Abstract
Background/Aim: The benefit of neoadjuvant (chemo) radiotherapy for locally advanced upper rectal tumors remains controversial. Thus, we aimed to evaluate the outcome of patients with stage II or-III upper rectal cancer undergoing neoadjuvant (chemo) radiotherapy followed by total mesorectal excision in our institution. Patients and Methods: From April 2004 to October 2019, all patients with stage II or III upper rectal cancer treated with neoadjuvant (chemo) radiotherapy followed by total mesorectal excision were identified from our database. Overall survival, progression-free survival, and local recurrence were assessed using the Kaplan–Meier method. Acute and late treatment-related toxicities were recorded according to the CTCAE-5 version. Results: The study group consisted of 106 patients. Respectively, 36% and 61% of patients had stage II and stage III upper rectal cancer. The median follow-up period was 4.4 ± 3.4 years. Five-year overall survival and progression-free survival were respectively 78% [95% confidence interval (CI)=69.2-88] and 76.8% (95%CI=68.4-86.2). The rate of local recurrence at 5 years was 3.78% (95%CI=0-7.98). Forty-two percent of patients presented early toxicities and 27.4% of patients experienced early surgical complications. Late toxicities and surgical complications occurred in 24.5% and 9.4% of patients, respectively. Conclusion: Neoadjuvant (chemo) radiotherapy followed by total mesorectal excision of stage II-III upper rectal cancer is effective and safe.
Rectal cancer is the 8th most common cancer worldwide with around 704,000 new cases every year (1). In the United States, rectal cancer is diagnosed in 35% of patients at a locally advanced stage (2). In this patient group, neoadjuvant (chemo) radiotherapy followed by total mesorectum excision (TME) has been established as the standard treatment by randomized trials (3-7).
Yet, whether it applies to all tumor locations is still debated. In several randomized trials, neoadjuvant (chemo) radiotherapy was associated with improved local control in low and middle locally advanced rectal cancer and led to higher sphincter preservation (5). However, there were contrasting results regarding upper rectal tumors (8, 9). Currently, neoadjuvant (chemo) radiotherapy for locally advanced upper rectal cancer is not recommended by current guidelines (10, 11). Nonetheless, it remains controversial and is used heterogeneously worldwide depending on cancer centers’ practice.
In our institution, patients with locally advanced upper rectal cancer are treated with preoperative (chemo) radiotherapy. We here report on our experience.
Patients and Methods
Patients. All patients, with pathologically proven rectal carcinoma and stage II-III rectal cancer treated with (chemo) radiotherapy followed by TME were included. In accordance with ESMO guidelines, upper rectal tumors were defined by a distance between the anal verge and the distal tumor end of 10 cm to 15 cm. Local staging was assessed in 98 (92.5%) patients, with magnetic resonance imaging (MRI) in 60.4% of patients and/or with endoscopy and endoscopic ultrasound (EUS) in 80.2% of patients. One hundred and two (96.2%) patients were staged for distant metastasis, mostly (87.7%) with a chest-abdomen-pelvic computed tomography (CAP CT). Staging was determined according to the 7th edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual (12).
Radiotherapy. Patients received long-course radiotherapy (45 Gy in 25 fractions/44 Gy in 22 fractions) or short-course radiotherapy (25 Gy in five fractions) to the mesorectum and pelvic lymph nodes. A number of patients received a radiation boost of 5.4-6 Gy following long-course radiotherapy.
Patients were treated with three-dimensional conformal radiation therapy (3DCT) or with intensity modulated radiation therapy (IMRT). Treatments were delivered through an Elekta Synergy® linear accelerator (Elekta AB, Stockholm, Sweden) or a TomoTherapy® treatment planning system (Accuray Incorporated, Sunnyvale, CA, USA).
Chemotherapy. Neoadjuvant chemotherapy was delivered concurrently with long course radiotherapy. Adjuvant chemotherapy was recommended for (y) pT4 tumors or in case of pathological lymph node involvement.
