Panitumumab in combination with infusional oxaliplatin and oral capecitabine for conversion therapy in patients with colon cancer and advanced liver metastases

The MetaPan study


  • We would like to thank Radhika Srinivasan, PhD, for the extensive revision of the article.



Preoperative chemotherapy improves the outcome in patients with colorectal cancer with liver metastases. In the current study, the authors evaluated the activity of a conversion treatment with the combination of capecitabine plus oxaliplatin (XELOX) used in association with panitumumab in patients with unresectable, liver-only, metastatic colon cancer.


Chemotherapy-naive patients with unresectable liver metastases from colon cancer with no other metastatic disease sites were enrolled. All patients received upfront therapy with XELOX plus panitumumab (P-XELOX) and were reevaluated for resectability every 4 cycles. The primary endpoint was the objective response rate (ORR). Secondary endpoints were overall survival (OS), progression-free survival, the percentage of patients whose disease became radically resectable, and the safety of the P-XELOX combination.


A total of 49 patients were recruited, 35 of whom had wild-type KRAS (wtKRAS) and 14 of whom (who were enrolled before study amendment) had unknown (9 patients) or mutated (5 patients) KRAS mutational status. Forty-six patients were evaluable for response. After conversion P-XELOX therapy, the ORR in the general population was 54%, with 2 complete responses, 23 partial responses, and 14 cases of stable disease. In patients with wtKRAS, the ORR of the patients reached 65% (2 CRs and 19 PRs), which allowed 15 patients with initial unresectable liver metastasis to be reclassified as having resectable disease. Survival analysis demonstrated a median progression-free survival of 8.5 months and a median OS of 21.9 months. Patients who underwent surgery were found to have a significantly better OS when compared with those who did not undergo surgery (P < .001). Overall, toxicities were found to be predictable and manageable, with the most common being cutaneous, gastrointestinal, and neurologic toxicities.


Conversion P-XELOX therapy yields high response and resectability rates for patients with metastatic colon cancer with extensive liver involvement. Cancer 2013;119:3429–3435.. © 2013 American Cancer Society.


Colorectal cancer (CRC) is the third most common cancer worldwide, with up to 1 million new cases diagnosed each year.[1] The standard treatment for metastatic disease involves chemotherapy based on fluoropyrimidines, oxaliplatin, and irinotecan (used in combination and sequentially), and monoclonal antibodies targeting vascular endothelial growth factor and epidermal growth factor receptor (EGFR).[2-7]

In selected patients, surgery can be included in the treatment plan for metastatic CRC (mCRC ); in particular, the resection of hepatic metastases improves progression-free survival (PFS) and offers the chance for cure in approximately 10% to 20% of patients.[8, 9] In patients with extensive metastatic involvement of the liver, neoadjuvant chemotherapy can result in tumor downsizing and lead to the reclassification of a percentage of patients considered to be unresectable to resectable. However, the classification of liver metastases into clearly resectable, potentially resectable, or definitely unresectable is influenced by several variables, such as surgeon experience and innovative radiological and surgical techniques. Therefore, the criteria for defining resectability often differ in clinical studies and, as a consequence, the percentage of inoperable patients who undergo radical resection after frontline chemotherapy varies between 13% and 40%.[10-12] The optimization of the preoperative regimens is critical to the success of the curative strategy. In fact, a strong relationship between objective tumor response rate (ORR) and resection rates has been documented in patients with mCRC.[13]

A previous report has demonstrated that for patients with liver metastases who are refractory to conventional chemotherapy, combination therapy with cetuximab(Erbitux; ImClone LLC a wholly-owned subsidiary of Eli Lilly and Company, Branchburg, NJ) increases resectability rates.[14] In a phase 2 trial of cetuximab in combination with 5-fluorouracil and oxaliplatin as the first-line treatment of patients with mCRC, an ORR of 79% was reportedly obtained.[15]

Panitumumab (Vectibix; Amgen-Dompe, Milan, Italy) is a fully human antibody that binds to EGFR and prevents receptor dimerization, tyrosine autophosphorylation of EGFR, and the activation of downstream signaling molecules.[16] To our knowledge, the impact of panitumumab as a part of a conversion therapy in the metastatic setting has not been investigated to date.

