Recent advances in (chemo-)radiation therapy for rectal cancer: a comprehensive review

Rectal cancer represents approximately one-third of all colorectal cancer with the second highest incidence and the second highest cause of cancer death in the western society [1]. Considering the restricted role of radiotherapy in the treatment of colon disease, we’ll focus our review mostly on rectal cancer where radiotherapy has a leading position in combination with both surgery and chemotherapy.

During the last 3 decades, the role of radiation therapy in the management of locally advanced rectal cancers, has been gradually modified. Starting in the ‘80 s with a prevalent adjuvant role due to its potential in reducing pelvic recurrence after surgical resection and increasing survival rates when combined with 5-FU based chemotherapy [2], radiotherapy was challenged, in the early ‘90 s, with the introduction of total mesorectal excision (TME) that significantly decreased locoregional recurrence (LRR) by itself, questioning the necessity of radiotherapy before or after surgery [3]. Several short course (5 Gy × 5 days) randomized trials [4‐9] have demonstrated the importance of preoperative RT plus TME in reducing LRR, in stage II and III rectal cancer patients. The assumption that adding chemotherapy to long course (45–50 Gy) preoperative radiotherapy could increase the local effect of radiotherapy, led to the comparison between radiotherapy and radiochemotherapy as neoadjuvant regimen [10]. The addition of concomitant chemotherapy to preoperative radiotherapy resulted in a significant increase in local control while only slightly increasing acute toxicity, without affecting adherence to radiotherapy, feasibility of surgery (with no increase of postoperative morbitidy), or adherence to adjuvant chemotherapy. However, no significant improvement in overall survival was observed in any single trial.

Over the time, the availability of different treatment options (including radiotherapy and chemotherapy) and the possibility to use different regimens (pre- and postoperative), has resulted in an increasing demand of reliable preoperative staging. Different imaging techniques have been used to locally stage rectal cancer with variable sensitivities and specificities [11]. High-resolution MRI has been shown to be superior to clinical examination, computer tomography and endoluminal ultrasound (EUS) for rectal cancer staging [12]. The possibility to have more accurate information related to the pelvic structures as the possibility to distinguish a tumor from rectal wall, to depict the mesorectal fascia [13], to identify anatomical structures useful to support an optimal surgical technique [14] and to better characterize suspicious lymph nodes [15], made MRI the principal imaging technique in the assessment of a rectal cancer. Based on MRI imaging, able to identify poor prognostic factors preoperatively, it was possible to divide rectal cancer patients into three groups (“good”, “bad” and “ugly”), according to their local and systemic failure’s risks [16].

For “good” tumors, surgery alone is the mainstay of treatment. Only for tumors located in the distal rectum, radiochemotherapy can be considered with a neo-adjuvant/definitive intent to increase either sphincter preservation or achieve organ preservation by omission of surgery or a combination with local excision in selected cases.

Considering patients having “bad” MRI features, neoadjuvant treatment has been established to reduce both the risks of LLR and distant metastases. Two different regimens have been tested in those patients: conventional long-course radiochemotherapy (LCRT: 45–50 Gy with 1.8–2 Gy fractions over 5–6 weeks), mostly used in South Europe and in the United States and short-course radiotherapy (SCRT: 5 Gy × 5 fractions) without preoperative chemotherapy, mostly used in the North of Europe. Several studies [7‐9, 17, 18] have investigated the two regimens in the past, even if the enrolled population was not completely comparable considering that the SCRT regimens included patients with early tumor (stage T1–T2 and some resectable T3), while LCRT studies considered mainly more locally advanced rectal cancer patients (T3, T4 and unresectable tumors). Although LCRT was expected to have advantages of higher sphincter preservation and lower complication rates, several phase III randomized studies [19, 20] have found no difference in oncological outcomes (DFS, OS, local relapse-free survival). However, LCRT schedule showed higher pathological complete response (pCR) rate and clear resection margin. Similar results were obtained in the study of Ngan et al. [21] with a trend of better local control rate of LCRT in distal rectal lesions. Stating that there is not a real evidence to recommend one treatment modality over the other, the results of Ngan et al. justify the common practice to treat a tumor that is located in the distal rectum, close to the anal sphincter and/or locally advanced cT4 or CRM-positive tumors with a LCRT schedule.

In order to test if the longer interval between the end of radiotherapy and surgery was responsible for the higher pCR rate of the LC schedule, a SCRT regimen with delayed surgery (between 6 and 8 weeks after radiotherapy) was tested. A higher pCR rate was reported for delayed as compared with immediate surgery after SCRT [22]. Similar oncological outcomes were observed for early and delayed SCRT, with a higher acute radiation toxicity rate in the delayed and a significantly higher rate of postoperative complications in the early surgery group.

