Introduction

Rectal cancer is a common malignancy in both men and women, with an estimated 43,030 new cases diagnosed in the USA in 2018 [1]. Over the past several decades, local recurrence rates have declined dramatically due to improved surgical techniques and the use of neoadjuvant therapy [2,3,4,5,6,7]. As a result, the focus of clinical research has shifted to reducing the risk of distant metastatic disease, a significant source of morbidity and mortality for these patients [2, 4,5,6, 8]. The most common treatment paradigm in the USA employs neoadjuvant chemoradiation, surgery, and adjuvant chemotherapy for patients diagnosed with stage II–III disease. The use of adjuvant chemotherapy in rectal cancer is primarily an extrapolation from its use in patients with colon cancer [9, 10] and has been widely implemented in clinical practice despite limited data. An emerging and increasingly utilized approach is described as “total neoadjuvant therapy” or TNT. This approach transitions the use of systemically dosed chemotherapy to the neoadjuvant setting and has shown significant promise in multiple clinical trials. Here, we highlight the prominent literature evaluating this approach, including a review of its merits and potential shortcomings.

Historical Context

Recurrence patterns for rectal cancer have shifted over the past 30 years with improved local control rates attributable to implementation of total mesorectal excision (TME) and chemoradiation or short-course radiation therapy in the neoadjuvant setting. TME was first described in the 1930s but gained widespread attention in the late 1970s and early 1980s. With TME, a precise, sharp dissection between the visceral and parietal layers of the endopelvic fascia is performed. This includes en bloc removal of the mesorectal envelope as a single packet of tissue, with the surrounding lymphatics, vascular and perineural tumor deposits, with the primary tumor encompassed inside. Additionally, this approach is more likely to preserve autonomic nerve function in the pelvis and reduces the risk of post-operative bleeding. This technique replaced the historical blunt dissection, which often violated the mesorectal envelope leaving residual disease in the pelvis [13, 14]. TME has reduced the risk of both involved surgical margins and local recurrences in the pelvis [3, 4].

For patients with locally advanced rectal cancer, including those with T3-T4 and/or node-positive disease, the use of neoadjuvant therapy improves both local control and disease-free survival [5, 15]. Multiple trials support the current treatment paradigm in the USA [5, 15,16,17,18,19], which includes the use of 5-fluoropyrimidine (5-FU)-based chemotherapy given concurrently with neoadjuvant radiation therapy utilizing a “long-course” approach (treating to approximately 50.4 Gy over 5.5 weeks). Alternatively, “short-course” radiation therapy alone (treating to 25 Gy in 5 fractions) is more widely used in other parts of the world, including Northern Europe [20, 21].

In the past, local failure was a significant source of morbidity and mortality for patients with rectal cancer. However, with the aforementioned enhancements in locoregional therapy, the risk of developing a local recurrence has decreased from > 25% to approximately 5–10% [2,3,4,5,6,7]. This has shifted the leading cause of morbidity and mortality for these patients to distant metastases, which occurs in approximately 30% of patients [2, 4,5,6, 8]. Historical trials of the 1970s and 1980s demonstrated an improvement in overall survival in patients with rectal cancer receiving adjuvant therapy, including the use of adjuvant chemotherapy [22]. There are also prospective, randomized trial data supporting the use of adjuvant chemotherapy in patients with colon cancer, demonstrating an improvement in overall survival in those with node-positive disease [9, 10]. However, trials specifically evaluating the utility of adjuvant chemotherapy in patients with rectal cancer have largely failed to demonstrate benefit [2, 11•, 12, 13, 23, 24]. For example, one of the larger trials from the European Organization for Research and Treatment of Cancer (EORTC 22921) did not demonstrate a disease-free or overall survival benefit of adjuvant 5-FU therapy for patients with locally advanced disease. The risk of distant metastases remained approximately 30% for these patients at 10 years [2, 11•, 12]. Some of these trials were likely underpowered to detect a difference in survival due to their slow accrual and early closure [23, 24], although the EORTC and an Italian trial did not demonstrate benefit despite adequate patient sample size [2, 13]. Additionally, these trials primarily utilized 5-FU alone in the adjuvant setting, as opposed to oxaliplatin-based adjuvant therapy.

