The utility of short-course radiotherapy in a watch-and-wait strategy for rectal cancer – the need to measure the interval to tumour response assessment from the radiation start

The watch-and-wait strategy (w&w) is gaining momentum in rectal cancer [1–8]. Most of the w&w observations come from reports showing the outcomes after neoadjuvant long-course chemoradiation (CRT) [1,3,5,8]. The rate of patients undergoing successful w&w after short-course radiotherapy, i.e. 5 × 5 Gy, with delayed surgery (SCRT) or after SCRT combined with consolidation chemotherapy (SCRT + CT) is uncertain. Therefore, in a post hoc analysis of two pooled prospective studies on w&w, we aimed at addressing this issue.

The rate of complete tumour response to radiotherapy is increasing with extending the interval between radiation and tumour response assessment [9–12]. Traditionally, this interval is measured from the completion of radiotherapy. However, radiobiological findings indicate that it should rather be measured from its start. This is because each fraction of radiation sterilises the same proportion (not number) of surviving cancer cells [13]. Thus, the overwhelming cell-killing effect occurs during the first 1–3 fractions as shown in Supplementary Figure 1. Because of different treatment durations, this issue is important when comparing tumour response after SCRT with that after CRT.

The material and methods of both prospective studies were described in detail in our previous publications [6,14]. Briefly, the entry criteria were biopsy-confirmed adenocarcinoma of the rectum accessible to digital rectal examination (DRE) or adenocarcinoma of the anal canal (diagnosed when most of the tumour mass was located in the anal canal), and M0 status. Pelvic magnetic resonance imaging (MRI) was strongly recommended. The diagnosis of clinical complete response (cCR) had to be made using DRE, endoscopy and MRI. Criteria described in the literature were used to diagnose cCR or near-cCR [5,15]. In the patients with near-cCR or major tumour shrinkage, observation continued and the final decision on w&w or surgery had to be made after an additional 4–8 weeks. The first single-institutional study (trial A, ClinicalTrials.gov NCT01863862) evaluated w&w in consecutive elderly patients (≥70 years) with a small cancer (tumour length ≤5 cm and circumferential extent ≤60%). The primary end-point was the rate of regrowth in the patients achieving cCR and undergoing w&w. The second multi-institutional study (trial B, NCT04095468) evaluated w&w in consecutive patients receiving routine preoperative radio(chemo)therapy. The primary end-point was the frequency of achieving cCR.

The CRT schedule included 50 Gy given in 2 Gy per fraction over 5 weeks concomitantly with the Nordic schedule of chemotherapy that consisted of 5-fluorouracil 400 mg/m² in bolus and leucovorin 100 mg in short infusions on days 1, 2, 14, 15, 28 and 29. Only a few patients received capecitabine concomitantly. In the SCRT + CT schedule, only three chemotherapy cycles repeated every 2 weeks were given, starting one week after completion of radiation. The De Gramont schedule was most often delivered, i.e. leucovorin 200 mg/m²/day on days 1 and 2, and 5-fluorouracil bolus 400 mg/m²/day followed by continuous infusion 600 mg/m²/day on days 1 and 2.

The databases of the two trials were pooled. The time intervals were calculated from the start of radiotherapy. The percentage of cCR was used to evaluate treatment regimens for efficacy using a logit model. The schedule of radiotherapy and the variables listed in Table 1 were included in the logistic analysis. Categorical data were compared using the chi-squared test or the Fisher exact test. Cumulative incidence of regrowth was reported accounting for death as a competing risk; differences were compared using Gray’s test. All tests were two-sided. The data were analysed with IBM SPSS Statistics software version 23 for Linux (IBM Corp., Armonk, NY, USA) and R software (www.r-project.org).

Table 1. Patient and tumour characteristics in relation to the schedule of neoadjuvant radiotherapy.

