Authors
No. of pts
RT dose Gy/fractions/ days
T2
Chemo
Procedure
cCR
No failing
Local failure
Habr-Gama et al. [24]
118
50.4/28/38
Yes
Bolus FUFA
No surgery
32 %
8/30
Not fully stated
27 %
Rossi et al. [55]
16
50.4/28/38 + 30Gy b
Not stated
FUFA
No surgery
6/16
5/6
5/6
38 %
83 %
83 %
Birnbaum et al. [56]
72
45/25/33 + 30Gy brachy
Yes
No chemo
No surgery
Not stated
26/72
26/72
36 %
36 %
Nagakawa et al. [57]
52
45–50.4/28/38
No
FUFA
No surgery
19 %
Not stated
Not fully stated
Gerard et al. [58]
29
70 contact + 39/13/17 EBRT
Yes T1–3
No chemo
No surgery
Not stated
8/29
8/29
28 %
28 %
Gerard et al. [54]
63
80 contact + 39/13/17 EBRT
Yes T2/3
No chemo
No surgery
58/63
17/63
37 %
92 %
16 %
Habr-Gama et al. [13]
260
50.4/28/38
Yes 20 %
FUFA
No surgery
71/260
5/70
4.2 %
28 %
7 %
Wang et al. [10]
271
40–52/20
Yes
No chemo
No surgery
80/271 30 %
Not stated
78 %
Lim et al. [59]
48
Variable, mean 50Gy/25 fractions
Yes T1 and T2 (33 %)
PVI 5FU 92 %
No surgery
56 %
18/48
11/48
37 %
23 %
Hughes et al. [15]
58
45Gy/25/33
No 50 % T4
FUFA
No surgery
10/58
6/10
6/10
17 %
60 %
60 %
The shortcomings of these series are that patients in the Habr-Gama series were clinically staged without MRI and often without TRUS. It is stated that 20 % were clinically staged as T2N0. Many others are likely to have been small tumours as the median size of the patients who failed to achieve cCR was only 4 cm. One could speculate that this approach is more suitable for small cT2 tumours – as in the recent ACCORD study where pCR was common in cT2 patients (Table 16.2). In addition, most published studies have up to 10 % of patients lost to follow-up. The Habr-Gama series does not appear to have any patients lost to follow-up. Also the original denominator – i.e. how many patients after completion of chemoradiation at 6 weeks achieve a complete clinical response – remains unclear. The use of a delayed time point, for assessing response, loses important information on those patients, who relapse within the first year. Unless we have definitive information on the downside of this approach in terms of those patients who have an initial complete clinical response, who then recur locally within the first year or subsequently develop metastatic disease, it is not possible to assess the validity of the watch-and-wait approach.
Table 16.2
ACCORD-12 PRODIGE trial: percentage of ypCR according to clinical T stage in both arms [60]
T2 (%) | T3 (%) | T4 (%) | |
|---|---|---|---|
CAPOX 50Gy | 47 | 18 | 13 |
CAP 45Gy | 33 | 13 | 7 |
What Are the Risks to Wait and See?
Recurrent rectal cancer is often a diffuse pelvic process, particularly following radiotherapy, and is often associated with metastatic disease. For these reasons, local recurrence is rarely surgically salvaged and is associated with poor outcomes particularly after preoperative radiotherapy [61]. Patients risk being left with very difficult to manage pelvic symptoms, which can be very challenging to palliate adequately.
In the Habr-Gama series [28], patients with apparent cCR where a delay in surgery was instituted did not appear to experience a negative impact regarding survival. In the original Habr-Gama data, only 2 patients developed a local pelvic relapse. When further updated in 2005 and 2006 with an additional 28 patients (i.e. 99 patients) being identified as having had a cCR (99/360), only 3 more local relapses were identified. However, this finding is not necessarily extrapolated to locally advanced T3/T4 tumours where up to 70 % are node positive.
Sterilisation of Pelvic Lymph Nodes
The second major concern regarding chemoradiotherapy as a definitive treatment especially in locally advanced disease is the ability of chemoradiotherapy to adequately sterilise the pelvic lymph nodes. Residual viable disease within the lymph nodes leaves a potential source of pelvic relapse. Chemoradiotherapy does result in the downstaging of pelvic lymph node disease in locally advanced disease; however, the concordance between the eradication of disease within the primary and within the lymph nodes is not absolute with some patients appearing to have obtained pathological complete responses within the primary but not in the lymph nodes. Pelvic residual disease is obviously more likely in advanced lesions where CRT is administered for a bulky tumour which threatens the CRM than a small T1/T2 tumour [62]. Tumours which are clinically and radiologically node negative are more likely to achieve a pCR [63].
The difficulty in predicting the status of pelvic lymph nodes remains a major issue. To date imaging has proved insufficiently accurate. We have ‘a chicken or an egg’ situation because to date the best method of predicting lymph node status is the clinical and pathological response to radiotherapy and chemoradiation. For this reason, it may be possible to predict retrospectively the initial status of the lymph nodes by the pathological response to chemoradiation in the primary tumour.
