9. Management of Patients with Metastases and in the Palliative Situation – Colorectal Cancer

The first decision level focuses on the patient. Presuming the patient is motivated to undergo treatment, this level looks at the question of therapeutic tolerability. Here, the patient’s overall health and suitability for effective (intensive) therapy is at the foreground. This assessment is to be understood against the backdrop of the populations enrolled in the studies with a median age of 60-65 years and a predominantly good performance status (ECOG 0-1). The choice of the optimal treatment for the individual patient is thus made on the basis of the following categorisations:

1. Not suitable for intensive therapy

2. Suitable for intensive therapy

Additional details are provided in sections 9.6 and 9.7.

The second level of decision making relates to the spread of the metastatic disease. It focuses on local, locoregional or primarily systemic treatment approaches. The following categorisations are therefore helpful when choosing the optimal treatment strategy:

1. Availability of a curative treatment option

a. Resectable disease

b. Potentially resectable disease

2. Mostly palliative therapy concept

a. Oligometastases

b. Presumably not/never resectable, disseminated disease

The criteria for categorisation are listed in 9.7.2.

Sufficient knowledge of the molecular biology of the tumour provides the basis for an adequate therapy decision regarding the optimal targeted therapy.

Additional details are provided in section 9.8.

Background

Determining the status of RAS and BRAF mutations prior to initiating first-line therapy ensures that the patients receive the most effective therapy. The molecular pathological evaluation is thus an absolute prerequisite for treatment within the specified therapy algorithm.

Background

The enzyme UDP-glucuronosyltransferase is responsible for the glucuronidation of SN- 38, the active metabolite of irinotecan, to the inactive metabolite SN-38G. Benign unconjugated hyperbilirubinaemia (Gilbert’s syndrome) is caused mainly by a variant in the UGT1A1 gene. More than 60 polymorphisms of the UGT1A21 gene have been identified to date. UGT1A1*1 is the wild type allele, while *28, *93, *60, and *6 are among the most common polymorphic variants. The most common variant found in Caucasians is the UGT1A1*28 allele, which causes a reduction in gene expression of up to 70% and results in a marked increase in irinotecan toxicity.

The frequency of the *28 allele is 39% in Europeans, 16% in Asians and 43% in African patients. Around 10–20% Caucasians and African Americans are homozygous for *28, while this is the case for less than 5% of Asians [Perera, M. A. et al. 2008].

Increased rates of side effects, in particular neutropenia and diarrhoea, are observed in patients with a decreased metabolisation of SN-38 undergoing irinotecan therapy.

The currently available data are insufficient to support a general recommendation of pretherapeutic UGT1A1 genotyping ([978]). In a meta-analysis of 12 clinical studies, UGT1A1*28 polymorphism was not identified as a reliable predictor of treatment effectiveness (response, PFS). When the UGT1A1*28 allele was present, no statistically significant result, but a trend towards a higher mortality risk was found in two models [978].

In case of spontaneously increased serum bilirubin (especially with low conjugated bilirubin), Gilbert’s syndrome may be present and, with this, an increased risk of irinotecan-associated side effects. According to the prescribing information for irinotecan, patients known to be homozygous for UGT1A1*28 should receive the typically indicated initial dose (taking into account the bilirubin value), but must be monitored for signs of haematological toxicity. In patients who show signs of haematological toxicity under the previous therapy, a reduced initial dose of irinotecan is to be considered.

In mCRC, MSI testing is helpful, on the one hand, to establish an indication for human genetic counselling. On the other hand, it can be useful with regard to the treatment with immune checkpoint inhibitors when determining a strategy spanning various therapy lines.

While the guideline was being created, the MSI status was of no relevance yet for the choice of first-line therapy. However, data are available on pretreated patients which suggest a high efficacy of a treatment with the checkpoint inhibitor pembrolizumab in patients with microsatellite instability (MSI-H). However, in 82% (9/11) of the examined patients, a hereditary component was found (Lynch syndrome), so that the transferability of these findings to patients with a sporadic MMR defect still has to be demonstrated [1012].

Background

In patients exhibiting signs of familial colon cancer, the microsatellite analysis (through an analysis of the immunohistochemical expression of the DNA mismatch repair proteins MLH1, MSH2, MSH6 and PMS2 or microsatellite instability testing) and, where necessary, coupled with a methylation analysis of the MLH1 promoter/exon 1 area, or, if the latter is not available, by means of BRAF diagnostics (hereditary vs. sporadic) is imperative.

The selectivity of a BRAF mutation analysis to distinguish between spontaneous and hereditary colorectal cancer is around 50% [1010]; the selectivity of a methylation analysis is complete (~100%) [Funkhouser, W. K., Jr. et al. 2012].

Background

For patients in whom intensive therapy is not a primary option, the possibility of administering bridging therapies with minimal side effects, e.g. therapy with a fluoropyrimidine, should be reviewed. Bevacizumab can be added to the therapy if rapid surgery is not the primary objective. For RAS wild type tumours, an anti-EGFR monotherapy is also an option.

As a general rule, the aim is to restore the patient’s ability to undergo therapy and to achieve a status in which more intensive combination therapies are tolerated. It is important to regularly review and adapt the treatment concept based on the patient’s overall health and motivation and on the tolerability and efficacy.