	The percentages of patients (number of patients)
	5 × 5 Gy N = 199	5 × 5 Gy +  chemotherapy N = 197	Chemoradiation N = 94	p-value
Diagnosis				0.04
Rectal cancer	99% (196)	99% (194)	94% (88)
Anal canal adenocarcinoma	2% (3)	2% (3)	6% (6)
Age in years				<0.001
<60	20% (39)	32% (62)	26% (24)
60-66	19% (38)	31% (60)	31% (29)
67-74	22% (44)	27% (54)	26% (24)
≥75	39% (78)	11% (21)	18% (17)
Participating institution				<0.001
1	55% (109)	84% (166)	31% (29)
2	7% (13)	3% (6)	0
3	13% (26)	5% (9)	1% (1)
4	16% (31)	6% (12)	36% (34)
5	5% (9)	1% (1)	6% (6)
6	6% (11)	2% (3)	26% (24)
Distance between tumour and the anal verge in cm				0.025
0-2	17% (33)	19% (38)	30% (28)
3-4	32% (63)	31% (61)	38% (36)
5	25% (49)	21% (41)	11% (10)
6-8	27% (53)	29% (57)	21% (20)
Unknown (number)	1
cT category				<0.001
cT1	1% (1)	1% (1)	2% (2)
cT2	19% (38)	8% (16)	22% (21)
cT3	74% (148)	68% (134)	57% (54)
cT4	6% (12)	23% (46)	18% (17)
cN category
N0	54% (106)	28% (55)	35% (33)
N+	47% (92)	72% (142)	65% (61)
Unknown (number)	1
Tumour length in cm				<0.001
1-3	16% (32)	12% (23)	22% (21)
4	30% (59)	17% (34)	22% (21)
5-6	35% (69)	38% (75)	36% (34)
≥7	19% (37)	33% (64)	19% (18)
unknown (number)	2	1
% of circumferential extent				<0.001
≤40	22% (44)	17% (33)	34% (32)
50	33% (66)	26% (51)	22% (21)
60-90	23% (45)	20% (40)	31% (29)
100	22% (43)	37% (72)	13% (12)
unknown	2	1
Mesorectal fascia threatened				<0.001
No	76% (129)	24% (45)	54% (43)
Yes	24% (41)	76% (146)	46% (37)
unknown	29	6	14
Tumour mobility				<0.001
Movable	89% (163)	71% (136)	92% (81)
Fixed	11% (21)	29% (55)	8% (7)
Unknown (number)	15	6	6
Interval in weeks between start of RT and the first response evaluation				<0.001
<9.1	34% (68)	23% (45)	4% (4)
9.1 − 11.4	27% (54)	35% (69)	5% (5)
10.4 − 12.0	23% (45)	31% (60)	18% (17)
>12.0	16% (32)	12% (23)	72% (68)
Study				<0.001
Study A	17% (33)	2% (4)	31% (29)
Study B	83% (166)	98% (193)	69% (65)

Study A analysed 66 consecutive patients who met the entry criteria in the single institution. Study B registered 424 consecutive patients from six regional tertiary-care institutions. In total, there were 490 patients. SCRT, SCRT + CT and CRT were received by 41% (n = 199), 40% (n = 197) and 19% (n = 94) of patients, respectively. There was an apparent selection of the patients to each of the three radiotherapy schedules (Table 1). Most often SCRT was given in elderly patients or in patients with a small cancer, whereas SCRT + CT was given in advanced cancers. The application of CRT varied between the participating institutions. The median intervals from the radiation start to the first tumour response assessment after SCRT, SCRT + CT and CRT were 10.0, 10.4 and 13.2 weeks, respectively.

In total, 73 (15%) patients achieved cCR. In 51% of them, the diagnosis was made at the first assessment and in 49% after additional waiting. The multivariable logistic analysis showed that only tumour length, circumferential extent within the rectal wall and cN status were significant independent predictors for cCR (Supplementary table 1). The odds ratios of cCR after SCRT and SCRT + CT were 0.53 (95% confidence interval [CI] 0.26–1.09) and 0.83 (95% CI 0.40–1.75), respectively, compared to CRT. Using the tumour length and the circumferential extent, as indicated by the multivariable analysis, cancers were categorised as having a high, intermediate or low chance for cCR. Detailed data justifying such stratification were presented in our previous report (6). Table 2 shows the proportions of cCR achieved after SCRT, SCRT + CT and CRT in the entire cohort and in these three tumour categories. For example, in the high-chance category (≤4 cm tumour’s length and ≤50% of circumferential extent), the proportions of cCR achieved after SCRT, SCRT + CT and CRT were 26% (95% CI 16%–37%), 38% (95% CI 24%–52%) and 45% (95% CI 29%–61%), respectively.