Long-Term Follow-up Is Required
Data on long-term outcome of 271 patients with rectal cancer suggest radical external beam radiotherapy is a reasonable management option in rectal carcinoma for patients who are not fit surgical candidates or refuse surgery [10]. However, despite achieving a cCR, the majority of patients in this series eventually failed locally. Habr-Gama’s data show recurrences occur late, and this observation is supported by other small series [15]. Hence, small studies with short follow-up do not add to our knowledge of the risks of this approach [64]. Current prospective studies will require a costly long-term follow-up programme of sequential MRIs and PET scans.
Increasing the Rate of Pathological Complete Response
Can we increase the likelihood of patients with rectal cancer achieving a pCR with more aggressive neoadjuvant treatment, such as neoadjuvant chemotherapy, biological agents, higher local radiation doses, more potent cytotoxic radiosensitisers or brachytherapy boosts on the primary tumour? To date these strategies have not been very effective [60, 65, 66]. A different option to increase pCR may be to lengthen the interval between chemoradiation and surgery. Other strategies are to increase the radiation dose to the primary with brachytherapy boost [67]. Habr-Gama recently reported that by extending the duration of the chemotherapy post-chemoradiation, a cCR of 48 % and an overall complete response (i.e. including cCR and pCR) could be increased to an astonishing 65 % [68].
This strategy could potentially be extrapolated to a full course of post-chemoradiation adjuvant chemotherapy using 5FU and oxaliplatin to consolidate the local response and address the potential for distant metastases. However, we are not aware of any randomised studies in this setting.
Retrospective studies reporting a ‘watch-and-wait’ approach have limitations because they are hampered by the landmark method, which omits important data on outcome for patients who fail in the first year. Habr-Gama attempts to provide information on this group [28]. In 23 such patients failing in the first year, the recurrence rate was 34 %. The authors suggest the outcomes of 5-year OS and DFS rates were 84 and 51.6 %, respectively, and not significantly different to the remaining patients who did not achieve a cCR and proceeded directly to radical surgery. Yet this is a flawed comparison, because the downstaging required to achieve a cCR (even if not sustained for 12 months) should select out a group with a more favourable prognosis.
The Role of Local Excision After Chemoradiation
Local excision and transanal endoscopic microsurgical resection (TEMS) are attractive alternative techniques to radical surgery because of the low morbidity and mortality and better functional outcome. However, the primary curative goal of surgery cannot always be achieved through these techniques, as the pelvic lymph nodes are not resected [69]. The stumbling block for accepting that in fit patients local excision is curative therapy is the challenge of developing and validating selection criteria that identify those patients where local excision alone is safe and does not compromise cure. Surgical series [70] suggest that the likelihood of having microscopic peri-rectal or mesorectal nodal involvement would be rare in T1 tumours but 10–30 % for T2 tumours and as high as 60 % risk in T3 tumours. For this reason, most surgeons would not accept that T3 tumours can be treated by local excision alone.
Retrospective studies demonstrate that tumours <3 cm in size, which are limited to superficial layers of the muscularis propria and are well or moderately well differentiated (as opposed to poorly differentiated), without lymphovascular invasion or extramural vascular invasion are the most reliable clinical and pathological features whereby the risk of lymph node involvement is predicted to be small and a local excision alone may prove sufficient treatment. Tumours with these characteristics are associated with a low risk of microscopic lymph node involvement and a late local recurrence rate of <10 % provided adequate surgical margins can be achieved. In principle chemoradiotherapy is likely to be more effective and pCR more frequent in such smaller earlier T stage tumours.
Hence, local excision might be a viable treatment option in more advanced tumours if patients are selected for avoidance of radical surgery by their response to preoperative CRT [18]. Following chemoradiation, patients who achieved a clinical good response had the residual tumour resected. With a mean follow-up at 24 months, there were no recurrences in patients who had a complete pathological response. In a similar study but using radiotherapy alone, an Italian group performed transanal endoscopic microsurgery (TEM) on the residual [71]. The local recurrence rate was only 2.85 % with a median follow-up of 38 months (range 24–96). However, few studies accurately document the rate of subsequent local recurrences that are amenable to salvage by AP excision of the rectum [72]. In summary, the advantage of preoperative chemoradiation is that the subsequent histopathology and regression can categorise patients between very high and very low risk.
In a study of 272 patients receiving preoperative chemoradiation [73], there were positive nodes in 1.6 % of ypT0, 6.3 % of ypT1 and 24 % of patients achieving tumour regression grade 2 (TRG2). A pooled analysis showed that after CRT, the local recurrence for patients achieving pCR was only 1 %, for ypT1 8 % and for ypT2 11 %, respectively [74].