The term oligometastasis describes a limited spread of a potentially resectable or locally interventionally treatable metastasis formation, in which the spread is generally limited to e.g. 1-5 metastases and few organ systems (1-3 organs). Owing to the markedly poorer prognosis, metastases in lymph nodes, the brain or bones are not included in this categorisation.

No consensus on the definition and treatment of oligometastatic disease has been reached yet due to the lack of reliable data.

Background

A binding definition of oligometastasis is not available at the present time. In addition to the spread of metastasis, the concept also takes into account the possibility of local ablative or locoregional measures to treat the tumour in particular. Under favourable conditions, a curative treatment approach can also be considered in patients with oligometastases. This definition is based on the following assumptions:

a) The specific tumour biology appears to suggest a course in which the oligometastatic process shows a limited metastasis formation, at least for a relevant interval.

b) Surgical, local ablative or locoregional measures to treat the tumour can be administered in addition to systemic therapy.

c) The control of the localisation of the metastases is relevant for the prognosis, and the use of local ablative or resecting procedures may enable an interruption of the systemic therapy.

Where possible and reasonable, surgical resection should be the primary approach for locally treatable metastases. Local ablative procedures include: thermal ablation (RFA, MWA), radiotherapy (e.g. SBRT, brachytherapy) or electroporation (see section 9.13 Local Ablative Procedures).

Locoregional procedures include intra-arterial chemotherapy of the liver (HAI) or selective intra-arterial radiotherapy (SIRT) as well as cytoreduction with hyperthermal chemotherapy (HIPEC) (see also 9.14.1-9.14.3). Ultimately, a combination of surgical and ablative procedures is also possible. The indication for local ablative procedures should be established in multidisciplinary tumour conferences (see 9.14 Locoregional Procedures).

Background

A direct comparison of anti-EGFR antibodies (cetuximab and panitumumab) with the anti- VEGF antibody bevacizumab was carried out in three randomised studies (FIRE-3, PEAK, CALGB 80405) in the first-line therapy of mCRC [Stintzing, S. et al. 2016], [1049], [Lenz HJ, Niedzwiecki D, Innocenti F et al. et al. 2014]. The FIRE-3 study (phase III) compared FOLFIRI plus cetuximab with FOLFIRI plus bevacizumab and evaluated ORR as the primary endpoint. The evaluation of the RAS wild type (wt) population showed no significant difference in ORR and PFS. In contrast, a marked survival benefit was observed in the cetuximab arm (HR 0.70, p=0.0059) [Stintzing, S. et al. 2016].

The PEAK study was conducted as a randomised phase II study and compared FOLFOX plus panitumumab to FOLFOX plus bevacizumab. Here, PFS was evaluated as the primary target criterion. In the subgroup of patients with RAS-wt tumours, the panitumumab arm was found to be superior with regard to PFS (HR 0.65, p=0.029) and OS (41.3 vs. 28.9 months). Given the low event rate (<50%), no statistical significance was achieved with regard to OS (HR 0.63, 0.058) [1049].

The CALGB/SWOG-80405 study (phase III) carried out a randomised comparison of (any) chemotherapy plus cetuximab versus chemotherapy plus bevacizumab. The choice of chemotherapy regimen, FOLFOX (76%) or FOLFIRI (24%), was left up to the participating centres. This study reported a significantly higher ORR in the cetuximab arm (69% vs. 54%, p<0.01). In contrast, no significant difference between the treatment arms was established with regard to PFS and OS [Lenz HJ, Niedzwiecki D, Innocenti F et al. et al. 2014].

A meta-analysis based on the published results of the three comparative studies corroborated the superiority of the anti-EGFR treatment regarding ORR (odds ratio 1.46, p=0.004) and overall survival (HR 0.77, p=0.016). The PFS (HR 0.92, p=0.5) between the treatment arms was similar [999].

As a therapeutic alternative associated with more side effects, treatment with the chemotherapy triplet with FOLFOXIRI (possibly + bevacizumab) can also be considered in these patients. The TRIBE study compared FOLFIRI plus bevacizumab to FOLFOXIRI plus bevacizumab in unselected patients. Compared to doublet chemotherapy, the triplet therapy was found to be superior with regard to ORR (65% vs. 54%; OR 1.59, p=0.013), PFS (12.3 vs. 9.7 months; HR 0.77, p=0.006) and OS (29.8 vs. 25.8 months; HR 0.80, p=0.03) [1008]. As anticipated, the treatment with FOLFOXIRI plus bevacizumab was accompanied by a significant increase in grade 3-4 side effects such as neutropenia, diarrhoea and peripheral neuropathy compared to the administration of FOLFIRI plus bevacizumab [1015].

While phase II studies suggest a high efficacy of triplet plus anti-EGFR therapy, there are no corresponding results available from phase III studies.