Rationale for TNT

One of the most prominent challenges in assessing the efficacy of adjuvant chemotherapy is poor patient tolerance. For example, in the EORTC 22921 trial, only 73% of patients randomized to receive adjuvant chemotherapy actually received any and only 43% received the majority of planned doses [2]. The previously mentioned Italian trial also showed poor patient compliance with adjuvant therapy with only 58% of patients receiving at least half of the six planned cycles of 5-FU-based chemotherapy [13]. The most common reasons for not completing intended therapy in these trials include disease progression, patient refusal, and post-operative morbidity [2, 13]. The need to reduce the risk of distant progression and poor compliance with adjuvant therapy necessitates the development of new treatment strategies to address this clinical scenario.

Investigators have touted the use of systemically dosed chemotherapy in the neoadjuvant setting, with the goal of avoiding some of the pitfalls associated with adjuvant therapy. It would seem likely that patient compliance would improve in the neoadjuvant setting, as patients would not be handicapped by post-operative morbidity and lower performance status. Given the risk for mortality associated with distant metastatic disease, there is potential benefit in treating micrometastatic disease earlier in a patient’s disease course. Additionally, less treatment-related toxicity in the post-operative setting may allow appropriate patients to undergo ostomy reversal at an earlier time point following low anterior resection (LAR) [8]. For those patients with locally advanced disease (T3-T4 and/or node-positive), chemotherapy in the upfront setting may allow for enhanced tumor response, potential downstaging, and higher rates of complete (R0) resection. Tumor downstaging may permit more patients the opportunity for organ preservation, both from a surgical perspective (low anterior resection (LAR) in lieu of abdominoperineal resection (APR)) and utilization of the emerging “watch-and-wait” approach. The watch-and-wait approach seeks to identify those patients who are good candidates for omission of surgery, in favor of a strict surveillance program [25].

Figure 1 depicts representative timelines for the treatment paradigms utilized in the management of rectal cancer.

Fig. 1
figure 1

Example timelines for each of the following treatment paradigms. a Total neoadjuvant therapy (TNT) with long-course chemoradiation (CRT) followed by neoadjuvant chemotherapy (NAC). b Long-course CRT. c TNT with short-course radiation followed by NAC. d Short-course radiation. e TNT with NAC followed by long-course CRT. f long-course CRT with adjuvant chemotherapy

Full size image

Potential Shortcomings

While there are many potential benefits of TNT, rigorous evaluation of disease-related outcomes is needed before its widespread adoption. One potential disadvantage to its use is the delay to definitive surgical resection. In those patients who do not adequately respond to neoadjuvant chemotherapy, local progression is possible. Upfront systemic therapy and the associated toxicities also have the potential to impact a patient’s ability to undergo a definitive surgical procedure by negatively impacting performance status.

Clinical Outcomes: Is TNT Beneficial?

In recent years, multiple trials have published on the outcomes of patients treated with TNT. Two general approaches have been used: [1] neoadjuvant radiotherapy (±concurrent chemotherapy) followed by chemotherapy or [2] neoadjuvant chemotherapy followed by chemoradiation. The primary endpoints of these prospective trials have focused almost exclusively on short-term endpoints, including pathologic complete response (pCR) and R0 resection rates. This allows for early assessment of patient outcomes and may be a surrogate for long-term outcomes. However, the efficacy of TNT on long-term patient outcomes needs confirmation through prospective phase 3 trials, as it has yet to be conclusively demonstrated that pCR following neoadjuvant therapy for rectal cancer meets formal criteria to act as a surrogate for local control and/or overall survival [26, 27].

Radiotherapy (±Concurrent Chemotherapy) Followed by Chemotherapy

Garcia-Aguilar et al. published a multi-institutional phase 2 trial in which patients were assigned to receive chemoradiation with 5-FU followed by increasing cycles of neoadjuvant chemotherapy prior to surgical resection. Four treatment arms were included based on the number of neoadjuvant chemotherapy cycles given prior to TME, including 0, 2, 4, or 6 cycles of modified FOLFOX (5-FU/oxaliplatin/leucovorin). The primary endpoint was pCR, which rose significantly with increasing cycles of FOLFOX. For example, the pCR rate was 38% for those patients that received 6 cycles compared to 18% for those that did not receive any neoadjuvant chemotherapy. Importantly, none of the patients in the trial experienced disease progression during neoadjuvant treatment, regardless of treatment arm [28••]. A potential confounder when interpreting the results of this trial is the interaction between rate of pCR and time from neoadjuvant treatment to surgical resection. Multiple series have suggested that delaying the time to surgery to more than 7–8 weeks after completion of chemoradiation improves the rate of pCR [29, 30], with all patients in the Garcia-Aguilar trial having a median interval between chemoradiation and definitive surgery of > 8 weeks [28••]. It should be noted, however, that other randomized trials and retrospective series have failed to show an impact on pCR for intervals > 12 weeks [31, 32].

