Commentary
Downstage Migration After Neoadjuvant Chemoradiotherapy for Rectal Cancer: The Reverse of the Will Rogers Phenomenon? Emmanouil Fokas, MD, DPhil1; Torsten Liersch, MD2; Rainer Fietkau, MD3; Werner Hohenberger, MD4; € del, MD1 Clemens Hess, MD, PhD5; Heinz Becker, MD2†; Rolf Sauer, MD3; Christian Wittekind, MD6; and Claus Ro
Downstaging after neoadjuvant treatment is increasingly used as a prognostic factor and surrogate endpoint in clinical trials. However, in recent trials of neoadjuvant 5-fluorouracil–based chemoradiotherapy for rectal cancer, downstaging did not translate into a benefit with regard to either disease-free survival (DFS) or overall survival. By analyzing the 10-year outcome data of the German CAO/ARO/AIO-94 phase 3 trial, the authors demonstrated that significantly fewer patients had poor prognostic features (eg, ypT3-4, ypN1-2) after preoperative 5-fluorouracil–based chemoradiotherapy. Nevertheless, these patients with International Union for Cancer Control stage II disease were found to be at a higher risk of developing distant metastases and had poorer DFS compared with patients with corresponding TNM tumor (sub)groups in the postoperative treatment arm, whereas patients with International Union for Cancer Control stage III disease demonstrated a nonsignificant trend toward a worse outcome after preoperative treatment. Overall, DFS remained identical in both treatment arms. Thus, “downstage migration” after neoadjuvant treatment resembles the reverse of the Will Rogers phenomenon and thereC 2015 American Cancer Society. fore may not be a reliable endpoint for long-term outcomes. Cancer 2015;000:000-000. V KEYWORDS: downstage migration, rectal cancer, phase 3 trial, chemoradiotherapy, long-term follow-up, reverse Will Rogers phenomenon.
INTRODUCTION Historically, postoperative 5-fluorouracil (5-FU)-based chemoradiotherapy (5-FU-CRT) has been shown to improve survival over surgery alone among patients with histopathologically confirmed International Union for Cancer Control (UICC) stage II and III rectal cancer.1,2 The CAO/AIO/ARO-94 phase 3 trial of the German Rectal Cancer Study Group demonstrated the superiority of preoperative versus postoperative 5-FU-CRT with regard to local control and treatment compliance.3 Recently, the 10-year follow-up outcomes of this trial confirmed the original findings.4 Two other randomized trials from the European Organization for Research and Treatment of Cancer (EORTC 22921) and the Federation Francophone de Cancerologie Digestive (FFCD 9293) reported the superiority of preoperative 5-FU-CRT over preoperative radiotherapy with respect to local control rates.5,6 A common finding of these 3 large randomized trials of preoperative 5-FU-CRT in patients with clinically staged UICC stage II and II rectal cancer was a statistically significant UICC downstaging effect after neoadjuvant 5-FU-CRT compared with primary surgery or preoperative radiotherapy alone before surgery. Patients who experience downstaging are reported to have a more favorable long-term prognosis.7 However, it is interesting to note that in none of the 3 recent trials did this effect translate into an improved disease-free survival (DFS) or overall survival for the entire treatment arm. Herein, we propose an explanation of this finding that resembles the reverse of the Will Rogers stage migration phenomenon. In general, stage migration occurs with careful staging of cancer. For example, upstaging of UICC stages I and II to stage III colorectal cancer may occur by increasing the number of lymph nodes examined, as well as through the advent of
*Emmanouil Fokas’s current address: Department of Oncology, Oxford Institute of Radiation Oncology, University of Oxford, Oxford, United Kingdom. Corresponding author: Emmanouil Fokas, MD, DPhil, Department of Radiotherapy and Oncology, Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany; Fax: (011) 0049 696301-5091; [email protected] 1 Department of Radiotherapy and Oncology, University of Frankfurt, Frankfurt, Germany; 2Department of General Surgery, University Medical Center Gottingen, Gottingen, Germany; 3Department of Radiation Therapy, University of Erlangen-Nuremberg, Erlangen, Germany; 4Department of Surgery, University of ErlangenNuremberg, Erlangen, Germany; 5Department of Radiotherapy, University Medical Center Gottingen, Gottingen, Germany; 6Institute of Pathology, University Hospital Leipzig, Leipzig, Germany
†Deceased. DOI: 10.1002/cncr.29260, Received: November 29, 2014; Revised: December 28, 2014; Accepted: January 5, 2015, Published online Month 00, 2015 in Wiley Online Library (wileyonlinelibrary.com)
Cancer
Month 00, 2015
1
Commentary TABLE 1. Ten-Year Cumulative Incidence of Local And Distant Disease Recurrence and Disease-Free Survival for the Subgroup of Patients Who Received Protocol-Specified Treatment Before or After Curative Surgery (R0,M0) According to the 2002 UICC TNM (Sub)Staging System No. of Patients (%)
