Rectal cancer is one of the more frequent human malignancies accounting for 3.9% of all new cancer cases causing 3.2% of all cancer-related deaths globally in 2018 [1]. The survival of the disease increased in the last years reaching an overall 5-year rate of 67% [2]. This could be related to the medical advances in rectal cancer treatment, increased awareness and screening [3].

Surgery is still the mainstay of treatment for rectal cancer since the introduction of abdomino-perineal resection techniques in 1908 [4]. Since then advances progressed the surgical platform and total mesorectal excision (TME) became the standard surgical approach ever since its introduction by Heald and Ryall [5] in 1982.

Neoadjuvant and adjuvant therapies have been repeatedly tested in many trials to improve the outcomes of rectal cancer patients in terms of local recurrence and survival. The addition of radiotherapy (RT), in addition to the improvement in surgical techniques, demonstrated improvement in local recurrence [6,7]. Despite that, other studies reported that around one-fifth of the cases will have no response to RT. Other workers reported that the contribution of neoadjuvant RT to pathological complete response in terms of survival was of limited or minimal benefit [8].

In view of the above changes, the real challenge in rectal cancer treatment is to identify patients who would fully benefit from the new treatment protocols which involve RT. The aim of this article is to present the different histological systems used to classify the response rate of the patients and to discuss the different types of biomarkers used to predict of the response of patients to RT/chemoradiotherapy (CRT).

Oxidative stress damage is extensive in cells exposed to ionising radiation. This damage mainly targets nuclear DNA through inhibiting DNA transcription in both direct and indirect pathways [9]. DNA damage is caused by the direct targeting nuclear chromatin. This damage will find the way through generating intra- and inter-strand chromosomal cross linkages along with breaks and mutations of DNA material. Direct damage to plasma membrane happens by disrupting both electrical gradient and the rigidity of the two layers of phospholipids membranes affecting the cell integrity. The indirect pathway is through free radicals from the ionisation of H₂O₂ molecules [10].

Altering the coordinated repair of DNA processes that repair damage induced by ionising radiation improves outcome. Cell repair, at low radiation level, can accommodate injuries such as double strand breaks and with increasing level the inflicted damage on the cell promote the cells to enter apoptosis and eventual cell death. This sensitivity to the dose of radiation that inflicts these injuries differs from tumor to tumor and the surrounding non-tumorous tissues [11]. Cells with high turnover are typically the first to respond to radiotherapy treatment thus fulfilling the aim of RT which is to eradicate or at least debulk the tumor so it becomes more amenable to surgical excision [12].

The addition of chemotherapy to the treatment plan has additional effects on cell’s response to radiation and this combined treatment leads to either delay or abolition of the reparative process. These chemotherapeutic agents will target the cell cycle at more sensitive phase [13]. The flurouracil-based neoadjuvant CRT (nCRT) was further improved by adding oxaliplatin. This regimen significantly improved disease-free survival of patients with stages cT3-4 or cN1-2 rectal cancer in comparison to the older regimen [14].

In 2004, Habr-Gama et al. [15] have pioneered the concept of “wait and see” for rectal cancer as they have found significant number of patients who have received CRT responded very well and staged as complete pathological responders. As those patients have an excellent prognosis according to Habr-Gama et al. [15], they proposed that carefully selected rectal cancer patients responding to CRT should have a nonsurgical management. Despite the argument against wait and see [16] which stated that the rationale of a “wait and see” policy when complete clinical response status is achieved relies on retrospective observations, and data are currently insufficient to support this policy except in patients who are recognized to be unfit for or refuse radical surgery. The “wait and see” policy however is gaining ground and many centres in the world are following the Sao Paulo regime with great success as it has shown excellent outcomes in 25%–40% of carefully selected and staged patients [17].

The challenge we are facing now is how to accurately predict pathological complete response (pCR) to change the management plan before radical surgery. An increased interest in assessment of nCRT response is well documented, but this did not show much success yet. In the “watch-and-wait” approach, patients deemed to have achieved clinical complete response (cCR) must meet strict inclusion criteria. This includes digital rectal examination, direct visualization by colonoscopy, and biopsying suspicious lesions in the irradiated tumor bed [18]. Difficulty in distinguishing fibrosis from residual tumor and the distribution of residual rectal cancer within different layers of bowel wall after nCRT poses a significant difficulty in assessing complete response [19]. The use of positron emission tomography (PET) as an addition for the clinical assessment of cCR is a refinement that improved the overall accuracy from 91% to 96% [20].

In 1989, a prospective study of 186 rectal carcinoma patients was carried out, in which, 97 were randomized to surgery alone and 89 to surgery with preoperative radiotherapy [21]. The study showed that radiation under-staged the disease biologically but over graded it histologically in the sense that the appearances of individual cells may appear more bizarre after CRT than before treatment. It also suggested that any clinical pathology staging should include the presence or absence of radiation. To that extent the minimum dataset (MDS) which is used for all cancers in UK stipulated that the form should state whether the tumor received radiation therapy and if so whether there is response. The response is determined by looking into the volume of the neoplasm versus fibrosis. The MDS categorise the response rate as follows: no residual tumor/mucous lakes only, minimal residual tumor or no marked regression [22].

