig. We thank all patients and their families for their participation in this study and Anett Reiche for technical assistance.

Author Contributions

Gunnar Wichmann conceived and designed the experiments; Susan Cedra and Gunnar Wichmann performed the experiments; Susan Cedra and Gunnar Wichmann analyzed the data; Susan Cedra, Marlen Kolb, Susanne Wiegand, Andreas Dietz, and Gunnar Wichmann wrote the paper.

Conflicts of Interest

GW received funding for the project “Cilengitide in HNSCC ex vivo” and honoraria and travel expenses from Merck Serono GmbH. The other authors all declare no conflict of interest. The founding sponsor Merck has no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.


  1. Ahmedah, H.T.; Patterson, L.H.; Shnyder, S.D.; Sheldrake, H.M. RGD-binding integrins in head and neck cancers. Cancers2017, 9, 56. [Google Scholar] [CrossRef] [PubMed]
  2. Mas-Moruno, C.; Rechenmacher, F.; Kessler, H. Cilengitide: The first anti-angiogenic small molecule drug candidate design, synthesis and clinical evaluation. Anti-Cancer Agents Med. Chem.2010, 10, 753–768. [Google Scholar] [CrossRef]
  3. Goodman, S.L.; Holzemann, G.; Sulyok, G.A.G.; Kessler, H. Nanomolar small molecule inhibitors for alphav(beta)6, alphav(beta)5, and alphav(beta)3 integrins. J. Med. Chem.2002, 45, 1045–1051. [Google Scholar] [CrossRef] [PubMed]

2. Strategies for Cancer Cell Targeting Peptides Discovery

2.1. Phage Display

In 1985, Smith showed that filamentous bacteriophages (bf) can be genetically engineered to express foreign peptide sequences on their surface (protein coat) [4]. This led to the development of libraries of phages in which each clone displays a peptide of combinatorial sequence [5,6]. These libraries can be used in vitro to screen for and isolate clones displaying peptides with high selectivity for almost any desired target. They can also be injected into an animal to isolate clones that bind to a desired target tissue [7]. Multiple repetitions on these selection and amplification steps, termed panning, bring about enrichment of the clones with highest affinity to a specific target.

In the literature, there are two approaches to isolating target specific clones; one approach is to expose the total phage pool to the target, for example, an antibody. The binding clones are then recovered and exposed to a non-relevant protein, such as a non-relevant antibody in order to recover only those clones specific to the target antibody’s paratope. The alternate approach proposes that the original phage pool contains mostly non-specific clones that indeed may compete or interfere with the binding of specific clones to the target. Therefore, in this approach, the phage pool is first exposed to one or several non-specific targets and non-binding clones are recovered. These clones are then exposed to the desired target and specific binders isolated.

Since being introduced by Smith [4], this technology—known today as phage display peptide libraries—has become a powerful tool for the discovery of specific ligands [6,8] with high receptor affinity [6,9].

Notwithstanding the contribution of phage-display to the discovery of targeting peptides the technique has some critical disadvantages. One of these relates to the method of recovery of phage positive clones. Traditionally, recovered phage are titrated on bacterial lawns, however it is technically challenging to retrieve and sequence the peptide inserts of more than 10–20 such clones. Thus, many potentially interesting clones are missed. The introduction of Next Generation Sequencing (NGS) can now alleviate this problem [10]. While NGS is more expensive and time consuming, it allows analysis of the entire pool of target positive phage. Another disadvantage is that the technology has limited, in that it produces peptides with predetermined length and only from natural amino acids. These disadvantages can be overcome using synthetic combinatorial methods, which are described below.

2.2. Synthetic Peptide Libraries—OBOC

In 1984, Geysen et al. [11] introduced a combinatorial approach for the segmental epitope mapping of the VP1 protein from the foot and mouth disease virus. A library of 208 overlapping hexapeptides, each peptide differing at one amino acid position, was synthesized, covering the whole 213 amino acid sequence of the protein. The peptide library was synthesized on a solid support, enabling its rapid and feasible immunological screening [11]. In its original form this combinatorial method, later termed “multipin technology”, utilized polyethylene pins covered with acrylic acid (for the formation of a polar support with improved solubility in polar solvents) as a solid support. These pins were adapted to a 96 well polypropylene plate, where each well served as a separate reaction vessel. At the end of the synthesis the peptides are directly subjected to a biological screening or are first removed from the pins by enzymatic [12] or chemical [13] cleavage. Several different combinatorial methods and techniques have since been introduced [14,15,16], and the field of synthetic peptide libraries has now become a powerful tool for drug discovery [17,18,19] as well as for fundamental biological investigation [20,21]. In the context of synthetic homing peptides, the one-bead-one-compound (OBOC) method has made a particular impact [22].

