The treatment of locally advanced rectal cancer (LARC) is multidisciplinary and it is developed from clinical trials evidence [
]. The trimodality approach, including neoadjuvant 5-fluorouracil (5FU)-based chemoradiotherapy (CRT), surgery and adjuvant chemotherapy, is considered the standard of care in this setting of patients [
]. Despite improvement in surgical techniques as well as advances in radiation therapy techniques and chemotherapeutic agents, nowadays the vast majority of recurrences in LARC are systemic [
]. Distant metastasis rate remains high (approximately 40%) and becomes the prevailing method of treatment failure.
Based on the efficacy demonstrated in colon cancer trials, the intensification regimen with oxaliplatin (OXP) as radiation-sensitizing agent in neoadjuvant CRT has been tested in several large phase III studies – STAR-01, ACCORD 12, NSAPB R-04, CAO/ARO/AIO- 04 and Chinese trial – but definitive conclusions are still pending [
Recently update of results have been presented and in light of these data we performed a meta-analysis. The aim of this meta-analysis was to evaluate whether the addition of OXP in neoadjuvant treatment for LARC could be superior to standard CRT, in term of overall survival (OS), disease free survival (DFS), local failure (LF) and distant failure (DF).
Selection of trials
The preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines were followed to perform search strategy and selection processes. The meta-analysis included trials, written in English, without any restrictions on publication date. The last search was carried out on May 2016. Systematic literature electronic search was conducted in Pubmed, Medline and Scopus databases. The search term used were “radiotherapy”, “chemoradiotherapy”, “chemotherapy”, “oxaliplatin/OXP”, “fluorouracil/5FU/fluoropyrimidine”, “capecitabine/xeloda”, “rectal cancer/locally advanced rectal cancer”, “neoadjuvant”, “randomized” and “clinical trials” in the title.
To be eligible for this meta-analysis, trials needed to compare the addition of OXP to neoadjuvant CRT with standard 5FU-based CRT in LARC. To reduce publication bias, data from all clinical randomized trials, both abstract and full-text paper, were included using literature electronic databases searching (Pubmed, Medline and Scopus) and hand searching (meeting proceedings of European Society for Radiotherapy & Oncology, European Society of Medical Oncology and American Society of Clinical Oncology). Reference lists of previously published reviews and meta-analysis were explored to identify relevant citations. Trials were eligible if participants were newly diagnosed, with histologically proven adenocarcinoma of the rectum at study entry. In closer evaluation of potentially eligible articles, when two articles appeared to report results with overlapping data, only the data representing the most recent publication were included in the meta-analysis.
Extracted data were recorded into standardized database according to the following parameters: first author’s surname, year of publication, trial acronym, sample size of experimental and standard group, chemotherapy regimen, drug and dosage, radiotherapy total dose and single fraction.
The intent of the analysis was to evaluate disease free survival (DFS), overall survival (OS), local failure (LF) and distant failure (DF). DFS was defined as the time from the date of randomization to last follow-up, death or disease progression. OS was defined as the time from the date of randomization to last follow-up or death. LF was defined as the time from random assignment to local recurrence within the pelvis. DF was defined as the time from random assignment to distant metastasis occurrence, irrespective of whether this was a first event or not.
The number of events (death, progression and distant metastasis), when available, were derived from each study. At least one of these two outcomes should have been assessed and reported in the trial to be included in the present analysis.
Statistical analysis was performed using Review Manager 5.0 (
). It was based on both abstract (thus analysis of the full-text results was not performed) and full-text paper results. The analysis used odds ratio (OR) to compare results for OXP-5FU group to control patients. The pooled OR was calculated using a random-effects model. Forest plots were used for graphical representation of each study and pooled analysis. OR, variance, 95% confidence interval (CI), log [risk ratio] and standard error for each study were calculated, based on Tierney et al. method [
]. A significant two-way
value for comparison was defined as
< 0.05. Statistical heterogeneity between studies was investigated using Cochrane Q statistic (significant at
< 0.1) and the I
value (significant heterogeneity if >50%) [
]. Publication bias was examined using Egger et al. [
] and Begg et al. [
We performed a meta-analysis to compare the efficacy of neoadjuvant CRT with OXP plus 5FU to 5FU alone in LARC. The main result of this meta-analysis was that addition of OXP to standard 5FU-based CRT was related to significant clinical benefit in term of DF. Whereas, there was no significant increase in OS, as well as no lower DFS and LF rates were observed between groups, although globally a higher proportion of events was recorded in patients treated with standard CRT.
