Key points

Introduction

Colorectal cancer (CRC) is one of the most frequent and deadliest cancers worldwide, with an estimated 136,830 new cases and an estimated 50,310 deaths in 2014 in the United States. Survival times in metastatic disease have significantly improved over the past several decades, lately reaching 24 to 30 months in clinical trials. However, 5-year survival rates remain at a low of about 8%. The therapeutic regimen evolved from single-agent 5-fluorouracil (5-FU), with response rates of about 20% to 25%, to chemotherapeutic combination of leucovorin and 5-FU with oxaliplatin (FOLFOX) or irinotecan (FOLFIRI), with response rates of around 40% to 55%. Infusional 5-FU can be substituted by the rational designed drug capecitabine, which is a prodrug to 5-FU. In metastatic CRC (mCRC), most therapeutic regimens are based on 5-FU. Combination regimens such as FOLFIRI and FOLFOX reach median overall survival (OS) times of about 20 months. During the 2000s, the introduction of antibody treatment against the vascular endothelial growth factor (VEGF) using bevacizumab and the epidermal growth factor receptor (EGFR) using cetuximab or panitumumab has further increased response rates and survival times. The antiangiogenic treatment principle has, in addition to the VEGF-A antibody bevacizumab, guided the development of aflibercept, a recombinant fusion protein to target VEGF-A, VEGF-B, and placental growth factor (PlGF). The multikinase inhibitor regorafenib is also targeting angiogenesis and has a relative specificity to VEGF receptors 1 to 3, platelet-derived growth factor receptor (PDGFR), and other protein kinases. Regorafenib statistically significantly prolonged survival by 1.4 months in a phase III placebo-controlled trial, and is approved for further-line treatment of mCRC. The VELOUR study demonstrated efficacy for aflibercept in combination with FOLFIRI in second-line treatment of mCRC after FOLFOX failure by statistically significantly prolonging survival by 1.4 months. For a selected patient population with younger age and good performance status, trials suggest that a combination of leucovorin, 5-FU, irinotecan, and oxaliplatin (FOLFOXIRI) may further extend the survival benefit reached by cytotoxic agents in mCRC.

The development of new drugs and the introduction of combination regimens have prolonged survival from 6 months to 24 to 30 months. The decision made in selection of first-line treatment is important for OS, but current guidelines are rather nonspecific and offer a panoply of possibilities. In recent years it has become increasingly clear that certain subpopulations of mCRC may derive significant advantage from targeted treatments, whereas other patients do not benefit or might even be harmed by certain combinations. As technologies such as next-generation sequencing (NGS)-based screening for mutations and polymorphism and gene-expression arrays become more accessible, new prognostic and predictive subpopulations of mCRC will be defined. This advance will open up new opportunities to predict treatment response, resistance, and toxicity, and will stimulate drug development by defining novel druggable targets. Approaches such as liquid biopsies will further help to achieve a better understanding of the development of treatment resistance and the best treatment sequence for the individual patient.

Introduction

Colorectal cancer (CRC) is one of the most frequent and deadliest cancers worldwide, with an estimated 136,830 new cases and an estimated 50,310 deaths in 2014 in the United States. Survival times in metastatic disease have significantly improved over the past several decades, lately reaching 24 to 30 months in clinical trials. However, 5-year survival rates remain at a low of about 8%. The therapeutic regimen evolved from single-agent 5-fluorouracil (5-FU), with response rates of about 20% to 25%, to chemotherapeutic combination of leucovorin and 5-FU with oxaliplatin (FOLFOX) or irinotecan (FOLFIRI), with response rates of around 40% to 55%. Infusional 5-FU can be substituted by the rational designed drug capecitabine, which is a prodrug to 5-FU. In metastatic CRC (mCRC), most therapeutic regimens are based on 5-FU. Combination regimens such as FOLFIRI and FOLFOX reach median overall survival (OS) times of about 20 months. During the 2000s, the introduction of antibody treatment against the vascular endothelial growth factor (VEGF) using bevacizumab and the epidermal growth factor receptor (EGFR) using cetuximab or panitumumab has further increased response rates and survival times. The antiangiogenic treatment principle has, in addition to the VEGF-A antibody bevacizumab, guided the development of aflibercept, a recombinant fusion protein to target VEGF-A, VEGF-B, and placental growth factor (PlGF). The multikinase inhibitor regorafenib is also targeting angiogenesis and has a relative specificity to VEGF receptors 1 to 3, platelet-derived growth factor receptor (PDGFR), and other protein kinases. Regorafenib statistically significantly prolonged survival by 1.4 months in a phase III placebo-controlled trial, and is approved for further-line treatment of mCRC. The VELOUR study demonstrated efficacy for aflibercept in combination with FOLFIRI in second-line treatment of mCRC after FOLFOX failure by statistically significantly prolonging survival by 1.4 months. For a selected patient population with younger age and good performance status, trials suggest that a combination of leucovorin, 5-FU, irinotecan, and oxaliplatin (FOLFOXIRI) may further extend the survival benefit reached by cytotoxic agents in mCRC.