Retrospective analyses of clinical studies suggest that the localisation of the primary tumour is not only of relevance for the prognosis, but is also an important determinant in the therapeutic efficacy and should therefore be considered when making therapy decisions [1000], [1001], [1002]. Preclinical analyses support the various patterns of gene mutation and gene expression in right- and left-sided tumours [1003]. [1004]. The majority of analyses describe the splenic flexure as the dividing line [1005], [1006]. In principle, the line dividing right- and left-sided tumours would be drawn between the proximal two thirds and the distal third of the transverse colon. However, in light of the retrospective evaluations, the splenic flexure was used as the dividing line in the majority of analyses for pragmatic reasons [1005], [1006]. Correspondingly, the caecum, ascending colon and transverse colon are considered part of the right hemicolon, while tumours of the descending colon, sigmoid colon and rectum are considered left-sided. Right-sided tumours are less common than left-sided tumours (30% versus 70%), and more female and elderly patients are affected. They are characterised by a higher tumour mutation burden and a higher immunogenicity. From a molecular biological perspective, right-sided tumours exhibit a higher rate of CIMP (CPG-island methylation phenotype), BRAF mutations and microsatellite instability [1000].

The currently available data suggest that left-sided tumours largely benefit from a treatment with anti-EGFR substances. In first-line therapy, especially doublet chemotherapy regimens have been assessed. In combination with anti-EGFR substances, these regimens were shown to be markedly more effective in left-sided tumour in terms of ORR, PFS and OS than comparable combinations with bevacizumab or without monoclonal antibodies [1006], [1007].

Right-sided tumours, on the other hand, are characterised by a less favourable prognosis with a poorer response to standard therapies and anti-EGFR antibodies. In two studies (FIRE-3, CALGB 80405), the combination of doublet chemotherapy with the anti-EGFR antibody cetuximab was less effective than the combination of the same chemotherapy with bevacizumab [1006], [1005]. Due to the low sample size, the significance level was not achieved in the subgroup analysis.

In contrast to EGFR antibodies, there are no signs to suggest an efficacy of bevacizumab depending on the tumour localisation. In the NO16966 and AVF2107g studies, the addition of bevacizumab achieved longer survival periods than chemotherapy alone for right-sided tumours. The interaction tests for tumour localisation were negative [1002].

Based on the currently available data, doublet or triplet (+/- Bev) chemotherapy is recommended for right-sided primary tumours in the first-line therapy of the metastatic disease.

Background

Anti-EGFR antibodies did not prove effective in patients with a RAS mutation and should therefore not be used [1020], [Van Cutsem, E. et al. 2015], [Bokemeyer, C. et al. 2015]. Prospective studies assessing the effectiveness of anti-VEGF substances in the first-line therapy of RAS-mutated mutation tumours are not available. A retrospective subgroup analysis of the FIRE-3 study showed comparable survival periods in patients with a KRAS mutation who received FOLFIRI plus bevacizumab or FOLFIRI plus cetuximab [Stintzing, S. et al. 2012]. In contrast, a retrospective analysis of the TML study regarding negative interaction tests suggests that the effectiveness of the treatment with bevacizumab in the second line of therapy was independent of the KRAS status [Kubicka, S. et al. 2013].

A subgroup analysis of the TRIBE study compared the effectiveness of triplet chemotherapy (FOLFOXIRI plus bevacizumab) with that of doublet chemotherapy (FOLFIRI plus bevacizumab) in patients with RAS-mutated tumours [1008]. Here, a higher effectiveness was reported for the triplet therapy with regard to overall survival (HR 0.88; 95% CI: 0.65-1.18), PFS (HR 0.78; 95% CI: 0.60-1.02) and the response rate (OR 1.55; 95% CI: 0.91-2.62). However, the significance level was not reached in any of the effectiveness measures. Consequently, a clear recommendation for the use of triplet chemotherapy for RAS-mutated tumours cannot be given.

Background

A BRAF V600 mutation is observed in 8-12% of mCRC patients. Women are more commonly affected, and the age of onset is usually higher. In around two thirds of affected patients, the tumour is located in the right hemicolon; increased mucinous subtypes are observed histologically.

A higher rate of lymph node metastases and peritoneal cancer is reported clinically. From a molecular pathological perspective, microsatellite instability and a “methylator phenotype” are common [Samowitz, W. S. et al. 2006], [972]. The prognosis of patients with a BRAF V600 mutation is exceedingly poor; numerous studies have reported median PFS intervals of less than 6 months and median survival periods of less than one year [974].

Triplet chemotherapy with the FOLFOXIRI regimen is currently recommended for patients with a BRAF V600 mutation. However, this recommendation is based on a subgroup analysis of only 28 patients with a BRAF mutation treated within the scope of the TRIBE study. Under treatment with FOLFOXIRI plus bevacizumab (n=16) versus FOLFIRI plus bevacizumab (n=12), considerably more favourable outcome data were observed in these patients: markedly prolonged OS (19.0 vs. 10.7 months; HR 0.54), longer PFS (7.5 vs. 5.5 months; HR 0.57) and a higher remission rate (56% vs. 42%; OR 1.87) [977]. On the one hand, the results of this analysis can only be regarded as a basis for generating hypotheses owing to the low sample size; on the other hand, further analyses of the same working group have been published which support the effectiveness of FOLFOXIRI plus bevacizumab in patients with a BRAF mutation [1009].

