Adjuvant chemotherapy for patients with rectal cancer – will the controversy be resolved?

In gastrointestinal oncology adjuvant chemotherapy after radical rectal cancer surgery has been one of the most controversial issues during the past decades or basically since 1990 when the first evidence that adjuvant chemotherapy prevented recurrences and improved survival in colon cancer was published [1]. One year later, a National Cancer Institute (NCI) statement further emphasized that adjuvant chemotherapy and radiotherapy had the same effects in rectal cancer as in colon cancer, in spite of much less evidence, and that these treatments belong to the standard of care [2]. Since then, adjuvant chemotherapy in rectal cancer has not been an issue in the USA, whereas it has in Europe [3–7].

During the celebration of the first 50 years of Acta Oncologica in 2013 [8,9], controversial issues were identified prior to a young scientist workshop with the aim to bring further light on the selected issues. In all areas, a systematic review of the literature should be performed. Adjuvant therapy of rectal cancer was one of those issues. Two similar reports also dealt with other controversial issues in the management of colorectal cancer [10,11]. A group of young scientists under the lead, in this case, of several senior scientists publish their systematic review in this issue of Acta Oncologica [12]. The authors conclude that for patients treated with preoperative (chemo)radiotherapy, available data do not support the use of adjuvant chemotherapy, whereas they do if the patients were not preoperatively treated. Different recommendations between guidelines, not only between USA and Europe but also within European countries [5–7], seem to depend on if the preoperative therapy and the effects of this are important for the choice and if the results from the colon cancer studies are transferable to rectal cancer patients.

Benford's law of controversy

In 1980, the US professor in astrophysics, Gregory Benford published the science fiction novel Time scape, containing an adage later known as Benfords’ law of controversy. It states that “passion is inversely proportional to the amount of real information available”. In other words, the less we know, the more room for controversies. The novel has received several awards, e.g. the Nebula Award and John W. Campbell Memorial Award. My own memories of reading the novel more than 30 years ago have faded, although after finding it in the book shell, I think I enjoyed reading it although I remember that some passages in the book did not bring the story forward and were thus not really needed. Scientists in 1998, or 18 years ahead of 1980 would give advice to scientists in 1962, or 18 years back in time from when the book was written, so that history could be changed. Among other things I remember in the book, John F. Kennedy did not die from the Lee Harvey Oswald shooting.

Level of evidence in colon and rectal cancer

Alluding to the Benford's law of controversy, the scientific level of evidence for beneficial effect in colon cancer is based upon multiple randomized trials. The evidence in colon cancer stage III has been summarized in many reviews and they uniformly tell that there is a 30–40% relative reduction in risk of recurrence and a 25% reduction in mortality using a fluoropyrimidine alone [1,13]. The addition of oxaliplatin reduces the risk of recurrence further by about 18–20% [14,15]. The trials in stage III have, however, not been subject to a systematic meta-analysis of all available evidence. This has been the case in stage II where the relative gain after systemic administration of 5-FU is comparably much smaller or in the order of 17% in disease-free survival (DFS) and 6% in overall survival (OS) [16]. Adjuvant!Online, one widely used web-based tool to assess the risk of recurrence and the gains from adjuvant therapy [17] uses in stage II risk reductions of 20% for DFS and 18% for OS and in stage III 42% and 38%, respectively. In rectal cancer fewer trials have been done and the results are not as clear as also stated in the review by Poulsen et al. [12] and in the recent editorial [18]. In rectal cancer patients chiefly treated without preoperative radiotherapy or chemoradiotherapy, there appears to be a gain numerically in between that in colon cancer stages II and III, or according to a Cochrane analysis [19], a 25% reduction in DFS (HR 0.75, 95% CI 0.68–0.83) and a 17% reduction in OS (HR 0.83, 95% CI 0.76–0.91).

During late 2014 and early 2015, the evidence has increased in rectal cancer thanks to an update of one large trial [20] and the reporting of results from three other trials [21–23]. That information is included in the overview by Poulsen et al. [12] as study data were reported as abstracts at scientific meetings, although the final publications were not available. The trial data has also already been subject to systematic overviews/meta-analysis, including one based upon individual patient data from four trials and 1196 patients [24,25]. The meta-analysis found no significant differences in OS (HR 0.91, 95% CI 0.81–1.17), DFS (HR 0.91, 95% CI 0.77–1.07 or distant recurrences (HR 0.94, 95% CI 0.78–1.14). The overall finding of no significant gain from adjuvant chemotherapy in rectal cancer after preoperative (chemo)radiotherapy thus remains, with a possible exception of high rectal cancers (10–15 cm from the anal verge or closest to colon) (HR for DFS 0.59, 95% CI 0.40–0.85) [24]. In one other meta-analysis, based upon published data (Bujko, K, pers. commun.), a gain was seen for all rectal cancers in the trials where randomization was done postoperatively. As not all patients randomized preoperatively, like in [20] will have radical surgery and are fit to start adjuvant chemotherapy postoperatively, any gain, if true, will be more difficult to show than if randomization was done postoperatively. Compliance with the adjuvant chemotherapy was also much lower in the trials with preoperative randomization (about 50% vs. 75%).

