Rectal sparing is an option for some rectal cancers with complete or good response after chemoradiotherapy (CRT); however, it has never been evaluated in patients with metastases. We assessed long-term outcomes of a rectal-sparing approach in a liver-first strategy for patients with rectal cancer with resectable liver metastases.

We examined patients who underwent an organ-sparing approach for rectal cancer with synchronous liver metastases using a liver-first strategy during 2010-2015 (n = 8). Patients received primary chemotherapy and pelvic CRT. Liver surgery was performed during the interval between CRT completion and rectal tumor re-evaluation. Clinical and oncological characteristics and long-term outcomes were assessed.

All patients underwent liver metastatic resection with curative intent. The R0 rate was 100%. Six and two patients underwent local excision and a watch-and-wait (WW) approach, respectively. All patients had T3N1 tumors at diagnosis and had good clinical response after CRT. The median survival time was 60 (range, 14-127) mo. Three patients were disease free for 5, 8, and 10 years after the procedure. Five patients developed metastatic recurrence in the liver (n = 5) and/or lungs (n = 2). Only one patient developed local recurrence concurrent with metastatic recurrence 24 mo after the WW approach. Two patients died during follow-up.

The results suggest good local control in patients undergoing organ-sparing strategies for rectal cancer with synchronous liver metastasis. Prospective trials are required to validate these data and identify good candidates for these strategies.

Rectal cancer affects nearly 10000 new patients every year in France, among whom 20%-25% present with synchronous liver metastases. Despite oncological advances, the only potentially curative therapy remains surgical resection or destruction of lesions at both sites[1]. Rectal and liver resections can achieve 5-year survival rates of > 50%[1,2] compared with only approximately 5% for patients treated with palliative intent[3].

Because the prognosis of these patients is directly related to the presence of liver metastases and because complications of rectal surgery are common after chemoradiotherapy (CRT) and may therefore delay the start of appropriate metastatic treatments, the liver-first approach has been proposed for patients with locally advanced rectal cancer and synchronous liver metastases[4-7]. Thus, patients receive computed tomography (CT) first, followed by liver surgery, before and/or after CRT depending on the team. Triplet CT and newer targeted therapies such as cetuximab and bevacizumab have led to improved response rates at both sites[8] and conversion rates to hepatic resectability[9,10].

Rectal pathological complete response has been observed in 15%-20% of patients after standard CRT[11] and in up to one-third of cases after adding triplet CT, following the same pattern as that for patients with metastases[12].

In these conditions, the question of whether to maintain the indication for radical surgery or total mesorectal excision (TME) has been raised by several therapeutic trials evaluating rectal-sparing strategies in patients without metastasis[13,14]. In France, the most widely evaluated strategy is local excision (LE) via the transanal approach. This strategy is reserved for patients with an initially favorable lesion (T2 or low T3 of less than 40 mm). The rationale of this strategy compared to radical surgery is based on the preservation of quality of life (QoL) and digestive and urogenital functions with identical oncological efficacy owing to rectal preservation and the absence of surgical nerve damage[15,16]. Recent studies have shown that LE is a safe alternative for TME for patients who are good responders after CRT for T2T3N0-1 mid-to-low rectal cancer[13,17] with a 5-year local recurrence rate of 7%. Although this strategy has not been evaluated in patients with metastases, the rationale remains similar, i.e., to improve the QoL of patients whose prognosis is related to a higher risk of hepatic recurrence than the risk of local recurrence. Thus, this study aimed to assess long-term outcomes of a rectal-sparing approach in a liver-first strategy for selected patients with rectal cancer with resectable liver metastases.

The oncological strategy was chosen as a function of the overall condition of the patient and the resectability of the liver metastasis and rectal tumor in our multidisciplinary meetings (including liver surgeons, rectal surgeons, oncologists, radiotherapists, radiologists, and pathologists).

Rectal sparing within a liver-first strategy for rectal cancer with resectable liver metastases.

Currently, the treatment of colorectal liver metastases (CRLM) remains a major clinical challenge without a consensus[20]. The case-by-case treatment strategy is determined according to: (1) Tumor and disease-related characteristics, patient-related factors, and treatment-related factors such as toxicity and main oncological problems; (2) presence or absence of predictive factors for rectal and liver resection morbidity; and (3) response to initial CT. New regional and systemic chemotherapies associated with biological agents combined with technical advances in liver surgery have made it possible to broaden indications for CRLM resection by offering personalized treatment.

For rectal tumors, TME remains the only available treatment option with curative intent in patients with metastatic rectal cancer, regardless of the response to neoadjuvant therapy. However, a complete clinical response or a very good response is observed in 15%-20% of patients after standard CRT and in up to one-third of cases after addicting CT, as suggested by a recent randomized controlled trial (RCT) in patients without metastasis[12,21].