Whether anti-EGFR substances are effective against BRAF mutations is the subject of controversial discussions. In this regard, two meta-analyses came to a different evaluation: In their analysis, Pietrantonio et al. found no significant increase in PFS (HR 0.88, p=0.33) or OS (HR 0.91, p=0.63) with the addition of anti-EGFR antibodies [974]. Rowland et al., on the other hand, argue that the evidence is insufficient to definitively rule out that anti-EGFR antibodies have a different treatment effect in BRAF mutations than in BRAF wild type tumours [Rowland, A. et al. 2015].

Ultimately, the available analyses are characterised by small sample sizes which do not allow for drawing definitive conclusions either in the individual nor in the joint meta- analysis.

The question of the significance of bevacizumab-based therapy versus cetuximab-based therapy was also addressed in a subgroup analysis of the FIRE-3 study. In 48 evaluable patients with RAS-wt/BRAF-mut mCRC, OS was short and comparable in both therapy arms (median 12.3 vs. 13.7 months), regardless of whether cetuximab or bevacizumab had been given in combination with FOLFIRI [Stintzing S, Jung, L. Rossius, D.P. et al. et al. 2014]. This analysis led to the hypothesis that, equally, neither an anti-EGFR nor an anti-VGEF strategy has the ability to improve the therapeutic outcome.

Owing to the poor prognosis of BRAF-mutated tumours, individual (as yet unauthorised) therapeutic approaches, for example with a BRAF inhibitor, MEK inhibitor and anti-EGFR antibody or, where possible, within the scope of a clinical study, can be considered for the second-line therapy [Kopetz, S. et al. 2015].

Immune checkpoint inhibitors have been shown to be active in first clinical investigations in pretreated mCRC patients with microsatellite instability (MSI). In view of the currently limited data, however, a first-line therapy according to the RAS mutation status is first recommended in patients with MSI. In later therapy lines, the possibility of treatment with checkpoint inhibitors should be assessed.

Background

Mutations in the mismatch repair genes (MLH1, MSH2, MSH6 and PMS2) lead to faulty DNA replication, which in turn manifests as microsatellite instability (MSI) in variable lengths of the microsatellite DNA. Compared to proficient mismatch repair (MMRp), defective mismatch repair (MMRd) is responsible for a 10- to 100-fold increase in the mutation rate ([1010]). This causes increased immunogenicity and ultimately leads to the markedly increased lymphocytic infiltration of MSI tumours ([1011]). Upregulation of immune checkpoints, such as the programmed death (PD-1) pathway, has been shown especially for MSI tumours in the sense of an immune escape mechanism.

In colorectal cancer, MMRd occurs both within the scope of germ line mutations in one of the four mismatch repair genes (hereditary nonpolyposis colorectal cancer (HNPCC or Lynch syndrome) as well as by somatic mutations or epigenetic silencing [1012].

First clinical data confirm the hypothesis that MSI tumours – in contrast to MSS tumours – respond well to a PD-1 blockade. A phase II study assessed 32 patients who had received at least 2 prior chemotherapy regimens [1012]. Of these, 11 patients were classified as MMRd and 21 as MMRp. Treated with the PD-1 inhibitor pembrolizumab, MMRd patients showed a significantly higher ORR (40% vs. 0%) and SD rate (50% vs. 11%) as well as a highly significant prolongation of PFS and OS compared to MMRp patients (medians were not achieved).

Immunological checkpoint inhibitors were not yet authorised for the treatment of mCRC at the time of creating this guideline.

Background

A good overall condition of the patient is a prerequisite for intensive combination chemotherapy. Since the option of a less effective fluoropyrimidine monotherapy plus bevacizumab with markedly fewer side effects is theoretically available, it is important to involve the patients in the discussion and to define their motivation regarding a potential prolongation of survival [1013], [1014].

Background

The AVEX study assessed the effectiveness of a combination of capecitabine plus bevacizumab in elderly patients (≥70 years) and compared it with capecitabine monotherapy within the scope of a phase III design [1014]. Mainly patients with a performance status ECOG 0 (46%) and ECOG 1 (45%) were enrolled. Patients with ECOG ≥2 accounted for less than 10% of the patient population. PFS was assessed as the primary endpoint. Treated with capecitabine plus bevacizumab, a significant prolongation of PFS (9.1 vs. 5.1 months; HR 0.53, p<0.0001; primary study endpoint) and an increase in ORR (19% vs. 10%; p=0.04) was achieved. Overall survival (secondary endpoint) was 16.8 months in the control arm and 20.7 months in the group treated with capecitabine and bevacizumab ((HR 0.79, 95% CI: 0.57–1.09; p=0.18); this difference was not significant. The tolerability of the treatment can generally be rated as good. The frequency of serious adverse events (SAEs) was 31% in the group receiving capecitabine monotherapy and 30% in the group receiving the combination of capecitabine and bevacizumab. Quality of life analyses were not carried out [1014].

