2Department of Radiation Oncology, American Hospital, UT MD Anderson Radiation Treatment Center, Istanbul-Turkey
3Department of Radiation Oncology, HospitalUniversity of Texas M.D. Anderson Cancer Center, Houston,TX-ABD DOI : 10.5505/tjo.2018.1796
Summary
Rectal cancer management can be defined as maximizing local tumor control and overall survival while preserving anal sphincter, bladder, and sexual functions or improving the quality of life with an overall reduction in morbidity. Appropriate preoperative or postoperative therapy is required to minimize the risks of both local and distant recurrence. Preoperative radiotherapy is the current standard for treating patients with high-risk rectal cancer owing to lower rates of local relapse and toxicity. Modern radiotherapy capabilities are well suited for any short- or long-course protocol with decreased toxicity in irradiated structures such as the small intestine, bladder, or femoral heads. As clinicians and researchers, we must aim to establish tailored treatments for these patients based on the most suitable evidence based ground in a multidisciplinary environment regarding the expectations of both our patients and team physicians. Herein, we present a review of ongoing clinical trials in order to shed light on the current debates of standard approaches for treating rectal cancer.Introduction
Rectal cancer management can be simply defined as maximizing local tumor control and overall survival while preserving anal sphincter, bladder, and sexual functions or improving the quality of life with an overall reduction in morbidity.It is well known that different risks of both local and distant recurrence mandate a tailored approach, including appropriate preoperative or postoperative therapy; therefore, all subsequent modalities and their timings must be optimized according to prognostic evaluation. The prognosis of patients with rectal cancer is determined mostly based on defined factors. The most important prognostic criteria are provided by histopathology. TNM staging is independent prognostic factor according to multivariate analysis, whereas size, differentiation, and vascular invasion are independent prognostic factors according to univariate analysis. The following parameters should be assessed when initially deciding the local recurrence risk: T stage with depth of extramural spread in mm, N stage with lymph node involvement load, extramural vascular invasion, circumferential resection margin status, and peritoneal perforation caused by the tumor. Tumor stage T3 increases local recurrence risk, thereby resulting in prognostic inhomogeneity. Patients with T3 tumors have been demonstrated to have a significantly diflonger cancer-specific survival if extramural invasion is less than 5 mm in pathology (5-year survival rate: 85% vs. 54%).[1] The risk propagates with increasing tumor invasion into the perirectal fat to increase nodal involvement.[2] Lymph node involvement has long been identified as an independent adverse prognostic factor.[3] The extent of nodal involvement with four or more tumor-positive nodes has started to be more relevant for pelvic recurrence after total mesorectal excision (TME) in comparison to any nodal involvement predicting recurrence before TME.[4]
The involvement of circumferential resection margin (CRM) is also an independent prognostic factor,2 and its significance for local recurrence, distant metastases, and survival persists despite TME.[5,6] TME is the current standard of surgery as the middle and lower rectum are resected together with the mesorectum,[7-9] and this surgical approach is said to decrease local recurrence,[7-12] along with a decrease in positive CRM. Therefore, patients with potential for CRM positivity should be given neoadjuvant treatment to decrease related risks. The incidence of positive circumferential radial margins in a Polish study was found to be lower after long-course chemoradiotherapy in comparison to short-course radiotherapy (4% vs 13%, P=0.017).[13] If the final pathology defines an involved CRM, postoperative treatment unfortunately appears to have a limited ability to compensate, as reported in a subset analysis of the Dutch CKVO trial, which demonstrated the inefficacy of postoperative long-course radiation alone to decrease local recurrences,[14] and in the MRC CR-07 trial, which revealed a local recurrence rate of 11% despite postoperative long-course chemoradiotherapy.[15]
Evolution of Preoperative Radiotherapy
We have shed light on the discussion of preoperative
versus postoperative radiotherapy first with two phase
3 trials defining decreasing local recurrence and complication
rates with preoperative radiotherapy.[16,17]
A Swedish trial of 5×5 Gy preoperative radiotherapy
alone versus 60 Gy postoperative radiotherapy randomizing
471 patients reported a significant decrease
in local recurrence (5-year recurrence: preop 13%,
postop 22%) and complication rates related with obstruction
(5-year: preop 5%, postop 11%).[16] These
results were confirmed 13 years later with a German
chemoradiotherapy trial of preoperative (50.4 Gy and
PVI 1 mg/m2/day 5FU 1st, 5th weeks) versus postoperative
(50.4 Gy+5.4 Gy boost and PVI 1 mg/m2/day
5FU 1st, 5th weeks) radiotherapy randomizing 823 patients.[17] This trial delineated the superiority of the
preoperative approach with a decrease in local recurrence
(5-year recurrence: preop 6%, postop 13%) and
acute and late complication rates (acute: preop 27%,
postop 40%: late: preop 14%, postop 24%). The overall
survival did not differ in both trials.
