Prognosis and Quality of Life Analysis After Neoadjuvant Chemoradiotherapy and Primary Surgery for Low-lying Locally Advanced Rectal Cancer: Can Patients Benefit From Neoadjuvant Therapy?


 BackgroundNeoadjuvant Chemoradiotherapy (nCRT) is a widely accepted regimen for patients with locally advanced rectal cancer (LARC). This compared the long-term prognosis and postoperative quality of life (QoL) between patients with low-lying LARC receiving nCRT and primary surgery.MethodPatients underwent nCRT or primary surgery for low-lying LARC between 2010 and 2016 were identified. Five-year local recurrence (LR) and disease free survival (DFS) were compared between groups. Quality of life (QoL) of patients who were disease-free was investigated using European Organization for Research and Treatment of Cancer QoL questionnaire core-30 (EORTC QLQ-C30) and QLQ-Colorectal Cancer module (CR29).ResultsA total of 304 patients were included in this study. Differences in 5-year LR and DFS between groups showed no statistical significance. In terms of QoL, apart from less stoma care problem, nCRT patients showed unsatisfactory social function and worse symptoms including diarrhoea, financial difficulties, buttock pain, fecal incontinence, embarrassment and impotence compared with primary surgery group. Intergroup analysis indicated that the QoL of patients receiving nCRT with preserved sphincter was relatively inferior compared with other subgroups, as reflected in higher symptom scores including financial difficulties and those related to low anterior resection syndrome (LARS) such as diarrhea, stool frequency, flatulence and fecal incontinence.ConclusionFor patients with low-lying LARC, nCRT has no advantage in terms of 5-year survival and QoL. NCRT with sphincter preserving surgery should be conducted meticulously considering its limited benefits for patients.


Introduction
For patients with locally advanced rectal cancer (LARC), neoadjuvant chemoradiotherapy (nCRT), currently recommended by the National Comprehensive Cancer Network (NCCN), 1 is routinely carried out in multiple centers in order to control local recurrence (LR) and achieve tumor down-staging and anal preservation. [2][3][4][5][6] Thanks to the development of laparoscopic surgery and the concept of total mesorectal resection (TME), 7 together with the tumor regression effect after nCRT, sphincter preserving surgery (SPS) has now become an alternative of abdominoperineal resection (APR) for low-lying LARC, sparing patients from permanent stoma.
Nevertheless, despite favorable LR rate, long-term effect of nCRT remains controversial due to the inconclusive improvement of overall survival and disease free survival (DFS) for LARC patients. 8-10 Much as nCRT can increase anal preserving rate, 80% of patients underwent SPS suffered from low anterior resection syndrome (LARS) and it was reported that up to 19% of patients require permanent colostomy due to severe LARS, [11][12][13] thus making it a disputable matter whether anal preservation equates to favorable quality of life (QoL). [14][15][16] For patients with low-lying tumor location, QoL could be worsened due to lower tumor distance form the anal verge and the toxic effect of radiotherapy, [17][18][19][20] compromising the bene t of sphincter preservation earned by nCRT. Such is a dilemma that clinical practitioners are often confronted with, yet a lack of evidence exists considering the bene t of nCRT for low-lying LARC patients in current literature, which are often limited by small sample sizes, short-term follow-ups or omissions of QoL and the impact of nCRT [21][22][23][24] . In this study, longterm oncological outcome and QoL of low-lying LARC were compared between patients with or without nCRT, strati ed by different surgical approaches including APR and SPS.
2 Materials And Methods 2.1 Patients and treatment ileostomy and subsequent stoma reversal were conducted for patients received SPS, while APR was done with de nitive colostomy. All patients received and completed chemotherapy before or after surgery in this cohort.

QoL assessment
QoL assessment was carried out among patients who achieved disease free survival 5 year after surgery. Eligible patients were asked to complete European Organization for Research and Treatment of Cancer QoL core-30 (EORTC QLQ-C30) and QLQ-Colorectal module (CR29) questionnaires, which were validated for the use in our country, 25-27 during out patient clinic visit, telephone follow-up or via mail 5 years postoperatively. The EORTC qlq-C30 questionnaire covers ve functional domains (physical, role, emotional, cognitive, and social functioning), nine symptom scales (fatigue, nausea/vomiting, pain, constipation, diarrhea, loss of appetite, insomnia, dyspnea, and nancial di culties) and a global health status. 25 The EORTC qlq-CR29 questionnaire is an instrument specialized for colorectal cancer, that includes 5 functional scales (body image, anxiety, weight, men/women sexual interest) and 18 symptom scales (urinary frequency, blood and mucus in stool, stool frequency, urinary incontinence, dysuria, abdominal pain, buttock pain, dry mouth, hair loss, taste, atulence, fecal incontinence, sore skin, embarrassment, stoma care problems, impotence and dyspareunia). 28 For both instruments, higher functional scores indicated better function, and higher symptom scores signi ed more severe symptoms.

