Short-course radiation followed by mFOLFOX-6 plus avelumab for locally-advanced rectal adenocarcinoma

Rectal cancer continues to take a high toll in morbidity and mortality worldwide [1, 2]. In 2017, an estimated 39,910 new rectal cancer cases were diagnosed in the US [3]. The incidence of rectal cancer in the European Union is between 15 and 25 cases in a population of 100,000, and around 33% of these end up with death each year [4]. Over the last 20 years, stage II and III rectal cancers have shown a relatively steady 5-year OS of approximately 65% [5]. Despite the fact that the widespread use of screening has resulted in earlier identification and treatment of premalignant lesions and a decrease in incidence of rectal cancer [6], the SEER data estimated that by 2030 the colorectal incidence rate for the age group between 20 and 34 years will increase by 124.2% based on the previous colorectal data of cases under 50 years of age for the period between 1974 to 2010 of colorectal cancer [7].

Current standard of care for LARCs supports the results of the 5-year German (CAO/ARO/AIO) 94 trial and entails a multidisciplinary approach, where oxaliplatin was added to the preoperative CRT and adjuvant chemotherapy regimen, even though the concurrent use of oxaliplatin with chemoradiation is not the standard practice. This approach led to induce tumor regression, to increase the surgical negative margins (R0), and to decrease the local recurrence risk rate [8]. However, the EORTC 22921 trial follow up results showed no significant effect of adjuvant chemotherapy on both DFS and OS rates, adding the low compliance rate to chemotherapy with 82% pre-operatively and just 42.9% post-operatively [9, 10].

As a result of the low compliance and high complication rates observed with adjuvant chemotherapy, recent shifts toward a total neoadjuvant approach, where neoadjuvant instead of adjuvant chemotherapy is administered in an attempt to increase the pCR rate, which is of major prognostic impact in rectal cancer [11].

The recent concept of TNT in locally advanced rectal cancer was first tested in a clinical trial by Chau et al. which showed an 88% objective tumor control rate with neoadjuvant capecitabine/oxaliplatin [12, 13]. Recent trials have also explored this concept with results revealing promising pCR rates using TNT. Garcia-Aguilar et al. concluded in a phase II trial that patients who received 6 cycles of mFOLFOX6 between CRT and TME procedure resulted in an increase in pCR, reaching 38%, with better compliance to chemotherapy and without a significant increase in surgical complications compared to other groups who received less or no chemotherapy cycles in between [14]. Cercek et al. compared 320 patients receiving regular CRT to 308 patients receiving TNT, the CR rate was 36% in the latter group as compared to 21% in the first one [15]. Moreover, Bahadoer et al. showed in the randomized RAPIDO trial that the TNT regimen is superior to the standard preoperative CRT and surgery in pCR rates [16]. Furthermore, Garcia-Aguilar et al. showed in the ORPA trial that the TNT regimen can form the basis of organ preservation in rectal cancer management [17]. These substantial rates of tumor regression and pCR suggest that it’s more favorable to use of this treatment modality i.e., Chemotherapy and CRT before planned surgery of LARC. This could be secondary to the better compliance to chemotherapy experienced with TNT and to the > 8 weeks-delay between radiotheraoy and surgery, that increased the odds of pCR [18]. Findings of these studies provided additional support for the NCCN guidelines that classify TNT as an acceptable treatment modality for rectal cancer [19].

Although preoperative CRT and neoadjuvant chemotherapy followed by TME seems promising in locally advanced rectal cancer patients, the superiority in OS of short course radiation followed by neoadjuvant chemotherapy over preoperative CRT was not maintained after 8 years [20]. outcomes remain poor with a 3-year DFS of approximately 50%. Also, although this strategy greatly decreases the risk of loco-regional recurrence, the fact that late development of distant metastatic tumor spread is still common, requires innovative systemic strategies to overcome disease progression. For instance, while local recurrence rates for locally advanced rectal cancers were stable at 5 to 6% following the combination strategy used, distant recurrence rates were found to be around 25%, and metastases were considered the main cause of death [21, 22].