The utility of radiotherapy in rectal cancers was questioned via two important meta-analyses, and its robust role has been verified.[18,19] The Colorectal Cancer Collaborative Group evaluated 22 trials, including 8500 cases, and concluded that both preoperative (46% decrement in local recurrence) and postoperative (37% decrement in local recurrence) radiotherapy provides local control benefit over surgery alone.[18] The Swedish Council of Technology Assessment in Health Care reported their analysis of 42 randomized studies, 3 meta-analyses, and 36 prospective and 7 retrospective studies (including 25000 cases) and concluded that preoperative radiotherapy ensures better local control in comparison to postoperative radiotherapy.[19] It should be noted that preoperative single-modality radiotherapy significantly increased overall survival by 10%, whereas postoperative radiotherapy failed to reach significance without chemotherapy.
Sphincter preservation was also an issue to be questioned in preoperative chemoradiotherapy trials. The two randomized trials of conventionally fractionated preoperative versus postoperative chemoradiotherapy for clinically resectable rectal cancer reported contradictory results for sphincter preservation: sphincter preservation significantly increased in the German trial (39% vs 20%, P=0.004)17; no significance was reported in the NSABP R-03 trial (48% vs 39%).[20] As the NSABP trial had limitations in statistical power due to low accrual (267 patients of the 900 planned), the German trial delineates the standard for sphincter preservation using preoperative chemoradiotherapy.
Although 18% of patients clinically staged as T3N0 in the German trial who underwent initial surgery without neoadjuvant treatment were found to be pathologically T1-2N0 with overtreatment debates, the Memorial Sloan-Kettering Cancer Center data revealed that 22% of clinically staged T3N0 patients who completed neoadjuvant chemoradiotherapy were proved to be ypN+.[21]
Short Course versus Long Course
Preoperative radiotherapy has been shown to be preferable
to postoperative radiotherapy with lower rates
of local relapse and toxicity.[16,17] The regimens diflongerfer in the preoperative radiotherapy approach, whereas
short course is preferred in Northern Europe and
long course in Southern Europe and America. Neoadjuvant
radiotherapy has been developed to offer
two regimens that could be accepted as standards for
resectable rectal cancer: short course 25 Gy (5×5 Gy)
radiation therapy alone and long-course chemoradiation
therapy
Three major studies have shaped the literature on the use of preoperative radiotherapy alone: Rotterdam? Holland, EORTC, and Swedish trials.[22-24] The Rotterdam and EORTC trials evaluated 34.5 Gy (2.3Gy/fraction/day) preoperative radiotherapy and revealed local control benefit besides subgroup overall survival benefit for patients with T3-4 tumors resected curatively.[22,23] The Swedish trial was the first randomized study demonstrating overall survival benefit for all cohorts with preoperative radiotherapy. [24] A total of 1168 patients with clinically resectable rectal cancer were randomized to 25 Gy (5Gy/fraction/ day) preoperative radiotherapy and immediate surgery in 1 week versus surgery alone; local tumor control and overall survival benefit was obtained with preoperative radiotherapy. In the TME era, a Dutch trial evaluated the same protocol and indicated a significant decrease in local failures (surgery alone: 8.2% vs. radiotherapy+surgery: 2.4%) and noted a longer follow-up requirement for survival.[25]
Randomized trials testing short-course neoadjuvant radiotherapy accumulated evidence to be safe and efficient (Stockholm 1 [26], Stockholm 2 [27], Swedish trial [28], 5×5 Gy with immediate surgery vs surgery alone; Uppsala, 5×5.1 Gy with immediate surgery vs postoperative radiotherapy, 60 Gy, 2 Gy per fraction [16]; Dutch TME trial [14,29], MRC CR07 [15], 5×5 Gy with immediate surgery vs surgery alone or postoperative radiotherapy for high-risk patients; Stockholm III trial [30], 5×5 Gy with immediate surgery as group 1 vs 5×5 Gy with surgery after 4?8 weeks as group 2 vs 25×2 Gy with surgery after 4?8 weeks as group 3). The Stockholm III trial recently concluded that shortcourse radiotherapy with delay in surgery appeared to be a convenient alternative to conventional short-course radiotherapy with immediate surgery.[31]
Swedish trials [26-28] comparing surgery alone with neoadjuvant short-course radiotherapy and immediate surgery identified that radiotherapy reduced local recurrence ranging from 52% to 65% as well as resulted in an absolute overall survival benefit of 8% at 13 years.[28] Even in the TME era, randomized studies of the Dutch TME trial [14<,r29>] and Medical Research Council (MRC) CR07 trial [15] showed an approximately 50%-60% relative reduction of local recurrence after short-course preoperative radiotherapy with an absolute local control benefit of 5%-6%, whereas no overall survival benefit with radiation was found this time.
