A systematic review of treatment guidelines for metastatic colorectal cancer

Aim A systematic review of treatment guidelines for metastatic colorectal cancer (mCRC) was performed to assess recommendations for monoclonal antibody therapy in these guidelines. Method Relevant papers were identified through electronic searches of MEDLINE, MEDLINE In Process, EMBASE and the Cochrane Library; through manual searches of reference lists; and by searching the Internet. Results A total of 57 relevant guidelines were identified, 32 through electronic database searches and 25 through the website searches. The majority of guidelines were published between 2004 and 2010. The country publishing the most guidelines was the USA (12), followed by the UK (10), Canada (eight), France (eight), Germany (three), Australia (two), Spain (two) and Italy (one). In addition, eight European and three international guidelines were identified. As monoclonal antibody therapy for mCRC was not introduced until 2004, no firm recommendations for monoclonal antibody therapy were made in guidelines published between 2004 and 2006. Recommendations for monoclonal antibody therapy first appeared in 2007 and evolved as more data became available. The most recent international, European and US guidelines recommend combination chemotherapy with the addition of a monoclonal antibody for the first-line treatment of mCRC. Second-line treatment depends on the first-line regimen used. For chemoresistant mCRC, cetuximab or panitumumab are recommended as monotherapy in patients with wild-type KRAS tumours. Conclusion The study indicates that recent treatment guidelines have recognized the role of monoclonal antibodies in the management of mCRC, and that treatment guidelines should be updated in a timely manner to reflect the most recently available data.


Introduction Background
Colorectal cancer (CRC) is currently the third most common cancer worldwide [1]. Approximately 20-25% of patients with the disease already have metastases at the time of diagnosis and 50-60% of the remainder will develop metastases [2,3]. For most patients with metastatic CRC (mCRC), treatment is palliative rather than curative. The goals of systemic treatment in these patients are to prolong survival and to maintain quality of life for as long as possible. However, a small proportion of patients with mCRC (e.g. those whose metastases are confined to the liver) can be converted to a potentially curable state through surgical resection of the metastases after systemic therapy. For these patients, the goal of systemic treatment is to shrink the metastases [4].
The monoclonal antibodies bevacizumab, cetuximab and panitumumab are more recent additions to the list of systemic drugs available for the treatment of mCRC. Bevacizumab, an antibody against vascular endothelial growth factor (VEGF), was first approved as a treatment for mCRC in 2004, followed by cetuximab (also in 2004) and panitumumab (2006). Cetuximab and panitumumab both target the epidermal growth factor receptor (EGFR) and are effective only in patients with wild-type KRAS mCRC [8,9,12]. Panitumumab is the only approved fully human anti-EGFR monoclonal antibody, while cetuximab is a chimeric antibody and bevacizumab is a humanized monoclonal antibody.
The current indications for monoclonal antibody therapy in mCRC differ in Europe and the USA and between the three monoclonal antibodies. Bevacizumab is indicated for the first-and second-line treatment of mCRC in combination with fluoropyrimidine-based chemotherapy. Cetuximab and panitumumab are indicated for wild-type KRAS mCRC as monotherapy, and cetuximab is also indicated in combination with chemotherapy in Europe and in combination with irinotecan in irinotecan-refractory wild-type KRAS mCRC in the USA (Table 1). However, the optimal use of these agents in the treatment of mCRC is still evolving as new data become available [10,11,13].

Objectives
Many guidelines for the treatment of mCRC have been published. As new treatments for the disease become available, the complexity of treatment increases and it is therefore important that these guidelines provide appropriate guidance to clinicians for the treatment of mCRC. A systematic review was performed to identify treatment guidelines for mCRC and to assess the recommendations for monoclonal antibody therapy in these guidelines.