Table 1 Summary of select trials evaluating total neoadjuvant therapy (TNT) for patients with locally advanced rectal cancer
Full size table

The Polish II trial also provided useful information on long-term patient outcomes with this approach. In this trial, patients received either short-course radiation therapy and neoadjuvant chemotherapy versus long-course chemoradiation. Following neoadjuvant treatment patients proceeded to surgery after a median of 12 weeks. The authors found no difference in rate of R0 resection, local control, or disease-free survival. Interestingly, patients who received short-course radiation plus neoadjuvant chemotherapy had improved overall survival at 3 years compared to the long-course arm (73 vs. 65%, p = 0.046); however, this improvement in overall survival is discordant with the similar rates of disease control in each arm. The authors speculate that the higher dose per fraction in the short-course arm may result in a more prevalent antitumor immune response [33••].

Neoadjuvant Chemotherapy Followed by Chemoradiation

In contrast to the trials evaluating radiation first followed by chemotherapy, studies evaluating the converse have been less fruitful. The CONTRE study most closely mirrors the trial by Garcia-Aguilar et al. In this single arm, prospective trial, patients received 8 cycles of modified FOLFOX prior to chemoradiation, followed by surgery 6–10 weeks later with a 33% pCR rate [36].

Analogous to the Polish II trial, the Spanish Grupo Cancer de Recto 3 (GCR-3) trial has published 5-year outcomes. This phase 2 trial randomized patients with locally advanced rectal cancer to receive 4 cycles of CAPOX (capecitabine/oxaliplatin) either before neoadjuvant chemoradiation or after surgery and reported no difference in disease-free or overall survival at 5 years. It is important to emphasize this trial was not powered to detect differences in long-term outcomes. Nonetheless, the authors did not find a difference in pCR rates, which was 13–14% regardless of treatment arm. They postulate this may be related to underlying differences in the baseline characteristics between the two treatment arms, including a higher percentage of patients with a threatened circumferential margin in the neoadjuvant chemotherapy group [34••, 38].

Patient Tolerance and Organ Preservation

As one might anticipate, the rates of therapy completion or compliance appear improved in patients who receive neoadjuvant chemotherapy versus adjuvant chemotherapy. For the Garcia-Aguilar trial, treatment compliance rates were 77–82%, depending upon the treatment arm. This compares favorably to the 43–58% noted in previous adjuvant trials [2, 13]. Similarly, rates of post-operative complications were not increased by the addition of chemotherapy prior to surgical resection. In the Polish II trial, the rate of compliance with neoadjuvant therapy and the toxicity profile was superior in patients on the short-course radiotherapy plus neoadjuvant chemotherapy arm compared to long-course chemoradiation (72 versus 64% compliance, respectively). Several smaller, single-arm prospective trials have also demonstrated > 90% compliance with a chemoradiation followed by neoadjuvant chemotherapy approach [39,40,41]. Compliance rates are similarly improved in those patients who receive neoadjuvant chemotherapy followed by chemoradiation. In the GCR-3 trial, patients in the neoadjuvant chemotherapy arm had a compliance rate of 94 versus 57% in the adjuvant arm (p = 0.0001). Importantly, patients receiving neoadjuvant therapy also had a lower risk of developing high-grade toxicity (19 versus 54%, p = 0.004) [34••, 38]. The EXPERT and EXPERT-C trials examined the role of 4 cycles of neoadjuvant CAPOX ± cetuximab followed by chemoradiation and then surgery. Study results demonstrated an 89% or greater rate of compliance with neoadjuvant therapy [42•, 43•]. A pooled analysis of these trials (PAN-EX) reported a compliance of 91% with neoadjuvant therapy [35••]. Again, several small prospective trials examining this treatment paradigm also show high rates of treatment compliance [36, 39, 44,45,46,47,48,49].