The OBOC method is based on a “mix and split” technique and enables preparation of peptide libraries with 106–108 different peptides [23]. The peptides are coated on to 100 µm diameter polymeric beads, each bead covered with about 1013 copies of the same peptide [15]. The library can be screened against various biological targets, including intact cells or specific receptors. Beads, which physically bind to the target of interest, are isolated and the structure of the coating compound can then be elucidated [15,24]. Recently a novel screening method for identification of targeting peptides derived from a “mix and split” library was reported [25]. This method involves encoding of each member of the library with a unique peptide nucleic acid (PNA) [26,27] which is biologically stable [28], yet suitable for DNA microarray assays [29,30]. PNA encoded peptide libraries are appropriate for the identification of targeting peptides to any biological target of interest, either in vitro or in vivo due to the stability of the PNAs in biological environments.

The synthetic flexibility of the OBOC method, and the size of its libraries make it an ideal optimization tool for peptide leads previously discovered by phage display or any other methodology [23].

2.3. SPOT-Synthesis

In 1992, Frank introduced a method using cellulose membranes as the solid support for peptide synthesis [31]. In the SPOT synthesis, the peptides are synthesized on pre-functionalized cellulose, which enables the attachment of activated amino acids. “Spotting” small volumes of reagents at defined positions on the cellulose support actually results in a creation of microreactors, whose size is defined by the volumes dispensed and the physical properties of the solid support. The scale of the reaction, as well as the number of synthesized peptides are directly derived both from the size of separate spots on the membrane sheet. Removal of the protecting groups and washings are performed by dipping the sheet in the appropriate solution. After the accomplishment of the synthetic procedure, the peptides can be assayed while attached to the solid support, or can be cleaved for further performance of bioassays in solution. The advantages of SPOT synthesis are that it is a flexible, simple and cheap method, it yields sufficient amounts of product [32], and can be applicable for various biological [33,34] and synthetic [35,36,37] applications. Since the invention of this parallel synthetic technology, it has been developed further by several groups, including the introduction of new polymeric solid supports [38,39,40], linker anchors [35,41,42] and automated systems [42,43].

In contrast to biological combinatorial method such as phage display, synthetic combinatorial methods have the advantage that they can incorporate d-amino acids, unnatural amino acids and non-amino acid building units into the combinatorial sequence [22]. The increased stability of these types of compounds in the proteolytic environment of biological fluids—compared to natural l-amino acids—enhances the half-life of the targeting ligand and can contribute to increased efficacy of the TDD system [44,45,46].

2.4. Rationally Designed Peptides

Multistep syntheses and the need for exhausting screening of random combinatorial peptide libraries consisting of millions of different compounds are motivating the design of more target oriented peptide libraries. Rational design of peptide ligands generally depends on bioinformatic databases and a resolved crystallographic structure of the target–ligand complex together with computational methods, in order to design more appropriate binding compounds [47].

One sophisticated approach is based on homology modeling. In this approach the design of new ligands to target is performed by using known targets as structural templates. As we have recently described, the ligand interactions with the different targets can be studied by a stepwise energy evaluation in which the effects of ligand mutations and different residues of the target are examined [48,49]. Such an approach provides a valuable alternative to a costly and time-consuming combinatorial approach since it can dramatically decrease the number of candidate peptides to be synthesized and tested.

Rational design of peptides is usually validated by two optimization methodologies: cyclo scanning (CYCLOSCAN) and positional scanning. These methods are also useful in phage display and OBOC for optimizing peptide hits. Notably, classical modes of cyclization include the formation of a lactam bridge through carboxyl and amino functional groups, or disulfide bridges through thiol groups leading to side-chain-to-side-chain bridge formation (Figure 2). Two main drawbacks to this classical mode were reported: (i) cyclization may lead to a loss of biological activity, due to the involvement of side chain groups crucial for bioactivity; and (ii) the number of cyclization possibilities is limited. If the linear peptide does not contain the appropriate amino acid to allow classical cyclization, various amino acids in the native sequence need be replaced by amino acids bearing amine (Lys, Orn), carboxyl (Glu, Asp) or thiol (cysteine), an operation that may lead to loss of biological activity [50,51].

