Delphi consensus for the third-line treatment of metastatic colorectal cancer

Colorectal cancer (CRC) is a public health problem. It is the third most common cancer in men and the second in women and represents the second leading cause of death by cancer worldwide [1]. In Spain, CRC represents the second most common tumor in both genders [2]. It is believed that in the next 30 years, the mortality rate for colon cancer will be reduced, but there will be an increase for rectal cancer [3].

Standard first- and second-line treatments are based on combinations of fluoropyrimidine plus oxaliplatin or irinotecan, associated with an anti-angiogenic monoclonal antibody or anti-epidermal growth factor receptor (EGFR), which is chosen based on the RAS mutational status, although the optimal sequence is still unknown [4, 5]. Although the prognosis for metastatic CRC (mCRC) is poor, these treatments have led to an increase in the survival of patients, reaching a median of 25 to 30 months [4, 5]. Nevertheless, there is a growing number of patients with mCRC being candidates for a third line (3L) or beyond, although the optimal drug regimen and sequence are still unknown in this setting.

Only the presence of RAS-activating mutations (KRAS/NRAS), present in 30% to 45% of mCRC, has proven to be a negative predictive biomarker of response to anti-EGFR [6]. Although no other validated predictive biomarkers have been described, there are other biomarkers of special interest, such as BRAF mutation (present in 8% to 10% of mCRC), human epidermal growth factor receptor 2 (HER2) amplification, high microsatellite instability (MSI-H), and ALK/ROS1/NTRK fusion/rearrangements (occur in 0.2–2.4% of mCRC) [7‐11].

Taking in mind these biomarkers, several alternatives have been proposed in the 3L for patients who have not responded or are refractory to the previous lines: (1) sequential administration of the two oral drugs approved in this indication, trifluridine/tipiracil and regorafenib, which have shown a statistically significant benefit in progression-free survival (PFS) and overall survival (OS) with a different toxicity profile both in clinical trials and in real-life studies [12, 13]; (2) administration of an anti-EGFR, such as cetuximab or panitumumab, in treatment-naive patients with RAS wild type, which is increasingly rare because these drugs are usually indicated in first or second line [14]; (3) reuse drugs already administered that were discontinued owing to toxicity or progression (oxaliplatin, irinotecan, fluoropyrimidine, anti-angiogenics, and anti-EGFR [if RAS wild type]). In this case, high-quality evidence is limited, but this strategy is often used in routine clinical practice in the absence of alternative therapies, especially in patients with good performance status [15, 16]; (4) specific treatments for selected populations, such as dual inhibition of HER2 in HER2-positive CRC (i.e., tucatinib plus trastuzumab), immunotherapy in MSI-H, and intrahepatic therapies in limited disease or primarily located in the liver, although the main recommendation is to include patients in clinical trials [17].

Although there are different therapeutic options for the 3L treatment of mCRC, the optimal sequence of these therapies and the management of these patients in 3L are unknown. For this reason, the aim of this work is to develop a consensus to know the opinion of a panel of Spanish experts on the management of patients with mCRC in the 3L and to clarify certain points regarding clinical decisions in this setting.

The 67 experts consulted completed the 2 rounds of the Delphi consensus. Of the 21 items proposed in the first round, consensus was reached on 16 (76%), all of them in agreement. The five non-consensus items were sent to the panelists to be assessed in a second round. Of these, consensus was reached only in one item (in agreement) and four did not reach consensus. After the 2 rounds, with a total of 21 items, 17 reached consensus in agreement (81%) and 4 did not reach consensus (19%) (Fig. 1). Table 1 shows in detail the degree of agreement reached with each item after the two rounds.

Although the optimal pharmacologic regimen and sequence for patients with mCRC who are candidates for 3L or subsequent treatment are still unknown, the experts consulted showed a high degree of agreement with the proposed items (81%). However, there was no clearly defined position on 19% of the items, indicating there was no consensus. This shows that in certain specific cases of mCRC management in the 3L, the opinions and criteria of the experts on the same issues differ considerably.