All ypT0N0 UICC stage I (y)pT1N0 (y)pT2N0 UICC stage II IIA (y)pT3N0 IIB (y)pT4N0 UICC stage III ypT0N1-2 IIIA (y)pT1-2N1 IIIB (y)pT3-4N1 IIIC (y)pTXN2
10-Year Cumulative Incidence of Local Recurrence
10-Year Cumulative Incidence of Distant Metastases
10-Year DFS
Preop CRT
Postop CRT
Preop CRT, %
Postop CRT, %
Pa
Preop CRT, %
Postop CRT, %
Pa
Preop CRT, %
Postop CRT, %
Pa
361 (100) 36 (10) 108 (29.9) 16 (4.4) 92 (25.5) 115 (31.8) 111 (30.7) 4 (1.1) 98 (27.1) 4 (1.1) 23 (6.4) 35 (9.7) 40 (11.1)
298 (100) 0 (0) 72 (24.1) 3 (1) 69 (23.1) 85 (28.5) 85 (28.5) 0 (0.3) 141 (47.3) 0 (0) 14 (4.7) 62 (20.8) 65 (21.8)
5.5 2.9 3.4 0 4 4.2 4.3 0 10.3 0 0 16.1 10.5
7.5 NA 7.5b 0 7.8 2.4 2.4 0 10.8 0 0 6.3 19.7
.199 .082 .117 .606 .576 .979 .178 .545
25.1 11.8 11.5 6.2 12.4 19 18.8 25 51.6 0 17.8 50.6 71.8
24.7 0 9b 0 9.4 8.5 8.5 0 42.3 0 36.5 34.1 52.5
.972 .809 .665 .117 .037 .042 .204 .219 .142 .093
73.3 88.2 85.3 93.8 83.9 80.1 80.3 75 47.4 100 82.2 46.9 25
73.2 86.6 0 86 90.3 90.3 56.2 63.5 65.9 43.7
.803 .897 .665 .978 .050 .056 .221 .219 .095 .126
Abbreviations: CRT, chemoradiotherapy; DFS, disease-free survival; NA, not applicable; Postop, postoperative; Preop, preoperative; UICC, International Union for Cancer Control. a Bold type indicates statistical significance. b Note that protocol-specified treatment for patients with pT1-2N0 disease after primary surgery was observation only.
modern methods,8,9 which in turn may improve the prognosis of patients with UICC stages I/II and III disease without any changes in individual outcomes or the prognosis of the entire group.10,11 Conversely, one could ask if “downstage migration,” as observed after preoperative 5FU-CRT, is associated with a worse outcome for corresponding ypTN categories and UICC stages without altering the individual outcomes or the outcome for the entire treatment arm? Or, in other words, is the outcome of a patient with, for example, ypT3N0 disease after neoadjuvant 5-FU-CRT different from that of a patient who received postoperative 5-FU-CRT for a pT3N0 tumor? In the CAO/AIO/ARO-94 phase 3 trial, patients with cT3-4 and/or cN1 rectal cancer were included after clinical staging with endorectal ultrasound and computed tomography. A total of 799 patients were randomized to receive preoperative 5-FU-CRT followed by total mesorectal excision surgery or immediate total mesorectal excision and postoperative 5-FU-CRT. Patients were well matched with regard to pretreatment cT and cN categories.3,4 To assess downstaging, we used the chi-square test to compare percentages (ie, the T, N and UICC stages of disease). Preoperative 5-FU-CRT resulted in significant downshifts of the ypT and ypN categories and, consequently, UICC downstaging compared with primary surgery: (y)pT0, (y)pT1, (y)pT2, (y)pT3, and (y)pT4: 10%, 6%, 29%, 51%, and 4% versus 0%, 2%, 25%, 68%, and 2
5%, respectively (40 patients, 23 patients, 117 patients, 203 patients, and15 patients vs 0 patients, 7 patients, 96 patients, 266 patients, and 19 patients, respectively; P
Downstage Migration After Neoadjuvant Chemoradiotherapy for Rectal Cancer: The Reverse of the Will Rogers Phenomenon? Emmanouil Fokas, MD, DPhil1; Torsten Liersch, MD2; Rainer Fietkau, MD3; Werner Hohenberger, MD4; € del, MD1 Clemens Hess, MD, PhD5; Heinz Becker, MD2†; Rolf Sauer, MD3; Christian Wittekind, MD6; and Claus Ro
Downstaging after neoadjuvant treatment is increasingly used as a prognostic factor and surrogate endpoint in clinical trials. However, in recent trials of neoadjuvant 5-fluorouracil–based chemoradiotherapy for rectal cancer, downstaging did not translate into a benefit with regard to either disease-free survival (DFS) or overall survival. By analyzing the 10-year outcome data of the German CAO/ARO/AIO-94 phase 3 trial, the authors demonstrated that significantly fewer patients had poor prognostic features (eg, ypT3-4, ypN1-2) after preoperative 5-fluorouracil–based chemoradiotherapy. Nevertheless, these patients with International Union for Cancer Control stage II disease were found to be at a higher risk of developing distant metastases and had poorer DFS compared with patients with corresponding TNM tumor (sub)groups in the postoperative treatment arm, whereas patients with International Union for Cancer Control stage III disease demonstrated a nonsignificant trend toward a worse outcome after preoperative treatment. Overall, DFS remained identical in both treatment arms. Thus, “downstage migration” after neoadjuvant treatment resembles the reverse of the Will Rogers phenomenon and thereC 2015 American Cancer Society. fore may not be a reliable endpoint for long-term outcomes. Cancer 2015;000:000-000. V KEYWORDS: downstage migration, rectal cancer, phase 3 trial, chemoradiotherapy, long-term follow-up, reverse Will Rogers phenomenon.
