1 Disease Background

Despite major improvements during the last 20 years, the management of patients with locally advanced rectal cancer is far from satisfactory, and mortality rates remain high. Local recurrences have been reduced to less than 10% with the combination of optimized surgery, pre- or postoperative chemoradiation, and sequential adjuvant chemotherapy even though they remain a major problem in subsets of patients with more advanced/aggressive disease. In contrast, distant metastases still develop in approximately 30% to 40% of rectal cancer patients. This figure has remained substantially unchanged over the last 20 years with only minor, and controversial, improvements achieved with postoperative fluorouracil-based chemotherapy [3, 4]. Distant recurrence has thus become the major form of failure in these patients. In addition, more than 20% of clinically resectable patients are found to actually have a positive circumferential resection margin after surgery with a significantly increased risk of local recurrence and distant metastases and reduced 5-year survival rates. Of note, this negative prognostic effect is particularly marked for patients with a positive circumferential resection margin after neoadjuvant treatment [20].

2 Rationale for Incorporating Oxaliplatin in the Preoperative Treatment of Rectal Cancer

5-fluorouracil and, more recently, capecitabine have been the mainstays of both chemoradiation regimens and adjuvant chemotherapy programs for rectal cancer. Potentiation of these regimens by incorporation of new drugs known to be active in metastatic colorectal cancer may be helpful to control the development of distant metastases in locally advanced rectal cancer.

Oxaliplatin is a diaminocyclohexane (DACH), platinum compound, active in several solid tumor types, including some cisplatin/carboplatin refractory neoplasm as colorectal cancer. The main mechanism of action, like cisplatin, is mediated through the formation of DNA adducts, but DACH-platinum adducts are bulkier and more hydrophobic than cisplatin adducts.

The improvement in disease-free and overall survival achieved when oxaliplatin was added to 5-fluorouracil or capecitabine as adjuvant treatment for radically resected, early-stage intraperitoneal colon cancer [2, 25, 28] suggests that incorporating this agent in the treatment programs for rectal cancer may improve the control of micrometastases at distant sites.

Oxaliplatin also increases the antitumor activity of FPs in metastatic colorectal cancer [9], and it has been shown to be a radiation sensitizer in preclinical models with optimal efficacy when it is given concomitantly to radiation, presumably by inhibiting postradiation DNA repair [7]. Its combination with FPs and radiation in the preoperative setting may thus improve also tumor shrinkage before surgery. This may be particularly important in low-lying tumors when sphincter preservation is attempted and in patients with bulky T3, T4, and tethered tumors where R0 resections are difficult to achieve unless the tumor is adequately shrunk before the operation.

Oxaliplatin also has a toxicity profile with a relatively low incidence of gastrointestinal toxicity that makes this agent a good candidate for combination with pelvic radiotherapy and concurrent infused 5-fluorouracil.

Based on these premises, a large investigational program has been developed in the last 15 years with multiple phase I to III studies in Europe, the USA, and Asia.

3 Phase I–II Studies

Week 1 and 5 schedules of oxaliplatin administration have been initially evaluated in a series of phase I studies in combination with 5-fluorouracil and radiation [12, 14, 26]. These studies demonstrated the feasibility of adding oxaliplatin to 5-fluorouracil and standard preoperative pelvic radiation with a high-delivered dose intensity, acceptably low gastrointestinal toxicity, neutropenia as well as early postoperative complications, and rates of pathologic complete responses (pCRs) generally above 25%.

A weekly schedule of oxaliplatin administration concomitant to continuous infusion 5-fluorouracil and preoperative chemoradiation has been developed by our group in Italy [3,4,5]. Weekly administration of oxaliplatin in combination with pelvic radiotherapy may in fact result in reduced acute toxicity, thanks to dose fractionation, and improved control of side effects, with better modulation of drug delivery, compared to administration every 3–4 weeks. The high-dose density of a weekly schedule of administration may also be optimal to maximize the inhibition of sublethal, radiation-induced DNA damage repair (advocated as the main mechanism for the radiosensitizing effect of oxaliplatin) [8].

