Is there a need for neoadjuvant short-course radiotherapy in T3 rectal cancer with positive lymph node involvement? A single-center retrospective cohort study

Local recurrence was a major problem in rectal cancer before the introduction of the total mesorectal excision (TME). Swedish rectal cancer trial showed that neoadjuvant short-course radiotherapy (RT) reduced the local recurrence risk and improved survival in patients operated on before the adoption of the TME technique [1, 2]. Dutch trial found that short-course radiotherapy reduced the risk of local recurrence when combined with TME surgery, however, without any beneficial effect on survival [3‐5]. It was reported that there was 10.6% local recurrence risk in stage III disease after short-course RT combined with TME and 20.6% local recurrence risk in those treated with TME alone [4]. On the other hand, a local recurrence rate of 4% with meticulous TME surgery alone has been reported and no difference in local recurrence rate between stage II and III disease was noted in that series [6‐8]. Efforts have also been made to recognize the so-called “good” prognosis T3 cancers and separate them from “poor” prognosis T3 cancers [8, 9]. However, we do not know if the patients with ugly T3 cancer benefit from short-course RT.

Short-course RT can cause harm. It increases the leakage risk of colorectal anastomosis, and it can cause bowel, urinary, and sexual dysfunction [10‐14]. It is also suspected to have a higher incidence of secondary malignancies after radiotherapy [15].

At the present, there are remarkable differences in preoperative treatment strategies of rectal cancer in European countries [16]; Sweden and the Netherlands use neoadjuvant short-course RT most often, but in Norway and Denmark, short-course RT is little used as a treatment option. Despite this, there are no remarkable differences between the oncological treatment results of these countries.

The treatment strategy of rectal cancer in our unit is very similar to that of Sweden; most patients with T3 tumors and a suspicion of positive lymph node involvement receive short-course neoadjuvant RT. However, diagnosis of positive lymph node involvement in rectal MRI is difficult, as shown by high-quality units [8], and some patients are not suitable for RT for other reasons; therefore, there are still substantial numbers of patients with pT3N1-2 rectal cancer whom have not undergone short-course RT preoperatively.

The first aim of this study was to evaluate the effect of short-course RT on local recurrence risk in patients that have pT3N1-2 rectal cancer in pathological specimen. The second aim was to compare risk factors between patients that had short-course neoadjuvant RT before TME surgery and those that had TME surgery alone.

We found that neoadjuvant short-course RT was not a preventive factor for local recurrence in patients operated for pT3N1-2M0 rectal cancer in our dataset. Anyhow, all surgical procedures executed in the present study used the TME-technique, which is fundamental for the adequate dissection in a rectal cancer operation [7], and has better local control compared to blunt dissection [18]. Our data comprised only of patients with stage III diseases assessed by pathology report postoperatively. The total local recurrence rate for all the patients was 11.3%, for the RT plus surgery group 9.6% and for the surgery only group 14.0%. Thus, these findings are in line with a previously published local recurrence rate [19] and the difference between groups was not significant. A Dutch trial found that the local recurrence rate at 5 years was 9.3% for the RT plus surgery group and 19.4% for the surgery only group in stage III disease [20]. A Norwegian study, reporting the results of T3 rectal cancers treated with curative intention with TME without neoadjuvant treatment, observed a 3-year local recurrence rate to be about 18% for T3N1 tumors and about 24% for T3N2 tumors [9]. Possibly, the quality of TME in the early phase of that study, i.e., when TME was novel, in some of the centers might not have been at as high level as would be the case nowadays. Moreover, a series that included patients operated on before and after the adoption of the TME technique found the local recurrence risk for T3N1-2 patients to range between 34.5 and 42.9% [21].

The median time to local recurrence in our series seemed to be shorter in the surgery only group than for the RT plus surgery group (2.0 vs. 2.8 years). A similar pattern was also reported in the Dutch TME trial [20]. However, more patients in the RT plus surgery group in our study received adjuvant chemotherapy than in the surgery only group (87% vs. 61%), which may also have delayed the emergence of local recurrence. In our unit, patients with evidence of positive lymph node involvement in their specimen are evaluated by an oncologist for possible suitable adjuvant treatment, but not everyone receives chemotherapy because of patient-related factors, such as fragileness or prolonged recovery or refusal of such treatment by some patients.