The development of new drugs and the introduction of combination regimens have prolonged survival from 6 months to 24 to 30 months. The decision made in selection of first-line treatment is important for OS, but current guidelines are rather nonspecific and offer a panoply of possibilities. In recent years it has become increasingly clear that certain subpopulations of mCRC may derive significant advantage from targeted treatments, whereas other patients do not benefit or might even be harmed by certain combinations. As technologies such as next-generation sequencing (NGS)-based screening for mutations and polymorphism and gene-expression arrays become more accessible, new prognostic and predictive subpopulations of mCRC will be defined. This advance will open up new opportunities to predict treatment response, resistance, and toxicity, and will stimulate drug development by defining novel druggable targets. Approaches such as liquid biopsies will further help to achieve a better understanding of the development of treatment resistance and the best treatment sequence for the individual patient.

Predictive biomarkers for anti–epidermal growth factor receptor

The Rat Sarcoma Story

As of now, the only established and widely accepted predictive biomarker in the treatment of mCRC is mutation of the RAS proto-oncogene. The anti-EGFR antibodies cetuximab and panitumumab have been shown to prolong survival in further-line treatment when applied as monotherapy or in combination with irinotecan. The predictive value of KRAS exon 2 (codons 12 and 13) mutations for cetuximab was first described in a small retrospective analysis of 30 treated patients. The first-line phase III trial testing FOLFIRI plus cetuximab against FOLFIRI alone (CRYSTAL study) confirmed the negative predictive value of KRAS exon 2 mutations for cetuximab treatment. The phase II OPUS study testing FOLFOX against FOLFOX plus cetuximab was the first to reveal a potential harm regarding progression-free survival (PFS) and OS in patients with KRAS exon 2 mutant tumors treated with cetuximab. Accordingly the approval for cetuximab was subsequently restricted to KRAS exon 2 wild-type patients. The PRIME study, a first-line phase III study testing FOLFOX plus panitumumab versus FOLFOX, confirmed the negative predictive value of KRAS exon 2 mutations for the use of panitumumab. All 3 studies showed improved efficacy for anti-EGFR plus chemotherapeutic treatment when compared with chemotherapy alone only in the KRAS exon 2 wild-type population. Further analysis suggested that different mutations might have different predictive value and that exon 2 codon 13 (p.G13D) mutations might have a less detrimental effect on cetuximab treatment, although numbers were too small to establish a clear conclusion.

With further analyses of the PRIME study it has become clear that all RAS mutations, including not only all KRAS exon 2 but also KRAS mutations within exon 3 (codons 59 and 61) and exon 4 (codons 117 and 146), in addition to mutations within the NRAS (neuroblastoma RAS) exon 2 (codons 12 and 13), exon 3 (exons 59 and 61), and exon 4 (codons 117 and 146), were more frequent in mCRC than previously described and, when combined, have a negative predictive value comparable with what has been seen in KRAS exon 2 mutant tumors. NRAS shares more than 90% of the KRAS protein and has similar functions. The extended RAS analysis revealed another 8% to 35% of RAS (KRAS and NRAS) mutations within the KRAS exon 2 wild-type population ( Fig. 1 ). The population defined by the extended RAS analysis as “RAS wild-type” had a statistically significant OS benefit when treated with panitumumab and FOLFOX in comparison with FOLFOX alone, and the hazard ratio in favor for the panitumumab arm improved from 0.83 to 0.78. The influence of the extended RAS analyses on study end points of the OPUS, PEAK, FIRE-3, PICCOLO, and CORE-2 studies are summarized in Table 1 . As a consequence of these data, the use of anti-EGFR antibodies in mCRC is now restricted to RAS wild-type patients.