Results from the pivotal phase I–II studies of this regimen have shown that the recommended weekly dose of oxaliplatin (given for six times on the first day of each week of treatment at escalating doses from 25 up to 60 mg/msq) in combination with pelvic radiotherapy (fixed standard dose of 50.4 Gy in 28 fractions) and infused 5-fluorouracil (administered for the entire duration of radiotherapy at the fixed dose of 225 mg/msq/day) is 60 mg/msq given weekly ×6 [5]. The feasibility of this regimen and the recommended dose of weekly oxaliplatin established in our study (60 mg/msq/week) were then confirmed in CALGB 89901 [24].

At the recommended dose, this regimen allows to deliver a total dose of oxaliplatin (360 mg/msq of over a 6-week time concomitant to radiation) substantially higher compared to other oxaliplatin-based chemoradiation programs [12, 14, 26]. In addition, this oxaliplatin dose is potentially active also systemically with dose intensity approximately 50% higher even in comparison to the regimens commonly used in the treatment of metastatic colorectal cancer [9]. This schedule has therefore an added advantage over a week 1 and 5 chemotherapy program in that it incorporates a dose and schedule of both oxaliplatin and 5-fluorouracil known to be systemically active. This may be particularly important to optimize both local control, with an enhanced chemoradiation regimen, and the control of (micro)metastases at distant sites with the concomitant administration of systemically active combination chemotherapy.

A regimen of weekly oxaliplatin combined with capecitabine and radiotherapy has also been developed with an oxaliplatin break on day 15 (third week of radiation) in the attempt to reduce toxicity [21].

4 Phase III Studies

There are seven completed phase III randomized clinical trials assessing the value of oxaliplatin combined with fluoropyrimidines (FPs) and preoperative external beam pelvic radiation in locally advanced rectal cancer (Tabl 35.1). Infused 5-fluorouracil was the FP backbone in two studies (STAR-01 and CAO/ARO/AIO) [6, 22, 23], while oxaliplatin was combined with capecitabine in the French ACCORD trial [15], PETACC-6 [25], Chinese study [17], and INTERACT trial [27]. Either capecitabine or infused 5-fluorouracil was used in NSABP R-04 [1] that entailed a 2 × 2 factorial design.

Table 35.1 Phase III randomized studies of FP-based preoperative chemoradiation ± oxaliplatin
Full size table

A weekly schedule of oxaliplatin administration, with minor variations in the weekly dose and/or a break in the third week of chemoradiation treatment, was tested in six of these studies in combination with either infusional FU or oral capecitabine. Only INTERACT entailed a triweekly oxaliplatin schedule (130 mg/mq on day 1, 19, 38).

As to control arms, STAR-01, NSABP R-04, PETACC-6, ACCORD 0405, and the Chinese trial used exactly the same FP dose and schedule as in the experimental arm, while a shorter FU infusion with a higher daily dose (similar to the chemoradiation regimens used in the treatment of anal cancer) was used in the control arm of the German study and a lower capecitabine dose, combined to an intensified radiotherapy regimen, was used in the INTERACT trial.

Radiotherapy intensification was also pursued in the experimental arm of the French study (45 Gy in the control arm vs 50 Gy in the experimental arm), while the planned radiotherapy dose was 50.4 Gy for both arms in the other studies.

Data on surgical and activity endpoints have been reported for all of these studies (Table 35.2) and safety data published for five out of seven (Tables 35.3 and 35.4).

Table 35.2 Response to preoperative chemoradiation ± oxaliplatin in randomized phase III studies
Full size table
Table 35.3 Acute toxicity in published randomized trials of FU-based preoperative chemoradiation +/− Oxaliplatin
Full size table
Table 35.4 Surgical mortality/morbility in randomized trials of FU-based preoperative chemoradiation +/− Oxaliplatin
Full size table