Method Search strategy
The review question was to describe treatment guidelines for mCRC. The PICOS elements were as follows: participants, patients with mCRC; interventions, the search was divorced from interventions; comparisons, the search was divorced from comparisons; outcomes, treatment guidelines for mCRC; and study design, the search was divorced from study design and instead was based on disease state (mCRC) and treatment guidelines.
Relevant papers were identified through electronic searches of MEDLINE, MEDLINE In Process, the Excerpta Medica Database (EMBASE) and the Cochrane

Selection criteria
Citations ⁄ abstracts of identified studies were reviewed and assessed for relevance by two independent researchers. Full paper copies of studies considered to be relevant were then reassessed for inclusion against the criteria below. Disagreements between the two researchers, which were rare, were resolved by discussion until a consensus was reached. Inclusion criteria for papers were global, national or regional treatment guidelines for mCRC from Australia, Canada, France, Germany, Italy, Spain, the UK or the

Data collection and analysis
Data from relevant publications were extracted into a data extraction table by three researchers, with the principal researcher overseeing all extraction. The following data were extracted: author, publication year, title, organizational body, country ⁄ region, publication type, language, target population, treatments included in guideline, and monoclonal antibody therapy guidance. Findings from the guidelines were summarized in tabular format. No statistical analyses were performed.

Results
A total of 1633 citations ⁄ abstracts were identified in the initial searches. Of these, 1542 were excluded (reasons for exclusion are summarized in Fig. 1), leaving 91 for full-paper review. After analysis of the 91 full papers, 59 were excluded (reasons for exclusion are summarized in Fig. 1), leaving 32 relevant papers. In addition, 25 papers were identified from manual and website searches, giving a total of 57 relevant guidelines (see the Appendix).

Timeframes
Of the 57 guidelines, only 11 were published prior to 2004 (the year in which monoclonal antibodies were introduced). The main treatment options discussed in these publications were surgery, chemotherapy and radiotherapy. As expected, none of the publications mentioned targeted therapy. Twenty-four guidelines were published between 2004 and 2006. Over this period, the main treatment options discussed were again surgery, chemotherapy and radiotherapy. However, targeted therapy started to receive a mention, as did regional treatments, such as hepatic arterial infusion. No firm recommendations for targeted therapy were made apart from those in a consensus statement arising from a consensus conference sponsored by the American Hepato-Pancreato-Biliary Association (Table 3 Table 3 (country specific guidelines other than the US are not presented here).
When the recommendations for targeted therapy from international, European and US guidelines were grouped according to line of therapy and date of publication, a clear pattern was seen (Table 4). For first-line therapy, recommendations started in 2006 for combination chemotherapy plus bevacizumab. Cetuximab did not appear in recommendations until 2009, when it was recommended for use in patients with wild-type KRAS tumours.
The most recent European guidelines from the European Society of Medical Oncology (ESMO) recognized the roles of bevacizumab and cetuximab (wild-type KRAS mCRC only) in the first-line treatment of mCRC, as well as cetuximab and panitumumab monotherapy in patients with chemo-refractory and wild-type KRAS mCRC (Table 4; Glimelius and Oliveira 2009 [19], Van Cutsem and Oliveira 2009 [2]). The most recent US guidelines from the National Comprehensive Cancer Network [33,34] recognized the role of all three monoclonal antibodies in the early-line treatment of mCRC. These guidelines recommended that initial therapy for mCRC should consist of FOLFOX, CapeOX, FOLFIRI, 5-FU ⁄ leucovorin, FOLFOXIRI or capecitabine, with bevacizumab, cetuximab or panitumumab being added to FOLFIRI or FOLFOX (wild-type KRAS mCRC only for cetuximab and panitumumab), and bevacizumab being added to CapeOX, 5-FU ⁄ leucovorin or capecitabine.
For second-line therapy, the guidelines from the US National Comprehensive Cancer Network [33,34] stated that possible treatment options following first progression in patients with wild-type KRAS mCRC were FOLFIRI plus cetuximab or panitumumab; cetuximab plus irinotecan; or single-agent cetuximab or panitumumab. Treatment options after second progression in patients with wild-type KRAS mCRC included cetuximab plus irinotecan, or single-agent cetuximab or panitumumab. If bevacizumab was used in a first-line regimen, it was not recommended for second or subsequent lines of therapy. Similarly, if cetuximab or panitumumab were used as part of the initial treatment regimen, neither agent was recommended in second or subsequent lines of therapy.
For chemoresistant mCRC, European guidelines also recognized the effectiveness of cetuximab or panitumumab as single agents, or irinotecan plus cetuximab, in patients with wild-type KRAS tumours [2].