We conducted this prospective, phase 2, multicenter study to investigate the activity, in terms of ORRs and conversion to resectable status, of the combination of panitumumab and capecitabine plus oxaliplatin (XELOX) chemotherapy (P-XELOX) in patients with colon cancer (CC) and synchronous unresectable liver metastases.


Eligible patients had histologically confirmed metastatic, unresectable CC with liver involvement with or without involved perivisceral lymph nodes. Only patients with synchronous metastatic disease were enrolled to minimize the impact of previous adjuvant chemotherapy and to ensure a more homogenous population. Subjects with rectal cancer diagnosed within 12 cm from the anal verge were excluded because of the different therapeutic approach that is required in these cases.

Since November 2008, after it was discovered that wild-type KRAS (wtKRAS) was required for the clinical activity of panitumumab in patients with metastatic CC (mCC),[17] only patients bearing wtKRAS on codon 12 and 13 of exon 1 were included in the current study.

Liver metastases were considered unresectable in the simultaneous presence of multiple and bilateral lesions, with < 50% of the remnant healthy liver, specifically defined as > 3 liver metastases with > 50% hepatic involvement and requiring a major hepatectomy with contralateral wedge resection. In addition, any metastasis requiring resection that did not preserve at least 2 contiguous hepatic segments or adequate vascular inflow and outflow as well as biliary drainage was considered unresectable.

Other inclusion criteria were a minimum age of 18 years and an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of ≤ 2. In addition, adequate hematologic, hepatic, and renal function were required.

Patients with extrahepatic metastasis, other than that of regional lymph nodes, were excluded.

The protocol was approved by the Institutional Review Boards at each participating institution and the study was performed in accordance with the Declaration of Helsinki. All patients provided written informed consent.

Treatment and Assessment

All patients received panitumumab administered by intravenous infusion at a dose of 9 mg/kg once every 3 weeks, in combination with oxaliplatin at a dose of 130 mg/m2 on day 1 and capecitabine at a dose of 1000 mg/m2 twice daily from days 1 to 14. Dose reductions or delays for both panitumumab and chemotherapy were provided in cases of toxicity according to the study protocol.

After 4 cycles of chemotherapy, patients were evaluated for tumor response. Those who experienced stable disease (SD) or a partial response (PR) but were still considered unresectable after the first course of therapy received a second course of treatment (4 cycles) and a surgical reevaluation. In the case of a complete response (CR), patients continued chemotherapy for 8 cycles.

Patients who were considered eligible for surgery after treatment proceeded to undergo curative metastasectomy and received 4 additional postoperative cycles of P-XELOX chemotherapy.

Tumor response was assessed using modified Response Evaluation Criteria in Solid Tumors (m-RECIST). At baseline, computed tomography scans of the chest, abdomen, and pelvis, along with appropriate imaging of all other disease sites, were evaluated.

Safety was assessed on the basis of reports of adverse events, laboratory test results, and vital sign measurements. Toxicities were graded using version 3.0 of the National Cancer Institute's Common Toxicity Criteria (NCI-CTC), with the exception of skin-related or nail-related toxicities, which were graded using NCI-CTC version 3.0 with modifications.

Statistical Analysis

This phase 2 trial was designed using a Fleming single-stage design to assess the ORR. The null hypothesis is that the ORR is ≤ 25%. The alternative hypothesis is that the ORR is ≥ 45%. Requiring a 90% power against the alternative hypothesis and using a 5% significance level, the ‘Hearn method indicates that 54 patients with wtKRAS are required.[18] Thus, patients with wtKRAS tumors who were enrolled before the approval of the amended protocol version 3 of 30 (November 2008) were considered part of the sample.

Efficacy analyses were performed according to the intent-to-treat principle.

Both PFS and overall survival (OS) were estimated using the Kaplan-Meier method.


Baseline Characteristics

From November 2007 to January 2011, 49 patients were enrolled in 6 different centers in Italy. The baseline characteristics of the patients are summarized in Table 1. Data collection was censored in August 2011.

Table 1. Patient Characteristics at Baseline
Characteristic No.%
  1. Abbreviation: ECOG PS, Eastern Cooperative Oncology Group performance status.