Most of the panelists considered that the main objective of third-line is to equally increase survival and improve patients’ quality of life, but preferably the quality of life. Currently, all lines of treatment are required to increase overall survival, but prioritizing patient’s quality of life. Panelists argued that it would be equally valid to increase survival even at the cost of an “acceptable” deterioration in quality of life.

Regarding treatment options in the 3L, the option most agreed by the panelists was that patients preferred active treatment over symptomatic treatment. Although they considered trifluridine/tipiracil to be the best oral treatment available to patients with mCRC in 3L, there was no consensus on regorafenib. The panelists argued that although there is a group of patients who could benefit from regorafenib, this drug has adverse events that often require dose adjustments [21]. In relation to the maintenance of anti-angiogenics in 3L to improve chemotherapy, despite being feasible due to their mechanism of action, the panelists considered that more clinical evidence is still needed [15]. Even some commented that it is not necessary to continue using the anti-angiogenic drug when there are alternatives such as regorafenib. At the time of this Delphi consensus, the SUNLIGHT study had not been published and was, therefore, not part of the questionnaire. The results of this study support that trifluridine/tipiracil plus bevacizumab may represent a new standard of care for the treatment of patients with refractory mCRC who have progressed after two lines of therapy [22]. Even more, the ESMO living guidelines for mCRC include this therapeutic combination for the 3L and beyond [23]. However, even if the study and ESMO recommendations had been published prior to our Delphi consensus, this combination still does not have a reimbursement price in Spain and cannot be used, so health-care professionals would continue to administer the 3L with the available options, options that would remain valid for those who are not candidates for trifluridine/tipiracil plus bevacizumab combination therapy.

In a specific group of patients, panelists mostly prefer targeted treatments, such as cetuximab together with encorafenib in patients with BRAF V600E, and immunotherapy in patients with MSI-H or dMMR. Clinical evidence with these therapeutic strategies seems to be valid to the experts [14, 24‐27]. Even so, they also consider that trifluridine/tipiracil offers better outcomes in patients with ECOG PS 0–1, one or two metastatic sites, time from first metastasis  ≥ 18 months, and those without liver metastases. In addition, they considered the option of rechallenge with anti-EGFR in patients with RAS/BRAF wild-type mCRC, PFS ≥ 4–6 months to this therapy in first line, with an anti-EGFR-free interval of at least 4 months, and remain wild type after liquid biopsy [28].

Panelists agreed the use of a therapeutic sequence that not only increases outcomes but also allows patients to be treated later. However, it is interesting to note one of the most widely agreed items, the use of as many lines as possible to improve patient survival [29].

It has been described that the mechanism of action of trifluridine/tipiracil is different from that of the other fluoropyrimidine. This allows this drug to be used in those patients refractory to previous treatment with 5-fluorouracil or in whom this treatment cannot be administered [30, 31]. Panelists agreed with this information. Despite the scarce evidence (LONGBOARD trial, Clinicaltrials.gov ID: NCT04166604), opinions on the use of G-CSF after trifluridine/tipiracil administration are noteworthy. Some panelists commented that although there is no study to support it, the use of G-CSF is common due to the experience acquired with other chemotherapies, to avoid delays and to obtain the maximum benefit from the molecule. Other panelists explained that neutropenia caused by trifluridine/tipiracil is not severe and can be managed with delay in treatment by maintaining the dose without the need to use G-CSF.

Despite the benefits of this methodology, this work has some limitations. The selection of panelists was neither systematic nor randomized; recruitment was based on their clinical expertise in mCRC management. Panelists were recruited from different Spanish regions, although not all regions were represented. Finally, the results may have been influenced by ambiguity in the phrasing of some of the statements. In addition, panelists received honoraria from the sponsor. The fact that the survey was online had the advantage of anonymity, but it could lead to erroneous interpretations of the statements by the panelists, which could influence the result.

This Delphi consensus is of particular relevance and highlights expert opinions on the treatment of patients with mCRC in 3L. The experts agreed with most of the proposed items on 3L treatment of mCRC, prioritizing therapeutic options that increase survival and preserve quality of life, while facilitating the possibility that patients can continue to be treated later. The lack of consensus observed in some items suggests the need to improve knowledge about them.