To overcome these pitfalls, Gilon et al., applied a new facile approach of backbone cyclization in which any two backbone nitrogens are connected through bridges of various sizes and chemical nature [52,53,54,55]. In this method, side-chains are not altered, and a highly variable and large set of different cyclizations can be applied to any linear sequence [56]. Thus, one can generate conformational libraries in which many diverse amino acid sequences share the same structure, thereby enabling the optimization of a known three-dimensional biological motif, or libraries in which a single sequence is contained within a large variety of conformations, thereby identifying the active conformation of a biologically active sequence.

Positional scanning for peptide sequence begins with identification of an amino acid of interest at a single position, followed by sequential substitution with other amino acids. Increased bioactivity of the peptide indicates the preferred amino acids at altered positions in the sequence [57,58].

3. Targeted Delivery of Chemotherapeutics Based on Clinically Investigated Peptides: Arginine-Glycine-Aspartic Acid (RGD), Somatostatin, Luteinizing Hormone-Releasing Hormone (LHRH) and Bombesin

The success of using peptides to target over or exclusively expressed receptors in cancer cells, including those that are associated with tumor angiogenesis, serves as a basis for the creation of targeted drug delivery (TDD) systems. These systems are generally constructed with a peptide as a targeting moiety, a linker moiety and a therapeutic agent, as schematically presented in Figure 1. While a number of targeting peptides clinical applications have been isolated and are being developed), the tripeptidic sequences—RGD—have received significant attention.

3.1. RGD

In their pioneering work in the mid-1980s, Ruoslahti and Pierschbacher reported on the importance of the Arginine-Glycine-Aspartic acid (RGD) tri-sequence in fibronectin as an essential cell recognition site for integrin α5β1 [59]. This observation has rapidly led to other evidence for the central role of RGD as a general ligand for additional proteins [60,61,62,63,64,65].

In parallel, a family of glycoprotein cell surface receptors was discovered that were recognized by the RGD sequence, [66,67,68,69,70,71,72]. This class of cell surface receptors, consisting of two subunits in mammalian cells [73,74], were for the first time termed “Integrins” in 1986 for their role as “an integral membrane complex involved in the transmembrane association between the extracellular matrix and the cytoskeleton” [75].

Since integrins are involved in processes which are usually associated with tumor progression such as angiogenesis, invasion and metastasis, and since the RGD peptidic ligand selectively targets them, integrins have attracted special focus. Currently, 24 distinct integrins are known [76] and they have been shown to play key roles in many processes including cell adhesion, migration, and proliferation [77]. Enhanced expression of specific integrins in cancer cells is crucial for promoting metastasis [78,79,80], angiogenesis [81], proliferation [82,83], migration [84,85,86] and invasion [87,88], as well as for the proteolytic destruction of extra cellular matrix (ECM) [87], all essential components in the process of tumor progression [89]. The variety of essential functions attributed to the different members of this receptor family in the neoplastic process have been comprehensively reviewed [89,90]. The over expression of integrins and their important role in different cancers, make them an obvious target for therapeutic intervention [91,92], as well as for the selective delivery of chemotherapeutics [93,94,95,96], nanoparticles [97], and imaging agents [98,99]. RGD has for several reasons often been selected as an attractive targeting ligand in many peptide–drug conjugates: it is recognized by 8–12 of the 24 known integrins [93], and there is confirmation of the recognition of RGD by integrins on the structural basis [100], which is also supported by the crystal structure of αvβ3 integrin with the RGD analog Celingetide [101]. For example, Ruoslahti et al published a work in which RGD peptides were used to selectively deliver cytotoxic compounds to cancer cells. The researchers showed that doxorubicin covalently conjugated to the nonapeptide CDCRGDCFC considerably improved survival rates of mice bearing human MDA-MB-435 breast carcinomas [94]. In another paper, Sherz and coworkers reported on selective accumulation of the cyclic RGD analog conjugated to the fluorescent bacteriochlorophyll analog in the tumor necrotic domain in MDA-MB-231-RFP bearing mice, allowing early detection of tumor proliferation [102]. Conjugation of highly potent microtubulin poison paclitaxel to the bicyclic RGD peptide Ec(RGDyK)2 resulted in increased drug efficacy towards tumor cells and decrease in off-target toxicity [96].

Other highly effective RGD analogs include 9-RGD [103], iRGD [104], and the cyclic penta-peptide Cilengitide, the latter being developed by Kessler and co-workers [105,106]. Cilengitide has reached phase III clinical trials for the treatment of glioblastoma [107] and phase II clinical trials for some other tumors [92]. In all these cases Cilengitide acts as a highly specific antagonist of αvβ3 and αvβ5


Leave a Reply

Your email address will not be published. Required fields are marked *