Surgery. All patients underwent TME after (chemo) radiotherapy. TME was performed one to two weeks after short course radiotherapy and six to eight weeks after long course (chemo) radiotherapy. Negative surgical margins notified as R0 were defined as a circumferential resection margin >1 mm.
Histopathological assessment. Tumor and/or lymph node downstaging were respectively defined as the reduction of at least one tumor/lymph node stage between the clinical assessment and pathological assessment.
Treatment and secondary effects assessment. As per our institution’s protocol, patients were assessed weekly during treatment. After treatment, patients were assessed every 3 months for the first two years and every 6 months after. Generally, CAP CT was performed every 6 months during the first two years, then yearly. Colonoscopy was performed every three years. If clinical suspicion of progression was present, the appropriate imaging was performed. Acute toxicities were evaluated during treatment and late toxicities, defined as the persistence of side effects ≥6 months after treatment, were assessed using the last clinical report available. All toxicities were graded according to the CTCAE-5 criteria.
Statistical analysis.
Data collection. The first radiotherapy day served as the start date for survival analyses. Additionally, progression, local or distant recurrence and death were retrieved from our database. We encountered 3 cases of incomplete data with accessible months and years but missing days and decided to convert the date to the middle of the month. One patient had accessible year but missing month and day, we choose to convert the date to the middle of the year and month.
Data analysis. All calculations were made using R (13) and more specifically the survival package. Every survival analysis was performed using the Kaplan–Meier method. In case the cause of death was not known, patients with missing information (n=3 out of 106) were not considered for cancer-specific survival analyses. Finally, predicted survival rates were given with a 95% confidence interval.
Results
Patient and treatment characteristics. From April 2004 to October 2019, 106 patients were treated with neoadjuvant (chemo) radiotherapy followed by total mesorectum excision for stage II-III rectal cancer. Mean age was 67.3 years, with rectal cases higher in men (62.3%) than women (37.7%). Nearly one-third (61.3%) of patients were diagnosed with stage III. Sixty-four (60.4%) patients received neoadjuvant chemoradiotherapy with the remainder receiving short-course radiotherapy (18.9%) or long-course radiotherapy without chemotherapy (20.8%). Out of 86 (81.1%) patients who received long-course radiotherapy, 53 (50.0%) patients received a radiation boost. The majority of patients (88.7%) were treated with 3DCT. More than half of patients (33%) received the CAPECITABIN regimen followed by the TEGAFUR–URACIL regimen (22%). Patients’ and treatment characteristics are described in Table I.
Patient, tumor, and treatment characteristics.
Postoperatively, 67.2% of patients who initially had positive lymph nodes experienced downstaging. Tumor downstaging occurred in 33.0% of patients. Pathological complete response was found in 7.5% of patients. Three (2.8%) patients had positive surgical margins and 39 (36.8%) patients had lymphovascular or/and perineural invasion. Twenty-six (24.5%) patients received adjuvant chemotherapy, either with the modified FOLFOX 6 regimen (n=22) or with the LV5FU2 regimen (n=3). Five (4.7%) patients did not receive adjuvant chemotherapy due to surgical complications (n=3) and chemotherapy-related cardiac side effects (n=2).
Early (chemo) radiotherapy side effects. Forty-four (41.5%) patients had early (chemo) radiotherapy-related side effects. The majority of patients (24%) had grade 1 side effects. Four (3.8%) patients experienced grade 3 side effects, in the form of diarrhea (n=2), bowel obstruction (n=1), and myocardial infarction (n=1). Diarrhea was the most common gastrointestinal side effect (90.3%). Chemotherapy-related side effects included leukopenia (n=1), thrombocytopenia (n=2), myocardial infarction (n=2) and fatigue (n=7). Early (chemo) radiotherapy side effects are detailed in Table II.
Early secondary effects.