Conventionally, oxaliplatin was known to induce DNA damage and crosslinks ultimately leading to apoptosis. Furthermore, in mice, oxaliplatin was found to stimulate an increase in cytotoxic T cells and support an anti-tumor immune response. The immunogenic cell death is the interaction between chemotherapy and immune response, where oxaliplatin induces a change in the tumor microenvironment through activating cytotoxic T cells and cancer cells expression of MHC class I and immune checkpoint molecules. This leads to the recognition of cancer cells by the immune response [23, 24]. Moreover, adding 5-FU to oxaliplatin augment anticancer immune response through deactivating the myeloid derived suppressor cells that in turn downregulate tumor cell growth and angiogenesis [25].

In a multicenter phase II trial, Marco et al. demonstrated that delaying the chemoradiation-to-surgery period and increasing the number of neoadjuvant chemotherapy cycles caused an increase in the pCR rates [26]. Furthermore, a recent phase III randomized trial explored the effect of delaying surgery beyond 8 weeks from finishing radiation on pCR rates in LARC patients. A significant increase in the pCR rate with delayed TME (> 8 weeks from RT) was noted, reaching 28.6%, in comparison to 10% in the other group (< 8 weeks from RT) [27]. Finally, in the Stockholm III trial, short-course radiotherapy with delayed surgery resulted in no significant difference in locoregional disease control, pCR rates, distant recurrence and OS in comparison to long-course radiotherapy [28]. This advocates the use of SCRT and to benefit from the abscopal effect. The abscopal effect is the elimination of distant metastatic tumor cells by an immune response triggered by targeting the tumor with radiation locally. It was hypothesized that the destruction of tumor cells by radiation causes the release of tumor antigens that induce an immune response locally and distally [29]. Moreover, a correlation was noted between radiation dose per fraction and the strength of the abscopal effect. In mice, Camphausen et al. showed that 50 Gy in 5 fractions had a stronger abscopal effect than 24 Gy in 12 fractions [30]. In another study, 16 Gy in 2 fractions caused a higher increase in interferon levels and PD-L1 expression in comparison to 20 Gy in 10 fractions. This advocates for the use of high dose per fraction (hypofractionated) radiation with immune checkpoint inhibitors [31].

In an attempt to further increase the pCR and improve outcomes in rectal cancer patients, we proposed to investigate a novel approach of combining immune checkpoint inhibitors with systemic chemotherapy after SCRT. Baeten et al. showed that short course radiotherapy was significantly superior to long course chemoradiation in increasing cytotoxic T cells in the tumor biopsy post treatment in patients with rectal cancer [32]. Furthermore, chemoradiation was found to increase PD-L1 expression, in the tumor and its invasive margin, in patients with rectal adenocarcinoma [33]. As a result of this increase in both, PD-L1 and its PD-1 receptor can be utilized as a potential therapeutic target of immunotherapy through proposing to use PD-1/PD-L1-inhibiting drugs.

Preliminary pharmacokinetic and clinical safety data from cohorts of phase I trials, the phase II pivotal trial (EMR100070–003), and the ongoing phase III trials support the efficacy of avelumab in combination with other treatment such as Axitinib in Renal Cell carcinoma (RCC). In addition, Avelumab is now approved in Merckel-Cell carcinoma (MCC) and studies are still ongoing in Lung cancer. Responses with such combinations were observed early during treatment and appeared to be durable in nature, lasting for > 1 year in several of the cohorts [34, 35]. This further supports the idea of combining PD-1/PD-L1 inhibitors with CRT for improved clinical efficacy in the management of rectal cancer patients.

Based on the above, we initiated this phase II clinical study to evaluate the pCR rate following 5 fractions of SCRT (for a total of 25 Gy), followed by 6 cycles of mFOLFOX-6 chemotherapy combined with avelumab immunotherapy (given every 2 weeks), and then TME procedure in patients with LARC.

Combination therapy with preoperative CRT, followed by TME procedure, and then adjuvant chemotherapy comprises the cornerstone of treatment in stages II and III rectal cancer. Although this has resulted in an improvement in tumor regression and a decreasein the risk of local recurrence [44], use of adjuvant chemotherapy has been accompanied by low compliance rates, manifested by patients not receiving or completing the planned treatment.

