Panitumumab in the management of patients with KRAS wild-type metastatic colorectal cancer

Abstract

The past 15 years has seen a marked increase in available therapeutic options for patients with metastatic colorectal cancer resulting in improvements in median survival from 12 to 24 months. One of these new options is panitumumab, which is a fully humanized monoclonal antibody that binds to the epidermal growth factor receptor of tumor cells and inhibits downstream cell signaling with antitumor effects of inhibition of tumor growth, induction of apoptosis and inhibition of angiogenesis. Large randomized clinical trials have demonstrated significant improvements in tumor response rates and progression-free survival when panitumumab is combined with chemotherapy and as monotherapy in chemorefractory metastatic colorectal cancer. Clinical benefit with panitumumab is limited to patients with nonmutated KRAS tumors. Rash is a common toxicity of panitumumab treatment but can potentially be ameliorated with the use of prophylactic strategies. The role of panitumumab in the overall treatment of metastatic colorectal cancer is evolving and future clinical trials will focus on improved patient selection through use of novel predictive biomarkers, and the optimal timing of treatment.

Introduction

Colorectal cancer (CRC) is the third most commonly diagnosed cancer worldwide, and the second leading cause of cancer mortality (behind lung cancer) in the developed world [Jemal et al. 2011]. While the incidence of CRC is reducing or stabilizing in historically high incident areas such as North America, Europe and Australia, there has been an increase in incidence rates in Spain, Eastern Europe and Eastern Asia, due to dietary changes and higher rates of obesity and smoking [Center et al. 2009a, 2009b]. Outcomes in patients with metastatic CRC (mCRC) have improved significantly over the last 15 years largely due to increased adoption of surgical resection of liver-limited metastasis and improved systemic therapeutic options [Kopetz et al. 2009]. In the late 1990s 5-fluourouracil (5-FU) was the only available systemic therapy for mCRC, a contrast to today where nine drugs with significant antitumor activity are approved by the US Food and Drug Administration (FDA) for use in mCRC (Table 1). The availability of these agents has seen median overall survival (OS) in patients with mCRC improve from 8–12 months in 1998 to 18–21 months with the availability of capecitabine, oxaliplatin and irinotecan [Grothey et al. 2004]. The addition of targeted therapies has further improved reported OS in mCRC to 21–24 months [Douillard et al. 2010; Saltz et al. 2008; Van Cutsem et al. 2009].

Table 1.

US Food and Drug Administration approved agents for metastatic colorectal cancer.

Chemotherapy
Fluoropyrimidines (5-fluourouracil, capecitabine)
Oxaliplatin
Irinotecan
Targeted Agents
Vascular endothelial growth factor
Bevacizumab
Aflibercept
Regorafenib
Epidermal growth factor receptor
Cetuximab
Panitumumab

Despite these improvements, long-term survival of patients with unresectable mCRC remains poor with 5-year survival of approximately 10% [Kopetz et al. 2009; Peeters and Price, 2012].

Approximately 60–70% of patients with mCRC receive a biological agent during their treatment course targeting one of two main pathways; vascular endothelial growth factor (VEGF) and epidermal growth factor receptor (EGFR) [Peeters and Price, 2012].

Bevacizumab (Avastin®, F. Hoffmann-La Roche Ltd, Basel, Switzerland), a monoclonal antibody (MoAb) targeting VEGF, prolongs progression-free survival (PFS) and OS when used in the first and second lines of treatment in combination with chemotherapy [Peeters and Price, 2012]. A median OS benefit of 1.4 months (21.3 versus 19.9 months) was demonstrated in the NO16966 trial with the addition of bevacizumab to first-line infusional 5-FU/folinic acid/oxaliplatin (FOLFOX) chemotherapy [Saltz et al. 2008]. Aflibercept (Zaltrap, Sanofi U.S., Inc.), also targeting VEGF, was approved by the FDA in August 2012 for second-line treatment of mCRC in combination with infusional 5-FU/folinic acid/irinotecan (FOLFIRI) chemotherapy after demonstrating improved PFS and OS (13.5 versus 12.1 months) in this setting [Van Cutsem et al. 2012]. Regorafenib, an oral kinase inhibitor of angiogenic signaling (including VEGF) has demonstrated improved OS (6.4 versus 5.0 months) and PFS as monotherapy in treatment refractory mCRC [Grothey et al. 2013]. FDA approval of Regorafenib in September 2012 provides a further treatment option for patients with mCRC.

Two EGFR inhibitors are available for treatment of mCRC: cetuximab (Erbitux®, Merck KgaA, Darmstadt, Germany) and panitumumab (Vectibix®, Amgen Inc., CA, USA). The efficacy of these drugs is limited to patients with wild-type (nonmutated) Kirsten Rat Sarcoma (KRAS WT) tumors. Evidence supports use of these agents as monotherapy in chemotherapy refractory mCRC and in first and second lines of treatment in combination with chemotherapy.

The markedly increased number of active agents and plethora of recent clinical trial data have made the overall treatment algorithm for mCRC more complex. There is a lack of randomized trial data evaluating the optimal sequencing and combinations of available agents, and apart from KRAS mutation status there is a lack of predictive biomarkers available to support drug selection for individual patients. Until such data are available the clinician relies on an understanding of data relating to each class of agent with the goal of exposing a patient to all available active agents to achieve improved survival and optimal disease control [Kopetz et al. 2009].

The following paper examines the clinical trial data relating to panitumumab in the treatment of KRAS WT mCRC and provides a commentary for clinicians on how panitumumab may be best used in the clinic until further data becomes available.

Pharmacology of panitumumab

The EGFR is a transmembrane cell surface glycoprotein and a member of the human epithelial growth factor receptor (HER) family of tyrosine kinases [El Zouhairi et al. 2011]. Binding of ligands to the EGFR leads to homodimerization or heterodimerization of HER family members and results in autophosphorylation of intracellular tyrosine residues, initiating several key cell-signaling pathways including Ras/Raf/MAPK and PI3K/AKT pathways [El Zouhairi et al. 2011; Jean and Shah, 2008; Yang et al. 2010]. Binding to EGFR enhances processes critical to tumor growth and progression such as angiogenesis, apoptosis inhibition, tumor invasiveness and metastatic spread, making EGFR inhibition an important therapeutic target in cancer [Toffoli et al. 2007; Yang et al. 2010].