Compliance

Overall, findings from these studies confirm that weekly oxaliplatin can be added to FP-based preoperative chemoradiation even though toxicity is increased. Indeed, unexpected adverse events were not reported, and there was no excess treatment-related mortality in the oxaliplatin arms; over 95% of patients were operated as planned within 6–8 weeks from the end of chemoradiation, independent of the preoperative treatment they had received, and surgical mortality was also low and similar with or without oxaliplatin. Consistently, postoperative morbidity was not affected by the use of oxaliplatin along with preoperative radiation. Compliance to radiotherapy was slightly reduced with the addition of oxaliplatin with an apparent relationship between the dose of oxaliplatin delivered concomitant to radiation and the proportion of patients receiving a reduced radiotherapy dose (6% in the German study where 200 mg/msq of oxaliplatin were delivered concomitant to radiation, 7% in the Chinese study with 240 mg/msq, 13% in the French study with 250 mg/msq, and 16% in the STAR-01 study with 360 mg/msq of oxaliplatin added to 5-fluorouracil and radiation) (Table 35.5). However, even in the STAR study that entailed the highest oxaliplatin dose, more than 90% of the patients in the oxaliplatin arm received at least 45 Gy, and more than 80% had at least five weekly administrations of oxaliplatin.

Table 35.5 Treatment compliance in published randomized trials of FU-based preoperative chemoradiation +/− Oxaliplatin
Full size table

Toxicity

The addition of oxaliplatin resulted in more toxicity with grade III–IV events of any type reported in more than 20% of patients and rates of grade III–IV diarrhea between 12% and 15% in four studies with published toxicity data (Table 35.3). This increase in toxicity appears to be less pronounced in the German study, possibly because of the lower cumulative dose of oxaliplatin administered along with radiation. The higher rate of severe toxicity in the control arm of this study (22% compared to less than 10% in the Italian and French studies), however, provides an alternative explanation for the less marked toxicity increment observed with the addition of oxaliplatin.

Activity

Despite the promising phase II data (Table 35.6), these randomized studies consistently showed that adding oxaliplatin to preoperative fluoropyrimidine-based chemoradiation does not improve primary tumor response. pCR rates were nearly identical in the experimental compared to the control arms in the STAR-01, NSABP-R04, PETACC-6, and INTERACT studies and numerically, but not significantly, increased in the French study (that also entailed intensified radiotherapy in the oxaliplatin containing arm) and Chinese trial.

Table 35.6 Phase I-II studies of oxaliplatin and pelvic radiation therapy
Full size table

Only CAO/ARO/AIO-04 showed a limited, but statistically significant, difference in pCR rate [22]. The clinical relevance of this difference, however, appears to be questionable (13% vs 17%) (Table 35.2). The less pronounced toxicity and higher compliance to radiation therapy achieved in this study (possibly related to the lower total oxaliplatin dose and/or the oxaliplatin break on week 3) has been proposed as the main explanation for the significantly improved antitumor activity. However, on speculative ground, statistical significance in this trial also reflects the larger study population compared to other trials.

Other measures of response to preoperative chemoradiation (including tumor regression grade, the proportion of patients with residual tumor confined to the muscular layer of the rectal wall, the rates of node-positive disease found at surgery, and the incidence of circumferential resection margin positivity) were also similarly distributed among patients treated with or without oxaliplatin in all of these studies, ruling out any clinically relevant effect on primary tumor response (Table 35.2).

Consistently, no impact was observed in the proportion of patients receiving sphincter-saving surgery that were similar with or without oxaliplatin (Table 35.2).

The lower pCR rate observed with OXA in these randomized studies (compared to higher figures in phase II trials) probably reflects the better control for the multiple factors affecting response to preoperative chemoradiation and the larger sample size compared to phase II assessment. This underlines the crucial importance of properly designed, adequately powered randomized multidisciplinary studies with standardized and optimized radiological staging, radiotherapy, TME surgery, and pathologic evaluation to assess the effect of new treatment strategies in rectal cancer.

These results therefore indicate that oxaliplatin is not a clinically effective radiation sensitizer in rectal cancer or, at least, that this platinum derivative does not increase radiosensitization already achievable with 5-fluorouracil or capecitabine administration concomitant to radiation.

Of note, the pCR rates in the control arms of these seven studies [1, 6, 15, 17, 22, 23, 25, 27] appear to be higher than those reported in less recent studies, probably thanks to optimized radiotherapy techniques and higher radiation doses. It is thus conceivable that optimal radiotherapy, with optimal radiosensitization by FPs, already maximizes primary tumor response with little or no room for further improvement with additional radiation sensitizers.