Age, yMedian (range)60 (37–76)
ECOG PS03469.4
Surgery of primary tumor 2857.1

Since November 2008, the protocol has been amended and wtKRAS included as a selection criterion. KRAS mutational status was performed, wherever possible, for patients who had been included in the study before the protocol amendment. In 9 cases, the KRAS mutational status was unknown. Of the remaining patients, 35 were classified as having wtKRAS mutational status and 5 patients who were enrolled before the protocol amendment were classified as having mutated KRAS. The median age of the patients was 60 years (range, 37 years-76 years). All patients had a good PS defined as ECOG 0 (34 patients) or ECOG 1 (15 patients); no patients with ECOG 2 were enrolled, although this was permitted by inclusion criteria. In 28 cases, patients had undergone colon surgery before study enrollment.

The median follow-up was 16.6 months (range, 1.1 months-38.9 months).

All patients received at least 1 cycle of P-XELOX, with a median of 6.3 cycles (range, 1 cycle-12 cycles).


The primary endpoint of the current study was ORR. All data were analyzed separately for the entire study population and for those patients who had wtKRAS disease. The results are shown in Table 2. Tumor response was assessed in 46 patients (32 with wtKRAS mutational status) as the best response achieved during the treatment plan, excluding responses achieved with surgery. An ORR was recorded in 54% of patients with 2 CRs and 23 PRs. Fourteen patients (30.4%) achieved SD and 7 (15.2%) developed progressive disease.

Table 2. Response Rates
Response RatesTotal Population (n = 49)wtKRAS (n = 35)
EvaluableNo.%95% CINo.%95% CI
  1. Abbreviations: 95% CI, 95% confidence interval; wt, wild-type.

Complete response24.30–10.326.32.6–15.1
Partial response235035–651959.441.4–77.4
Stable disease1430.416.6–44.2721.96.7–37
Progressive disease715.24.2–26412.50–24.6
Total4693.8 3291.4 

As expected, the results from patients with wtKRAS demonstrated a better outcome. The overall disease control rate for patients with wtKRAS mutational status was 87.5%, with 2 CRs, 19 PRs, and 7 cases of SD.

Fifteen patients, all of whom had wtKRAS tumors, underwent surgical resection of their liver metastasis: 10 patients after the first radiological assessment and 5 patients after the second. In 11 cases, a major hepatectomy, defined as a resection of ≥ 3 Couinaud segments, was performed. Ten patients received a median of 3.5 postoperative treatment cycles (range, 1 cycle-4 cycles). Among the patients who underwent surgery, R0 resection, defined as radical resection with histologically negative margins, was achieved in 10 of 15 patients. Colon surgery was performed at the same time as liver resection, if it had not been performed previously. No surgery-related deaths were recorded, nor were any second surgeries required due to complications. Details regarding the type of resection, number of perioperative chemotherapy cycles, and outcomes are shown in Table 3.

Table 3. Approach Used and Outcome of Patients Treated With Liver Surgery
Patient No.Previous Colon SurgeryType of Liver SurgeryMarginsNo. of Preoperative CyclesNo. of Postoperative CyclesSite of Disease RecurrencePFS, Months
  1. Abbreviations: PFS, progression-free survival; UNK, unknown.

1YesMinorR024Lymph nodes20.9
3NoMinorR044Lung and liver11.3
6NoMajorUNK70Lymph nodes8.3
13YesMajorR084Lung and liver14.7

In the absence of radiologically visible disease, 2 patients with a radiological CR to chemotherapy did not proceed to liver resection.

Patients were evaluated for survival analysis; the median PFS in the entire population was 8.4 months (95% confidence interval, 6.8 months-9.9 months) and the median OS was 21.9 months (95% confidence interval, 13.3 months-30.6 months) (Fig. 1 Top and Bottom).

Figure 1.

(Top) Overall survival and (Bottom) progression-free survival in the entire study population are shown.

Patients who had undergone surgery for liver metastasis had a better survival outcome when compared with those who were treated only with palliative chemotherapy; the median PFS was 14.7 months for patients undergoing surgery and 7.3 months for unresected patients (P = .079). The median OS was not reached for resected patients and was 17.1 months for unresected patients (P < .001) (Fig. 2 Top and Bottom).