Panitumumab is a fully humanized IgG2 monoclonal antibody that binds to the extracellular domain of the EGFR of both tumor and normal cells with high affinity. By competing with endogenous ligand binding panitumumab inhibits receptor phosphorylation and activation of EGFR associated cell signaling (see figure 1) [Yang et al. 2010]. This also results in downregulation of EGF receptors through receptor internalization, induction of apoptosis, autophagy and inhibition of angiogenesis [Giannopoulou et al. 2009; Keating, 2010; Yang et al. 1999].

Figure 1.

Panitumumab and epidermal growth factor receptor (EGFR) signaling pathway. (Illustration courtesy of Alessandro Baliani. Copyright © 2013.)

Panitumumab is administered intravenously at a recommended dosing schedule of 6 mg/kg once every 2 weeks [Doi et al. 2009; Stephenson et al. 2009]. This schedule results in steady-state concentrations by the third infusion [Jean and Shah, 2008]. The elimination half-life of panitumumab is approximately 7 days, longer than cetuximab where the half-life (following 400 mg/kg infusion) is 4 days, providing the rationale for the weekly schedule of cetuximab versus the alternate weekly administration schedule for panitumumab [Jean and Shah, 2008].

Panitumumab has dual clearance mechanisms. Specific clearance of panitumumab involves binding to EGFR, internalization into the cell and degradation in lysosomes. This mechanism of clearance plays an important role of panitumumab clearance but is capacity limited resulting in nonlinear clearance. At doses above 2 mg/kg clearance is dependent on nonspecific mechanisms mediated by the reticuloendothelial system (RES) [Yang et al. 2010].

Predictive markers of response to panitumumab

KRAS exon 2 mutations

Tumor KRAS gene mutation predicts for resistance to anti-EGFR antibodies and remains the only prospectively validated predictive biomarker to assist choice of systemic treatment in mCRC.

KRAS is a small serine-threonine kinase activated just downstream of activation of the EGFR (or other tyrosine kinase receptors) and propagates further signaling events. Gain of function mutations in KRAS result in a constitutively active protein, regardless of EGFR activation, resulting in increased proliferation, tumor angiogenesis, metastasis, and inhibition of apoptosis, which support continued cancer cell survival [Peeters et al. 2009a]. This provides the biological rationale for the finding that KRAS mutations predict for resistance to anti-EGFR antibody therapy [Yang et al. 2010].

KRAS mutations are found in 30–45% of CRC tumors, with high concordance between primary and metastatic lesions [Maughan et al. 2011; Peeters et al. 2012b; Price et al. 2011; Van Cutsem et al. 2011]. Mutations are most common in exon 2 (codon 12 and 13) but also occur in exon 3 (codon 61) and exon 4 (codon 117 and 146) [De Roock et al. 2010b; Oliner et al. 2013; Peeters et al. 2012b].

Clinical evidence for the predictive role of KRAS exon 2 mutation first came from retrospective analysis of phase III trials evaluating anti-EGFR antibodies as monotherapy in the chemorefractory setting [Amado et al. 2008; Karapetis et al. 2008]. Subsequently trials of anti-EGFR antibodies in combination with chemotherapy have confirmed, prospectively, the negative predictive role of KRAS exon 2 mutation on clinical outcomes [Peeters et al. 2010].

Evidence from retrospective analysis of patients treated with cetuximab suggests patients with G13D KRAS mutations may benefit from cetuximab therapy to a similar degree as KRAS WT patients [De Roock et al. 2010b; Tejpar et al. 2012]. This effect, however, was not seen in a retrospective pooled analysis of the three phase III trials of panitumumab therapy [Douillard et al. 2010; Peeters et al. 2010; Van Cutsem et al. 2007]. In 1053 KRAS MT patients treated with panitumumab, no specific exon 2 KRAS MT allele demonstrated predictive value for outcome [Peeters et al. 2012b]. Whether there is a difference between the mechanism of action of the drugs remains unclear and these results highlight the difficulties of retrospective analysis.

Ultimately analysis of KRAS mutation status is required prior to treatment with an anti-EGFR antibody in mCRC with regulating bodies restricting use of panitumumab to patients with confirmed KRAS WT mCRC, and this is unlikely to change unless there is prospective data relating to G13D.

Predictive markers beyond KRAS exon 2

Importantly, not all KRAS WT patients respond to anti-EGFR antibodies, highlighting the need to identify additional predictive biomarkers to better assist patient selection for treatment with this class of drug. Mutations in KRAS exon 3 and 4 have since been analyzed retrospectively in phase III randomized trials of panitumumab [Oliner et al. 2013; Peeters et al. 2013]. While the incidence of KRAS mutations beyond exon 2 is relatively small, exon 3 MT (2–4%) and exon 4 MT (5–6%) appear to confirm resistance to panitumumab and exclusion of these patients improves the relative benefit of panitumumab therapy [Oliner et al. 2013; Peeters et al. 2013; Seymour et al. 2013].

Other downstream effector genes, BRAF, NRAS and PIK3CA, and reduced PTEN protein expression have been postulated as conferring resistance to anti-EGFR antibodies though results, largely based on retrospective subgroup analysis have been conflicting [De Roock et al. 2010a; Di Nicolantonio et al. 2008; Oliner et al. 2013; Peeters et al. 2013; Tol et al. 2010; Van Cutsem et al. 2011].

NRAS mutation, found in 5–7% of CRC specimens, is emerging as a consistent negative predictive biomarker for panitumumab therapy with supporting evidence from both retrospective and now prospective large phase III panitumumab trials [Oliner et al. 2013; Peeters et al. 2013; Seymour et al. 2013]. Mutation in tumor BRAF gene occurs in approximately 10% of mCRC, is mutually exclusive of KRAS mutation and an independent poor prognostic marker in mCRC [Price et al. 2011]. BRAF mutation may provide an explanation for an additional population of patients resistant to anti-EGFR antibodies, although conflicting trial results have failed to establish BRAF firmly as a predictive biomarker. BRAF requires further validation in large cohorts of mCRC patients [Bardelli and Siena, 2010; Tie et al. 2011].