These data are supported by a randomised phase II study conducted in patients for whom first-line therapy with irinotecan was not an option [1013]. The proportion of patients with ECOG 2 was again below 10% in this study. In this population, bevacizumab plus 5-FU/LV was compared to 5-FU/LV therapy alone. The addition of bevacizumab to 5-FU/LV led to a significant increase of PFS (9.2 vs. 5.5 months; HR 0.50, p=0.0002) and to a non-significant increase of OS (16.6 vs. 12.9 months; HR 0.79, p=0.16) and ORR (26.0% vs. 15.2%, p=0.055). The evaluation of the FACT-C score showed no negative effect of bevacizumab on the quality of life (QOL). The median time to worsening of the quality of life was 3.2 months in the bevacizumab arm and 2.3 months in the placebo arm (HR 0.66; p=0.016).

In summary, these study results suggest that first-line chemotherapy with fluoropyrimidine and bevacizumab is effective in elderly patients and in patients unsuitable for initial irinotecan-based therapy, and is thus a expendient therapeutic option for this patient population.

Background

A “meta”-analysis of two prospective, randomised studies, as well as two additional prospective, randomised studies are available [1016], [1017], [1018], [1015], [1008], [1019]. The meta-analysis compared FOLFOXIRI to FOLFIRI; the other two studies compared FOLFOXIRI + bevacizumab to FOLFIRI + bevacizumab. The data consistently show a higher response rate, prolonged PFS and OS, and in the studies that assessed this, a higher secondary R0 resection rate of liver metastases. The rate of grade 3-4 side effects was significantly higher in the group of patients treated with FOLFOXIRI plus bevacizumab than in the group treated with FOLFIRI plus bevacizumab. The most common side effects were neutropenia (50% vs. 20.5%, p<0.001), diarrhoea (18.8 vs. 10.6, p=0.01) and peripheral neuropathy (5.2 vs. 0%, p<0,001).

Background

Monoclonal anti-EGFR antibodies such as cetuximab or panitumumab are only effective in patients with RAS wild type tumours (Table 7).

In a randomised phase III study (CRYSTAL) with RAS-wt patients, the addition of cetuximab to FOLFIRI chemotherapy led to a prolongation of median survival from 20.2 to 28.4 months (HR 0.69, p=0.0024) [Van Cutsem, E. et al. 2015]. Comparable data were described in the PRIME study. In this randomised phase III study, the addition of panitumumab to the FOLFOX regimen resulted in a prolongation of OS from 20.2 to 26.0 months (HR 0.78, p=0.04).

In a meta-analytical evaluation (see Guideline Report) of the available studies, the addition of anti-EGFR antibodies to combination therapies in patients with KRAS wild type tumours resulted in a significant prolongation of PFS (HR 0.83, p=0.0001) and OS (HR 0.89, p=0.02).

In view of the study results, it was retrospectively postulated that anti-EGFR substances fail to achieve adequate efficacy in combination with a 5-FU bolus regimen or oral fluoropyrimidines. If one takes these studies (COIN and NORDIC) out of the meta- analytical evaluation, the therapeutic effects due to the addition of anti-EGFR antibodies are considerably stronger for PFS (HR 0.72, p<0.00001) and OS (HR 0.79, p=0.0003).

Compared to FOLFOX monotherapy, the use of panitumumab in the PRIME study in combination with FOLFOX led to a marked increase in grade 3-4 side effects, such as skin toxicity (36% vs. 2%), diarrhoea (18% vs. 9%), fatigue (9% vs. 3%) or hypomagnesaemia (6% vs. <1%). The rate of grade 3 infusion reactions after administration of the human IgG2 antibody panitumumab was 0.3% [Douillard, J. Y. et al. 2010]. Similar side effects were also reported in the CRYSTAL study. Compared to the administration of FOLFIRI alone, the combination of FOLFIRI plus cetuximab led to an increase in grade 3- 4 side effects, such as acneiform exanthema (16.2% vs. 0%) or diarrhoea (15.7% vs. 10.5%). The rate of grade 3-4 infusion reactions after administration of the chimeric IgG1 antibody cetuximab was 2.5% [881].

Regarding the use of anti-EGFR substances depending on the localisation of the primary tumour, see 9.8.2.

Background

The effectiveness of an anti-VGEF treatment with bevacizumab was assessed in numerous studies on the first-line therapy of mCRC (Table 15). The study population included patients with an unselected molecular pathology. In the meta-analytical evaluation of all available studies, the addition of bevacizumab to fluoropyrimidine- based chemotherapy (monotherapy or combination therapy) led to a significant increase of PFS (HR 0.71, p<0.00001) and OS (HR 0.85, p=0.0008).

When limiting the analysis to fluoropyrimidine monotherapy (5-FU bolus or infusional 5- FU, capecitabine), the addition of bevacizumab resulted in a highly significant prolongation of PFS (HR 0.57, p<0.00001) and OS (HR 0.83, p=0.03).

If one excludes the no longer common 5-FU bolus regimen (IFL) from the analysis and focuses the evaluation on infusional combination chemotherapies, the addition of bevacizumab led to a significant prolongation of PFS (HR 0.79, p<0.0001), but not of overall survival (HR 0.92, p=0.18).

In summary, it can be said that an increase of overall survival was achieved in the meta- analytical evaluation of the available studies when bevacizumab was added to fluoropyrimidine monotherapy; this was not the case when it was added to infusional combination chemotherapy.