Early surgical complications. Twenty-nine (27.4%) patients experienced early surgical complications. In 18 (17.0%) patients, surgical complications were limited to grade 2 according to the Clavien–Dindo classification (14). Seven (6.6%) patients had grade 3 complications and one (0.9%) patient had grade 4b complications. Two patients died following brain stroke and multiple organ failure. Early surgical complications are presented in Table II.
Late side effects and surgical complications. Sixteen (15.1%) patients did not undergo rectal reconstruction and eight patients had restoration of gastrointestinal continuity but no further assessment (n=4) or a follow-up ≤6 months (n=4). These patients were not taken into account for late side effects assessment. In patients with rectal reconstruction, late side effects occurred in 26 (24.5%) patients and were limited to digestive side effects. Late incontinence was present in 10 (9.4%) patients including fecal incontinence (n=8) and flatus incontinence (n=2). Late side effects are presented in Table III.
Late secondary effects.
Late surgical complications occurred in 10 (9.4%) patients, including anastomotic stenosis (n=4), fistula (n=4), anastomotic leakage (n=1) and abscess (n=1). Late surgical complications are presented in Table III.
Survival outcomes. At the time of analysis, out of 19 patients with disease recurrence, two (1.9%) had local recurrence, 15 (14.2%) had metastatic recurrence and two (1.9%) had both local and metastatic recurrence. Most recurrences occurred in patients with stage III rectal cancer (89.5%) rather than patients with stage II (10.5%) rectal cancer. Out of 17 patients with cN2 disease, 10 (58.8%) experienced recurrence. In terms of mortality, twenty-seven (25.5%) deaths occurred. Thirteen (12.3%) were cancer-related or treatment-related, 11 (10.4%) were associated with other causes, and three (2.8%) were not specified. Five-year overall survival and progression-free survival were, respectively, 78% [69.2-88] and 76.8% [68.4-86.2] with a median follow-up time of 4.47 (±3.43) and 3.92 (±3.2) years, respectively. Removing unknown cause of deaths, five-year cancer-specific survival was 85.5% [77.2-94.6] with a median follow-up time of 4.51 (±3.41) years. Local recurrence at 5 years was 3.82% [0-7.98] and 7.67% [0-15.7] at 10 years. Population analyzed was N=106 except for the study of cancer-specific survival (N=103) due to missing data. Overall survival and local recurrence are presented in Figure 1.
Overall survival (OS) and local recurrence (LR) among 106 patients who underwent preoperative (chemo) radiotherapy for locally advanced upper rectal cancer. The 5-year OS and 5-year LR rates were 78% and 3.8%, respectively. The shaded area represents the 95% confidence intervals that were determined using the Kaplan–Meier method.
Discussion
Our study retrospectively evaluated neoadjuvant (chemo) radiotherapy for stage II-III upper rectal cancer. Our population was composed of a substantial proportion of patients with advanced stage and high risk factors (15). In comparison, the Swedish rectal cancer trial and the Dutch TME-trial included 31% and 35% of patients with stage III rectal cancer, respectively. In the MRC-CR7/NCIC-CTG-C016 trial (8), patients with stage III rectal cancer accounted for 40% in the preoperative group. A total of 7.5% patients achieved complete pathological response, a result similar to the German CAO/ARO/AIO-94 study (9%) (9).
In terms of survival outcome, our study showed excellent local control with a five-year local recurrence of 3.78% [0-7.98]. This rate is in accordance with the MRC-CR7/NCIC-CTG-C016 and the CAO/ARO/AIO-94 trials, which reported a five-year upper rectal cancer local recurrence of 4.7% and 2.5%, respectively. In addition, in a retrospective study, Huang et al. (16) found a five-year upper rectal cancer local recurrence rate of 8.6%.
Of note, our study found that a considerable number of patients that experienced progression were initially diagnosed with stage III and cN2 disease. Because this study was initiated before the results of the PRODIGE 23 and RAPIDO trials (17, 18), these patients did not receive total neoadjuvant therapy.