The relatively low incidence of these mutations has reduced the power of retrospective analysis to draw conclusions about the predictive role of individual mutations. Some authors have suggested selection of ‘triple wild-type’ patients (KRAS/NRAS/BRAF) for anti-EGFR antibody treatments may improve patient selection, response rates and outcome, although given the low prevalence of BRAF and NRAS mutations, further validation, probably by way of meta-analysis, is required [Bardelli and Siena, 2010; De Roock et al. 2010a].

Preclinical trials have classified CRC into subgroups based on extended gene expression profiles [De Sousa et al. 2013; Sadanandam et al. 2013]. In vitro resistance to anti-EGFR antibodies has been demonstrated in specific subgroups and the development of a clinically useful assay may allow improved patient selection for these therapies.

Inferior outcome in RAS MT patients treated with panitumumab

The predictive biomarkers discussed above identify subgroups of patients who do not benefit from anti-EGFR antibody therapy. Of concern is the emerging evidence regarding the potential harm of these subgroups of patients when receiving such therapy.

The first-line PRIME trial of FOLFOX chemotherapy with or without panitumumab was the first trial to demonstrate an inferior outcome in KRAS MT patients treated with panitumumab (PFS 7.3 versus 8.8 months, hazard ratio [HR] 1.29, p = 0.02; and OS 15.5 versus 19.3 month, HR 1.24, p = 0.068). Inferior PFS and OS were also seen in KRAS MT patients in the first-line OPUS trial, evaluating cetuximab and FOLFOX in mCRC [Bokemeyer et al. 2011]. A prespecified retrospective analysis of the PRIME trial assessing RAS mutations beyond KRAS exon 2 was recently presented [Oliner et al. 2013]. With exon 3 and 4 KRAS MT and NRAS MT included in the analysis, a further 17% of patients were identified as RAS MT. This group had a numerically inferior OS when treated with panitumumab. When included in the updated analysis, patients with any RAS MT had inferior PFS (HR 1.31, 95% confidence interval [CI] 1.07–1.60) and OS (HR 1.25, CI 1.02–1.55) when treated with panitumumab in combination with FOLFOX compared with FOLFOX alone. The finding of inferior outcome in RAS MT patients with panitumumab was also demonstrated in the second-line phase III PICCOLO trial randomizing patients to irinotecan versus irinotecan and panitumumab [Seymour et al. 2013]. In the analysis of patients with any mutation (KRAS/NRAS/BRAF/PIK3CA) a statistically inferior OS was seen in the panitumumab-treated patients versus irinotecan alone (HR 1.64, 95% CI 1.14–2.34). Given the low prevalence of each individual mutation statistical significance of inferiority was not demonstrated individually, except for BRAF MT which predicted for inferior OS with panitumumab (HR 1.84, 95% CI 1.10–3.08). Interestingly inferior outcomes were not observed in KRAS MT patients in the other phase III trials of panitumumab [Amado et al. 2008; Peeters et al. 2010]. The evidence of a potentially harmful effect in RAS MT patients treated with oxaliplatin in combination has lead to a recommendation that RAS testing beyond exon 2 should occur in clinical practice.

The mechanism for the observed adverse outcome in KRAS MT patients treated with anti-EGFR antibodies remains unclear. A recent hypothesis suggests nonmutant isoforms of RAS may play a role in suppressing the mutant RAS isoform. For example, in the setting of mutant KRAS, wild-type NRAS and HRAS have been shown to continue to play an important role in modulating downstream signaling. Furthermore wild-type RAS isoforms have an inhibitory effect on oncogenic mutant KRAS [Young et al. 2013]. Inhibition of EGFR (i.e. with panitumumab) inhibits the function of wild-type RAS isoforms and may remove its inhibitory effect on mutant RAS, paradoxically leading to increase in MAPK signaling and cell proliferation.

Evidence for clinical efficacy of panitumumab

First-line trials

The evidence to support the use of panitumumab in the first-line treatment setting comes from the phase III PRIME trial [Douillard et al. 2010] and two single-arm phase II trials [Berlin et al. 2007; Kohne et al. 2012] (Table 2).

Table 2.

Progression-free survival and overall survival in first-line trials of panitumumab.

Trials	KRAS status	Treatment regimen (Control)	PFS (months)	HR (p values)	Median OS (months)	HR (p values)	ORR
Phase III trials
PRIME	KRAS WT	FOLFOX4 + P	9.6	0.8 (p=0.02)	23.9	0.83 (p=0.072)	55%
(n = 1183)		(FOLFOX4)	8		19.7		48%
	KRAS MT	FOLFOX4 + P	7.3	1.29 (p=0.02)	15.5	1.24 (p=0.068)	40%
		(FOLFOX4)	8.8		19.3		40%
Phase II trials
PEAK	KRAS WT	FOLFOX + P	10.9	HR 0.87 (CI 0.65–1.17)	Not Reached	HR 0.72 (CI 0.47–1.11)	58%
(n=285)		FOLFOX + Bev	10.1		25.4		54%
[Kohne et al. 2012]	Unselected	FOLFIRI + P	7.6		NR		48%
(n=154)	KRAS WT	FOLFIRI + P	8.9		NR		56%
	KRAS MT	FOLFIRI + P	7.2		NR		38%
[Berlin et al. 2007]	Unselected	IFL + P	5.6		17		46%
(n=43)	Unselected	FOLFIRI + P	10.9		22.5		42%

Bev, bevacizumab; CI, 95% confidence interval; FOLFIRI, infusional 5-fluorouracil/folinic acid/irinotecan; FOLFOX, infusional 5-fluorouracil/folinic acid/oxaliplatin; HR, hazard ratio; IFL, bolus 5-fluorouracil/folinic acid/irinotecan; KRAS, Kirsten rat sarcoma gene; MT, mutated; n, number of patients; NR, not reported; ORR, objective response rate; OS, overall survival; P, panitumumab; PFS, progression-free survival; WT, wild-type (nonmutated).