Many phase II and ongoing phase III trials adapted the strategy of administering neoadjuvant, rather than adjuvant, chemotherapy, and results have revealed marked improvement in pCR in TNT arms when compared to the regular CRT arms, reaching up to 38% in some cases [8, 14, 15, 21]. In addition to the results of the randomized RAPIDO trial that showed a doubling in the pCR rates with the TNT regimen in comparison to preoperative CRT [16]. This led to recognition of TNT as a valid treatment approach in rectal cancer by the NCCN [19].

Based on our experience with TNT at AUBMC, outcomes seem promising. Of the total 16 patients who underwent TNT since 2016, 10 have received 3–6 cycles of FOLFOX followed by long course CRT (Capecitabine + XRT 28 fractions), 3 have received 3–6 cycles of XELOX followed by long course CRT (Capecitabibne + XRT 28 fractions), 2 have received short course CRT followed by 6 cycles of FOLFOX, and 1 has received short course IMRT (5 fractions) followed by 2 cycles of XELOX. 44% of patients (7 of 16) had complete response (CR), 12% of patients (2 of 16) had down-staging from T3N1M0 to pT1-2 N0 and T4bN1M0 to pT1-2 N0, respectively, and 44% of patients (7 of 16) had incomplete response. In addition, only 25% of patients (4 of 16) had relapse later on.

Although TNT seems promising in LARC patients, the POLISH II trial after 8 years showed no superiority of TNT over CRT in survival and pCR rates. This may be due to the initial design of the study where patients only received 3 cycles of neoadjuvant chemotherapy [20]. Moreover, Fokas et al. showed in a randomized trial that sequence of TNT affects the clinical outcomes of patients, where CRT followed by neoadjuvant chemotherapy results in a higher pCR rate [45]. In an attempt to further increase the pCR and improve outcomes in rectal cancer patients, we proposed the use of a novel approach of combining immune checkpoint inhibitors with systemic chemotherapy after radiation which may overcome these limitations [46‐49]. We initiated this phase II clinical study to evaluate the pCR rate following SCRT (5 fractions), followed by mFOLFOX-6 chemotherapy combined with 10 mg/kg of avelumab immunotherapy (6 cycles), and then a TME procedure in patients with LARC.

Innovative studies aimed at testing the efficacy of using anti PD-1/PD-L1 in combination with CRT in cancer patients showed promising result [46‐49]. Indeed, PD-L1 expression and CD8+ TILs density significantly increased after neoadjuvant CRT in a matched comparison analysis of preoperative CRT-induced alterations on pre-CRT biopsy and post-CRT resected specimens of rectal cancer patients. It was also noted that patients whose PD-L1 expression was consistently high both, before and after CRT, experienced less elevation in CD8+ TILs compared to the rest of the groups [46‐48]. These results suggest that CRT causes an up-regulation of anticancer immunity by increasing the CD8+ TIL density. They also support the idea that this increase in TIL density can be the reason behind the elevation in PD-L1 expression in tumor cells [50].

In the light of the marked increase in both after CRT, PD-L1 and its PD-1 receptor can be viewed as potential therapeutic targets, and through proposing using PD-1/PD-L1-inhibiting drugs. This is supposed to enhance the long-term antitumor effects of therapy in cancer patients [46‐49], especially that the use of inhibitors of these cell surface markers in cholangiocarcinoma have played a potential therapeutic role in enhancing the immune killing of cancer cells by removing the inhibitory function of PD-1 on T cells [49]. Although previous trials involving the use of PD-1/PDL1 checkpoint inhibitors in colorectal cancer patients could not demonstrate a clear benefit [51], we rely on the fact that avelumab 10 mg/kg once every 2 weeks has demonstrated meaningful clinical activity across various treatment settings and tumor types, including MCC and RCC. For instance, preliminary pharmacokinetic and clinical safety data from cohorts of phase I trials, the phase II pivotal trial (EMR100070–003), and the ongoing phase III trials support the use of this dose of avelumab in combination with other treatment options, with responses being observed early during treatment and appearing to be durable in nature, lasting for > 1 year in several of the cohorts [34, 35]. As a result, a dose of 10 mg/kg IV once every 2 weeks was considered to have a favorable risk benefit profile for our present study.