Considering the essential improvement in LC reached with the modern neoadjuvant treatments and surgical TME technique, the reduction of treatment-related side-effects and postoperative complications is now a priority. In LCRT and SCRT with delay, a 6–8 weeks break after RT is considered standard, with a higher surgical morbidity when surgery is delayed for 11 weeks after LCRT according to a randomized trial [23], although this finding could not be confirmed in other prospective trials [24, 25]

For the “ugly” group, characterized by a high risk for local recurrence and distant metastasis, neoadjuvant LCRT is recommended. To optimize the treatment, systemic chemotherapy can be administered either before or after neoadjuvant LCRT/SCRT, referred to as Total neoadjuvant therapy (TNT) [26].

Finally, the benefit of adjuvant chemotherapy in patients with locally advanced rectal cancer treated with neoadjuvant SCRT or LCRT is still highly controversial [27].

Although RT is accepted as an essential component of multidisciplinary treatment (MDT), specific issues still remain unaddressed. It’s well defined that neoadjuvant LCRT followed by TME surgery is recommended for locally advanced rectal cancer, but what about early rectal cancer located in the lower rectum? And more, what is the role of the TNT and the chemotherapy intensification in rectal cancer? This review will recall the controversial issues and analyze the recent advances in the radiation therapy field.

During the last decades, an increasing interest in intensified treatment has been paid, mainly focused on locally advanced and metastatic rectal cancer. Using the standard chemoradiation approach, only about 11–18% of patients will achieve a pathological complete remission (pCR) [28‐33]. Because this small group of patients shows a clearly improved overall prognosis compared to patients with less or no response [34], several strategies have been explored to improve the pCR rate or even omit surgery in selected cases. Those include the use of more intensive chemotherapy regimens concurrent to radiation, addition of targeting agents to concurrent chemoradiation, escalation of radiation dose or the use of altered fractionations, and the sequential use of (chemo)radiation and (intensified) induction or consolidation chemotherapy regimes in the neoadjuvant setting (total neoadjuvant therapy, TNT).

The highest level of evidence exists for adding oxaliplatin to standard 5-FU based neoadjuvant chemoradiation. Results from six phase III trials ([28‐33, 35, 36], Table 1) addressing this issue have been published so far. Although two reported significantly increased pCR rates [28, 33] and one significantly improved disease-free-survival (DFS) [28], all others failed to show any significant improvements in major oncological endpoints while reporting increased toxicities [29‐32, 35]. Taken the slightly different treatment schedules into account, there might be a (small) benefit for the addition of (dose-dense) oxaliplatin, however its efficacy does not seem to be high enough as a sole strategy. Nevertheless, adding oxaliplatin to concurrent chemoradiation might be still considered in patients with urgent need to downsizing or embedded in TNT- or non-operative management (NOM)-approaches.

Irinotecan was also tested early as an adjunct to standard 5-FU based chemoradiation based on its known activity in metastatic colorectal cancer [37]. Several phase I-II trials and one phase III trial ([38‐56], Table 2) evaluating different dose schedules have reported conflicting results with pCR rates of 10–38% (weighted average 24%). While the only phase III trial showed a significantly improved pCR rate at the cost of increased toxicity [55], the only randomized phase II trial failed to show any benefit of the addition of irinotecan to standard 5-FU based chemoradiation [51, 52]. Moreover, another randomized phase II trial comparing the addition of Irinotecan or Oxaliplatin to standard chemoradiation reported a pCR rate of only 12% with Irinotecan but 23% with Oxaliplatin with similar toxicity [49, 50]. There is some evidence that patients with certain UGT1A1 genotypes may respond better to irinotecan-based therapies [53‐55], which may allow a better patient selection in the future. Until then, the results from the ongoing British phase III ARISTOTLE trial are awaited.

Bevacizumab, a monoclonal antibody targeting vascular epithelial growth factor (VEGF), is part of most current standard first-line multidrug regimens used in metastatic colorectal cancer [57, 58]. Therefore it has been evaluated as an adjunct to 5-FU or 5-FU/Oxaliplatin based chemoradiation in numerous phase I and II studies for rectal cancer ([59‐77], Table 3). Reported pCR rates range from 8 to 40% with a weighted average of 19%. The only randomized phase II trial [75] found a small but significant benefit in terms of pCR compared to standard chemoradiation with no significant increase in acute side effects or postoperative morbidity. However, several others have reported high rates of severe acute toxicities and postoperative complications mainly in terms of impaired or delayed wound healing [64, 69] which led to delayed or omitted adjuvant chemotherapy in a high percentage of patients. Therefore Bevacizumab does not seem to be an ideal candidate for treatment intensification at least in neoadjuvant settings outside a NOM approach.

