“Near complete” response after total neoadjuvant therapy in patients with rectal cancer: close, but not close enough?

Total Neoadjuvant Therapy (TNT) for the treatment of locally advanced rectal cancer was incorporated into National Cancer Clinical Network (NCCN) guidelines in 2015, and has since changed the landscape of rectal cancer management worldwide. ¹ The rationale behind a TNT treatment regime hinges on improved chemotherapy compliance by front‐loading this treatment resulting in improved disease‐free survival, increased primary tumour complete response rates, and less time with a covering stoma. ² Importantly, TNT is also targeted at improving clinical downstaging, facilitating organ preservation via non‐operative management (NOM) and resulting lower rates of permanent stoma. ² TNT is not yet considered standard of care in Australia. However, a personalized TNT protocol was implemented at the Royal Adelaide Hospital (RAH) in 2019, with treatment sequencing based on oncological risk at presentation, as well as local ethics approval and prospective data collection. Published short‐term outcomes indicated that 40.5% of patients responded to treatment with a clinical complete response (cCR) using strictly defined criteria. ³ , ⁴

As we have embarked on this journey, however, it has become evident that the binary outcome of demonstrating a cCR or not is interjected with a grey area of what is commonly referred to as a ‘near’ complete response (nCR). With the push to organ preservation, we have had to repeatedly consider this entity in our discussions with patients. However, unlike criteria for a cCR, there is currently no clear consensus on the definition of nCR. ⁵ , ⁶ Patients categorized as having an nCR show considerable heterogeneity within their response to TNT with individual or a combination of features derived from each of the response modalities of digital rectal examination (DRE), endoscopy and medical imaging. This inconsistency was recently described in a systematic review including fibrosis or tissue hardening on DRE, presence of an ulcer, mucosal abnormalities or superficial ulceration, and small residual lesions observed via endoscopy. ⁷ Most studies included in the review utilized magnetic resonance imaging (MRI) and the extent of tumour downstaging with abnormal residual fibrosis and magnetic resonance tumour regression grading (mrTRG) 1–2 or mrTRG2, low or moderate residual signal were frequently indications of nCR. While lymph node status is not routinely considered in response evaluation to treatment, the degree of response is variable. Indicators of nCR in this context are reduction in short axis measurements or significant disease regression but still with enlarged lymph nodes. ⁷ After a local and statewide consensus meeting, and review of the published literature we have now conservatively defined nCR in our pTNT protocol as seen in Table 1.

Table 1.

Clinical criteria for complete and near‐complete clinical response (cCR, nCR)

	cCR	nCR
Digital rectal exam	No palpable tumour	Minor palpable irregularity
Flexi‐sigmoidoscopy	No visible tumour AND White scar Negative biopsy (not mandatory)	Residual flat ulcer Negative biopsy (mandatory)
MRI pelvis	TRG 1 or 2 Downsizing with no residual tumour OR Residual fibrosis only OR Residual wall thickening due to oedema with fibrosis AND No suspicious lymph nodes/EMVI	TRG 1 or 2 Downsizing with no residual tumour OR Residual fibrosis only OR Residual wall thickening due to oedema with fibrosis AND No suspicious lymph nodes/EMVI

It is not just the criteria that need to be defined, but the timepoint at which these are valid. Historically, timing of cCR assessment was based on maximal tumour response at ~8–10 weeks after long course chemo‐radiotherapy (LC‐CRTx). However, when LC‐CRTx is incorporated into a TNT protocol, response is assessed at the conclusion of all neoadjuvant treatment, including chemotherapy. In this setting, this timing will vary dependent on TNT sequencing administration. For patients receiving induction chemotherapy, 8–10 weeks post CRTx coincides with completion of all neoadjuvant treatment. Whereas, when a consolidation chemotherapy pathway is administered, formal assessment of cCR can equate to as much as 24 weeks after CRTx completion. The response to TNT treatment is reported to be time‐dependent and raises the question, how long do we ‘wait and see’ if conversion occurs from nCR to cCR? ⁵ For rectal cancer patients managed at the RAH, the decision for further surveillance (MRI and flexible sigmoidoscopy at 3‐monthly intervals) or proceeding with surgery is determined by MDT discussion in conjunction with patient wishes as part of a shared decision‐making process. In cases where patients do not demonstrate conversion from nCR to cCR the delay to surgery has been suggested to increase surgical complexity. ⁸ This is our experience in three specific cases where surgery was delayed by as much as 14 months and surgical resection was noted to be particularly difficult, mostly due to fibrosis. While our small sample size must be interpreted with caution, we do now consider delay to surgery as one of the risks of managing patients with nCR non‐operatively.

It is debated whether increasing time between treatment (particularly CRTx) and surgery is opening the door to local tumour regrowth or recurrence. ² This is particularly relevant when an nCR is initially observed and time to surgery is extended while ‘waiting’ for a definitive assessment of response to treatment. In cases of cCR and NOM, salvage surgery for local regrowth is suggested to increase the risk of distant disease failure, but local regrowth not amenable to successful surgical outcomes is uncommon with up to 90% of cases resulting in R0 resections. ² , ⁹ These observations are derived from a large international watch and wait database where patients were not managed with TNT. In a study that included patients who received consolidation chemotherapy, surgical outcomes were not compromised when surgery was delayed after cCR and NOM but only 42% of the population received chemotherapy and not analyzed separately, presumably due to small numbers. ¹⁰ A more recent review of a TNT approach to rectal cancer management indicated that whether patients completing TNT undergo surgery immediately after restaging (incomplete response) or surgery is delayed due to surveillance for conversion after nCR, local recurrence rates are similar and no different to standard treatment of chemoradiotherapy alone. ⁶

In our own database, 44% (11/25) of patients initially demonstrating a nCR have converted to cCR. Thus far, no local regrowth has occurred in these patients, which is consistent with the OPRA trial findings. ¹¹ Importantly, all patients with nCR who did not convert to cCR underwent successful interval salvage surgery with R0 resections (>1 mm margins), despite notable increased difficulty in some cases requiring wider margins to achieve this.

The benefits of a TNT approach to rectal cancer treatment are undeniable with increased cCR, potential for NOM and improved survival. Consequently, the management of locally advanced rectal cancer patients at the RAH underwent a significant change in 2019 when a TNT treatment strategy was introduced. The assessment of short‐term outcomes and continued prospective data collection has led to the necessity for nCR criteria to be incorporated into the protocol and a delicate balance that needs to be considered as to whether NOM is offered in selected cases of nCR. The answer to how long do we ‘wait and see’ remains unanswered and significant consideration must be given to whether a delay to surgery will compromise curative intent, particularly in medically fit patients. ⁹ Further, while the jury is still out on optimal sequencing administration, standardizing the reassessment period after nCR is met with more difficulty. The prospect of potentially avoiding surgery, leading to better quality of life and less time with a stoma, may be an attractive option for further surveillance but must be carefully considered against the impact on surgical complexity (if conversion does not occur), risk of regrowth, influence on distant disease failure and long‐term survival outcomes.

With data maturity, it is hoped that the personalized TNT protocol implemented at the RAH incorporating a definition of nCR will provide insight into the timing of reassessment after nCR and conversion rates to cCR and continue to improve rectal cancer patient outcomes.

Author contributions

Tracy R. Fitzsimmons: Writing – original draft. Tarik Sammour: Supervision; writing – review and editing.