Specific patient populations
The two most recent international treatment guidelines addressed specific patient populations. The guidelines from an international panel of 21 colorectal oncology experts [22] focused on patients with colorectal liver metastases, and stated that FOLFIRI, FOLFOX or

US National Comprehensive Cancer Network
Following first progression in wild-type KRAS mCRC: FOLFIRI + cetuximab or panitumumab; cetuximab plus irinotecan; or single-agent cetuximab or panitumumab After second progression in wild-type KRAS mCRC: cetuximab plus irinotecan, or single-agent cetuximab or panitumumab If bevacizumab, cetuximab or panitumumab are used for first-line therapy, they are not recommended for second or subsequent lines of therapy If bevacizumab is not used for initial therapy, it can be used following disease progression Chemoresistant mCRC

Van Cutsem and
Oliveira 2008 [20] ESMO Guidelines Working Group Cetuximab or panitumumab as single agents Irinotecan + cetuximab XELOX with the addition of cetuximab was a valuable first-line therapy in patients with wild-type KRAS mCRC, while FOLFIRI, FOLFOX or XELOX with bevacizumab was valuable in unselected patients. The task force of the International Society of Geriatric Oncology [23] developed guidelines for the treatment of elderly colorectal cancer patients. It recommended that combination chemotherapy with or without bevacizumab should be the treatment of choice; cetuximab and panitumumab should be used within the context of their licensed indications in patients with wild-type KRAS mCRC. These guidelines also considered safety in the elderly patient population, stating that monoclonal antibodies are generally safe; however, bevacizumab, in particular, has a side-effect profile that includes hypertension (the most frequent side-effect), proteinuria, thromboembolic events, bleeding, wound healing complications and bowel perforation. These possible sideeffects require careful consideration when treating elderly CRC patients. In particular, arterial thromboembolic events following bevacizumab were more likely to occur in patients over 65 years of age or in those who had a previous history (> 18%) of such events [23].

Other guidelines
National Institute for Health and Clinical Excellence (NICE) in the UK published guidance on cetuximab for the first-line treatment of mCRC [39]. This singletechnology appraisal recognized the role of cetuximab in the treatment of mCRC for a confined sub-patient population. The guidelines recommended cetuximab in combination with FOLFOX (or FOLFIRI in patients who were unable to tolerate or had contraindications to oxaliplatin), within its licensed indication, only when all of the following criteria were met: the primary colorectal tumour had been resected or was potentially operable; the metastatic disease was confined to the liver and was unresectable; and the patient was fit enough to undergo surgery to resect the primary colorectal tumour and to undergo liver surgery if the metastases became resectable after treatment with cetuximab.
The most recent Canadian guidelines [35] recommended that patients with mCRC and an Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1 or 2 should be offered palliative chemotherapy. Whether treatment was with combination chemotherapy or sequential monotherapy (with or without bevacizumab) depended upon the patient's treatment goals, their physical status and other life circumstances, as assessed by the treating oncologist. The guidelines recognized that cetuximab and panitumumab delay disease progression compared with best supportive care in patients with KRAS wild-type mCRC who are refractory or intolerant to fluoropyrimidine, irinotecan and oxaliplatin. However, no firm recommendations for their use were made.
According to recent international, European and US treatment guidelines, the generally agreed recommendations for chemotherapy regimens in the early-line treatment of mCRC are for 5-FU ⁄ leucovorin in various combinations and schedules with oxaliplatin or irinotecan, with all guidelines recommending FOLFOX and FOLF-IRI and a few also including XELOX [2,22,23,33,34,39].