Cetuximab, a monoclonal antibody targeting the Epidermal growth factor (EGFR), is part of the current standard multidrug regimen in metastatic KRAS wild-type colorectal cancer [57, 58]. Therefore it (as well as other anti-EGRF antibodies like Panitumumab or downstream tyrosinkinase inhibitors like Gefitinib) has been evaluated in combination with standard 5-FU based-chemoradiation in various phase I and II trials ([78‐95], Table 4) also for rectal cancer. Although most of them showed a modest toxicity profile, results in terms of pCR rates were mainly disappointing (pCR rates 0–27%, weighted average 14%). Two randomized phase II trials comparing Capecitabine- and CAPOX-based chemoradiation with or without anti-EGFR agents did not observe a significant benefit for their addition [88, 89]. Moreover, neither KRAS status nor EGFR-expression seems a robust predictor of pCR [88].

Several other agents have been tested in phase I or II trials concurrent to chemoradiation based on more or less robust preclinical and/or clinical evidence either for their activity in colorectal cancer or for enhancing radiation effects. Those include classic chemotherapy agents like Cisplatin [96], Mitomycin C [97], or Temozolomide [98], COX-2-inhibitors like Celecoxib [99‐101], proteasome inhibitors like Bortezomib [102], PI3K-akt-inhibitors like Nelfinavir [103], phosphatidylserine-antibodies like Bavituximab [104], Multi-Tyrosine-Kinase-Inhibitors like Sorafenib [105, 106], PARP-Inhibitors like Veliparib [107] and fusion proteins like Aflibercept [108] (listed in Table 5). Reported pCR rates varied from 7 to 39%. Although the addition of some agents resulted in promising pCR rates with acceptable toxicities, these findings should be interpreted as preliminary and further research is warranted.

Another possibility of improving chemoradiation effects is simply to increase radiation dose. Since Appelt et al. [109] provided clear evidence for a dose–response relationship between 50,4 and 70 Gy dependent on pretreatment T- and N-category [109], various prospective observational and phase I/II trial have evaluated dose escalation in the mentioned range within different concurrent chemotherapy regimens ([51, 52, 99, 110‐142], Table 6). Dose escalation was achieved by either adding more fractions in conventional fractionation, using altered fractionation regimes or by adding a brachytherapy boost. Reported pCR rates ranged from 0 to 50% with a weighted average of 22% (excluding the population based trial [137]). Mohiuddin et al. [51, 52] reported a comparative study using conventional fractionation to either standard dose (45–50 Gy) or escalated dose (55–60 Gy) and observed a significantly increased pCR rate with dose escalation (13% vs 44%). Regarding brachytherapy boosts, one phase II trial with a matched cohort found a significant increase in pCR rates (12% vs 29%) [134] while a Danish phase III trial did not observe a significant difference in pCR rates [128]. As pCR necessitates complete remission of primary tumor and lymph nodes (with the latter usually not affected by brachytherapy), approaches using external beam techniques for dose escalation seem more meaningful. Of note, none of the mentioned trials reported excessive grade 3+ late toxicity (0–11%). Therefore, further evaluation of moderate radiation dose escalation in larger trials seems to be one reasonable strategy to improve pCR rates.