The same author has reported an early analysis of a prospective study using short-course radiation in small rectal cancers prior to local excision. The rate of pCR was 41 %, and the strategy appears feasible [75]. A further small randomised study of chemoradiation prior to radical surgery or TEM in a highly selected group of small early tumours suggests that the results of conservative and radical surgery are equivalent if the tumour is downstaged to pCR or ypT1 [76].
Taking the assumption that the characteristics of patients in large randomised studies are likely to be the same in both arms, and that the populations of these studies reflect mainly patients with cT3/T4 stage, there were 33.3 % node-positive patients in the preoperative arm of the NSABP R03 compared with 47.5 % in the postoperative control arm [77]. This nodal sterilisation is almost identical in the POLISH study with 32 % in the preoperative chemoradiation arm versus 48 % in the SCPRT (where the histology will not have changed because insufficient time has elapsed for downstaging), respectively. In the German AIO trial [78], the node positivity was 25 % versus 40 % for the preoperative and postoperative arms. Data from these three trials of preoperative chemoradiation consistently suggest that if patients with cT3/T4 stage are selected, only approximately one third of involved lymph nodes will be sterilised by chemoradiation.
However, clinical nodal status may also impact on the chance of achieving a pCR. In a small retrospective study [63], clinically N0 tumours had a rate high rate of pCR – (p = 0.02); in contrast, only 3/33 (9 %) cN1–N2 patients responded with a pCR. So, node-negative patients may be more likely to achieve a pCR.
The most convincing evidence that patients with early stage presumed cN0 can be safely selected for a neoadjuvant, and local excision approach comes from the preliminary results of the ACOSOG Z6041 trial [79]. Patients with ultrasound-defined uT2N0 were treated with capecitabine and oxaliplatin and radiotherapy and local excision. In all, 36/90 achieved a pCR. Only six patients (6 %) had ypT3 tumours, and of the five local excision specimens, which contained lymph nodes, only one (a ypT3 tumour) had a positive node. The exciting results of this highly selected study demonstrated concordance between cCR and pCR in 31 of 36 patients.
So Where Do We Go from Here?
Some have tried to use the available data to develop a decision-analytic model examining the relative benefits of surgery versus observation in rectal cancer patients who achieve clinical complete response after neoadjuvant chemoradiation [80]. It seems unlikely that we can enhance the algorithm by achieving a well-designed non-inferiority randomised controlled trial on this question. Randomisation would probably be unacceptable to patients, and the numbers required would prove enormous.
We would therefore encourage careful observational studies in this setting. We are aware of several such studies. There is ongoing observational study at the Royal Marsden ‘Avoiding Surgery in Rectal Cancer After Pre-Operative Therapy’ (NCT01047969). The primary outcome measures are to estimate the percentage of patients who can safely omit surgery, defined as the percentage of patients at 2 years after end of CRT who have not had surgery and who are in CR (no detectable local disease), and also to prove the safety of deferred surgery, as measured by the percentage of patients who have local failure at 2 years, where local failure is defined as positive margin status of resected tumour or surgically unsalvageable disease. Many including the authors of this chapter feel that 2 years is a very short time frame for accepting the safety of this strategy.
There is also a Danish Colorectal Cancer Group Protocol (clinical trials gov. identification NCT00952926), which is a prospective observational study of patients with rectal cancer after concomitant radiation and chemotherapy. The objectives are to examine frequency of local recurrence at 1, 3 and 5 years after radiation and concomitant chemotherapy without subsequent operation patients with low rectal cancer. The first phase requires 30 patients. This study uses PET/CT as part of the follow-up evaluation.
A third study in the Netherlands (NCT00939666) follows complete responders, and some good responders are treated with a TEM. The United Kingdom in the North of England also has a registration programme run by surgeons. Finally in addition a formal European registry, the European Network for Watchful waiting (ENWW), has been initiated in Denmark – kfe.onk@slb.regionsyddanmark.dk
Conclusion
Controversial management options following CRT – such as a watch-and-wait approach (i.e. the omission of surgery) in case of a complete response, or local excision of the residual – remain experimental. These less invasive treatments have some obvious advantages, such as fewer potential surgical deaths, less morbidity in terms of urinary and faecal incontinence, sexual problems and fewer colostomies than after standard radical total mesorectal excision.
Yet the present limitations in assessing complete response still undermine our confidence in a watch-and-wait approach. Hence, methods to define pCR without resorting to radical surgery warrant further investigation and have been prioritised by the American College of Surgeons. Currently a local excision following chemoradiation is the most promising strategy. A local excision with complete or almost complete histopathological regression in the primary tumour may render radical pelvic surgery unnecessary. Selection of patients with cT2N0 tumours for this approach is more likely to demonstrate concordance between cCR and pCR than serendipitously observing more locally advanced tumours particularly cT3/T4 with an initial threatened circumferential margin. In addition, the role of additional ‘adjuvant’ chemotherapy needs to be defined.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