Figure 2.

(Top) Overall survival and (Bottom) progression-free survival are shown in resected versus unresected patients.

As expected, a survival benefit was noted for the wtKRAS subgroup when compared with those whose KRAS status was unknown or mutated (P < .001) (Fig. 3).

Figure 3.

Overall survival in patients with wild-type KRAS versus those with unknown or mutated KRAS mutational status is shown.


Adverse events that were noted are listed in Table 4. The most common toxicities reported were gastrointestinal, constitutional, cutaneous, and neurological. In particular, 25 patients experienced diarrhea, which reached grade 3 or higher in 10 cases. Nausea, vomiting, mucositis, and anorexia were reported in 23 patients, 7 patients, 8 patients, and 6 patients, respectively, but were generally of grade 1 to 2. Skin toxicity, mostly related to panitumumab, was commonly reported, occurring in up to 90% of patients; other forms of cutaneous toxicity that are typical of anti-EGFR agents were observed, in particular folliculitis (11 patients), pruritus (3 patients), conjunctivitis (5 patients), and ungual toxicity (3 patients).

Table 4. Adverse Eventsa
Adverse EventAny GradeGrade ≥3
  1. a

    Toxicities were graded using version 3 of the National Cancer Institute's Common Toxicity Criteria (NCI-CTC), with the exception of skin-related or nail-related toxicities, which were graded using NCI-CTC version 3.0 with modifications.

Anorexia6 (12.2 %)2 (4.1 %)
Mucositis8 (16.3 %)0
Nausea23 (46.9%)3 (6.1 %)
Vomiting7 (14.3 %)1 (2 %)
Abdominal pain5 (10.2 %)0
Constipation4 (8.2 %)0
Diarrhea25 (51 %)10 (20.4 %)
Asthenia29 (59.2 %)6 (12.2 %)
Neutropenia6 (12.2 %)0
Thrombocytopenia6 (12.2 %)1 (2 %)
Ungual toxicity3 (6.1 %)0
Cutaneous toxicity44 (89.8 %)11 (22.4 %)
Folliculitis10 (20.4 %)1 (2 %)
Itch3 (6.1 %)0
Hepatic toxicity5 (10.2 %)0
Neurological toxicity30 (61.2 %)3 (6.1 %)
Conjunctivitis5 (10.2 %)0
Allergic reaction7 (14.3 %)3 (6.1 %)

The overall incidence of oxaliplatin-specific neuropathy was relevant (present in 30 patients), but was only reported as being serious (grade 3 or 4) in 3 patients. It is interesting to note that 7 patients had allergic reactions to oxaliplatin that required treatment interruption in 4 cases.

Overall, hematological toxicity was a minor event, with 6 cases of mild (grade 1 or 2) neutropenia and thrombocytopenia noted. No febrile neutropenia was reported.

All toxicities were managed with dose reduction or treatment interruption according to the study protocol. Twenty-one patients required dose reductions for panitumumab, and 22 patients for capecitabine and oxaliplatin.

Nearly 50% of patients received < 8 cycles of chemotherapy; 8 patients developed progressive disease before the second radiological reevaluation. Treatment interruption was due to adverse events in 7 patients or related to surgical procedures in 6 patients (inadequate recovery, timing of surgery, and investigator choice).

Other causes of treatment discontinuation were consent withdrawal, medical decision, and death unrelated to disease progression or treatment.


To the best of our knowledge, the MetaPan study is, the first study of frontline treatment with the combination of a fluoropyrimidine, oxaliplatin, and panitumumab as neoadjuvant treatment in patients with liver-only, unresectable, synchronous mCC.

