Randomized study comparing full dose monotherapy (S-1 followed by irinotecan) and reduced dose combination therapy (S-1/oxaliplatin followed by S-1/irinotecan) as initial therapy for older patients with metastatic colorectal cancer: NORDIC 9

Survival of patients with metastatic colorectal cancer (mCRC) has increased considerably over the past several decades. Contributory factors for this improvement are surgical resection or other ablative techniques for metastatic disease, a more strategic approach to the delivery of systemic therapy with a continuum of care, better selection of patients to different treatments and more lines of effective anticancer drugs [1].

The backbone of medical treatment of mCRC is 5-fluorouracil (5-FU). Since about two decades, it is seldom given alone but biochemically modulated with calcium folinate due to higher response rates and likely improved survival [2]. The oral prodrugs of 5-FU (capecitabine, UFToral and S-1) are as efficient as modulated 5-FU and efficacy is further increased when 5-FU is combined with irinotecan or oxaliplatin [1, 3]. Combination regimens are often the best first-line choice, but monotherapy is an alternative in some clinical situations based upon the results of several trials. In both the CAIRO [4], FOCUS1 [5] and FFCD 2000–05 trials [6] unselected patients with previously untreated mCRC were randomized between a sequential strategy and combination therapy. In the FOCUS2 [7] and FFCD 2001–02 [8] trials, the doublet strategy in older patients with mCRC was explored, in FOCUS2 with a population of patients who were not candidates for full-dose chemotherapy. All studies showed that it is safe to start monotherapy if followed by new therapy upon progression.

Median overall survival (OS) often exceeds 24 months for mCRC patients included in clinical trials. In unselected populations of mCRC patients, however, median OS is only around 12 months [9‐11] although, in one of the population-based studies [12], median OS was 21.3 months in the subgroup of patients (36%) treated with combination chemotherapy as in a clinical trial. The short OS in general populations of mCRC patients is mainly for a short survival in patients above 70–75 years of age and in patients not receiving any chemotherapy due to a variety of reasons [12, 13]. The reported OS improvement over time in mCRC patients is mainly seen in younger patients and only minor improvements are found in older patients [14]. In a large community-based study, older patients (65+ years) were less likely to receive first-line doublet chemotherapy and also less likely to receive irinotecan, oxaliplatin, and bevacizumab during the entire course of the disease [15].

In a combined analysis of more than 2500 patients treated with different irinotecan/5-FU schedules in four first-line phase III trials, the authors concluded that older patients (70+ years) who fulfilled the inclusion criteria of the trials had similar benefits of treatment and similar risk of toxicity as younger patients, and these results have been confirmed in other studies including systematic reviews [16‐19]. A comparable outcome in patients below or above 70 years included in randomized trials is reasonably explained by inclusion of only the very fittest older patients, based on strict inclusion criteria, e.g. appropriate performance status and sufficient organ functions, but these patients are not representative of the majority of older patients cared for in the oncological clinics [20]. This has an impact on our perception of standard therapy, as colorectal cancer (CRC) is a disease of older age with approximately 50% of the patients being 70 years or older [9, 21]. From 2001 to 2005 the median age of patients with mCRC included in clinical trials was only 62 years [22], while the median age at mCRC diagnosis is 71–74 years according to cancer registries [14].

In clinical practice many oncologists recommend full-dose monotherapy or reduced dose combination therapy (without much evidence) in older and/or frail patients. In the FOCUS1 trial [5], where single agent therapy versus combination chemotherapy was explored, the median age was only 64 years, despite permissive entry criteria and no upper age limit, and a retrospective survey showed that almost twice as many patients had been treated off-trial during the same period, frequently using reduced-dose or single-agent schedules. The most frequent reasons for non-inclusion were physicians’ concerns about the adverse effects of standard-dose treatments and patients’ wishes to avoid toxic effects. The authors therefore designed FOCUS2 [7] for patients where the treating oncologist considered standard full-dose regimens to be unsuitable. A total of 459 patients (median age 74 years) were randomized to reduced dose monotherapy or reduced dose oxaliplatin combination therapy, with a planned dose escalation in patients with no or few side effects. However, dose escalation was only effectuated in one third of the patients. Combination therapy more than doubled response rate (from 13% to 35%) and marginally prolonged progression-free survival (PFS) (HR 0.84 (0.69–1.01), p = 0.07), but OS was not improved (HR 0.99 (0.81–1.18)).