The PRIME (Panitumumab Randomized Trial in Combination with Chemotherapy for Metastatic Colorectal Cancer to Determine Efficacy) trial randomized 1183 untreated patients with mCRC to FOLFOX chemotherapy (the control arm) versus FOLFOX chemotherapy and panitumumab. PFS in KRAS WT patients (the primary endpoint) was significantly improved in the panitumumab/FOLFOX arm versus FOLFOX alone (9.6 versus 8.0 months; p = 0.02). There was a trend to improvement in OS (23.9 versus 19.7 months; p = 0.072) and objective response rate (ORR) (55% versus 48%; p = 0.068) in the panitumumab arm although these did not reach statistical significance. Unexpectedly inferior outcomes were seen in KRAS MT patients treated with panitumumab is discussed in detail above (see the section on ‘Inferior outcome in RAS MT patients treated with panitumumab’).

Further evidence for the use of panitumumab in the first-line setting comes from three phase II trials (Table 2). The recently completed PEAK study, a randomized trial presented in abstract form, is the first head-to-head analysis of FOLFOX combined with either panitumumab or bevacizumab in KRAS WT mCRC [Schwartzberg et al. 2013a]. PFS was 10.9 versus 10.1 months (HR 0.87, 95% CI 0.65–1.17), OS not reached versus 25.4 months (HR 0.72, 95% CI 0.47–1.11) and RR 58% versus 54%, respectively. Extended RAS mutation analysis (KRAS exon 3 and 4 and NRAS MT) was also undertaken. In a finding similar to the PRIME trial, patients with WT KRAS exon 2 but MT in other RAS locus demonstrated inferior PFS with panitumumab therapy [Schwartzberg et al. 2013b]. These early results suggest that either biological agent may be considered in first-line therapy of RAS WT patients although larger phase III trials are awaited to confirm these results.

Kohne and colleagues evaluated FOLFIRI chemotherapy with panitumumab in 154 patients with ORR as the primary endpoint [Kohne et al. 2012]. When assessed by KRAS status there was improved ORR (56% versus 38%) and PFS (8.9 versus 7.2 months) in the KRAS WT compared with KRAS MT group with the degree of benefit consistent with the findings of the PRIME trial. Berlin and colleagues assessed panitumumab in combination with irinotecan-based chemotherapy in 43 patients with mCRC [Berlin et al. 2007]. KRAS status was not determined.

Summary

The PRIME study supports the use of first-line panitumumab in combination with FOLFOX for KRAS WT patients, with significant improvement in PFS and ORR demonstrated with the addition of panitumumab. The inferior outcome seen in the KRAS MT patients emphasizes the need to determine KRAS status prior to treatment with an anti-EGFR antibody, as treatment in an unselected population has the potential for inferior outcomes.

Second-line trials

The evidence supporting the use of panitumumab in second-line treatment of mCRC comes from two large phase III randomized trials, and two single-arm phase II trials. All of these trials assessed panitumumab in combination with irinotecan-based chemotherapy (Table 3).

Table 3.

Progression-free survival and overall survival in second-line trials of panitumumab.

Trials	KRAS status	Treatment regimen (Control)	PFS (months)	HR (p values)	Median OS (months)	HR (p values)	ORR	HR (p values)
Phase III trials
[Peeters et al. 2010]	KRAS WT	FOLFIRI + P	5.9	0.73 (p=0.004)	14.0	0.85 (p=0.12)	35%	p<0.001
(n=1186)		FOLFIRI	3.9		12.5		10%
	KRAS MT	FOLFIRI + P	5.0	0.85 (p=0.14)	11.8	0.94	13%
		FOLFIRI	4.9		11.1		14%
PICCOLO	KRAS WT	IrPan	NR	0.78 (p=0.015)	10.4	1.01 (p=0.91)	34%	p=0.0001
(n=460)		Irinotecan	NR		10.9		12%
Phase II trials
SPIRITT	KRAS WT	FOLFIRI + P	7.7	HR 1.01 (CI 0.68–1.50)	18	HR 1.06 (CI 0.75–1.49)	32%
(n=182)		FOLFIRI + Bev	9.2		21.4		19%
[Cohn et al. 2011]	KRAS WT	FOLFIRI + P	6.0	(KRAS WT versus KRAS MT)	11.5	(KRAS WT versus KRAS MT)	23%	(KRAS WT versus KRAS MT)
(n=109)	KRAS MT	FOLFIRI + P	4.3	HR 0.8 (CI 0.5–1.1)	7.1	HR 0.6 (CI 0.4-0.9)	16%	OR 1.6 (CI 0.6–4.3)
STEPP
(n=87)	KRAS WT	I/FOLFIRI + P	5.5		13.7		16%
	KRAS MT	I/FOLFIRI + P	3.3		13.3		8%

CI, 95% confidence interval; FOLFIRI, infusional 5-fluorouracil/folinic acid/irinotecan; HR, hazard ratio; I, irinotecan; IrPan, irinotecan and panitumumab; KRAS, Kirsten rat sarcoma gene; MT, mutated; P, panitumumab; PFS, progression-free survival; NR, not reported; ORR, objective response rate; OS, overall survival; WT, wild-type.

The large phase III trial conducted by Peeters and colleagues randomized 1345 patients to FOLFIRI chemotherapy alone or in combination with panitumumab for second-line treatment of mCRC [Peeters et al. 2010]. Statistically significant improvement in PFS was seen in KRAS WT patients treated with panitumumab compared with FOLFIRI alone (5.9 versus 3.9 months, HR 0.73, p = 0.004). A trend to improved OS was observed in KRAS WT patients (14.5 versus 12.5 months, HR 0.85, p = 0.12). No difference in PFS or OS was seen in KRAS MT patients. There was an imbalance in subsequent anti-EGFR use in the KRAS WT patients (9% in panitumumab group versus 31% control), which may have attenuated the treatment effect of panitumumab on OS. The reported ORR of 35% in panitumumab-treated KRAS WT patients is impressive and the highest reported in a second-line mCRC trial. No detrimental effect was observed in the KRAS MT patients treated with panitumumab and FOLFIRI. The results of this trial are consistent with findings of the second-line EPIC trial evaluating the addition of cetuximab to irinotecan, where a PFS improvement of 1.2 months was reported in the KRAS WT population (4.0 versus 2.8 months; HR 0.77; p = 0.095) [Sobrero et al. 2008]. The EPIC trial did not have the same methodological rigor, however, with the KRAS status analyzed retrospectively and in only 23% of patients.