Statistics
Raw data of the QoL questionnaires was linearly transform to a 0-100 score according to EORTC scoring methodology. 25 Continuous baseline characteristics and QoL scores were described using means±standard deviation (SD); categorical variables were presented as frequencies and percentages. Inter-group comparisons were made between patients who received surgery after nCRT treatment and those underwent primary surgery at the time of diagnosis (nCRT vs. Primary surgery). Subgroup strati cation took into account nCRT as well as surgical approaches: group A included patients received primary SPS; primary APR as group B; SPS after nCRT as group C and APR after nCRT as group D ( gure 1).
Analysis were performed with IBM SPSS Statistics, version 23.0. Student's t test and one-way ANOVA were used to compare the continuous baseline data; chi-squared test or Mann-Whitney U test were used for non-parametric data, when appropriate. Kruskal-Wallis test was performed to compare subgroup QoL scores. Kaplan-Meier survival estimates model were used to compare DFS and LR between groups. All p values were two-sided and p 0.05 was considered statistically signi cant; in subgroup analyses, we adjusted the two-sided α level to 0.0083(=0.05/6) by simple Bonferroni correction for multiple comparisons.

Patient characteristics
In total, 438 patients were enrolled in this study. After exclusion, 304 patients were eligible for further analysis ( gure 1). The baseline characteristics were similar except for sex (p 0.05) and age (p 0.05), showing that male and younger patients have better compliance to nCRT treatment. In the nCRT group, permanent stoma rate was signi cantly lower (p 0.001) but the 5-year total hospitalization cost was higher (p 0.001). Eight patients in the nCRT group failed to have stoma reversal after SPS and 2 patients demanded permanent stoma after reversal due to poor anal function. In primary surgery group, 9 patients failed to restore their stomas and 1 patient received permanent stoma because of poor function. Signi cant differences in sex, age, tumor height and pretreatment T stage were found among subgroups (Table 1). After adjusting the two-sided α level to 0.0083 by simple Bonferroni correction for multiple comparisons, signi cant differences were revealed mainly in sex, age, tumor height, pretreatment T stage and 5-year total hospitalization cost among subgroups (supplementary Table 1).

Prognostic analysis
Over a median follow-up period (interquartile range, IQR) of 57.0 months (43.3-71.8 months), 146 of 304 patients (48.0%) had clinical recurrence within 60 months postoperatively, the rates of DFS and LR did not differ between groups (supplementary Table 2). The result of Kaplan-Meier showed that the 5-year DFS (p=0.729) and LR (p=0.117) of primary surgery and nCRT group were not statistically different ( Figure 2). Subgroup prognostic analysis using Kaplan-Meier also showed no signi cant differences regarding DFS (p=0.210) and LR (p=0.465) (Figure 2).

QoL analysis
The baseline characteristics of the 132 patients enrolled in QoL analysis were similar among groups except for tumor height and pretreatment T stage according to subgroup comparison (supplementary Table 3,4). In QLQ-C30, statistically and clinically signi cant differences were presented for social function (13.0 points, moderate; p=0.034) in the functional scales and diarrhoea (17.4 points moderate; p=0.003) and nancial di culties (16.1 points, moderate; p 0.001) in symptom scales between primary surgery and nCRT groups ( gure 3)( Table 2). Subgroup analysis showed intergroup inequality in cognitive (p=0.038) and social function (p=0.001) in functional scales and constipation (p=0.031), diarrhoea (p=0.001) and nancial di culties (p 0.001) in symptom scales, indicating a general inferiority of QoL in group C ( gure 3)( Table 2)(supplementary table 5).