Several studies have tried to separate good prognosis T3 tumors from poor prognosis T3 disease, both before and after TME era [8, 9, 21]. Those T3 tumors, predicted to have less than 5 mm spread from the muscularis propria in MRI, have a minor risk for local recurrence after surgery without RT, regardless of lymph node involvement status [8, 21]. In the early days of this study period, the T3 tumors were not divided into four subclasses by an MRI radiologist; thus, we do not know the distribution to rT3a-d subclasses. However, there were more tumor perforations in the non-irradiated group, which suggests that tumors in this group were locally more advanced.

In our study, 38% of pT3N1-2 patients did not receive neoadjuvant RT, even though our treatment strategy has been to irradiate T3 tumors when there is suspicion of lymph node involvement upon MRI. There are several reasons for this. First, the MRI diagnosis of lymph node involvement is difficult. Only half of the “good” prognosis patients with positive lymph node involvement in the MERCURY study had been evaluated to have had positive lymph node involvement in the preoperative MRI [8]. Second, some of our patients had received previous pelvic irradiation for other reasons and the neoadjuvant RT for rectal cancer was thereby contraindicated. Third, some other patient-related factors, such as difficult dementia or otherwise poor cooperation, may have been a reason not to offer RT.

The main risk factors for local recurrence in this specific group, pT3N1-2M0 rectal cancer in our analysis, were as follows: involved CRM, tumor distance under 6 cm from anal verge, and tumor perforation. All these risk factors for local recurrence have already been acknowledged in many studies [3, 9, 22] and also in the data of our earlier published study [17]. A Dutch study reported the risk factors for local recurrence to be stage IV disease, T4 tumor, positive CRM in T3 or T4 tumors, and also N2 disease [20]. The preoperative RT did not prevent local recurrence in our study patients with stage III disease.

One reason for the critical evaluation of the usefulness of preoperative RT in this setting is its possible toxicity. Low anterior resection syndrome and poor functional outcome occur more often in patients who receive RT compared to patients whom have had surgery only [10, 13]. A recovery to daily activity is also slower after having received RT. Sexual dysfunction is a problem, as well [11, 12, 23]. Radiotherapy is found to be a risk factor for anastomotic leakage, which in itself may prevent adjuvant chemotherapy and also adversely affect bowel function [14]. The wound healing after abdominoperineal excision (APE) is slower after RT than without it [3, 24]. More secondary malignancies are reported to occur after RT (9.5% compared surgery only 4.3%) in irradiated areas [15]. More than 50% of patients report long-term side effects after having RT [23].

In the light of the data available including those of the present study, the efficacy of neoadjuvant short-course RT in T3N+ rectal cancer is uncertain and needs to be studied more. Radiotherapy decreased the amount of local recurrences before the TME era [1]. During the learning phases of TME surgery soon after its adoption, RT still had a beneficial effect on the local recurrence rate of tumors with positive lymph node involvement, at least in multicenter studies [3, 4]. It is also possible that multicenter studies [3, 4] also included centers of lower technical quality. However, other more current studies showed that even lower recurrence rates can be achieved by using meticulous TME surgery alone and the results were not poorer for lymph node positive disease, at least in good prognosis T3 tumors [7, 8]. The adoption of neoadjuvant short-course RT in rectal cancer treatment has taken place within the same time period as TME technique, and MRI-directed multidisciplinary teams in designing the treatment for each patient have become a standard, both of which have been shown to improve oncological results [2, 18, 22, 25]. T3c-d tumors seem to have poorer prognosis in terms of local recurrence, but it is still not known whether that subgroup of tumors would eventually benefit from preoperative short-course RT.

Our current data on pT3N1-2 tumor outcome suggests that the routine use of neoadjuvant short-course RT in T3N+ rectal cancer patients may not be justifiable for all patients. However, patient number in our study was small and patient groups may not be quite equivalent due to retrospective nature of the study. Thus, the possible beneficial effect of neoadjuvant RT may not manifest. Further, due to limited number of patients in our study, different T3 subgroups (T3a–c) as well as N1 and N2 patients could not be separately studied. Thus, some subgroup possibly having advantage of short-course radiotherapy could not be recognized.

In conclusion, neoadjuvant short-course RT did not beneficially affect the local recurrence rate of pT3N1-2M0 rectal cancer in our retrospective series. Since short-course RT also has potential side effects, its role as routine neoadjuvant treatment in T3N+ disease should be critically considered. However, further studies with larger patient numbers on pT3b-c rectal cancer and in N1 and N2 disease are needed to evaluate the value of neoadjuvant therapy in these particular groups.