Seymour and colleagues conducted a randomized phase III trial of irinotecan with or without panitumumab (IrPan) in mCRC patients progressing after fluoropyrimidine therapy [Seymour et al. 2013]. Panitumumab was administered at a schedule of 9 mg/kg 3-weekly. Of the 460 randomized exon 2 and 3 KRAS WT patients no difference was seen in the primary endpoint median OS in the IrPan-treated patients (10.4 versus 10.9 months, HR 1.01, 95% CI 0.83–1.23). IrPan did lead to prolonged PFS (HR 0.78, 95% CI 0.64–0.95) and improved ORR (34% versus 12%, p < 0.0001).

Cohn and colleagues evaluated FOLFIRI and panitumumab in 116 patients who had progressed following an oxaliplatin and bevacizumab containing first-line treatment for mCRC [Cohn et al. 2011]. KRAS mutation analysis was available in 109 patients with KRAS MT found in 41%. PFS was improved in KRAS WT patients compared with KRAS MT patients (6.0 versus 4.9 months), as were OS (11.5 versus 7.1 months) and ORR (23% versus 16%).

The STEPP trial specifically evaluated optimal management of skin toxicity with panitumumab treatment, but provides further evidence for use in combination with irinotecan-based chemotherapy in the second-line setting. A total of 98 patients treated with panitumumab plus either irinotecan or FOLFIRI (at investigators choice) were randomized to pre-emptive versus reactive treatment of skin toxicity [Lacouture et al. 2010; Mitchell et al. 2011]. KRAS MT were found in 44%. ORR (16% versus 8%), PFS (5.5 versus 3.3 months) and OS (13.7 versus 13.3 months) were higher in KRAS WT patients compared with KRAS MT. The results skin toxicity management is discussed below (see the section on ‘Tolerability’).

As is the case in first-line therapy, head-to-head trials of chemotherapy with either panitumumab or bevacizumab in KRAS WT mCRC are awaited. The recently reported randomized phase II SPIRITT trial randomized patients with KRAS WT mCRC who had received first-line FOLFOX and bevacizumab to second-line FOLFIRI with either panitumumab or bevacizumab [Hecht et al. 2013]. PFS (7.7 versus 9.2 months, HR 1.01, 95% CI 0.68–1.50) and OS (18 versus 21.4 months, HR 1.06, 95% CI 0.75–1.49) were not significantly different between treatment groups. ORR was higher with panitumumab (32% versus 19%) [Cohn et al. 2013].

Summary

Panitumumab in combination with FOLFIRI is associated with an improved PFS compared with FOLFIRI alone in second-line treatment of KRAS WT mCRC. Any difference in outcome between second-line/continuation bevacizumab versus panitumumab is awaited in phase III trials.

Third-line trials

The first phase III evidence of clinical efficacy of panitumumab in mCRC came from trials of panitumumab monotherapy in patients with chemorefractory disease. Van Cutsem and colleagues conducted a phase III trial, randomizing 463 patients to panitumumab plus best supportive care (BSC) versus BSC alone [Van Cutsem et al. 2007]. Panitumumab improved PFS (8.0 versus 7.3 weeks, HR 0.54, p < 0.0001) and ORR (10% versus 0%, p ≤ 0.0001) compared with BSC alone. No difference in OS was seen (Table 4). Retrospective analysis by tumor KRAS status helped define the role of KRAS as a predictor of response [Amado et al. 2008]. Exon 2 KRAS mutations were identified in 43% and the benefit of panitumumab was restricted to patients with KRAS WT tumors (PFS 12.3 versus 7.3 weeks, HR 0.45, 95% CI 0.34–0.59). No benefit was observed in KRAS MT patients treated with panitumumab (PFS 7.4 versus 7.3 weeks, HR 0.99, 95% CI 0.73–1.36). No improvement in OS was demonstrated irrespective of KRAS status. It is likely any panitumumab-related OS survival benefit was attenuated by the high crossover rate (76%) of BSC patients to panitumumab as part of a prespecified crossover extension study [Van Cutsem et al. 2008]. These results are consistent with the CO.17 trial of cetuximab monotherapy where efficacy was also limited to KRAS WT patients. In this trial crossover was not permitted and only 7% of patients randomized to BSC subsequently received cetuximab. OS was improved in cetuximab-treated KRAS WT patients compared with BSC [Jonker et al. 2007; Karapetis et al. 2008].

Table 4.

Progression-free survival and overall survival in third-line trials of panitumumab.

Trials	KRAS status	Treatment regimen (Control)	PFS (weeks)	HR (p values)	Median OS (months)	HR (p values)	ORR	HR (p values)
Phase III trials
[Van Cutsem et al. 2007]	Unselected	P + BSC	8	HR 0.54 (p<0.0001)	NR	HR 1.0	10%	p=0.0001
(n=463)		BSC	7.3		NR		0%
	KRAS WT	P + BSC	12.3	HR 0.34	8.1	HR 0.99	17%
		BSC	7.3		7.6		0%
	KRAS MT	P + BSC	7.4	HR 0.99	4.9	HR 1.02	0%
		BSC	7.3		4.4		0%
Phase II trials
[Hecht et al. 2007] (n=148)	Unselected	P	14		9		9%
[Muro et al. 2009] (n=52)	Unselected	P	8		9.3		14%

BSC, best supportive care; HR, hazard ratio; KRAS, Kirsten rat sarcoma gene; P, panitumumab; PFS, progression-free survival; NR, not reported; ORR, objective response rate; OS, overall survival.