All of the mentioned strategies aimed at enhancing either the chemo- or the radiation part during concurrent chemoradiation and therefore allowed only moderate escalations due to concerns of toxicity. However, dose-intense combination chemotherapy regimens alone may result in considerable rates of downstaging and pCR rates as indicated by several studies [143]. Therefore, it seems reasonable to combine chemoradiation with sequential dose-intense combination chemotherapy in the neoadjuvant setting to improve pCR rates (known as TNT). Similarly to other diseases, this should result also in enhanced treatment compliance compared to adjuvant chemotherapy and further targets the unsolved problem of high distant metastases rates in rectal cancer by early initiation of systemic treatment. Several trials have already reported encouraging results using different schedules of sequential radio(chemo)therapy and combination chemotherapy ([20, 24, 25, 64, 88, 144‐162], Table 7). Reported pCR rates ranged from 14 to 37% (weighted average 21%) in the TNT arms compared to 11–25% (weighted average 14%) in the standard chemoradiation arms of the comparative studies, indicating the superiority of the TNT approach. Moreover, the largest randomized trials observed significant benefits in terms of disease-free in the TNT arms mainly attributed to a reduction of distant failures [24, 156, 160‐162], although using slightly different approaches. The Timing of Rectal Cancer Response to Chemoradiation Consortium in the United States [24, 156] performed a sequential cohort phase II study including 259 patients with T3/4 or nodal positive patients. All received upfront long-course chemoradiation (50 Gy with 5-FU c.i.) and were sequentially scheduled to receive either no or 2–6 cycles of mFOLFOX6 consolidation chemotherapy prior to surgery [24]. Chemotherapy was completed postoperatively aiming at similar total numbers of chemotherapy cycles for all four arms. The pCR rate significantly increased with the number of consolidation chemotherapy cycles from 18% (none) to 38% (6 cycles) [24]. Three-year DFS rates were also significantly increased for all TNT arms compared to the standard arm, although it has to be noted that the mean number of total chemotherapy cycles was lower in the standard arm [156]. The Polish group [157] conducted a phase III trial randomizing 515 patients with fixed T3 or T4 tumors to long-course chemoradiation (50, 4 Gy with 5-FU, leucovorin and, partly, oxalipatin) or to short-course radiation (5 × 5 Gy) followed by 3 cycles of consolidation chemotherapy with FOLFOX prior to surgery [157]. They observed no significant differences in R0-resections rates (primary endpoint), pCR rates or DFS. The significant OS benefit at 3 years (73% vs 65%) [157] disappeared with longer-follow-up [20]. The RAPIDO group [160, 161] used a similar approach randomizing 911 patients with high risk rectal cancer (defined as cT4, cN2, EMVI+, MRF+ or positive lateral nodes) to either long-course chemoradiation (50,4 Gy + Capecitabine) or 5 × 5 Gy followed by six cycles of consolidation chemotherapy with CAPOX or nine cycles of FOLFOX [160]. They found significantly improved pCR rates (28% vs 14%) favoring the TNT arm, which came at the cost of significantly increased acute grade 3+ toxicity (48% vs 25%) [160, 161]. Moreover, they described a significant benefit for the TNT arm in terms of disease-related treatment-failure (24% vs 30%) [161]. Finally, the French Group [162] tested TNT using induction chemotherapy in 461 patients with T3/T4 lesions. The patients either received long-course chemoradiation (50 Gy + Capecitabine) followed by surgery and adjuvant chemotherapy (12 × mFOLFOX6 or 8 × Capox) or induction chemotherapy with 6 cycles of mFOLFIRINOX followed by chemoradiation, surgery and less intensive adjuvant chemotherapy (6x mFOLFOX6 or 4x CAPOX) [162]. Similar to the RAPIDO trial, they described significantly improved pCR rates (28% vs 12%) and 3-year-DFS rates (76% vs 69%) for the TNT arm [162].

Regarding the timing of chemoradiation and chemotherapy, both possible approaches (induction or consolidation chemotherapy) reached comparable results in terms of pCR rates and survival. The German CAO/ARO/AIO-12 trial [25], which directly compared induction and consolidation chemotherapy strictly using the same schedules during chemotherapy as well as chemoradiation in both arms, and thus achieving the same time interval from treatment start to surgery, found a non-significant but distinct difference in pCR rates favoring the consolidation arm (17% vs 25%). This might be explained by the longer time interval from chemoradiation to surgery, although the randomized GRECCAR-6 trial (without consolidation chemotherapy) was not able to confirm such an association [163].

In summary, several strategies to improve the pCR rate by neoadjuvant treatment intensification currently exist with the TNT approach probably being the most promising as it targets not only pCR rate but also seems to reduce distant failure rates with improved treatment compliance and acceptable toxicity. Although the most recent phase III trials (RAPIDO, PRODIGE 23) have been published only in abstract form so far, TNT will probably be the new standard of care for high-risk rectal cancer patients with the detailed treatment algorithm regarding to different subgroups yet to be defined.

Combination of the TNT approach with intensification of the concurrent treatment phase by moderate radiation dose escalation might be a reasonable further direction of research especially if non-operative management strategies (as addressed in the following part) are taken into account.

The role of radiation therapy and the further direction of treatment optimization have changed dramatically in the recent decade. The initial problem of high local recurrences rates after surgery, which prompted the introduction of radiation therapy into the treatment algorithm, seems to be solved for the vast majority of patients. Due to improvements in imaging and surgical technique we are now facing the contrast of evaluating the omission of radiation therapy in patients with low risk for local recurrence or the intensification of neoadjuvant (radiation therapy containing) approaches aiming at the improvement of oncological outcome in high risk patients or even attempting rates of complete remissions which allow non-operative organ-preserving approaches. Several strategies for treatment intensification exist, with the TNT approach seeming the most promising option due to its improvements not only in pCR and downstaging rates, but also in reduction of distant failures without distinct increases in treatment-related toxicity compared to standard chemoradiation or SCRT with adjuvant chemotherapy. Limited (organ-preserving) surgery in patients with good response to neoadjuvant CRT seems possible in selected patients but its role needs to be clarified in further trials. Preliminary results do further suggest, that organ-preserving non-operative approaches in patients with clinical complete responses to neoadjuvant therapy are safe if proper selection and meticulous follow-up examinations are performed. Future directions may include the development of tools for a more precise identification of suitable subgroups and/or the prediction of response as well as further treatment intensification by combining TNT approaches with intensified concurrent chemoradiation approaches.