The literature was lacking randomized trials comparing neoadjuvant short-course radiotherapy with neoadjuvant chemoradiotherapy for T3 cancers until the results of two trials questioning this dilemma: [1332-35] a Polish study, which evaluated differences in the rates of sphincter-preserving surgery between long-course chemoradiation and short-course radiotherapy, and an Australian study, which evaluated differences in local recurrence rates between these arms. Both trials demonstrated significantly increased early radiation toxicity in the chemoradiation group (grade 3?4 acute toxicity rates, Polish: 18% vs. 3%; Australian: 28% vs. 1.9%), which turned into improved adherence to the protocol in short-course radiation only arms. Interestingly, the sphincter-preservation rates were similar in both arms of the Polish trial (short: 61% and long: 58%), whereas the local recurrence rate was lower in the short arm (short: 10.6% and long: 15.6%). Although the follow-up is yet limited, no significant differences were observed between the randomized groups regarding survival, postoperative complications, late toxicity rates (severe late toxicity, Polish: 10.1% vs. 7.1%; Australian: 7.6% vs. 8.8%), quality of life, and anorectal and sexual functions.
There has been a debate about whether preoperative chemoradiotherapy or short-course radiotherapy is preferable for patients with threatened CRMs and those with low-lying cancer to increase the chance of anterior resection.[14,36] There is a consensus that neoadjuvant chemoradiotherapy is the treatment of choice for unresectable cancers, but the major conflict is for resectable lesions with threatened CRMs. The recent two trials did not exclude patients with resectable tumors having involved CRMs, and the subgroup analysis of Dutch TME (18.2% of enrolled patients) [14] and MRC CR07 trials (10.8% of enrolled patients) [15] revealed that the local recurrence rates were lower in neoadjuvant short-course radiotherapy arms in comparison to selective postoperative chemoradiotherapy arms (Dutch: 9.3% vs 16.4% and MRC: 13.8% vs 20.7%). Based on this data, short-course preoperative radiotherapy appeared to be more effective than selective conventionally fractionated postoperative radiotherapy or chemoradiotherapy in case of resectable cancer with threatened CRMs in preoperative paimaging. Besides, the Polish study [13] as well as the two systematic reviews [37,38] could not discover any improvement with neoadjuvant chemoradiotherapy in terms of sphincter preservation in comparison to neoadjuvant short-course radiotherapy alone despite the hypothesis of better reduction in tumor bulk after chemoradiotherapy for anterior resection to spare sphincter function.
Overall, the combined MRC CR07/NCIC-CTG C016 trial randomized 1350 patients with rectal cancer to preoperative short-course radiation of 25 Gy (5×5 Gy) or selective postoperative chemoradiation of 45 Gy (25×1.8 Gy) with concurrent 5FU for patients with an involved CRM.[15,39] Although no overall survival benefit was outlined between the two arms, this study noted 61% reduction in the relative risk of local recurrence with preoperative radiotherapy besides an absolute difference of 6.2% at 3 years (local recurrence with preoperative radiotherapy: 4.4%; postoperative chemoradiotherapy: 10.6%) and 24% improvement in disease-free survival with preoperative radiotherapy besides an absolute difference of 6% at 3 years (77.5% vs 71.5%).[15,39] The importance of achieving a negative CRM was also confirmed with special notification to the plane of surgery as an independent predictor of local recurrence according to multivariate analysis, and involved margins caused a three-fold increased risk of local recurrence (17% vs. 6%) and a reduced 3-year disease-free survival (50% vs. 79%).[15,39] This study delineated important aspects that affect the adjuvant treatment approach: local recurrence rates are lower in upper-third rectal cancers in comparison to lower-third tumors, and recurrence rates increase with more advanced TNM stage as well as with involvement of CRM. The CR07/C016 trial declared that surgery should be performed correctly, adjuvant radiation should be considered preoperatively, and short-term radiotherapy alone is safe and efficient.
