Keywords

1 Optimizing the Cost–Benefit Ratio of Treatment According to Evidence-Based Medicine

Since Hippocrates the aim of a curative medical treatment is to achieve the most efficient result against the disease and the less toxic effect for the patient. One the first curative treatments for rectal cancer was introduced by Miles in 1908 using “a radical abdomino-perineal resection” (APR) with an acceptable (although high) operative mortality. Since then surgery has been (and will remain) the cornerstone of the treatment of rectal cancer. The modern era of rectal surgery started with the introduction of the so-called “TME surgery” removing the mesorectum along the “holly plane” with sharp dissection under vision control (Heald and Ryall 1986). To improve local control and survival radiotherapy and chemotherapy have been used in association with surgery. Due to the many confounding factors, the results of such combined treatments can be evaluated only using randomized control trial. The ultimate aim is to reach 100 % cure with 0 % toxicity. Most of the new treatments aiming at better local control or survival use radiation dose escalation or more efficient multidrug medical treatments. The main limiting factor to this intensification is the induced toxicity. It is the merit of the “TME surgery” to be at the same time able to achieve a better local control by reducing the breaching of the rectal fascia and a lower toxicity by sparing the latero-pelvic nerves. It is probably the advantage of the laparoscopic approach to reach similar results by reducing further the operative health constraint for the patient (Panis et al. 2011). May be one of the most significant progress in the past decades impacting survival was the dramatic reduction in the rate of operative mortality. Intensive care, improved anesthesia, reduction of radiation toxicity with smaller irradiated volume, and better surgical bleeding and infection control have reduced the 60 days postoperative mortality from 10 % to close to 1 %. Only in elderly, frail patients surgery is remaining a significant trauma (Rutten et al. 2008).

When analyzing the benefits of the various combined multimodality treatments associated to surgery it is crucial to take into consideration the two aspects of the balance (benefit vs. cost) and to include in the cost all the sustainable aspects relevant to the patient, the healthcare system, and the society. Most probably in this subtle equilibrium between benefit and cost, toxicity is the main parameter because it is, since Hippocrates again, the key ethical message of medicine: “Primum non nocere” which in modern language may be assimilated to the “Principe de précaution.” In such a complex situation and as the improvements are generally (and unfortunately) small only RCT can bring reliable scientific evidence, which remains the best way leading to changes of practice in the medical community.

2 Recent Results Gained Through Randomized Trials

2.1 Quite Modest Even if Local Recurrence Rate Is Now Below 7 %

Until the end of the twentieth century radical surgery followed by adjuvant chemoradiotherapy (CRT) was the standard treatment for rectal cancer (stage M0) (Conference 1990; Krook et al. 1991; O’Connell et al. 1994) (Table 1). It was the merit of the Swedish and mainly German trials (Folkesson et al. 2005; Pahlman and Glimelius 1990; Frykholm et al. 1993; Sauer et al. 2004, 2012) to demonstrate that neoadjuvant CRT (nCRT) was more efficient and possibly less toxic than adjuvant CRT (Park et al. 2011). As “TME surgery” was introduced in early 2000, it was one of the main conclusion of the Dutch trial to show that even with a “TME surgery” preoperative radiotherapy (short course) was improving local control (Kapiteijn et al. 2001; van Gijn et al. 2011). Other more recent trials demonstrated that nCRT was more efficient for long-term local control than radiotherapy alone (Bosset et al. 2006, 2014; Gerard et al. 2006), that capecitabine was as efficient as Fluorouracile (5FU) (Gerard et al. 2010, 2012; Aschele et al. 2011; Schmoll et al 2013; Roh et al. 2011) that radiation dose escalation using external beam radiotherapy (EBRT) from 45 Gy/5 weeks up to 50 Gy/5w was producing more pathologic sterilization of the tumor (ypCR) without increase in 3-year toxicity, but without other clinical significant benefits (Gerard et al. 2012) and that oxaliplatine was not to be given concurrently with EBRT and Capecitabine (or 5FU) (Schmoll et al. 2013; Roh et al. 2011). As local control is at the present time close to 95 % in T2-3 M0 tumors the only way to improve survival is to find an efficient (and not too toxic) medical treatment. First-line chemotherapy has been proven possible without compromising nRT and surgery (Fernandez-Martos et al. 2010; Chua et al. 2010). With increasing knowledge about the molecular abnormalities driving cell growth, immune reaction, and tumor proliferation various molecular targeted drugs (MTD) have been tested. So far the results have not been totally convincing using as neoadjuvant treatment either anti EGFR MDT (Dewdney et al. 2012) or anti-VEGF concurrently with radiotherapy. Such MDT can provide sometimes increased toxicity (Crane et al. 2003) or decreased radiosensitivity (Machiels et al. 2007; Willett et al. 2009).