KRAS exon 2 versus extended RAS analysis.

( Data from Douillard JY, Oliner KS, Siena S, et al. Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer. N Engl J Med 2013;369(11):1023–34.)

Table 1

Studies published with extended RAS analyses using cetuximab/panitumumab in the treatment of metastatic colorectal cancer

Trial (KRASwt) (RASwt)	Treatment Arms	ORR (%)	P * OR	ORR (%) Extended RAS	P * OR	PFS (mo)	P ** HR	PFS (mo) Extended RAS	P ** HR	OS (mo)	P ** HR	OS (mo) Extended RAS	P ** HR
First-Line Treatment
PRIME
n = 656	FOLFOX	48	.068	NA	NA	8.0	.02	7.9	.004	19.4	.03	20.2	.009
n = 512	FOLFOX + Pan	55	1.35			9.6	0.80	10.1	0.72	23.8	0.82	25.8	0.78
OPUS
n = 179	FOLFOX	34.0	.002	30.4	.008	7.2	.006	5.8	.018	18.5	.39	17.8	.50
n = 82	FOLFOX + Cet	57.3	2.55	61.1	3.46	8.3	0.57	12.0	0.43	22.8	0.86	20.7	0.83
FIRE-3
n = 592	FOLFIRI + Cet	62.0	.18	65.5	.32	10.0	.547	10.4	.54	28.7	.017	33.1	.011
n = 342	FOLFIRI + Bev	58.0	1.18	59.6	1.28	10.3	1.06	10.2	0.93	25.0	0.77	25.6	0.70
PEAK
n = 285	FOLFOX + Pan	58	NA	64	NA	10.9	.22	13.0	.03	34.2	.009	41.3	.06
n = 170	FOLFOX + Bev	54		60		10.1	0.84	10.1	0.66	24.3	0.62	28.9	0.63
CORE-2
n = 152	FOLFOX + Cet	57.9	NA	61.3 a	NA	9.3	NA	9.7 a	NA	25.2	NA	28.5 a	NA
n = 124 a
Second-Line Treatment
PICCOLO
n = 597	FOLFIRI	10	.001	12	NA	3.9	.004	NA	.015	12.5	.12	NA	.91
n = 323	FOLFIRI + Pan	35	NA	34		5.9	0.73		0.78	14.5	0.85		1.01

Abbreviations: Bev, bevacizumab; Cet, cetuximab; HR, hazard ratio; KRASwt, number of patients with KRAs exon 2 wild-type tumors; NA, not available; OR, odds ratio; ORR, overall response rate; OS, overall survival; P *, Fisher’s exact test P ; P **, log-rank test P ; Pan, panitumumab; PFS, progression-free survival; RASwt, number of patients with extended RAS analyses wild-type tumors.

a In CORE-2 study, data for RAS and BRAF wild-type are given.

It is important to mention that different detection methods have different sensitivities, leading to different frequencies of RAS mutations. The methods used are Sanger-Sequencing, with a sensitivity of 10% to 20%, Pyrosequencing, with a sensitivity of less than 5%, WAVE based SURVEYOR Scan Kits (Transgenomic, Omaha, NE, USA), with a sensitivity of approximately 1%, and BEAMing technology (Sysmex Inostics, Inc, Baltimore, MD, USA), with a sensitivity of 0.01%. To date no cutoff analysis has been presented, and the question remains as to at what level of detection sensitivity patients will still benefit from anti-EGFR treatment. The detection of 0.1% or 0.01% RAS mutations may be critical to identify subclones that will outgrow anti-EGFR therapy but may be not sufficient to withhold therapy. For the analyses of KRAS exon 2 (codons 12 and13) a sensitivity between 1% and 5% has been regarded as sufficient. The widely used cobas KRAS mutation test (Roche Molecular Diagnostics, Pleasanton, CA, USA) has a sensitivity of less than 5%. The only test for KRAS exon 2 (codon 12 and 13) approved by the Food and Drug Administration has a sensitivity of 1% (TheraScreen KRAS test; Qiagen, Valencia, CA, USA).