Rare cases of rectal-sparing strategies in patients with metastases have been described: WW[22,23] and LE[4] in the liver-first strategy. A WW strategy was used in nine cases as a result of primary tumor disappearance after RCT[22-24]. Unfortunately, no study has specified the characteristics of rectal lesions or oncological outcomes of these patients. Mentha et al[4] and Buchs et al[25] reported two cases of LE with complete clinical response after RCT. One case in 2006[4] did not have any long-term data. Another case in 2015[25] had a confirmed pathological response after RCT but had recurrence 11 mo later and underwent abdominoperineal resection with a final staging of pT3Nx.

In a Dutch study[7], a rectal-sparing strategy could have been proposed in ten patients who had a complete response of their primary tumor after complete treatment according to a liver-first strategy, as introduced by Mentha et al[4]. This strategy involves systematic preoperative CT and resection of CRLM, followed by pelvic RCT and rectal resection. In our optimised liver-first strategy, liver surgery is performed at the interval between radiotherapy completion and rectal surgery. This strategy allows rectal re-evaluation without increasing the time without CT. Prolonging the interval between CRT completion and rectal staging increases the complete clinical response rate[26]. Thus, it allows for a better selection of patients who can benefit from a rectal-sparing strategy without increasing surgical morbidity[26,27].

Short-course radiotherapy followed by CT and delayed rectal surgery[21] is an option in the neoadjuvant setting of resectable rectal cancer that could potentially be adapted for patients with metastases[24]. This would make it possible to limit the time without CT while maintaining a delayed rectal reassessment and possibly proposing a rectal-sparing strategy in cases of good clinical response. Nevertheless, the oncological safety of this strategy has not been evaluated in specific studies.

It is important to note that we have a highly selected population after applying the two-stage selection criteria in the organ preservation for rectal cancer (GRECCAR 2) trial; we considered the initial rectal tumor characteristics and the clinical response to CRT. Seven of the eight patients studied had an initial N + tumor according to routine EUS and MRI. The initial lymph node involvement, especially the lymph node response after CT and RCT, is difficult to specify formally[28].

In addition to oncological multidisciplinary meetings, weekly meetings are organized with specialized radiologists and colorectal surgeons to review all examinations, including surveillance MRI, to improve our patient selection. Our results are consistent with those of GRECCAR 2 study[13], as we observed no lymph node recurrence among patients undergoing LE. Four patients had no residual tumor (ypT0), but two patients had residual ypT2 tumors equivalent to a risk of residual lymph node involvement evaluated at 8%. This risk is probably lower given the low TRG (TRG 1: few residual cells). Given the discordant results and the absence of validated criteria, the WW strategy seems to be reserved only for patients without residual scarring and is subject to very strict surveillance.

In patients without metastasis, the GRECCAR 2 trial’s 5-year results provide no evidence of differences in long-term survival (84% vs. 82%; P = 0.85) or cancer-specific mortality (7% vs. 10%; P = 0.53) between LE and TME[17].

In all cases, a favorable pathological response is associated with good prognosis and survival benefit[29]. Under these conditions, whether to maintain the indication for radical surgery in good responders or even in complete clinical responders is an issue that has never been raised in patients with metastases.

The oncological safety of rectal-sparing strategy has never been evaluated in patients with metastases but needs to be balanced with morbidity or functional benefits. Minimizing operative morbidity is a major issue for strategy treatment choice as it is an independent factor for overall survival and disease-free survival after CRLM resection[30]. The rectal-sparing strategy induces a more favorable global health status and bowel function than TME after CRT[16,31]. The effect of rectal cancer treatment on functional outcomes and patients’ QoL must now be considered in the decision-making process whenever possible.

To the best of our knowledge, this is the first study to provide detailed characteristics and long-term results of patients undergoing a rectal-sparing strategy for rectal cancer with synchronous liver metastasis. Our results are encouraging compared to the prognoses of patients with metastases in the literature because only one patient had a local rectal recurrence with concurrent hepatic recurrence using the WW strategy 3 years after liver surgery.

The present study has some limitations and caution must be exercised in interpreting its results given the small sample size. The rectal-sparing strategy requires coordinated action by a multidisciplinary team and depends on many criteria, including treatment times and tumor response to therapy. Moreover, patients are not always referred to our center at the time of diagnosis and have already started CRT, which does not allow for a first liver strategy and limits potential inclusions.

Second, this was a retrospective single-center study. In the absence of clear recommendations, practices vary widely from one center to another in the surgical and oncological management of CRLM, which hinders the realization of a multicenter study. Imposing the same protocol on several teams and institutions, with selection criteria often different from their usual practice, is an obstacle to its large-scale implementation.

In conclusion, although our findings should be interpreted with caution given the small sample size and high patient selection, we suggest that rectal-sparing strategies must become an option in expert centers to improve the QoL of patients with CRLM.