Discussion
The results of this systematic review show that recent international, European and US treatment guidelines for mCRC have recognized the role of and recommended use of monoclonal antibodies in the management of this disease. It is also noted that treatment guidelines, particularly those specific to individual countries, are outdated, as they do not reflect the most recently available data.
As expected, no firm recommendations for monoclonal antibody therapies were made over the period 2004-2006, a time when bevacizumab, cetuximab and panitumumab had just been introduced for the treatment of mCRC. Recommendations for the use of monoclonal antibody therapy did not appear in guidelines until around 2007. The guidelines started with recommendations for the addition of bevacizumab to combination chemotherapy for the first-line treatment of mCRC, and evolved into recommendations for the use of all three monoclonal antibodies as new data became available. The most recent ESMO treatment guidelines published in 2009 [2] more closely reflect the European Medicines Agency approved indications for the three monoclonal antibodies in mCRC, as their recommendations include bevacizumab in combination with chemotherapy in first-and second-line therapy; cetuximab combined with chemotherapy as early-line treatment for patients with wild-type KRAS mCRC; and both cetuximab and panitumumab as monotherapy in chemorefractory wild-type KRAS mCRC.
By contrast, the US National Comprehensive Cancer Network treatment guidelines [33,34] recognize the most recently available clinical data for cetuximab and panitumumab in combination with standard chemotherapy in the first-line and second-line treatment of wildtype KRAS mCRC, even though in the USA cetuximab is only indicated either in combination with irinotecan for irinotecan-refractory patients or as monotherapy for chemo-refractory patients, and panitumumab is only approved as a single agent in chemo-refractory mCRC.
For example, in the phase III PRIME trial, panitumumab in combination with FOLFOX4 significantly improved progression-free survival compared with FOL-FOX4 alone in the first-line treatment of KRAS wild-type mCRC [10]. Another phase III trial demonstrated that panitumumab in combination with FOLFIRI significantly improved progression-free survival compared with FOLF-IRI alone in the second-line treatment of wild-type KRAS mCRC [11]. Furthermore, panitumumab in combination with chemotherapy has also consistently demonstrated a trend in overall survival improvement, although not statistically significant, in both the first-line and secondline treatment of patients with wild-type KRAS mCRC.
Similarly, a recent trial of first-line treatment with cetuximab in combination with FOLFIRI showed that such treatment reduced the risk of disease progression compared with FOLFIRI alone in patients with KRAS wild-type tumours [8]. Another trial of first-line treatment in patients with wild-type KRAS mCRC showed that a combination of cetuximab and FOLFOX4 increased the likelihood of a response and was associated with a lower risk of disease progression than treatment with FOLFOX4 alone [40].
It is also worth noting that the recent international, European and US treatment guidelines [2,22,23,33,34] recommend FOLFOX and FOLFIRI, with some also including XELOX, as the main chemotherapy choice or as the chemotherapy backbones for combining with monoclonal antibodies in the early-line treatment of mCRC. Finally, these guidelines discourage using the same monoclonal antibody in subsequent lines of therapy. Specifically, if bevacizumab is used in a first-line regimen, it is not recommended for second or subsequent lines of therapy; similarly, if cetuximab or panitumumab is used as part of the initial treatment regimen, neither agent is recommended for second or subsequent lines of therapy.
A large proportion of the guidelines identified in this review were retrieved by manual searching, including a search of grey literature on the Internet (25 ⁄ 57; 44%). Possible reasons for the manual identification of such a large proportion of relevant papers include the fact that the searching of electronic databases relies on the correct indexing of papers and the use of appropriate key words in the titles ⁄ abstracts of papers. Since guidelines, expert consensus statements and recommendations do not always fulfil these criteria, they can prove difficult to retrieve from an electronic search. In addition, organizations often publish guidelines on the Internet rather than in journals, so the guidelines can only be retrieved through grey literature searches.
A limitation of this review is the fact that the content of non-English language guidelines was not assessed, apart from those identified in the Cochrane library. Thus, it was not possible to assess whether or not such guidelines reflect the most recent data. However, it is likely that non-English language treatment guidelines were specific to individual countries. The review was restricted to guidelines from eight countries only, another limitation. In future the review can expand to include Asia and other European countries. Finally the scientific validity and methodological quality of the included guidelines was not assessed. This was beyond the scope defined for this review, particularly as the aim was identification of the relevant guidelines rather than quality assessment of the guidelines; therefore the opportunity remains for this type of research in the future. There are various tools available for the quality assessment of guidelines such as the AGREE instrument [41] and the ADAPTE evaluation [42].
In conclusion, the findings from this study indicate that recent international, European and US treatment guidelines have recognized the role of monoclonal antibody agents in the management of mCRC. Since clinical data with monoclonal antibodies and other therapies in the treatment of mCRC are continually evolving, it is important that treatment guidelines are updated in a timely manner to reflect the most recently available clinical data.
funding from Amgen Inc. Zhongyun Zhao and Beth Barber are employees and stockholders of Amgen Inc.