Complete surgical resection (R0 resection) is the cornerstone of curative therapy for rectal cancer and is the goal of all operative intervention. Historically, treatment consisted of complete rectal resection using blunt dissection of the rectal fascia. However, this was associated with local recurrence rates up to 30 % [33]. This blunt dissection failed to obtain clear circumferential resection margins (CRM), which has since been shown to be an independent predictor of local recurrence and survival [9, 34].

The promotion of the concept of total mesorectal resection (TME) in 1982 drastically changed the techniques of rectal resection [35]. By utilizing sharp dissection of the avascular plane between the visceral and parietal endopelvic fascial layers, the entire mesorectum, including rectal lymphatic drainage, may be excised en bloc with the rectum itself. This paradigm shift in surgical approach drastically improved outcomes, with local recurrence rates dropping from 30 % to as low as 5 % and 5-year survival increasing from 50 % to nearly 75 % [36]. Mesorectal excision has been proven to be safe, with decreased morbidity and increased preservation of the pelvic autonomic nerves. TME, as part of a low anterior resection or abdominoperineal resection, is the operation of choice for tumors of the middle and lower rectum [12]. For tumors of the upper rectum, a complete TME may not be necessary. However, pathologic examinations have identified tumor cells in the mesorectum up to 4 cm distal to the primary tumor [37, 38]. As such, current recommendations are for a tumor-specific mesorectal resection for tumors of the upper rectum, extending no less than 5 cm distal to the lower tumor margin [12].

The level of optimal vascular ligation has been debated. Ligation at the level of superior rectal artery origin (“low tie”) has been shown to provide an appropriate proximal lymphatic resection [39]. A higher ligation including the inferior mesenteric artery (“high tie”) may increase lymph node yield, though there is no difference in survival when compared to a lower ligation [40]. Additionally, high ligation of the inferior mesenteric artery allows for improved mobilization to create a tension-free anastomosis without increased risk of anastomotic leak [41]. As such, although official recommendations indicate that a low tie is acceptable in the absence of clinically suspicious nodes above that level, many surgeons prefer and recommend a high tie in all patients [12].

Distal intramural spread is uncommon and rarely extends more than 1 cm [42, 43]. In light of this, a distal resection margin of 2 cm (down from an original minimum of 5 cm) is currently recommended [12]. For low lying tumors located at or below the mesorectal margin where a 2 cm margin would necessitate abdominoperineal resection, a 1 cm margin has been shown to be acceptable and may allow for increased rates of sphincter preservation through ultra-low anterior resection [44]. Meta-analysis has shown that early tumors, downstaging following neoadjuvant therapy, low serum CEA levels and good or moderately differentiated tumors are most suitable for distal resection margin <2 cm [45].

Sphincter preservation is an important part of rectal cancer surgery and should be considered when adequate margins are feasible. While abdominoperineal resection (APR) remains the treatment of choice when safe distal margins cannot be obtained, there is debate surrounding the possibility of increasing sphincter preservation in these patients. Some argue that neoadjuvant therapy may allow for tumor downsizing, which could permit low anterior resection in patients initially felt to require APR. There is no conclusive data regarding this subject. Despite some researchers, such as the German Rectal Cancer Study Group, demonstrating a statistically significant increase in the number of sphincter-sparing operations in patients who received neoaduvant therapy, others have shown no clear benefit [18, 23, 46]. There is currently no consensus or official recommendations regarding sphincter preservation, and additional studies are required. Patients with very low tumors, poor preoperative function, involvement of the levators and sphincter complex, and incontinence are best served with APR [47].

When sphincters can be preserved, the formation of a colonic reservoir should be considered to decrease the incidence of urgency, clustering, increased bowel frequency, and incontinence. Colonic J-pouch has been demonstrated to be superior to a straight coloanal anastomosis, with reduced urgency and bowel frequency [48, 49]. Defunctioning stoma, with loop ileostomy preferred, is recommended for all patients to reduce anastomotic leak and reoperation rates [50].

Surgical resection may be more extensive for patients with T4 tumors that invade into surrounding pelvic structures or with N1 tumors that have nodal involvement. To achieve an R0 resection, all involved structures should be resected en bloc along with the rectum and mesorectum [12]. This most often involves the bladder, prostate, vagina, uterus, ovaries, and ureters and a multi-disciplinary surgical team is advised when available. Partial resection of the involved organ may be feasible, however, complete resection is often required as part of a multivisceral resection. Resection of all direct adhesions to the tumor is also recommended, as it is impossible to assure no residual microscopic disease in these structures, especially following radiation to the area. For more extensive disease, and in patients with prostate and/or bladder involvement, a total pelvic exenteration is often indicated for optimal resection and reconstruction. Five-year survival and local recurrence rates are comparable to those of non-locally invasive tumors [51]. Lateral lymph node dissection should be performed in patients with clinically suspicious nodes. While Japanese studies have shown lateral lymphadenectomy in patients without apparent clinical involvement improved locoregional control, no oncologic benefit was found on meta-analysis when compared to conventional surgery [52, 53].

Tumor fixation to the bony and muscular structures of the lateral and posterior pelvis presents a difficult situation due to the difficulty in removing these structures to achieve R0 resection. Additionally, some tumors may have very narrow or involved margins that make complete resection challenging. Intraoperative radiotherapy (IORT), first developed in the 1960s in Japan, has been used as an adjunct to improve local control in these situations where there is expected microscopic or gross involvement [54, 55]. This technique allows for the delivery of focused radiation specifically to at-risk areas under direct visualization and with minimal exposure of surrounding structures. IORT is delivered using intraoperative electron beam radiotherapy (IOERT) or high-dose-rate brachytherapy (HDR-IORT), with no significant differences in outcomes between the methods [56, 57]. There have been no randomized trials investigating IORT, however, several series have demonstrated it to be safe with improvement in local control and overall survival [51, 55, 58–60]. These series have been relatively small, and some other studies have not been able to demonstrate significant survival benefits [58, 61–63]. Because of this IORT has not been adapted into current standards of care, however, it may be utilized as an adjuvant when available for difficult tumors that may otherwise be deemed unresectable due to margin involvement.