Table 2. Clinical complete response (cCR) in relation to the radiotherapy schedule and categories of tumour grouped by the chance of achieving cCR.

	% of patients with cCR (number of patients with cCR/ total number of patients in a subgroup) N = 490	95% confidence interval
Entire cohort
5 × 5 Gy	14% (27/199)	9%−19%
5 × 5 Gy + chemotherapy	13% (25/197)	8%−18%
Chemoradiation	22% (21/94)	14%−30%
Subgroup analyses^a
High chance for cCR
5 × 5 Gy	26% (18/69)	16%−37%
5 × 5 Gy + chemotherapy	38% (17/45)	24%−52%
Chemoradiation	45% (17/38)	29%−61%
Intermediate chance for cCR
5 × 5 Gy	10% (7/68)	3%−17%
5 × 5 Gy + chemotherapy	10% (5/51)	2%−18%
Chemoradiation	9% (3/32)	0%−19%
Low chance for cCR
5 × 5 Gy	3% (2/60)	0%−7%
5 × 5 Gy + chemotherapy	2% (2/100)	0%−5%
Chemoradiation	4% (1/24)	0%−12%

^aThe high chance for cCR was assigned to cancers involving ≤50% of the circumferential extent and ≤4 cm of the length. The low chance for cCR was assigned to circular cancers or with ≥7 cm tumour length. The intermediate-chance category includes all other cancers [6]. Three patients were excluded from the subgroups analyses because of the lack of data with respect to the tumour length or circumferential extent. The crude rate of distant metastases in patients who entered the watch-and-wait program and were assigned to the high chance category was 13% (7 of 52), intermediate-chance category 0% (0 of 15) and low chance category 1 of 4. The corresponding figures for the rate of all causes of deaths were 17% (9 of 52), 13% (2 of 15) and 1 of 4, and for the rates of deaths related to rectal cancer 8% (4 of 52), 0% (0 of 52) and 0 of 4.

Of the 73 patients who achieved cCR, 71 underwent w&w. In the patients undergoing w&w, the median follow-up times after SCRT, SCRT + CT and CRT were 28, 21 and 24 months, respectively. At 2 years, the regrowth rate was 13.0% (95% CI 4% − 22%) after SCRT + CT, which was numerically lower compared with the 30% (95% CI 18% − 43%) observed after SCRT or 36% (95% CI 22% − 50%) after CRT. However, after 3 years, the regrowth rates were similar and amounted to approximately 35% − 40% in all the three subgroups.

Our analysis showed that SCRT or SCRT + CT provided a clinically relevant chance for cCR (Table 2). In the patients undergoing w&w the rates of regrowth at 3 years after SCRT, SCRT + CT or CRT appear to be similar. To our knowledge, the current study is the first in the literature to show the cCR rates after SCRT and SCRT + CT compared with that achieved after CRT. However, the evidence level is low. This was an unplanned subgroup analysis. Thus, our results can only be used for hypothesis generation because of inherent bias resulting from a different selection of patients (Table 1). Moreover, our study was underpowered to detect significant differences in cCR. The strength of our study is the presentation of real-life data, because in the muli-institutional setting (study B) we included consecutive patients.

We performed a systematic search of the literature to gather evidence on cCR rate after SCRT, SCRT + CT or CRT. We also collected evidence on pCR after SCRT, SCRT + CT or CRT using the results of our previous systematic literature search [9] along with an updated search. We included all studies comparing the efficacy of different radio(chemo)therapy regimens in achieving cCR and randomised or population-based studies comparing the efficacy of different radio(chemo)therapy regimens in achieving pathologic complete response (pCR).