In the TME trial with a 12-year follow-up (19), patients with stage III rectal cancer, including upper rectal tumors, had an improved 10-year local recurrence with preoperative radiotherapy compared to TME only, while patients with stage II rectal cancer did not benefit from preoperative radiotherapy. In addition, when considering patients with a negative circumferential resection margin, 10-year overall survival was significantly higher in the preoperative radiotherapy group. Thus, we might argue that patients with stage III upper rectal cancer are likely to benefit more from preoperative chemotherapy compared to patients with stage II upper rectal tumors.
A few studies evaluated the rate of acute toxicities and acute surgical complications in stage II-III upper rectal cancer treated with preoperative (chemo) radiotherapy. In the Dutch TME-trial (20), which contained 30% of upper rectal tumors, digestive and genitourinary toxicities occurred in 14.9% and 2.8% of patients, respectively. However, all patients in this study were treated with short-course radiotherapy, which has been shown to produce less toxicities (6, 21, 22). Postoperatively, surgical complications included abscesses (5% of patients), fistulae (1%) and anastomotic leakage (11%). These results are overall in line with our study.
A number of trials have reported on late toxicities and late surgical complications in locally advanced rectal cancer, yet irrespectively of tumor location. Peeters et al. (23) evaluated late side effects of patients included in the Dutch TME-trial. They reported a significant increase in blood loss as well as fecal incontinence in the irradiated group compared to the TME-only group. They found that 52% of patients with upper rectal cancer presented fecal incontinence at day. As for Bosset et al. (6), they reported 9% of incontinence in patients with rectal reconstruction, 9.6% of ≥ grade 2 diarrhea and 3.1% of anastomotic stenosis. Furthermore, Azria et al. evaluated late toxicities in the ACCORD 12/0405-PRODIGE 02 (24) trial, and reported 15.2% of incontinence, 39.5% of late gastrointestinal side effects and 2.1% of anastomotic stenosis in patients treated with the CAP45 regimen. With the exception of the trial of Peeters et al., our results are overall in line with these studies.
In terms of surgery, total mesorectal excision remains the gold standard in low and middle locally advanced rectal cancer, however partial mesorectal (PME) excision has become an acceptable option for cancers located in the upper rectum as suggested by the ESMO and NCCN guidelines. Several studies have documented that TME produces more side effects than PME (25, 26) and is a risk factor for anastomotic failure (27), as well as incontinence (28, 29). Moreover, a study found a 14% mortality rate in elderly patients during the 6-months period after TME (30). With regard to PME, a number of studies have shown good results, particularly in terms of local control (26, 31). In conclusion, by excluding patients treated with PME, we might have overestimated surgical complication rates.
There are some limitations in our work. The retrospective nature of this study is associated with inherent bias. In addition, there is some heterogeneity in our patient group, especially in terms of preoperative treatment. Finally, the major limitation of this study lies in the absence of a comparative surgery group. In conclusion, our study demonstrated excellent local control associated with favorable survival outcomes and acceptable side effects.
Acknowledgements
The Authors would like to thank Mr Julien Fouret, researcher in bioinformatics, who performed statistical analysis.
Footnotes
Authors’ Contributions
Sandra Chomicki: Conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, validation, visualization, writing–original draft, writing–review, and editing. Sophie Chapet: Conceptualization, methodology, project administration, resources, supervision, validation, visualization, writing–review, and editing. Driffa Moussata: Validation, Visualization, writing-review, and editing. Mehdi Ouaissi: Validation, writing–review and editing. Pascal Bourlier: Validation, writing–review, and editing. Kamel Debbi: Validation, writing–review, and editing. Gokoulakrichenane Loganadane: Validation, writing–review, and editing. Gilles Calais: Conceptualization, methodology, project administration, supervision, validation, writing–review, and editing.
Conflicts of Interest
There are no conflicts of interest in relation to this study.
- Received July 28, 2022.
- Revision received August 21, 2022.
- Accepted August 24, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