The recent phase III Polish II trial for cT4 or fixed cT3 rectal cancer investigated the comparison of longcourse preoperative chemoradiation of 50.4 Gy in 28 fractions combined with two 5-day cycles of bolus 5-FU 325 mg/m2/day and leucovorin 20 mg/m2/day during the first and fifth week of irradiation along with five infusions of oxaliplatin 50 mg/m2 once weekly versus short-course 5×5 Gy and three cycles of consolidation FOLFOX4 chemotherapy.[40] Bujko et al. revealed no differences in local efficacy between both arms but reported an improved overall survival and lower acute toxicity for the 5×5 Gy schedule with consolidation chemotherapy.[40]
The RAPIDO phase 3 trial is open to accrual for locally advanced rectal cancer randomizing the standard arm of chemoradiation (1.8 Gy×25 or 2 Gy×25 with capecitabine) preoperatively followed by selective postoperative adjuvant chemotherapy of eight cycles of CAPOX versus short-course radiotherapy followed by six cycles of neo-adjuvant CAPOX chemotherapy.[41]
Time to Surgery
Interval to surgery after preoperative radiotherapy/
chemoradiotherapy is a direct correlation with tumor
down-staging and interval to surgery after radiotherapy,[42-44] and therefore, long-course chemoradiotherapy
has long been generally preferred in initially
unresectable tumors to gain downsizing.[13,45-48]
Prospective studies evaluating the rates of sphincter
preservation by neoadjuvant chemoradiotherapy determined
a change in the surgical approach from abdominoperineal
resection to sphincter preservation in
23%-85% (median: 75%), whereas the complete pathological
response rates were 9%?19% and local recurrence
rates were 0%?17%.[49-54] The interval to surgery
for postoperative radiotherapy was 2 weeks versus
6?8 weeks in a Lyon study and the complete surgical
response was 10% at 2 weeks and 26% at 6?8 weeks.
[42] However, the pathological response increased
when surgery was delayed even for short-course 5×5
Gy regimen in unresectable cancers.[30,55-56] The
appropriate interval after chemoradiation for surgery
timing was assessed in multiple retrospective database
studies and National Cancer Center Database (NCDB)
analyses, and it was found that intervals more than 8
weeks were associated with increased complete pathological
response rates without any increase in surgical
complications, [57] higher odds of positive margins
and plateauing pathological down-staging, [58] increased
odds of positive surgical margins in addition to
decreased rate of sphincter preservation, and increased
risk of death.[59] GRECCAR-6 is a multicenter, randomized,
controlled trial conducted for patients with
rectal cancer to evaluate the effect of interval between
neoadjuvant chemoradiation and surgery at 7 versus 11
weeks on complete pathological response; the results of
the trial revealed no difference in the complete pathological
response rates (15% v. 17.3%), whereas the 11-
week group had worse rate of complete TME and more
medical complications.[60]
Selective Use of Radiotherapy
The general hypothesis was to avoid radiotherapy after
an objective good clinical response in a subset of patients receiving neoadjuvant chemotherapy. The initial
effort was made in a pilot phase 2 MSKCC study enrolling
32 patients [61] who were candidates for low
anterior resection with TME and received six cycles of
FOLFOX, with bevacizumab included for cycles 1 to 4,
and were planned to undergo radiation before TME if
stable/progressive disease detected, whereas responders
were to undergo immediate TME; postoperative
radiation was administered if R0 resection could not
be achieved. All study participants underwent R0 resections,
whereas two were withdrawn due to cardiac
events during chemotherapy and received preoperative
chemoradiotherapy. The 4-year disease-free survival
was 84%, which did not seem to compromise
outcomes and encouraged neoadjuvant chemotherapy
and selective radiation for selected patients with clinically
staged II-III rectal cancer. Therefore, a randomized
phase III trial named "Preoperative Radiation or
Selective Preoperative Radiation and Evaluation Before
Chemotherapy and TME (PROSPECT)" is open
to validate this experience by comparing 5-fluorouracil
or capecitabine and 5.5 weeks of radiation therapy
followed by TME and adjuvant therapy of eight cycles
of FOLFOX versus completing FOLFOX chemotherapy
once every 2 weeks for six cycles over a total of 12
weeks and MRI or endorectal ultrasound response by
20% or more proceeding to TME or by less than 20%
receiving 5FU chemoradiotherapy then TME completed
with adjuvant FOLFOX after surgery.