We found no study on cCR after SCRT and only two small studies showing cCR after SCRT + CT [2,7]. In the prospective study (n = 19) on SCRT followed by 16 weeks of FOLFOX or CAPOX, Kim et al. [7] reported 74% of cCR and 57% (95% CI 36%–91%) of organ preservation at 2 years. In the another study (retrospective) by Jia et al. [2] on SCRT followed by 8 weeks of FOLFOX, 19 patients were evaluated by endoscopy after a median of 16  weeks from the radiation start. Nine (47%) of these patients achieved cCR. The above-mentioned cCR rates were much higher than the 13% obtained in our study after SCRT + CT (Table 2). This difference can be explained by the shorter consolidation chemotherapy in our study with corresponding shorter intervals to response assessment and by the larger tumour size, median 6 cm in our study vs. 4 cm in the trial by Kim et al. [7].

There were no reports of randomised trials comparing pCR rate after SCRT with that after SCRT + CT. Thus, there is no strong evidence that adding sequential chemotherapy to SCRT enhances the cell-killing effect. Three small randomised studies compared SCRT with CRT [16–18]. All these studies showed a statistically insignificant difference in the pCR rate, although numerically favouring CRT (Supplementary table 2). However, it should be taken into account that interval to surgery from the start of radiation was 4 weeks longer in the CRT groups than in the SCRT groups. This resulted in an inherent bias in favour of CRT.

Five randomised trials compared SCRT + CT with CRT [19–23] (Supplementary table 2). The Polish II study had the purest design because the duration of consolidation chemotherapy in the SCRT + CT group of only 6 weeks resulted in the identical interval (12 weeks) from the start of radiation to surgery in the SCRT + CT and CRT groups [16]. In this study, the pCR rates in the SCRT + CT group and in the CRT group were 16% and 12%, respectively, p = 0.17. In the RAPIDO trial, in which the duration of consolidation chemotherapy (18 weeks) and the interval to surgery were the longest, the corresponding values were 28% and 14%, p < 0.001 [20]. Supplementary table 2 shows that the pCR rate in the SCRT + CT groups tended to increase across the studies with the increasing duration of consolidation chemotherapy. It is impossible to resolve whether the increase in the percentage of pCR with the extension of the duration of consolidation chemotherapy in the SCRT + CT groups was caused by the longer chemotherapy itself, prolongation of the interval to surgery or both.

We found two non-randomised population-based studies showing pCR rate in relation to the radiotherapy schedule [4,24]. Hammarström et al. [4] in a Swedish population-based retrospective study compared the results between SCRT, SCRT + CT (12 or 16 weeks of CAPOX) and CRT. They compared the complete response (CR) occurrence defined as the sum of sustained cCR for at least one year after treatment in the patients undergoing w&w and pCR in those receiving surgery. The rates of patients with CR in the SCRT (n = 155), SCRT + CT (n = 94) and CRT (n = 123) groups were 8%, 27% and 18%, respectively. These differences between the SCRT group and the SCRT + CT or CRT groups were statistically significant (p < 0.001 and p = 0.01, respectively), whereas it was not between the SCRT + CT and CRT groups (p = 0.12). Rombouts et al. [24] in a Dutch population-based retrospective study reported that the pCR rates after SCRT (n = 764) or after CRT (n = 5070) was 9.3% and 17.5%, respectively; the odds ratio was 0.37 (95% CI 0.24–0.57). In both studies above, the patients receiving SCRT had less advanced cancers compared with those receiving SCRT + CT or CRT. This favoured SCRT. On the other hand, in the Swedish study, the 8–9-week interval between the start of radiation and surgery in the SCRT group was shorter than the 14 weeks in the CRT group and the 20 weeks in the SCRT + CT group. Thus, the duration of this interval favoured SCRT + CT or CRT over SCRT. Taking the current results and all the above literature data together, it seems that SCRT + CT and CRT provides similar cancer cell-killing effect. The currently on-going START-TREC randomised trial will show organ preservation rate after SCRT in relation to that after CRT [25].

Conclusions

SCRT or SCRT + CT provided a clinically relevant chance for cCR. Previously reported radiobiological findings indicate that the interval to response assessment should be measured from the start and not from the completion of radiation.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The database has been deposited in a repository and is available upon request.

Additional information

Funding

The study B was supported by sponsored by the Oncology 2025 Foundation and by the Polish Society for Surgical Oncology.