Nowadays, the treatment of LARC is multidisciplinary and it is developed from clinical trials evidence. The trimodality approach, including neoadjuvant CRT, total mesorectal excision surgery and adjuvant chemotherapy, represents the standard of care in this setting of patients, due to its value to improve local control up to 90% of cases [
]. However it does not decrease distant failures and year after year the concept of intensify treatment regimen has become progressively more common in order to improve systemic control. Based on the efficacy demonstrated in colon cancer patients [
], OXP-based intensification of neoadjuvant CRT has been tested in several randomized trials and, recently, a meta-analysis has evaluated its short term efficacy and toxicity results [
]. Briefly, OXP-5FU regimen increased pathologic complete response (OR = 1.20; 95% CI, 1.01 to 1.42) and reduced peri-operative metastasis incidence (OR = 0.51; 95% CI, 0.34 to 0.77), but increased severe toxicity rate (OR = 2.29; 95% CI, 1.31 to 4.0) compared to 5FU alone.
Given these results, the clinical outcomes analysis plays an important role in the treatment decision. Only one randomized trial demonstrated a survival benefit following addition of OXP, whereas the others failed to demonstrate it [
]. On this background, we performed a meta-analysis to investigate the important question of identifying the optimal concomitant chemotherapy regimen to use in neoadjuvant CRT treatment in LARC. No clear evidenced resulted in prolonging OS (OR = 0.84; 95% CI, 0.70 to 1.01) and DFS (OR = 0.97; 95% CI 0.72 to 1.30) compared with the standard CRT. But OXP patients had a 24% risk reduction of developed distant metastasis (
= 0.03) than patients treated with single agent CRT.
There are several considerations that should be made, because the trials included in the meta-analysis did not address exactly the same end-points, and treatments were not exactly similar.
The tested primary end-points were DFS in CAO/ARO/AIO-04 trial [
], DFS and OS in Chinese trial [
], pathologic complete response in ACCORD 12 trial [
] and loco-regional control in NSAPB R-04 trial [
]. These different primary endpoints delineated the lack of statistical power to assess the chosen endpoints. Considering that the vast majority of recurrences in LARC remain systemic, the theoretical advantage of adding OXP to standard CRT should be primarily to improve distant control. Thus, a DF analysis was performed. Based on this analysis, it is possible to hypothesize that the addition of OXP is an efficient regimen to reduce the risk of systemic metastasis in LARC (OR: 0.76; 95% CI, 0.60 to 0.97). The limitations of the analysis are related to the relative limited number of trials and the heterogeneity of primary end-points. In fact, one trial had no data on distant failure [
]. Moreover it should be noted that DFS was considered the primary end-point in two out of three trials included, and it maybe has reduced the power of the analysis, but do not bias it. In addition studies were conducted in different countries and probably quality of care might have influenced the results.
Indeed these trials included different surgical approach details, such as number of examined lymph nodes and type of surgery. Total mesorectal excision was performed in the vast majority of cases. Other types of surgery, including Hartmann procedure, intersphincteric resection, transanal local excision or low anterior resection were also recorded. But the absence of complete surgical data prevents from defining a specific population that would or would not benefit from OXP-5FU regimen. Thus whether the inclusion of OXP into 5FU-based CRT could reflect the new treatment option for patients with LARC cannot be resolved by our meta-analysis. Surely, the benefit of adding OXP on DF could justify the proven substantial increase in acute toxicity, especially severe diarrhea [
]. Among the trials included, patients received substantially similar OXP regimens – 50 mg/m
once a week or 60 mg/m
on days 1, 8, 22, 29 – but heterogeneous 5FU schedule. In CAO/ARO/AIO-04 trial [
], the 5FU schedules even differed in the two treatment groups – 250 mg/m
daily in experimental group versus 1000 mg/m
weeks 1 and 5 of RT in control group –, whereas the other trials used the same 5FU regimen for both treatment – 225 mg/m
daily or 1600 mg/m
daily – [
]. Considering that 5FU is probably the main responsible for gastrointestinal toxicity, including diarrhea, nausea and vomiting, we believe that acute toxicity could be easily minimize by decrease the individual 5FU doses (200 mg/m
) and increase the frequency of administration of weekly OXP (to a maximal total dose of 300 mg), maintaining therapeutic effectiveness [
]. In our published retrospective analysis, we found a considerable decrease in toxicity (17%), particularly in terms of severe diarrhea (4%) compared to grade 3/4 toxicity rate reported in randomized studies, ranged from 21.4% to 41.9% [
To our knowledge, this is the first meta-analysis with long-term follow-up data to show a DF benefit of adding OXP to 5FU-based CRT when compared with standard treatment with 5FU alone for patients with LARC. However results need to be interpreted with caution, because it is an abstract-based meta-analysis. Although abstract included statistical quality details, definitive full-text results could improve data analysis precision and produce more robust conclusions. In particular, subgroup analysis might reveal different results for some patients, such as node positive versus node negative patients. However, reduction in DF could be considered a sensible result, reflecting itself a relevant benefit for patients, regardless of survival improvement.