Recently it has been suggested that RAS mutations may have a predictive value for the use of oxaliplatin in first-line treatment of mCRC, as RAS mutant patients in one report had a longer PFS and OS when compared with RAS wild-type patients. Cell culture experiments revealed a higher efficacy for oxaliplatin in KRAS mutation transfected colorectal cancer cell lines by downregulating ERCC1 levels. In a German AIO study comparing CAPIRI (capecitabine plus irinotecan) against CAPOX (capecitabine plus oxaliplatin) both in combination with bevacizumab in first-line treatment of mCRC, a statistically significantly longer OS of the irinotecan arm in KRAS exon 2 mutant patients was demonstrated. More data are needed to better understand the role of RAS mutations and their chemosensitivity to oxaliplatin.

Other Mutations in the Epidermal Growth Factor Receptor Signaling Pathway

The relations and positions of the different predictive and prognostic markers in the EGFR-dependent signaling cascade are illustrated in Fig. 2 .

Simplified EGFR signaling pathway. AKT, AKR mouse thymoma inducing; AREG, amphiregulin; cSRC, cellular sarcoma inducing; EGFR, epithelial growth factor receptor; EREG, epiregulin; ERK, extracellular signal–regulated kinases; FAK, focal adhesion kinase; JNK, janus kinase; MEK, mitogen-activated protein kinase; mTOR, mammalian target of rapamycin; PI3K, phosphatidylinositol-3-kinase; PKC, protein kinase C; PLCy, phospholipase C; PTEN, phosphatase and tensin homologue; RAF, RAS-associated factor; RAS, rat sarcoma; STAT, signal transducer and activator of transcription. ^a Frequency of mutations in mCRC.

BRAF

BRAF is a member of the RAF (RAS-associated factor) gene family, which is downstream of RAS. Activating BRAF mutations (exon 15, V600E) are detected in about 10% to 15% of all mCRC cases. mCRC cases with both RAS and BRAF mutations are extremely rare, with an estimated frequency of 0.001%. Patients with BRAF mutant tumors have a poor prognosis, with short median OS times ranging from 9 to 14 months. Because of the small numbers of patients in individual trials, pooled analyses using data from several randomized trials were performed to determine the prognostic and/or predictive value of BRAF mutant tumors in patients with mCRC. Even then, no statistically significant statement could be made. However, there is a consensus that BRAF mutations are prognostic markers. There are data suggesting that patients with BRAF mutant mCRC may still benefit from first-line anti-EGFR therapy; however, some of the best clinical outcome data in these patients was shown using FOLFOXIRI plus bevacizumab in first-line treatment, reaching 11.8 months of PFS and 24.1 months of OS. Single-agent BRAF inhibitors have shown clinical efficacy in other tumor entities, but in mCRC single-agent activity was low. However, the combination of BRAF inhibitors and anti-EGFR antibodies is looking promising because of upregulated EGFR expression.

PIK3CA mutations

Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit α (PIK3CA) is subordinate to RAS and BRAF, activated through EGFR signaling (see Fig. 2 ). PIK3CA mutations can be detected in about 10% to 30% of mCRC samples. In almost 50% of cases, PIK3CA mutant tumors also harbor a concomitant BRAF mutation and in about 40% another RAS mutation ( Fig. 3 ), which makes it difficult to assess the predictive and prognostic value. PIK3CA mutations in CRC have gained more attention since a report claimed a higher survival rate in PIK3CA mutant adjuvant CRC patients taking aspirin, although a recently presented study was not able to confirm this.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Predictive and Prognostic markers in the treatment of metastatic colorectal cancer (mCRC) – personalized medicine at work – Surgical Management of Hepatic Metastases of Colorectal Cancer Is More Not Better? Diagnosis and Management of DNA Mismatch Repair-Deficient Colorectal Cancer Diet and Lifestyle in Survivors of Colorectal Cancer Is More Not Better?

Stay updated, free articles. Join our Telegram channel

Join