The use of TNT may also allow for increased likelihood of organ preservation in appropriately selected patients. Investigators from the Angelita and Joaquim Gama Institute at the University of São Paulo School of Medicine have championed the use of a watch-and-wait (or non-operative) management strategy in appropriately selected patients with locally advanced rectal cancer. With this approach, patients receive chemoradiation and for those who achieve a clinical complete response (cCR), surgery is deferred, and patients enter into a strict program of surveillance [25, 50]. With concurrent 5-FU and radiation therapy to 50.4 Gy, a cCR rate of 27% was reported [50]. These investigators have subsequently evaluated the addition of systemically dosed chemotherapy following chemoradiation over the 8–10-week interval before treatment response assessment, which increased the rate of cCR to 57%. Understanding the benefit of systemically dosed chemotherapy in this trial is somewhat obscured by the addition of another cycle of concurrent chemotherapy and increased radiation dose compared to their previous study [37••, 50]. Similar to other trials evaluating TNT, 97% of patients were able to complete neoadjuvant chemotherapy [51]. These investigators also reported improvement in radiographic response on positron emission tomography (PET) scan with the addition of chemotherapy following chemoradiation [52]. These data would suggest that it is not simply time to surgical resection in the trials evaluating TNT that led to improvement in pCR [28••, 37••, 52]. TNT may result in a higher percentage of patients being eligible for organ or sphincter preservation secondary to improvement in rates of downstaging Table 1 highlights selected prospective studies utilizing TNT.

Future Directions

Recent trials have focused on short-term pathologic endpoints such as pCR and R0 resection rates. Future trials will need to focus on long-term disease outcomes before TNT can be widely incorporated into treatment guidelines. The RAPIDO trial prospectively randomized patients to neoadjuvant chemoradiation, surgery, and adjuvant chemotherapy versus short-course radiation, neoadjuvant chemotherapy, and then surgery. The primary endpoint of this trial is 3-year disease-free survival [53]. The COPERNICUS trial evaluates neoadjuvant chemotherapy prior to short-course radiation therapy. Preliminary reports indicate high treatment compliance rates (95%), which is their primary endpoint [54]. The CREATE trial is a phase 3 trial evaluating the use of neoadjuvant chemotherapy prior to short-course or long-course radiation, with a primary end point of disease-free survival at 3 years [55].

With the exception of the GCR-3 trial, most TNT studies to date have not clearly specified the use or omission of adjuvant chemotherapy, often leaving its use to the discretion of the treating physician. The ongoing phase 3 trials include more specific guidelines regarding the use of adjuvant chemotherapy and should help to better define the risks and benefits of a strictly neoadjuvant approach.

Other groups are seeking to improve the efficacy of neoadjuvant therapies [8, 56, 57] or to include the use of novel radiosensitizing agents such as poly-ADP-ribose polymerase (PARP) inhibitors [58, 59]. Investigators from Memorial Sloan Kettering are currently enrolling patients on a randomized phase 2 trial evaluating the use of TNT for organ preservation. Patients will undergo chemotherapy followed by chemoradiation or chemoradiation followed by chemotherapy. Patients that achieve a cCR will proceed to a watch-and-wait or non-operative management approach with the primary endpoint being 3-year disease-free survival [60].

Finally, some have questioned the necessity of radiation in the neoadjuvant setting, speculating that the benefit demonstrated in previous trials was due to the use of non-TME surgical techniques. This topic is also an area of active investigation, including the ongoing randomized PROSPECT [61] and BACCHUS [62] trials.

Conclusions

A total neoadjuvant therapy (TNT) approach results in high rates of patient treatment compliance in locally advanced rectal cancer patients. TNT trials have focused primarily on short-term pathologic outcomes, which appear promising. However, ongoing randomized prospective trials are evaluating long-term disease-related outcomes for these patients. This will allow for a more critical assessment of this approach to determine if it should be routinely incorporated into treatment guidelines, as an alternative to current treatment paradigms. Ongoing trials are also seeking to refine the components of neoadjuvant therapy, which may allow for a higher proportion of patients being able to receive organ-preserving therapy.