Total Neoadjuvant Therapy
Neoadjuvant therapy has been also investigated as a
total preoperative therapy with upfront rather than adjuvant
chemotherapy to further improve outcomes by
addressing possible micrometastatic disease as well as
the primary tumor. Two phase II studies, UK and Spain
trials, have evaluated induction chemotherapy followed
by preoperative CRT in high-risk patients based
on MRI for assessing the extent of extramural tumor
invasion and risk of CRM positivity.[62,63] Induction
CAPOX chemotherapy before CRT in the UK EXPERT
and Spanish GCR-3 trials had similar pCR and complete
resection rates in comparison to postoperative
adjuvant CAPOX, whereas more favorable compliance
and toxicity profiles were achieved [63,64] Similarly, in
the MSKCC study, total neoadjuvant radiotherapy with
FOLFOX and chemoradiation followed by planned
TME resulted in a considerable rate of pathCR and delivery
of planned therapy in addition to offering a very
selective decent stand for possible non-operative management.[65] mFOLFOX6 chemotherapy after concurconcurrent
chemoradiation before TME has also shown to
potentially increase the pCR up to 38%.[66]
A recently proposed NRG-GI002 phase II clinical trial platform will be randomizing phase II modular clinical trials utilizing total neoadjuvant therapy with parallel experimental arms.[67] Besides Trial Evaluating 3-year Disease-Free Survival in Patients With Locally Advanced Rectal Cancer Treated With Chemoradiation Plus Induction or Consolidation Chemotherapy and Total Mesorectal Excision or Nonoperative Management
Non-operative Management (NOM)
Non-operative management (NOM) has been trending
as a paradigm shift to avoid surgery if possible
in case of complete clinical response to preoperative
treatment, following neoadjuvant Brazilian CRT data,
which define safe and good survival rates in a highly
selected group of patients without surgery because 99
of 122 patients with complete clinical response (cCR)
at first response assessment sustained cCR for a minimum
of 12 months were managed nonoperatively
with stage c0 and at a mean follow-up of 59.9 months,
13.1% recurred (5% endorectal, 7.1% systemic, 1%
combined).[68] An update of this data by Habr-Gama
in 2013 continued to encourage NOM in selected
patients.[69,70] Dutch data reinforced the NOM approach
with strict selection criteria and frequent follow-
up of endoscopy and MRI for organ preservation
as oncologically safe for selected cCR or near cCR after
neoadjuvant chemoradiation.[71,72] The 3-year overall
survival rates, distant metastasis-free survival rates,
local regrowth-free survival rates, and disease-free
survival rates were 96.6%, 96.8%, 84.6%, and 80.6%,
respectively.[72]
Because surgery is yet the only reliable method to detect a pCR and cCR does not mean pathologic response, an intensive effort to distinguish post-RT changes from residual disease, continuous evaluation of digital rectal examination, endoscopic assessment, endorectal ultrasound, MRI and PET, and methods to interpret post treatment biopsies must be prospectively investigated until NOM in rectal cancer can be considered a standard approach. NCT02008656 (Trial Evaluating 3-year Disease-Free Survival in Patients With Locally Advanced Rectal Cancer Treated With Chemoradiation Plus Induction or Consolidation Chemotherapy and Total Mesorectal Excision or Nonoperative Management) [73] and NCT02052921 (Observation Versus Surgical Resection in Patients With Rectal Cancer Who Achieved Complete Clinical Response After Neoadjuvant Chemoradiotherapy) [74] are ongoing for accrual.
Summary
It is well known that different risks of both local and
distant recurrence mandate an individualized approach,
including appropriate preoperative or postoperative
therapy. Preoperative radiotherapy is the current
standard for treating patients with high-risk rectal
cancer because of lower rates of local relapse and toxicity.
Modern radiotherapy capabilities are well suited
for any short- or long-course protocol with decreased
toxicity in irradiated structures such as the small intestine,
bladder, or femoral heads. As clinicians and researchers,
we must aim to establish tailored treatments
for these patients based on the most suitable evidence
based ground in a multidisciplinary environment regarding
the expectations of both our patients and team
physicians. In this review, we have shed light on the
current debates of a standard treatment approach for
rectal cancer in ongoing clinical trials.
Peer-review: Externally peer-reviewed.
Conflict of Interest: None declared.
Authorship contributions: Concept - U.S.; Design - D.S.;
Supervision - Y.B.; Data collection &/or processing - N.K.D.;
Analysis and/or interpretation - U.S.; Literature search -
N.K.D.; Writing - D.S.; Critical review - Y.B.