In the marketing authorisation study, the addition of bevacizumab to the IFL regimen led to an increase of grade 3-4 side effects such as diarrhoea (32.4% vs. 24.7%), hypertension (11% vs. 2.3%) or thrombotic events (19.4% vs. 16.2%) compared to treatment with IFL alone. An increase of grade 3-4 bleeding events (3.1% vs. 2.5%) and gastrointestinal perforations (1.5% vs. 0%) was additionally observed, however with an altogether lower frequency [1021]).

The following bevacizumab-specific side effects were described for the addition of bevacizumab to an oxaliplatin-based treatment regimen (FOLFOX or XELOX) in the NO16966 study: venous thromboembolic events (8% vs. 5%), arterial thromboembolic events (2% vs. 1%), bleeding events (2% vs. 1%), hypertension (4% vs. 1%) [1024].

Compared to capecitabine monotherapy, the addition of bevacizumab to treatment with capecitabine was associated with comparably fewer side effects. The following grade 3- 4 side effects were described in the AVEX study: hand-foot syndrome (16% vs. 7%), diarrhoea (7% vs. 6%), venous thromboembolic events (6% vs. 4%), bleeding events (0% vs. 0%), hypertension (0% vs. 1%) [1014].

Background

Three prospective studies consistently show that polychemotherapy in combination with anti-EGFR antibodies and with bevacizumab is associated with a reduced PFS and increased toxicity compared to polychemotherapy in combination with bevacizumab. Oxaliplatin-containing therapy was administered in all three studies; irinotecan-based therapy was additionally administered in one study [1028], [1029], [1030].

Background

Due to oxaliplatin-induced polyneuropathy, FOLFOX or CAPOX can generally not be administered for longer than 4-6 months. In the MACRO study and in the CAIRO3 study, 6 cycles of CAPOX (4.5 months) were administered prior to commencing maintenance therapy [Diaz-Rubio, E. et al. 2012], [Simkens, L. H. et al. 2015]. In the AIO KRK-0207 study, maintenance therapy was initiated after 6 months of induction therapy [1031]. In the TRIBE study, the duration of the induction therapy was also limited to a maximum of 12 cycles of the biweekly chemotherapy, following which maintenance therapy with 5-FU plus bevacizumab was administered and continued up to progression [1002].

The optimal duration of an induction therapy with FOLFIRI is not clear. For this reason, this treatment can be continued until an optimal response is achieved. In any case, an initial treatment duration of at least 4-6 months will also be aimed for here.

For patients receiving initial monochemotherapy with a fluoropyrimidine, this treatment should be continued until the disease progresses.

Background

After 4-6 months of oxaliplatin-based induction therapy, it should be verified whether maintenance therapy can be carried out to decrease toxicity and increase quality of life, or whether an interruption in therapy is possible.

Background

Where possible, metastatic CRC patients should be given access to all available medicines during the course of the treatment. The significance of an effective second-line therapy for overall survival has been analysed in several phase III studies. For example, a second-line therapy with irinotecan after failure of fluorouracil monotherapy resulted in a significant benefit in overall survival in comparison with BSC (OS 9.2 vs. 6.5 months) [1039] or infusional 5-FU/FA [1038] (OS 10.8 vs. 8.5 months).

In a comparison of irinotecan with infusional 5-FU, pain-free survival was longer under the treatment with irinotecan (10.3 vs. 8.5 months), with an otherwise comparable tolerability and quality of life in the two treatment arms [1038]. In the direct comparison with best supportive care, irinotecan significantly increased the time without worsening of the performance status. In the quality of life analysis, the treatment with irinotecan achieved better outcomes in all aspects (except diarrhoea) [1039].

Combination therapy with oxaliplatin and fluorouracil after failure of an irinotecan- containing protocol in first line-therapy was superior to 5-FU/FA or oxaliplatin monotherapy both in terms of the response rates achieved and time to progression [1037]. The combination therapy of 5-FU/FA with oxaliplatin and/or irinotecan, which show response rates (CR + PR) of 40-55% in first-line therapy, achieve response rates of 4% (FOLFIRI) to 15% (FOLFOX) and a progression-free survival of around 2.5-4.2 months in second-line therapy. The median survival of patients for both therapy sequences (FOLFOX followed by FOLFIRI or FOLFIRI followed by FOLFOX) was around 20 months [1040]. As anticipated, the toxicity profile differed depending on the medicines used. Quality of life analyses from these studies are not available. 

 Several randomised studies have confirmed the benefit of bevacizumab (E3200, TML, BEBYP), aflibercept (VELOUR) and ramucirumab (RAISE) in second-line therapy. The effects of the therapy are highly consistent. The evaluable studies consistently show that a significant increase of PFS and OS can be achieved by adding the anti-VEGF substances bevacizumab or aflibercept or the anti-VEGFR antibody ramucirumab to second-line chemotherapy. It should be noted, however, that the absolute increase in the survival is moderate in comparison with the median OS and generally ranges between 1-2 months.