With the promising results from trials that evaluate the combination of radiation therapy and immunotherapy in many cancer types, comes the question of the choice of when to start immunotherapy with respect to radiation therapy. Most reported cases of abscopal effects happened when radiation therapy was given either with or after cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) blockade [52‐55]. Data from the ProHA-TRAMP model reveals that the up-regulation of anti-tumor effects was most optimal when a tumor vaccine was given within a narrow window of 3–5 weeks after radiation therapy [56, 57]. Of note, while radiation therapy-induced immunogenic cell death occurs within 1–3 days in breast tumor cells, anti-tumor immunity to melanoma is improved when anti-CTLA-4 antibody precedes radiation therapy [52, 58, 59]. In our study protocol, immunotherapy will be concomitantly administered with mFOLFOX6 2 weeks from the start of SCRT as we assume that this provides enough time for the expected upregulation in PDL-1 expression and TIL densities to occur and synergize with immunotherapy.

Previous studies stressed on the role of RT as an immune adjuvant [60‐63]. For instance, CRT has been shown to increase the tumor antigen load, their related receptor molecules, and danger-related signal molecules as part of a long cascade of immune responses culminating in immunogenic cell death [64]. By increasing tumor cell lysis at the level of the localized treatment site, high doses of radiation result in the release of tumor-associated antigens (TAA) in the process. These antigens transported by antigen presenting cells (APCs) [65, 66], that once activated by pro-inflammatory cytokines, migrate to tumor-draining lymph nodes to activate cytotoxic T-lymphocytes (CTL) and mobilize them against tumor cells [67]. Indeed, abscopal effect is defined as the ability of radiation delivered to a local site to minimize or eradicate metastases at distant sites, outside the scope of the localized treatment [29]. This nonspecific eradication of distant tumors and metastases could be related to the increase in the levels of pro-inflammatory cytokines and chemokines released, the immune cells and tumor tissues in the system, following exposure to radiation [66].

Based on the results of several trials (Polish and Stockholm III) [20, 28], SCRT showed equivalent results to CRT for loco-regional control beside its potential immunogenicity. Also, current NCCN guidelines for the management of locally-advanced disease support the use of SCRT as a valid option, especially with data revealing tumor down-staging with the short-course [19, 68].

As for the timing of the TME procedure, Akgun et al. showed that randomized patients LARC localized within 12 cm of the anal verge into, an arm undergoing TME within 8 weeks and another arm undergoing the procedure after 8 weeks following CRT, showed improved stage regression and pCR when the interval between CRT and surgery exceeded 8 weeks [27]. Similarly, results from another study that randomized patients into two arms, with 6- and 12-week waiting periods, are available in an abstract form and show significantly higher pCR rates in the latter group [69]. Findings suggesting higher pCR and stage regression rates in long-interval groups are expected, especially that there is more time for the biological effects of radiation therapy to occur. For instance, despite the fact that DNA is damaged during irradiation, tumor cell lysis does not occur until weeks after irradiation [70‐72]. By planning the TME procedure to be performed around 16 weeks after the start of radiation therapy, we give enough time for these effects to take place.

Our primary endpoint is to evaluate the pCR rate following pre-operative treatment, especially with recent data suggesting it as being a surrogate for DFS in locally-advanced rectal cancer [73]. This will enable us to test the hypothesis that the addition of avelumab to mFOLFOX-6 chemotherapy, administered following SCRT for locally-advanced rectal cancer, will improve the post-operative outcomes of the disease.

The management of rectal cancer continues to be challenging. Despite the plethora of evidence suggesting significant improvement in local recurrence rates in rectal cancer patients with advances in preoperative chemotherapy and surgery, distant metastases continue to represent a major problem. In fact, the combined statistical analysis data from five European randomized controlled trials showed that the 5-year distant metastasis rate was 30.8% in 2759 recruited patients [22]. We hope that with the regimen suggested in this protocol, a significant improvement in pCR and outcome will be obtained.