Primary tumor response was the main endpoint in the French study and INTERACT trial, while the other studies were designed to test the impact of oxaliplatin on long-term efficacy endpoints (overall survival for STAR-01 and the Chinese study, local control in NSABP R04, and disease-free survival for the German and PETACC-6 trials). The lack of effect on primary tumor response does not preclude this hypothesis. In rectal cancer, there is in fact no evidence that differences in long-term outcome in a phase III trial are predicted comparing pCR rates between treatment arms [19]. The mechanisms of cytotoxicity and radiosensitization may in fact not be the same, and local tumor regression does not necessarily predict responsiveness of disseminated tumor cells toward the chemotherapy component of combined modality treatment programs.

Efficacy

Data on clinical outcome have now been published for four [1, 11, 15, 17, 22, 23] of these studies (Table 35.7).

Table 35.7 Clinical outcome data of published randomized trials of FU-based preoperative chemoradiation +/− oxaliplatin
Full size table

As to specific efficacy endpoints, none of them showed a difference in local control (local progression before surgery or local relapse in radically resected patients). Of note, this was the primary endpoint of NSABP R-04 that was thus a negative study.

Three of these studies showed slightly reduced proportions of distant metastases among patients receiving oxaliplatin added to preoperative chemoradiation, although statistical significance was reached only in the Chinese trial (Table 35.7). Although the low number of events may have precluded statistically significant results in the other studies, this apparent effect on distant metastases is plausible considering the systematically active doses of oxaliplatin used in these trials and consistent with the primary hypothesis of the investigational programs testing the incorporation of oxaliplatin in the neoadjuvant treatment of rectal cancer. We also reported a significant reduction in the rate of distant metastases at early time points (before surgery, intraoperatively or within 6 months after surgery) in the first analysis of STAR-01 [6, 18]. This finding remains to be confirmed and the meaning of an early systemic effect to be determined.

A numerical improvement in 3- or 5-year DFS, ranging from 4 to 5% in the three largest studies [1, 11, 15, 23] to 10% in the smaller Chinese trial [17], was observed in all the studies reported so far (Table 35.7).

Although a significant reduction in the risk of disease recurrence was achieved only in the German study, the observed absolute difference has a magnitude comparable to that observed with the incorporation of oxaliplatin in the adjuvant treatment of colon cancer (MOSAIC, C-07) [2, 28]. This lends support to the notion that OXA added to preoperative chemoradiation may reduce metastases at distant sites.

On statistical ground, CAO/ARO/AIO-04 [23] is therefore the only positive study, meeting its primary endpoint of a reduction in the risk of disease recurrence (DFS hazard ratio 0.79, 95% CI 0.64–0.98; p 0.03) with an improvement in the rates of DFS at 3 years from 71.2% in the control arm to 75.9% in the oxaliplatin containing arm. In this trial, however, oxaliplatin was administered also in the adjuvant setting (Table 35.1), making it impossible to establish a specific role for oxaliplatin in the preoperative setting in isolation. Direct evidence of the efficacy of oxaliplatin in the adjuvant treatment of LARC after preoperative chemoradiation was in fact recently provided [16]. The magnitude of the observed difference in 3-year or 5-year DFS (and in the rate of distant metastases), however, appears to be similar in the German study (Table 35.7), where OXA was added to standard treatment also in the postoperative adjuvant setting (Table 35.1), compared to the French and Chinese studies (Table 35.7), where the same postoperative treatment was given independent of the randomization arm (Table 35.1). One may thus speculate that the reduction in the rate of distant metastases and improvement in DFS observed in CAO/ARO/AIO-04 mainly depend on the addition of OXA to preoperative chemoradiation with a lesser role for its incorporation in the adjuvant postoperative setting.

Further light on the role of OXA added to preoperative chemoradiation in LARC will be provided by STAR-01 and PETACC-6 final data awaited for 2016.

Notwithstanding the results of the Italian and European studies that are pending and possible future pooled analyses of these trials, the small magnitude of the DFS improvement will have to be balanced against the increased toxicity resulting from the incorporation of oxaliplatin in the preoperative chemoradiation programs for LARC.

Maturation of OS data will be also important to draw final conclusions on the role of oxaliplatin given concomitantly to preoperative chemoradiation in locally advanced rectal cancer.

Meanwhile, alternative strategies are being developed including OXA-based neoadjuvant chemotherapy before chemoradiation [10] and consolidation OXA-based chemotherapy in the interval between chemoradiation and surgery [13].