Table 1 Overview of some of the main messages and conclusion derived from recent randomized control trials
Full size table

2.2 Does Neoadjuvant Treatment Reduce the Rate of Permanent Stoma? Surprisingly NO

One of the most common medical beliefs is that preoperative treatment especially nCRT with long interval will “downsize” the tumor (T2-T3-T4) and increase the chance of a conservative treatment, namely sphincter saving surgery (SSS) using either low anterior resection (LAR) or inter sphincteric resection (ISR) (Rullier et al. 2013). In fact, conservative treatment of rectal cancer is a very complex situation with the interaction of many multifactorial parameters related to the tumor, the patient, the surgeon, and the general culture of a specific country or area. Here, more than everywhere else, only randomized trials can give strong evidence regarding the benefit of any preoperative treatment in terms of conservative treatment. Two literature reviews have analyzed this question (Bujko et al. 2006; Gerard et al. 2012). Both authors came to the same conclusion: for T2-3 (4) rectal cancers the rate of permanent stoma (for distal and middle rectum) have dramatically decreased during the past decades from 70 to 25 % (and sometimes 10 %), but this increase in sphincter preservation was due to technical surgical innovation and new concepts regarding the distal margin to be respected (from 5 to 2 even 1 cm) (Pahlman et al. 2013). In none of the randomized trials the group using the experimental treatment showed a significant increase in the rate of SSS despite often an increase in sterilization (ypCR) of the operative specimen (Fig. 1). A recent randomized trial performed in South Korea (Park et al. 2011) compared (as in the German CAO/ARO trial) postoperative CRT with nCRT. Despite a highly significant difference in the rate of ypCR (0 vs. 20 %) there was a nonsignificant increase in the rate of sphinter saving surgery (62 vs. 70 %). In a Nordic trial (Braendengen et al. 2008; Braendengen et al. 2011) nCRT when compared to short course RT with immediate surgery for T4 tumor was able to increase R0 surgery, local control, and sphincter preservation.

Fig. 1
figure 1

Forest plot summarizing the results of recent randomized trials about sphincter preservation and showing the lack of benefit of neo adjuvant treatment (with the exception of the Lyon R 96-2 trial)

Full size image

2.3 The Clinical Complete Response Hypothesis

The Lyon R96-2 trial using sphincter preservation as the main endpoint was the only trial showing a benefit of the neoadjuvant treatment to improve the rate of conservative treatment (Table 2). The addition of a boost using contact Brachytherapy X Ray 50 kV (CBX) 90 Gy in 3 fractions to EBRT increased the SSS rate from 44 to 76 % without increase in toxicity and preservation of a good bowel function (Gerard et al. 2004). Two points were of interest in this trial: first the rate of clinical complete response (CCR) was increased in the CBX boost group from 2 to 29 % and this finding may explain why the surgeons were more in favor of a conservative approach in the boost group. When the surgeon see only a partial response he does not modify his surgical initial decision, but when he cannot see or palpate anymore a lesion he may reappraise his decision and perform a more conservative technique (Ortholan et al. 2006). Second, not only Anterior Resection was more frequent in the boost group, but also most of all “organ preservation” as 10 patients out of 45 in the CBX group were able to preserve the whole rectum after CCR using either a local excision or only a careful surveillance (Watch and wait). These data has been updated after 10 years of follow-up and the gain in stoma-free rate was maintained on the long-term without detrimental effect on local control or survival (Ortholan et al. 2012).