The role of anti-EGFR monoclonal antibodies in the induction of tumor shrinkage has become more defined, and currently this frontline treatment should preferably be performed in patients who might have a chance for resection. In patients with wtKRAS tumors, a high ORR is obtained using a treatment combining cetuximab with chemotherapy.[19-21] In particular, in the phase 2 randomized study of cetuximab plus FOLFOX or cetuximab plus FOLFIRI in the neoadjuvant treatment of nonresectable colorectal liver metastases (CELIM study), the ORRs were 68% and 57%, respectively, leading to a rate of R0 resection of 38%, and 30%, respectively, in patients with unresectable liver metastases.[19]

A phase 3 randomized study demonstrated that panitumumab, in combination with FOLFOX4, improves the median PFS when compared with treatment with FOLFOX4 alone (9.6 months vs 8.0 months; P = .02) in previously untreated patients with mCRC.[22] In addition, a survival advantage was observed in association with FOLFIRI as a second-line treatment or used alone in patients with chemotherapy-refractory mCRC.[23, 24]

It is difficult to compare our clinical results with those of other studies, mainly because of the different definitions of resectability. In fact, other studies[12, 20, 25] have considered the cutoff point of 5 liver metastases as the limit for defining resectability. However, the presence of large, multiple, or bilateral metastases is no longer considered an absolute contraindication to surgery, provided that radical resection is deemed possible. In the current study, a very strict definition of unresectable liver metastases was adopted to consider all enrolled patients as unequivocally excluded from upfront surgery. No upper limits to metastatic involvement were defined. Thus the population of patients with mCC enrolled in the current study included both patients potentially amenable to surgery once downsizing was obtained and those patients with unresectable disease independent of the degree of tumor shrinkage.

Consequently, it is not surprising that, unlike other studies of preoperative therapy, the rate of disease progression in the current study was 15%, and that in some cases a clinical progression anticipated the time of the first radiological evaluation. However, the P-XELOX combination resulted in a high ORR (54.3%) and an impressive rate of conversion from unresectable to resectable disease (30.6%). With regard to survival analysis, the results of 21.9 months for OS and 8.4 months for PFS are satisfactory if we consider the baseline characteristics of these patients. All patients had synchronous CC with major liver involvement, and were usually characterized as having a very poor prognosis.[8]

In patients who underwent surgery, the comparison between OS and PFS allowed us to support the use of surgery whenever feasible. Disease recurrence occurs after the first surgery in the majority of patients; however, these patients will most likely benefit from additional surgery or second-line chemotherapy.[26] This may explain why there was a statistically significant difference in the current study with regard to OS but not PFS between patients who did and those who did not undergo surgery.

Results according to KRAS mutational status appear to be clinically relevant. In the wtKRAS population, the ORR reached 65%, with a disease control rate of 87.5% and a conversion rate of 42.8%. However, given that this type of study did not include a comparison group, this observation suffers from selection biases, as do other noncomparative studies.

The small size of the population is a limiting factor of the current study. When it was designed, the role of KRAS mutational status as a predictive factor of response to anti-EGFR agents had not been clearly defined, and therefore was not required for enrollment. After the discovery that wtKRAS status is required for the determination of panitumumab activity in patients with mCRC,[17] an amendment to the study protocol was made, adding wtKRAS status as an inclusion criterion. This had significant consequences on patient accrual: the time needed to determine the KRAS mutational status occasionally conflicted with the urgent need for treatment among some patients. For this reason, the accrual proceeded slowly and the preplanned number of subjects was not reached.

With regard to safety, no unexpected toxicities were recorded and the treatment was globally manageable with dose reductions. Although toxicity can never be underestimated, in patients with exclusive metastatic liver involvement in whom surgery may play a critical role, the results in terms of ORR justify the use of an aggressive upfront treatment.

The P-XELOX combination has proven to be a promising frontline treatment in patients with extensive liver involvement from CC and deserves further investigation.


All investigators acknowledge Amgen for funding support. This work was supported by grants from “Progetti di Ricerca Rete Oncologia Piemonte-Valle d'Aosta” and “Associazione Italiana Ricerca sul Cancro-AIRC 5X1000” and “Progetto Terapia Molecolare dei Tumori della Oncologia Ca' Granda Onlus (OCGO) Fondazione.


Dr. Sartore-Bianchi is a member of the board of Aventis and has acted as a member of the Speakers' Bureau for Bayer, Roche, and Amgen. Dr. Siena is a member of the boards of Amgen, Sanofi-Aventis, Roche, AstraZeneca, Merck, and Bayer, and has acted as a member of the Speakers' Bureau for Bayer and Amgen.