An interesting subgroup analysis was reported from the original phase III panitumumab trial, evaluating the association between skin toxicity and efficacy outcomes [Peeters et al. 2009b]. In patients randomized to the panitumumab arm, PFS was improved in those who experienced grade 2–4 compared with grade 1 skin toxicity (HR 0.63, 95% CI 0.45–0.88). This benefit was limited to patients with KRAS WT tumors. Severe (grade 2–4) skin toxicity was also associated with improved OS (HR 0.60, 95% CI 0.43–0.85). This association has also been demonstrated retrospectively with cetuximab treatment [Cunningham et al. 2004; Saltz et al. 2004; Stintzing et al. 2013].

Two single-arm phase II trials have reported activity with panitumumab monotherapy although effect by KRAS status was not reported (Table 4) [Hecht et al. 2007; Muro et al. 2009]. The use of panitumumab following progression on cetuximab was evaluated in two small trials with conflicting results. A small number of responses were reported in one trial, with the other demonstrating no activity [Metges et al. 2010; Wadlow et al. 2012]. An EGFR mutation (S492R) arising following treatment with cetuximab and resulting in acquired resistance has been identified in vitro. This mutation was found in 2 out of 10 patients with cetuximab resistance and the 1 patient who then received panitumumab achieved a partial tumor response. This provides an explanation for a group of patients who remain sensitive to panitumumab despite progression with cetuximab [Montagut et al. 2012]. At present the use of an alternate anti-EGFR antibody following progression is not recommended.

Summary

Panitumumab significantly prolongs PFS when used as monotherapy in chemotherapy refractory mCRC. The benefit is restricted to KRAS WT patients. An overall survival benefit has not been demonstrated although it is likely the crossover in the pivotal panitumumab trial may have obscured the OS gain seen in the cetuximab trial. The severity of skin rash with panitumumab may predict likelihood of clinical benefit.

Meta-analysis

A recent meta-analysis provides further evidence for panitumumab and cetuximab in KRAS WT mCRC. Data was pooled from randomized trials where standard treatment was compared with standard treatment plus an anti-EGFR antibody [Petrelli et al. 2011]. Overall the relative risk (RR) of an objective response (OR) was 1.69 (p = 0.003), indicating patients with KRAS WT mCRC are 69% more likely to obtain a response with the addition of anti-EGFR antibody. HR for PFS and OS were 0.65 (p = 0.0006) and 0.84 (p = 0.03), respectively. When analyzed for each drug, only panitumumab resulted in a statistically improved PFS and OS compared with control. The addition of an anti-EGFR antibody in all lines of treatment led to a statistically significant improved likelihood of achieving an OR. This is particularly true when used as monotherapy where the RR of an OR is 33.84 (p = 0.0005). Significant HRs for PFS were seen in both first and subsequent lines of treatment, while OS difference was not significant in either first or subsequent lines of treatment.

In combination with bevacizumab

The large phase III Panitumumab Advanced Colorectal Cancer Evaluation (PACCE) trial investigated first-line dual VEGF and EGFR inhibition in combination with chemotherapy for mCRC [Hecht et al. 2009]. A total of 1043 patients were assigned to oxaliplatin-based (Ox-CT; n = 823) or irinotecan-based (Iri-CT; n = 230) chemotherapy cohorts by investigator choice in combination with bevacizumab (the control). Patients were then randomized to panitumumab versus no other treatment. The primary endpoint was PFS in the oxaliplatin chemotherapy cohort. Panitumumab was discontinued following a planned interim analysis demonstrating decreased PFS in the investigational arm. In an updated analysis a statistically significant inferior median PFS (the primary endpoint) was demonstrated in those treated with Ox-CT, bevacizumab and panitumumab versus Ox-CT and bevacizumab alone (10 versus 11.4 months, HR 1.27, 95% CI 1.06–1.52). There was a significantly inferior OS in the panitumumab-treated Ox-CT cohort.

Inferior outcomes with dual antibody inhibition were seen across all treatment cohorts irrespective of KRAS status. Severe (grade 3 and higher) toxicity was greater in the panitumumab arm compared with controls in both Ox-CT (90% versus 77%) and Iri-CT (90% versus 63%) cohorts. Mechanistic explanations for the inferior outcomes in this study are not known. Pharmacokinetic interactions between dual antibodies, reduced drug exposure due to higher toxicity or a pharmacodynamic interaction between dual antibodies and chemotherapy have been postulated.

The results of PACCE while unexpected and largely unexplained are consistent with other trials investigating dual EGFR/VEGF inhibition in mCRC in first line. The CAIRO2 study randomized 755 patients to capecitabine, oxaliplatin and bevacizumab with or without cetuximab [Tol et al. 2009]. In the KRAS unselected population PFS was lower in the cetuximab arm (10.7 versus 9.4 months; p = 0.01). No differences were seen in OS and RR. Toxicity with dual antibody treatment was significantly higher. Treatment with combined anti-EGRF and anti-VEGF antibodies in first-line therapy is not recommended outside of a clinical trial, noting that there may still be a role in late stage disease based on the BOND 2 data for cetuximab [Saltz et al. 2008]. The CALGB 80405 (first line) and SWOG S0600 trials (second line) are assessing combination of cetuximab and bevacizumab in KRAS WT mCRC and may provide further insight.

Summary

The PACCE trial demonstrates significant inferior outcomes with dual EGFR/VEGF inhibition and this strategy should be avoided outside a clinical trial. Further analysis of additional predictive markers may further enhance the individualization of panitumumab (and cetuximab) in the future.

Tolerability

While the rate of severe (grade 3 or higher) toxicity with panitumumab is low, mild to moderate toxicity, particularly dermatological, is common. This has significant quality of life (QOL) implications and close monitoring, early intervention and routine preventative measures are important.

Rash

The most common clinical manifestations are erythema, acneform dermatitis and pruritus, but can include skin desquamation, paronychia, skin fissures, pustular rash and skin infection. The distribution of rash typically involves the face and upper body [Lenz, 2006]. Exposure to sunlight can exacerbate skin toxicity [Amgen Inc., 2009].