Table 2 Lyon R96-02 randomized trial 1996–2001 (Géard 2004; Ortholan et al. 2012)
Full size table

2.4 How to Increase Safely the Rate of CCR to Perform More Conservative Treatment

There are mainly three ways to increase the CCR rate:

  1. (1)

    Increase the interval between the end of the neoadjuvant treatment and the surgery: the Lyon R 90-1 trial compared an interval of 2 versus 6 weeks. An increase in ypCR was observed (2 vs. 13 % p: 0.005), but it did not translate in a better SSS rate (Francois et al. 1999; Glehen et al. 2003). In Sao Paulo, Habr Gama has been for many years a strong advocate of a conservative treatment in case of CCR after nCRT. By increasing the interval before evaluation of the response from 5 to 12 weeks habr gama was able to increase the rate of CCR from 30 to 55 % (Habr-Gama et al. 2009; Habr-Gama et al. 2014). In the stockholm trial after short course an interval of 5 weeks (compared to immediate surgery) increased the ypCR without difference in toxicity, but without increase in SSS (Pettersson et al. 2010, 2012). It is probably after 2 to 3 months after the end of the nCRT that the optimal tumor response can be seen (Sloothaak et al. 2013; Wang et al. 2005).

  2. (2)

    Concurrent use of one or two chemotherapy with radiotherapy: The FFCD 9203 and EORTC 22921 trials have shown that the addition of 5FU to preoperative irradiation is increasing ypCR and most of all local control in T3-4 rectal cancer, but without gain in SSS (Bosset et al. 2006; Gerard et al. 2006). The addition of oxaliplatin to 5FU or Capecitabine is not adding any benefit to the patient and may increase the risk of diarrhea (Gerard et al. 2010; Aschele et al. 2011; Schmoll et al. 2013).

  3. (3)

    Radiation dose escalation is probably the most efficient way so far. In Toronto the dose escalation from 40 to 45 until 50 Gy was associated with a progressive increase of ypCR from 10 to 19 % (Wiltshire et al. 2006). In the ACCORD 12 trial a biological dose escalation of 15 % (from 45 to 50 Gy with the same protraction time of 5 weeks) increased the ypCR from 13 to 19 % (Gerard et al. 2010). The main limitation of such dose escalation, even with modern RT technique as IMRT or Proton therapy is the tolerance of the normal pelvic tissues and OAR (Gerard et al. 2004, 2003). The most efficient way to escalate the RT dose with regards to the “Toxicity/Benefit” ratio is using endocavitary irradiation, which can concentrate the dose to the primary tumor without arming too severely the surrounding OAR. With HDR Iridium combined with EBRT it is possible to achieve in T3 tumors a CCR rate of 70 % (Vuong et al. 2007). The randomized trial performed in Danemark (Jakobsen et al. 2006, 2012) despite an increase in ypCR in the group treated with Ir HDR did not show an increase in SSS and lead to more toxicity. Same findings in a phase III trial in Pakistan (Tunio et al. 2010). On the other hand as previously reported a safe dose escalation using CBX 50 kV was able to significantly increase CCR, sphincter saving, and most of all organ preservation. Unfortunately, this trial performed in a single institution in a limited number of patient has not influenced clinical practices (Gerard et al. 2004; Ortholan et al. 2012).

3 Aim of the Ongoing and Upcoming Randomized Trials

The most relevant and standard endpoint of CRT is overall survival, but in practice this endpoint is seldom used because it requires to include more than 1,000 patients and a very long follow-up. Disease Free Survival at 3 years is often the main endpoint of trial aiming at increasing survival. Other endpoints as ypCR, TRS (tumor regression score) (Taylor et al. 2011; Patel et al. 2011; Nougaret et al. 2013), R0 surgery may be used, but none can be considered as a robust surrogate endpoint of overall survival (Methy et al. 2010). Toxicity, rate of organ preservation, quality of life, and bowel or sexual functions are always major endpoints (Table 3). From a pragmatic point of view neo or adjuvant treatments are aiming at improved four clinical objectives:

Table 3 Ongoing or up coming randomized control trials in operable T2-3-4 Nx M0 rectal cancer
Full size table
  1. (1)

    Local Control In T2 T3 tumors it will be difficult to demonstrate a further improvement over 95 % local control. In T4 following the trial of Braendegen (Braendengen et al. 2008, 2011), the GRECCAR 4 trial (EUDRACT N° 1234556…) is comparing a standard radiation dose of 50 Gy combined with capecitabine to 60 Gy dose in a reduced boost volume. Different techniques of radiotherapy are tested to increase the tumor dose without increasing toxicity. Proton therapy if financially available is a promising technology (Thariat et al. 2013) to be used for rectal adenocarcinoma, which require dose above 90 Gy for 80–90 % rate of ypCR (Appelt et al. 2013).