In a recent French phase III trial (FFCD-2001-02), where 282 mCRC patients aged ≥75 years received either first-line monotherapy with 5-FU or FOLFIRI, no significant differences in PFS (5.2 months vs. FOLFIRI 7.3 months, HR 0.84 (0.66–1.07), p = 0.15) or OS (14.2 months vs. 13.3 months, HR 0.96 (0.75–1.24)) were found [8]. In line with these data, a recent meta-analysis showed that combination chemotherapy prolonged PFS but not OS with an increased risk of toxicity in older patients [23].

Further development of antineoplastic drugs may lead to agents with a potentially more favorable toxicity profile which may be of benefit especially for older cancer patients. S-1 (Teysuno®) is a third generation oral fluoropyrimidine comprised of tegafur, a pro-drug of 5-FU, and two modulators of 5-FU metabolism; gimeracil and oteracil [24]. S-1 provides sustained 5-FU plasma concentrations with reduced toxicities like hand-foot-syndrome (HFS) and probably cardiotoxicity and diarrhoea, due to the addition of the modulators [25‐27]. Randomized Asian studies in mCRC patients have shown that S-1 is as effective as 5-FU or capecitabine as monotherapy, and in combination with oxaliplatin [28, 29] or irinotecan [30]. In Japan S-1 is approved for the treatment of several cancers including gastric, colorectal and pancreatic cancers, but in the European Union presently only for gastric cancer [24].

Because of higher activity of CYP2A6 and thereby more effective conversion of S-1 to 5-FU in Caucasians, the optimal dose is lowered in Caucasians compared to Asians [24, 26]. In an observational chart review the use of S-1 as monotherapy (30 mg/m² twice daily on days 1–14 every 3 weeks) and in combination (25 mg/m² twice daily on days 1–14 every 3 weeks) with oxaliplatin or irinotecan, respectively, in 71 older mCRC patients was evaluated [26]. In all three settings S-1 was well-tolerated with a low rate of clinical and haematological adverse events, and especially a very low incidence of HFS was found. Furthermore the treatment indicated comparable efficacy to standard regimens in mCRC. However further prospective Western studies are needed to confirm efficacy and safety of S-1 in young and older populations.

Bevacizumab has in randomized studies, as a part of combination therapy improved efficacy in mCRC patients [31‐34]. Retrospective cohort studies indicate a similar benefit on a population level [35, 36]. In a recent study, S-1 plus bevacizumab therapy was effective and safe for older (median age 75 years) mCRC patients [37]. In the SALTO trial, S1 plus bevacizumab was as effective as capecitabine plus bevacizumab with a significantly lower incidence of HFS [38].

Based on these facts, the ongoing NORDIC 9 trial was initiated to compare full dose single agent S-1 therapy and dose-reduced combination therapy in older patients not considered candidates for standard combination therapy, hypothesizing that reduced combination therapy improves efficacy compared to fulldose monotherapy.

This study is conducted to evaluate two treatment strategies in older mCRC patients - full-dose single agent therapy or dose-reduced combination therapy. Patients are not candidates for full dose double or triple combination therapy either because of frailty, very old age, patient preference or because the treating physician recommends less intensive therapy [1], and thus constitute a heterogeneous group. With this study population we have tried to imitate the older population who often challenges physicians in clinical everyday life and where the decision on which treatment regimen to choose is not clear-cut.

As S-1 seems to have a better toxicity profile compared to capecitabine, especially because of a lower incidence of HFS [26], it may be a suitable drug for the study population.

A randomized phase II study, AVEX [32], showed that bevacizumab improved efficacy (RR and PFS) of single agent capecitabine in fit older patients. Therefore it is optional to add bevacizumab to chemotherapy at the discretion of the treating physician in the trial.

Standard treatment in patients with mCRC includes chemotherapy in several lines with an expanding range of treatment options [1]. Therefore we have planned a continuum of care/strategy for patients including both first and second line therapy as part of the trial.

If the study reaches the desired end-point for efficacy, it suggests that reduced combination therapy becomes the standard treatment for patients with mCRC who are not candidates for standard full-dose combination therapy. Future studies will be needed to confirm this and perhaps evaluate the potential benefit of adding bevacizumab to all patients in this particular study population.

SIOG recommends that future research focuses on the ability of screening tools to build clinical pathways and to predict different outcome parameters [48]. We have in the trial chosen to include several geriatric screening tools, which have sensitivity and specificity comparable to a full CGA, but are less time-consuming [48, 49]. Most of the studies performed on geriatric screening tools have been retrospective and performed in patient populations with heterogeneous cancers [39]. We will evaluate the chosen geriatric screening tools in a prospective study design in a population of older mCRC patients in order to predict efficacy and toxicity of the oncological treatment and search for at-risk individuals. If shown to be valid and relevant, the screening tools may in the future be part of routine screening in the treatment decision of geriatric patients [39].