The antiangiogenic therapy is associated with the typical side effects; for example, the following grade 3-4 side effects were observed with the addition of aflibercept to FOLFIRI compared to FOLFIRI monotherapy: hypertension (19.3% vs. 1.5%), haemorrhaging (2.9% vs. 1.7%), arterial thromboembolic events (1.8% vs. 0.5%) and venous thromboembolic events (7.9% vs. 6.3%). In addition, a potentiation of chemotherapy-associated toxicities, such as diarrhoea or stomatitis, was also observed in part [1041].

Similarly, an increase in side effects was also observed for the combination of the VEGFR inhibitor ramucirumab with FOLFIRI compared to FOLFIRI chemotherapy. The increase notably affected grade 3-4 side effects such as neutropenia (38% vs. 23%), hypertension (11% vs. 3%), haemorrhaging (1.9% vs. 1.5%) or gastrointestinal perforation (1.5% vs. 0.6%) [1042].

The clinical data of the TML study confirm that the continuation of an anti-VEGF therapy with bevacizumab after progression under a bevacizumab-based first-line therapy is an effective treatment strategy in unselected mCRC patients. In comparison with chemotherapy alone, patients who received bevacizumab plus chemotherapy in the second line of therapy showed a longer median overall survival of 11.2 months (95% CI: 10.4-12.2) for bevacizumab plus chemotherapy and 9.8 months (95% CI: 8.9-10.7) for chemotherapy alone (HR 0.81, p=0.0062) [1045].

The most commonly reported grade 3-5 side effects in the TML study were neutropenia (16% vs. 13%), diarrhoea (10% vs. 8%) and asthenia (6% vs. 4%). Under treatment with bevacizumab plus chemotherapy, the following grade 3-5 side effects were more common compared to chemotherapy alone: bleeding/haemorrhaging (2% vs. <1%), gastrointestinal perforation (2% vs. <1%) and venous thromboembolic events (5% vs. 3%).

Two large randomised studies compared a treatment with best supportive care (BSC) plus anti-EGFR antibodies versus BSC alone in patients with mCRC refractory to chemotherapy [Van Cutsem, E. et al. 2007], [Jonker D.J. , M.D., Chris J. O'Callaghan, Ph.D., Christos S. Karapetis, M.D., John R. Zalcberg, M.D., Dongsheng Tu, Ph.D., Heather-Jane Au, M.D., Scott R. Berry, M.D., Marianne Krahn, M.D., Timothy Price, M.D., R. John Simes, M.D., Niall C. Tebbutt, M.D., Guy van Hazel, M.D., Rafal Wierzbicki, M.D., Christiane Langer, M.D., and Malcolm J. Moore, M.D. et al. 2007]. A direct head-to-head comparison of cetuximab and panitumumab also yielded similar effectiveness and toxicity data [Price, T. J. et al. 2014]. Regarding the RAS status of unselected patients, the addition of panitumumab to BSC led to a significant improvement of the response rate (10% vs. 0%) and PFS (HR 0.54, p<0.0001). However, no increase in survival was achieved (HR 1.00), which was explained by the high proportion of cross-over patients (76%) [Van Cutsem, E. et al. 2007].

Another study evaluated the addition of cetuximab to BSC [Jonker D.J. , M.D., Chris J. O'Callaghan, Ph.D., Christos S. Karapetis, M.D., John R. Zalcberg, M.D., Dongsheng Tu, Ph.D., Heather-Jane Au, M.D., Scott R. Berry, M.D., Marianne Krahn, M.D., Timothy Price, M.D., R. John Simes, M.D., Niall C. Tebbutt, M.D., Guy van Hazel, M.D., Rafal Wierzbicki, M.D., Christiane Langer, M.D., and Malcolm J. Moore, M.D. et al. 2007]. In a retrospective analysis of patients with KRAS wild type tumours, the addition of cetuximab led to an ORR of 12.8% vs. 0% and a significant prolongation of PFS (3.7 vs. 1.9 months; HR 0.40, p<0.001) and OS (9.5 vs. 4.8 months; HR 0.55, p<0.001). In contrast, no benefit was observed in the subgroup of patients with KRAS-mutated tumours [Karapetis, C. S. et al. 2008]. With minor worsening of the physical function and global health status scores (both p<0.05), the quality of life in the cetuximab group was upheld for a significantly longer period [Jonker D.J. , M.D., Chris J. O'Callaghan, Ph.D., Christos S. Karapetis, M.D., John R. Zalcberg, M.D., Dongsheng Tu, Ph.D., Heather-Jane Au, M.D., Scott R. Berry, M.D., Marianne Krahn, M.D., Timothy Price, M.D., R. John Simes, M.D., Niall C. Tebbutt, M.D., Guy van Hazel, M.D., Rafal Wierzbicki, M.D., Christiane Langer, M.D., and Malcolm J. Moore, M.D. et al. 2007]. When comparing cetuximab to BSC alone, the following grade 3-4 side effects were observed: infusion reactions (4.5% vs. 0%), rash (11.8% vs. 0.4%), hypomagnesaemia (5.2% vs. 0%) [Jonker D.J. , M.D., Chris J. O'Callaghan, Ph.D., Christos S. Karapetis, M.D., John R. Zalcberg, M.D., Dongsheng Tu, Ph.D., Heather-Jane Au, M.D., Scott R. Berry, M.D., Marianne Krahn, M.D., Timothy Price, M.D., R. John Simes, M.D., Niall C. Tebbutt, M.D., Guy van Hazel, M.D., Rafal Wierzbicki, M.D., Christiane Langer, M.D., and Malcolm J. Moore, M.D. et al. 2007]).