  2. (2)

    Survival As none of the medical regimens tested in RCT has so far been able to increase survival (in opposition with colon cancers), various trials are testing different drugs combination and strategy to try to reduce the rate of distant metastases without detrimental effect on the local control and the overall toxicity rates. The Swedish RAPIDO trial is using a first-line chemotherapy with a short course (5 × 5 gy) versus a standard long course CRT in T3c-d T4 M0 tumors. The French Prodige 23 is comparing in T3T4 M0 a standard “Cap 50” regimen versus the same regimen preceded by four cycles of first-line chemotherapy using a Folfirinox regimen. The British Aristotle RCT is comparing Cap 45 versus the same treatment with the addition of concurrent irinotecan. With the growing development of targeted treatment toward specific molecular pathways some RCT intend to select patients according to some specific mutations and to use an adapted MTD.

  3. (3)

    Toxicity and constraint reduction in the French ACCORD 12 trial it was observed that in patients over 70 years of age the rate of treatment interruption, surgery not performed and postoperative toxicity was significantly increased using Capox 50 (or Cap 50) when compared to younger patients (Francois et al. 2014). A new Prodige randomized trial is upcoming who will test the hypothesis that after 70 years of age a short course radiotherapy (5 × 5 Gy in a small posterior pelvic volume) with delayed surgery will be better tolerated than the Standard Cap 50 and will lead to more patients able to undergo with reduced toxicity a TME surgery. No randomized trial has ever proved that in “goodT3 tumors” a surgery alone was as efficient as CRT (or 5 × 5 Gy) to achieve an optimal local control. Some institutions with a highly dedicated colorectal team tend to expect that TME alone can be proposed for these “good T3 tumors”. In Greccar 4 these patients staged with MRI are treated with a first-line chemotherapy using a Folfirinox regimen. In case of “good response” judged on MRI the patients are randomized between a standard Cap 50 nCRT or a TME without any a nCRT. So far no RCT is comparing the standard nCRT strategy versus TME surgery first-line. In MSKCC New York, following a phase II study using first-line Folfox and Bevacizumab with a 25 % ypCR in T3 tumors (Schrag et al. 2014) an upcoming phase III trial is testing the hypothesis that combined first-line chemotherapy and MTD may replace n CRT. It is not sure that replacing the toxicity of radiotherapy by the toxicity (cardio-vascular) of this new approach will benefit the patient.

    Following all recent trials capecitabine (oral) is replacing 5FU (iv), which is for the patients a benefit in terms of simplicity and toxic risk.

  4. (4)

    Conservative treatment This is possibly one of the most promising hypothesis. How to increase not so much Sphincter Saving using AR or ISR (Rullier et al. 2013), but organ preservation after CCR? The upcoming European trial OPERA (Organ Preservation for Early Rectal Adenocarcinoma) will include T2 T3a-b and after Cap 45 will compare a boost using EBRT (5.4 Gy) versus a boost using CBX (90 Gy). The hypothesis is that taking advantage of an increase in CCR in the CBX group the organ preservation rate will increase from 5 to 25 %. One still controversial question in case of CCR is to decide between local excision as proposed by Lezocche (Lezoche et al. 2012) or Garcia-Aguilar et al.(2012) (Sauer et al. 2004) and the GRECCAR group (Rullier et al. 2013) or close surveillance (Habr-Gama et al. 2014; Perez et al. 2013; Maas et al. 2011).

4 Conclusion

Important improvements have been made in the past decade in the treatment of rectal cancers and its overall prognosis is now slightly better than colon cancer. Chemotherapy and new molecular targeted drugs should be able to improve survival in the future. Robust prognostic and predictive markers will be necessary to tailor and optimize these “targeted” treatments for each individual patients. The growing development of colorectal screening will lead to the discovery of more early rectal cancers. In these patients and especially when elderly and frail, organ preservation (as for squamous cell carcinoma of the anal canal) should be an important step forward to better quality of life. Only well-conducted randomized control trials will bring strong enough evidence to influence and modify the clinical and surgical practices.