In the large phase III trial of panitumumab monotherapy the rate of skin toxicity was 90% but severe in only 16% of patients [Amgen Inc., 2009; Van Cutsem et al. 2007]. Skin toxicity from trials of panitumumab in combination with chemotherapy is also high, with reported rates of any grade skin toxicity of 95% with severe skin toxicity occurring in 35–37% of patients [Douillard et al. 2010; Hecht et al. 2009; Peeters et al. 2010].

The STEPP trial, described above, examined the benefit of pre-emptive skin treatment in 98 patients receiving panitumumab in combination with irinotecan based chemotherapy [Lacouture et al. 2010]. Pre-emptive skin treatment included use of skin moisturizers, sunscreen, topical corticosteroid and oral doxycycline. Pre-emptive skin treatment was associated with a significantly reduced rate of grade 2 or higher skin toxicity compared to reactive skin treatment (29% versus 62%, OR 0.3, 95% CI 0.1–0.6).

Pre-emptive supportive measures are advised for all patients commencing anti-EGFR antibodies and includes advice on sun-protective measures and avoidance of activities and products likely to contribute to skin dryness (long, hot showers and alcohol-based perfumed products) [Fakih and Vincent, 2010]. Canadian guidelines on the management of anti-EGFR antibody induced rash represent sound recommendations [Melosky et al. 2009]. Mild (grade 1) toxicity should be managed with a topical antibiotic and corticosteroid cream, moderate toxicity with an oral tetracycline in addition to topical therapy and severe toxicity necessitates an interruption to anti-EGFR treatment.

Hypomagnesemia

Hypomagnesemia occurred in 36% of patients receiving panitumumab monotherapy, which was severe in 3% of cases [Van Cutsem et al. 2007]. High EGFR expression at the ascending loop of Henle, the site of 70% of filtered magnesium resorption, is thought to result in damage to the renal tubule from anti-EGFR antibodies [Fakih and Vincent, 2010].

Serum magnesium levels should be closely monitored during treatment with panitumumab and continue for 8 weeks following discontinuation of treatment [Amgen Inc., 2009]. Patients with mild hypomagnesemia (lower level of normal (LLN) 1.2 mg/dl) are generally asymptomatic and do not require routine replacement therapy. Options for treatment of moderate hypomagnesemia (1.2–0.9 mg/dl) include close surveillance, oral supplementation or weekly intravenous magnesium infusion [Fakih, 2008]. Severe hypomagnesemia often leads to fatigue, muscle cramps and somnolence, which can be under-recognized [Tejpar et al. 2007]. All grade 3 or higher hypomagnesemia (Mg <0.9 mg/dl) should be corrected with intravenous magnesium and may be required several times per week and necessitate treatment interruption.

Diarrhea

Diarrhea can be problematic particularly when panitumumab is used in combination with chemotherapy. In trials of panitumumab monotherapy the rate of diarrhea has been reported at 21% but as high as 69–79% in combination with chemotherapy [Berlin et al. 2007; Cohn et al. 2011; Mitchell et al. 2011; Van Cutsem et al. 2007]. The rate of severe diarrhea with panitumumab monotherapy is very low (1%). Patient education, ensuring adequate oral fluid intake and the use of loperamide can prevent more significant complications and the need for intravenous fluid replacement.

Infusion reactions

The rate infusion reaction with panitumumab monotherapy is 3% with severe infusion reactions (anaphylaxis, bronchospasm, and hypotension) occurring in 1% of patients [Amgen Inc., 2009; Keating, 2010]. The rate of infusion reaction is significantly lower than with cetuximab, consistent with the fully human nature of panitumumab.

Pulmonary fibrosis

Pulmonary fibrosis has occurred in 2 of 1467 patients treated with panitumumab monotherapy in clinical trials. While the rate of pulmonary fibrosis appears low, prescribing information contains warnings of pulmonary fibrosis and interstitial lung disease (ILD) and suggests caution in patients with pre-existing ILD [Australian Government, 2013; Amgen Inc., 2009].

Regulating affairs

Approval of panitumumab and the indications for its use have been inconsistent between world regulating agencies. Panitumumab was first approved by the US FDA in September 2006 as monotherapy in patients with mCRC with disease progression following fluoropyrimidine, irinotecan and oxaliplatin chemotherapy [Amgen Inc., 2009]. In July 2009, following identification of the predictive value of KRAS, the product labeling was amended such that treatment with panitumumab was restricted to KRAS WT patients [Trotta et al. 2011].

The European Medicines Agency’s (EMA) Committee for Medicinal Products for Human Use (CMPH) initially rejected the application for panitumumab given the small absolute difference in median PFS (5 days) when used in KRAS unselected patients. When retrospective analysis demonstrated the benefit was restricted to patients with KRAS WT tumors approval was granted in December 2007 [Amado et al. 2008; Trotta et al. 2011].

The EMA and Australian Therapeutic Goods Association (TGA) have recently extended approval of panitumumab to include treatment of patients with KRAS WT mCRC as first-line therapy in combination with FOLFOX and as second-line therapy in combination with FOLFIRI [Australian Government, 2013; European Medicines Agency, 2013]. The US FDA does not carry this indication.

Role of panitumumab in the overall treatment of mCRC

The role of panitumumab in the treatment algorithm of patients with mCRC continues to evolve. While randomized clinical trial data, as described in this paper, demonstrates the activity of panitumumab in KRAS WT mCRC many important questions remain unanswered. The optimal timing of use of anti-EGFR antibodies, choice of first-line biological agent, optimal chemotherapy backbone, further predictive biomarkers and the clinical differences between panitumumab and cetuximab are all areas which require further study.

The data presented within this review suggests efficacy for anti-EGFR antibody treatment is best established in chemorefractory KRAS WT patients, and represents the standard of care in this group. Single agent panitumumab, single-agent cetuximab and combination cetuximab and irinotecan are all supported by phase III trials demonstrating improved outcomes in this setting [Amado et al. 2008; Cunningham et al. 2004; Karapetis et al. 2008].