Discussion
In this study, the 5-year DFS is similar between patients with and without nCRT (55.8% vs. 52.3%, p=0.558), a result complies with previous studies which nd it inconclusive that long-term survival can be improved by nCRT for LARC patients. 4, 6, 29-31 What's more, the superiority of LR in nCRT group is however statistically insigni cant (9.4% vs. 15.2%, p=0.147), which is different from concerned literature 24, 32 and thus expects veri cation by further prospective study with larger sample scale.
Which surgical approach can achieve better local control for low-lying LARC remains controversial. A systematic review found that rates of positive circumferential radial margins and local recurrence are signi cantly lower after SPS than those after APR. However, the author also noted that in the enrolled studies, tumor height is obviously lower and tumor stage is more advanced in APR patients, hence compromising the conclusion SPS is superior than APR regarding local control 33 . In a multicentre retrospective study based on propensity score matching, local recurrence rate was found similar between SPS and APR for low rectal cancer (11% vs. 13%, p=0.40) 23 , which parallels the results from our subgroup prognostic analysis (Figure 2). Importance should be attached on QoL evaluation as different treatment schemes showed no signi cant difference in long-term prognosis, especially when the adverse effect of nCRT on postoperative QoL in LARC patients is concerned. On one hand, radiotherapy can damage pelvic nerve and anal sphincter, induce brosis and decrease neorectal compliance, resulting in poor QoL and function especially in sex and defecation which was con rmed by various studies [34][35][36][37] . In our study, the damage effect was mainly presented in the decrease in social function and increase in some speci c symptom scores such as diarrhoea and fecal incontinence. On the other hand, as nCRT can increase anal preserving rate, it is preferred by patients who believed it can improve QoL. 38,39 It was showed in this study that nCRT group had better anal preserving rate than primary surgery group (43.7% vs 66.4%, p 0.01), which complies with previous RCT reports 10, 40 ; however, such matter is still controversial, for several systemic reviews and meta analysis pointed out that evidence is insu cient to support this conclusion 2, 41,42 . Despite that, the impact of SPS on QoL is to be discussed as 90% patients would suffer from LARS which worsens when tumor height is lower 11,43 ; in fact, it was reported that 19% patients even demanded permanent stoma due to severe LARS 12 . Currently, the difference of QoL between SPS and APR is found insigni cant by most reviews 16,44,45 . For patients with low-lying rectal cancer, those who underwent APR had better social and cognitive function and less symptoms compared with SPS group in a prospective study 46 . In a newest RCT study, different result was found as SPS showed better sexual-urinary functions in male patients 47 . According to the long-term QoL analysis in this study, although SPS patients had higher body image score, cognitive and social function are not satisfactory and LARS related symptoms such as diarrhoea and constipation are prominent. Hence, we suggest that the inferiority of APR to SPS is not justi ed.
Considering the fact that both nCRT and surgical approach can affect QoL, we divided patients into subgroups for further analysis. While QoL of APR patients showed similarity between group B and D, inferior social function and worse symptoms in diarrhoea and embarrassment were found in sphincter-preserved patients after nCRT treatment in the comparison of group A and C, indicating an adverse effect of nCRT which could be related to LARS 18 . Unlike primary surgery subgroups that both group A and B had their own advantages and defects in different QoL items, the comparison between group C and D showed superiority in most directions in patients underwent both nCRT and APR, especially in cognitive and social function and lowered LARS, suggesting a preventive effect of permanent stoma on such patients 48 . Generally, QoL scores in group C were inferior to those of other subgroups, which could be attributed to nCRT and the secondary impact of SPS.
In addition, economical burden caused by neoadjuvant radiotherapy cannot be overlooked. It needs to pay least €3200 for irradiation when receiving longcourse nCRT 49 , accounting for more than 20% of the total expenditure of each LARC patient 50 . In this study, the 5-year hospitalization cost of nCRT group was about 75% higher than that of primary surgery group, which explained its higher symptom score in nancial di culties. Further cost-effectiveness analysis should be carried out in the light of this issue.
Meticulous clinical decision and thorough informed consent are critical, considering the unsatisfactory QoL of low-lying LARC patients receiving both nCRT and SPS. Although 15% patients can achieve pathological complete response after nCRT 51 , most patients require surgical treatment and their QoL can be improved by primary surgery without compromising survival outcome in properly selected cases, which calls for a more accurate indication for nCRT. For patients who received nCRT, APR is recommended for better postoperative QoL. If permanent stoma is not accepted, patients should be informed of the possibility of low QoL and persisting LARS 34 . Moreover, it was reported that 3-year OS of neoadjuvant chemotherapy and nCRT were found similar 52 , but postoperative LARS and QoL can be improved by omitting radiotherapy, which is a promising alternative of traditional nCRT 36 .
To the best of our knowledge, it is the rst study which compared the oncological and QoL outcomes of low-lying LARC patients between primary surgery and nCRT treatment groups, which were further strati ed by surgical approaches with or without permanent stoma. According to our results, long-term DFS and LR of nCRT is similar to that of primary surgery, but it will signi cantly increase the cost of hospitalization. Although better sphincter preserving rate can be achieved by nCRT in low-lying LARC patients, no improvement was found in postoperative QoL. It should be conducted meticulously considering the unsatisfactory QoL of patients receiving both nCRT and SPS.

Declarations
Ethics approval and consent to participate This research has been performed in accordance with the Declaration of Helsinki and has been approved by the Medical Ethics Committee of the Sixth A liated Hospital, Sun Yat-sen University. Informed consent was obtained from the participants before their participation in this study as it is a retrospective study.

Consent for publication
Not applicable.

Availability of data materials
The data described in this data note is available and can be freely and openly accessed. The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
No external or internal source of funding was obtained for this study.      Comparison of the EORTC QLQ-C30 subscales between groups. A and C illustrate functional scales, where a higher score means better function; B and D illustrate symptom scales, where a higher score means worse symptom. A and B show the comparison between primary surgery and nCRT groups. **p 0.05 with moderate clinical signi cance (score difference: 10-20). C and D show the comparison between subgroups.

Figure 4
Comparison of the EORTC QLQ-CR29 subscales between groups. A and C illustrate functional scales, where a higher score means better function; B and D illustrate symptom scales, where a higher score means worse symptom. A and B show the comparison between primary surgery and nCRT groups. *p 0.05 with minor clinical signi cance (score difference: 5-10); **p 0.05 with moderate clinical signi cance (score difference: 10-20); ***p 0.05 with major clinical signi cance (score difference 20). C and D show the comparison between subgroups.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. supplementarytables.docx