Over the years, many systems for tumor regression grading with various scoring parameters were developed. The first was developed in 1994 by Mandard et al. [23], it was designed to assess tumor response in oesophageal cancer patients treated with nCRT. They used a five-tier system to classify regression from score 1 (no tumor cells; complete regression) to score 5 (no regression). The main advantage of the Mandard’s Tumor Regression Grade (TRG) system is comparing the proportion of the residual cancer cells to the amount of fibrosis. This was considered simple and was shown to be a reproducible method; 3 years later it was modified to colorectal cancer specimens by Dworak et al. [24] in 1997. They classified regression into five grades but the other way around from 0 (no regression) to 4 (total regression). In 2002, Wheeler et al. [25] devised a three-point grading system called rectal cancer regression grade, where grade 1 indicated complete response, grade 2 showed marked fibrosis despite persistent microscopic disease, and grade 3 showed no response with little or no fibrosis taking into consideration the macroscopic features of the specimen. This was later modified by Bateman et al. [26], in which both extremes were preserved; complete replacement of viable cancer cells by fibrosis and cancer cells persistence without fibrotic changes but subtle changes were added for groups of near-complete responders. Cutoffs of 0%–5%, 5%–50%, and >50% residual tumor were adopted. Although several grading systems for tumor response have been devised, a three-point TRG has been shown to provide good inter-observer reproducibility compared to the five grades, but both systems were shown to provide similar prognostic significance. The three-point system devised in Ryan et al. [27] has been shown to be reproducible and easy to use, with good inter-observer agreement. The most recently published classification by the Royal College of Pathologists dataset guidelines for colorectal cancer (2017) and the College of American Pathologists based on the AJCC/UICC, 8th edition [28] recommend a four-tier system with modification to the one described by Ryan et al. [27]. This is considered the best methodology and provided excellent inter-observer reproducibility compared with five-grade systems. It also showed similar prognostic significance. Table 1 summarizes staging systems by Wheeler et al. [25], Mandard et al. [23], and AJCC/UICC 8th edition. More recently however a group from South Korea compared Ryan et al. [27]’s, AJCC and modified Dworak staging systems and found that the modified Dworak (which included the assessment of mural tumor and the lymph node status) as a best system to reflect recurrence and disease free survival [29].

Mucin pools are a common feature during histopathological assessment of specimens treated with nCRT but the presence of acellular mucin following nCRT does not have a significant impact on patient outcome. Only the presence of malignant cells in the specimen is the considered criteria for staging of tumor following preoperative treatment. Haemorrhage, necrosis, inflammation and acellular mucin are not considered as part of staging process, they are considered only for studying regression [22].

In a study by Fokas et al. [30] in 2018, the prognostic role and surrogacy of the neoadjuvant rectal (NAR) score was examined for cT, ypT and ypN categories. NAR scores were categorised into three degrees; low when the score is below 8, intermediate (score between 8 and 16) and high in scores over 16 based on the observed scores in a clinical trial. In the previous study, NAR scores were capable of predicting treatment effects on clinical outcome and overall survival with lower NAR scores having better cumulative incidence of disease free survival, overall survival, and distant metastasis [30].

The commonest modalities for RT are short course (SC) followed by surgery after 5 days, SC with extended interval, long course (LC) combined with chemotherapy and the contact radiotherapy. Contact radiotherapy, also known as Papillon treatment, is a modality of radiotherapy given from inside the rectum in low doses reaching only few millimetres into the tissue. SC is given as 25 Gy over 5 days with a daily dose of 5 Gy. LC is given as a total dose of 45–50 Gy in 25–28 fractions [31].

Comparing SC to LC radiotherapy, both had similar outcomes in terms of sphincter preservation, local recurrences, late toxicity, the overall survival, disease-free survival and quality of life. However, pCR was higher in patients receiving LC of RT [32,33]. There is no uniformity of using SC or LC as practice varies widely but in many centres the use of LC is reserved when the circumferential resection margin (CRM) is threatened, otherwise SC is the preferred treatment. In other centres they only use SC or LC alone. Now preoperative RT or CRT is mostly used as it is less toxic and more efficient than postoperative RT [34]. There is very little experience on the histological changes after contact therapy or SC with extended interval.

Some tumors resist the treatment. Furthermore clinical, histopathological and radiological profiles have been unreliable in predicting which tumors would be responsive to the treatment. With regards to radiological modalities, some studies tried utilizing MRI TRG to assess patients preoperatively in order to modify the treatment strategy. They clustered the cases into good responders in which surgery may be delayed to avoid the associated mortality and morbidity whereas poor responders are advised to undergo additional treatment to help down-staging and early treatment systemic relapse risk [43]. Despite that, there is no consensus on this modality requiring more robust studies. Therefore, it is essential to look for predictive biomarkers that are highly sensitive and specific to avoid unnecessarily giving RT treatment to unresponsive tumors. Molecular biomarkers are used in clinical practice to predict treatment response in various malignancies. Several studies have investigated molecular markers in rectal carcinoma, and their ability to predict a response to CRT.

miRNA and protein biomarker are promising as predictive biomarkers for CRT in rectal carcinoma. However, none of these biomarkers have been employed in routine clinical practice. A single biomarker is probably unlikely to have the ability to be predictive with high sensitivity and specificity. A biomarker panel, which could include genetic, epigenetic and protein biomarkers, in combination with clinicopathological and radiological data need to be evaluated together to reach a robust method of predicting therapy outcome, which in turn predicts the prognosis, and informs therapeutic approach.