The use of anti-EGFR antibody treatment in combination with chemotherapy in earlier lines of treatment is a less common and more complex issue. For KRAS WT patients the optimal targeted agent in first line therapy remains undefined with the addition of either bevacizumab or an anti-EGFR antibody supported by multiple phase III trials. Several factors may lead the clinician to prefer an anti-EGFR antibody to bevacizumab. In patients with a contraindication to bevacizumab, such as refractory hypertension, proteinuria, recent surgery, major thrombosis or hemorrhage risk, the addition of anti-EGFR antibody to chemotherapy improves PFS and ORR compared with chemotherapy alone.

The use of first-line anti-EGFR inhibitor may be of particular relevance in the subset of patients with borderline resectable liver metastasis. The concept of preoperative ‘conversion therapy’ to allow borderline resectable patients to undergo curative intent resection has received increasing interest in the literature. At present the optimal conversion chemotherapy or biological agent is undefined, although given the importance of tumor response in this setting the ORR of 55% in KRAS WT patients treated with panitumumab/FOLFOX in the PRIME trial is comparable or better to any first-line phase III trial in mCRC and represents a reasonable approach in this subset of patients. Retrospective subset analysis from first line trials of bevacizumab, cetuximab and panitumumab in combination with chemotherapy have reported numerically improved attempted curative resection rates and rates of successful complete resection compared to chemotherapy alone [Douillard et al. 2010; Okines et al. 2009; Van Cutsem et al. 2011]. That said, the recently reported ‘new EPOC’ trial which reported inferior outcomes for KRAS WT ‘resectable’ patients, including those defined as borderline, who received oxaliplatin-based chemotherapy and cetuximab when compared with chemotherapy alone [Primrose et al. 2013]. As this was an abstract presentation we await the full data set and manuscript to fully understand the implications.

Head-to-head comparisons of biological agents in the first-line setting, assessing survival outcomes and rates of successful metastatic resection are needed. The German FIRE-3 trial, a phase III trial comparing first-line cetuximab with bevacizumab in combination with FOLFIRI in KRAS WT patients, was recently reported. The primary endpoint of ORR was improved in the cetuximab arm (ORR 72% versus 63%; p = 0.017) in those patients receiving at least three cycles of treatment [Heinemann et al. 2013]. While no difference was demonstrated in PFS, a significant OS benefit (median OS 28.7 versus 25.0 months, HR 0.77, 95% CI 0.62–0.96) was seen in cetuximab-treated patients. Information on subsequent anticancer therapy and differences in liver resection rates are awaited. Further insight will hopefully come from the CALGB 80405 trial also comparing first-line cetuximab with bevacizumab. As discussed earlier, the phase II PEAK trail also demonstrates a trend to improved OS with anti-EGFR antibody (panitumumab) compared with bevacizumab in the first-line setting in KRAS WT mCRC [Schwartzberg et al. 2013a].

Factors against first-line panitumumab include the potential treatment delay to obtain tumor KRAS mutation status and toxicity concerns (particularly dermatological). Access to panitumumab in the first-line setting may be problematic in some regions, with the US FDA currently restricting the use of panitumumab to the chemorefractory setting.

It is important to highlight a degree of uncertainty exists regarding the evidence for cetuximab in combination with chemotherapy. The MRC COIN and NORDIC VII trials were large first-line studies of cetuximab- and oxaliplatin-based chemotherapy which did not demonstrated a benefit of cetuximab irrespective of KRAS status [Maughan et al. 2011; Tveit et al. 2012]. The OPUS and CRYSTAL trials demonstrated a significant benefit [Bokemeyer et al. 2011; Van Cutsem et al. 2011]. Whether these conflicting results relate to an interaction between cetuximab and the chemotherapy backbone or whether improved biomarkers would allow better patient selection is unknown. The implications for panitumumab–chemotherapy combination is also unclear.

Future developments

Several upcoming clinical trials will hopefully provide answers regarding the optimal use of panitumumab for mCRC. The head-to-head phase III comparison trial, ASPECCT, of cetuximab versus panitumumab monotherapy may identify differences in efficacy or toxicity profile that have not been evident to date. The role of panitumumab in cetuximab refractory mCRC, the benefit of panitumumab in combination with irinotecan in irinotecan refractory mCRC, and tumor response with panitumumab in combination with 5-FU, folinic acid, oxaliplatin and irinotecan (FOLFOXIRI) are all being evaluated in current single-arm phase II clinical trials [Peeters et al. 2012a]. Panitumumab is also being assessed in combination with FOLFOX in the neoadjuvant setting prior to resection of high-risk primary CRC (T3/4). A phase II trial has reported the safety of this strategy and the extension phase III FOxTROT trial is currently recruiting [Foxtrot Collab-orative Group, 2012].

Predictive biomarkers, beyond KRAS, to guide patient selection for anti-EGFR antibody treatment is crucial to optimize outcomes and minimize toxicity in patients with mCRC. The role of other implicated biomarkers such as BRAF, NRAS and PIK3CA mutations and loss of PTEN expression will hopefully be clarified in coming years. Given the relatively low frequency of these mutations a carefully conducted meta-analysis of RAS mutations in panitumumab-treated patients is likely to be required. These identified markers and newly discovered ones are likely to be the focus of research over the next 5 years to further define the patient population best treated with this class of drugs.

Conclusion

The efficacy of panitumumab in KRAS WT mCRC has been demonstrated in multiple randomized phase III clinical trials [Douillard et al. 2010; Peeters et al. 2010; Seymour et al. 2013; Van Cutsem et al. 2007]. Panitumumab results in improved ORR and PFS the first- and second-line treatment settings in combination with chemotherapy and in chemorefractory patients as monotherapy. A statistically significant overall survival benefit has not been demonstrated in a phase III trial in any line of treatment. Defining optimal biomarker selection prior to panitumumab therapy is the immediate research priority, particularly given the possibility of adverse outcomes in subgroups of patients with mutations beyond KRAS, most notably NRAS [Oliner et al. 2013].