The knowledge base has thus increased, but not extensively since most of the trials were small and two of the four trials prematurely broken due to poor patient accrual [21,23]. Controversy will remain. Still we have to clinically act with these “statistically insignificant tendencies to small gains” in distant metastases risk, DFS and possibly OS, contradictory to those in colon cancer, when patients are informed and decisions made. This has been extensively discussed in the original articles, the meta-analyses and accompanying editorials [4,12,20–27].

An entirely different opinion, actually decreasing the controversy, is whether the gains in colon cancer are overemphasized. A benefit from adjuvant chemotherapy in colorectal cancer is thus present but it is not so large that it has been detected in the rectal cancer trials. I mentioned earlier that the results in colon cancer stage III has not been subject to systematic overviews, as, e.g. done by the Cochrane group, but limited to overviews of available trials, albeit some large within the ACCENT database [28]. Inclusion of “all” stage III trial patients may likely decrease the relative gain. In stage II, where a Cochrane-analysis has been done [16], the gain from adjuvant chemotherapy is statistically significant but not impressive [HR OS 0.94 (0.87–1.05) and recurrences HR 0.83 (0.75–0.91)] for prolonged systemically given biochemically modulated 5-FU.

Stage migration must modify are routines

I don't question the significant reduction in recurrences after surgery for colon cancer using adjuvant chemotherapy with either a fluoropyrimidine alone or with a fluoropyrimidine + oxaliplatin, but I am uncertain about the relative gain (overall and in the different substages). I am today even more uncertain about the absolute gain and whether it is worth it for many groups of patients. Due to markedly improved possibilities to stage, both pre- and postoperatively, and operate colorectal cancer patients together with the introduction of (national) guidelines, quality control of outcomes and multidisciplinary team discussions, a stage migration has likely occurred. The magnitude of this stage migration is poorly known [11,29,30], but means that the absolute risk of recurrence is much less than it was in the past when the adjuvant trials in colon cancer with a surgery alone group was run. It is possible that the relative gain from adjuvant chemotherapy is the same or even potentially slightly higher today than in the past, but the absolute gain is clearly less. For example, if the risk of recurrence for a particular stage like stage II with risk factors was 30%, it maybe 20% or even as low as 10% today [11]. With a relative gain for adjuvant chemotherapy of 20% (HR 0.80 like in stage II) this will mean that today two and not six patients per hundred will be rendered recurrence-free. Adding oxaliplatin would today decrease the risk by less than one patient per hundred as opposed to about 3–4 patients previously. The magnitude of a gain required for a patient to accept or the doctor to recommend adjuvant chemotherapy is not properly known. It may also have changed with time according to an investigation, where surgical and medical oncologists indicated that between 6% and 10% survival gain is the minimal percentage benefit that offsets the potential side effects due to treatment [31].

This situation is parallel to that seen in rectal cancer and the use of preoperative (chemo)radiotherapy. Better staging and more adequate surgery have resulted in less risk of local recurrences. The relative gain from preoperative radiotherapy is, if anything slightly better with better surgery (relative gains about 70% vs. 50%), but the absolute benefit is smaller [32,33] and by many (patients and doctors) not considered sufficient for use in many patients, considering the risk of late radiotherapy-induced consequences [34]. This is a natural consequence of development. Although it is said to be impossible to compare apples and pears, it is in my view more motivated to challenge the use of oxaliplatin in some patients in colon cancer stage III and most patients in colon cancer stage II than it is to omit preoperative short-course radiotherapy (or chemoradiotherapy) in intermediate risk (often designated locally advanced) rectal cancer [33]. The late neurological consequences are far from negligible after up to six months of oxaliplatin treatment. The risk maybe less if it turns out that three months of chemotherapy is not inferior to six months of chemotherapy [35].

Will the controversy be resolved?

Coming back to the question raised in the title, I don't think the question will be resolved by more trials with a no adjuvant treatment group after rectal cancer surgery, since experience has learnt that they are practically impossible to run, but rather that we reconsider the need for adjuvant chemotherapy in many colon cancer patients. Although molecular differences of relevance for chemotherapy sensitivity may exist between colon and rectal cancers, it is much more likely that in practice they react in the same way to adjuvant therapy. The trials in colon cancer were done when the recurrence risks were much larger than they are today, and it was thus methodologically easier to detect the benefit that likely is present in both colon and rectal cancers. As stated by Böckelman et al. [11] it is important to report absolute recurrence risks after colon cancer surgery for substages according to T, N, number of investigated nodes, vessel- and nerve invasion, differentiation, MSI and other relevant prognostic factors from unselected population-based materials where the quality of the medical care is up-to-date. The routine use of adjuvant chemotherapy for many groups will make the evaluations difficult, but not impossible [29].

Declaration of interest: The author reports no conflicts of interest. The author alone is responsible for the content and writing of the paper.