Background

Trifluridine/Tipiracil (TAS102) was assessed in two randomised studies (phase II and phase III) in pretreated CRC patients and was compared to placebo [1054], [1053]. One inclusion criterion was prior treatment with at least two chemotherapy regimens with monoclonal antibodies. Both studies showed an increase of overall survival (OS 1.8 and 2.4 months, respectively). In the phase III study, the hazard ratio for the improvement of overall survival was 0.68 (95% CI: 0.58-0.81, p<0.001). The therapy was also effective in patients who had previously undergone FU therapy. The rate of side effects was classified as moderate. The following grade 3-4 haematological side effects were mainly observed: neutropenia (38% vs. 0%), febrile neutropenia (4% vs. 0%), anaemia (18% vs. 3%) and thrombocytopenia (5% vs. <1%).

Background

Regorafenib is a multikinase inhibitor that was tested against placebo in pretreated mCRC patients within the scope of the CORRECT study [1056]. The study achieved its primary endpoint and showed a significant survival benefit for regorafenib (OS: 6.4 vs. 5.0 months; HR 0.77; 95% CI: 0.64–0.94). These results were confirmed by the CONCUR study, which was conducted exclusively in Asian patients [1055]. The use of regorafenib was seen critically owing to the toxicity associated with the treatment. The most significant grade 3-4 side effects include hand-foot syndrome (17% vs. <1%), fatigue (10% vs. 5%) and diarrhoea (7% vs. 1%) [1056].

The reinduction of antineoplastic substances whose efficacy was demonstrated in earlier therapy lines is a proven therapeutic strategy in oncology; however, evidence supporting the clinical effectiveness of this approach is limited.

Santini et al. analysed the effectiveness of a rechallenge with cetuximab in a small study (n=39). In patients who had responded to a cetuximab-based first-line therapy with a CR, PR or SD >6 months and who progressed under a cetuximab-free window therapy, an ORR of 54% and a PFS of 6.6 months was achieved with repeated cetuximab-based therapy [1057]. However, owing to the very small sample size, these data can only be considered as a basis for generating hypotheses. 

Background

In comparably small randomised studies, the combination of SIRT (selective intra-arterial radioembolisation with Yttrium-90 resin microspheres) with 5-FU was found to be more effective than chemotherapy alone. In patients refractory to chemotherapy who had undergone multiple prior therapies (n=44), the addition of SIRT to 5-FU monotherapy led to an improvement in the response rate (10% vs. 0%) and TTP (4.5 vs. 2.1 months, P<0.03) compared to 5-FU monotherapy [1069].

In the randomised comparison of SIRT plus 5-FU versus 5-FU, a significant increase in TTP (18.6 vs. 3.6 months, p<0.0005) and OS (29.4 vs. 12.8 months, p=0.025) was achieved in 21 patients in first-line therapy ([Van Hazel, G. et al. 2004]). Based on the data of a phase I study, the combination of SIRT with FOLFOX chemotherapy was assessed in three randomised multicentre studies (SIRFLOX, FOXFIRE and FOXFIRE global) [Sharma, R. A. et al. 2007].

In the SIRFLOX study, PFS (primary endpoint) was not improved by adding SIRT to the FOLFOX chemotherapy regimen in first-line therapy [van Hazel, G. A. et al. 2016]. Overall survival was analysed as the primary endpoint in the joint evaluation of all three studies under the blanket of the FOXFIRE global study. The results of the FOXFIRE global study were presented for the first time during the 2017 ASCO Annual Meeting. In the intention-to-treat population (n=1103), the addition of SIRT to an oxaliplatin-/5-FU-based first-line chemotherapy showed no advantage in relation to PFS or overall survival. However, the treatment with SIRT caused additional side effects [Sharma RA, Wasan HS, van Hazel GA et al. et al. 2017]. The high proportion of synchronous hepatic (87%) and extrahepatic metastases (35%), of 50-55% primary tumours in situ, and of 36% ECOG 1 patients in this study population is remarkable. These results should at least partly explain the modest effect of an isolated liver-associated therapy such as SIRT on overall survival. The addition of SIRT to 5-FU/oxaliplatin-based (+ bevacizumab) chemotherapy led to a significant improvement in the response rate (odds ratio 1.52, p=0.001) in the ITT population of the FOXFIRE global study. Even though a significant benefit of the SIRT therapy was also observed in the hepatic response, the resection rate in both treatment arms was almost identical (16% vs. 17%, p=0.669). Moreover, the secondary resectability was evaluated by a panel of hepatobiliary surgeons in the SIRFLOX study. In this blinded analysis of imaging at the time of the best response, a significantly higher potential resectability was found in the SIRT arm than in the control arm (Garlipp et al. ASCO 2017) [Garlipp B, Gibbs P, Van Hazel GA, et al. et al. 2017].