INTRODUCTION Historically, postoperative 5-fluorouracil (5-FU)-based chemoradiotherapy (5-FU-CRT) has been shown to improve survival over surgery alone among patients with histopathologically confirmed International Union for Cancer Control (UICC) stage II and III rectal cancer.1,2 The CAO/AIO/ARO-94 phase 3 trial of the German Rectal Cancer Study Group demonstrated the superiority of preoperative versus postoperative 5-FU-CRT with regard to local control and treatment compliance.3 Recently, the 10-year follow-up outcomes of this trial confirmed the original findings.4 Two other randomized trials from the European Organization for Research and Treatment of Cancer (EORTC 22921) and the Federation Francophone de Cancerologie Digestive (FFCD 9293) reported the superiority of preoperative 5-FU-CRT over preoperative radiotherapy with respect to local control rates.5,6 A common finding of these 3 large randomized trials of preoperative 5-FU-CRT in patients with clinically staged UICC stage II and II rectal cancer was a statistically significant UICC downstaging effect after neoadjuvant 5-FU-CRT compared with primary surgery or preoperative radiotherapy alone before surgery. Patients who experience downstaging are reported to have a more favorable long-term prognosis.7 However, it is interesting to note that in none of the 3 recent trials did this effect translate into an improved disease-free survival (DFS) or overall survival for the entire treatment arm. Herein, we propose an explanation of this finding that resembles the reverse of the Will Rogers stage migration phenomenon. In general, stage migration occurs with careful staging of cancer. For example, upstaging of UICC stages I and II to stage III colorectal cancer may occur by increasing the number of lymph nodes examined, as well as through the advent of
*Emmanouil Fokas’s current address: Department of Oncology, Oxford Institute of Radiation Oncology, University of Oxford, Oxford, United Kingdom. Corresponding author: Emmanouil Fokas, MD, DPhil, Department of Radiotherapy and Oncology, Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany; Fax: (011) 0049 696301-5091; [email protected] 1 Department of Radiotherapy and Oncology, University of Frankfurt, Frankfurt, Germany; 2Department of General Surgery, University Medical Center Gottingen, Gottingen, Germany; 3Department of Radiation Therapy, University of Erlangen-Nuremberg, Erlangen, Germany; 4Department of Surgery, University of ErlangenNuremberg, Erlangen, Germany; 5Department of Radiotherapy, University Medical Center Gottingen, Gottingen, Germany; 6Institute of Pathology, University Hospital Leipzig, Leipzig, Germany
†Deceased. DOI: 10.1002/cncr.29260, Received: November 29, 2014; Revised: December 28, 2014; Accepted: January 5, 2015, Published online Month 00, 2015 in Wiley Online Library (wileyonlinelibrary.com)
Cancer
Month 00, 2015
1
Commentary TABLE 1. Ten-Year Cumulative Incidence of Local And Distant Disease Recurrence and Disease-Free Survival for the Subgroup of Patients Who Received Protocol-Specified Treatment Before or After Curative Surgery (R0,M0) According to the 2002 UICC TNM (Sub)Staging System No. of Patients (%)
All ypT0N0 UICC stage I (y)pT1N0 (y)pT2N0 UICC stage II IIA (y)pT3N0 IIB (y)pT4N0 UICC stage III ypT0N1-2 IIIA (y)pT1-2N1 IIIB (y)pT3-4N1 IIIC (y)pTXN2
10-Year Cumulative Incidence of Local Recurrence
10-Year Cumulative Incidence of Distant Metastases
10-Year DFS
Preop CRT
Postop CRT
Preop CRT, %
Postop CRT, %
Pa
Preop CRT, %
Postop CRT, %
Pa
Preop CRT, %
Postop CRT, %
Pa
361 (100) 36 (10) 108 (29.9) 16 (4.4) 92 (25.5) 115 (31.8) 111 (30.7) 4 (1.1) 98 (27.1) 4 (1.1) 23 (6.4) 35 (9.7) 40 (11.1)
298 (100) 0 (0) 72 (24.1) 3 (1) 69 (23.1) 85 (28.5) 85 (28.5) 0 (0.3) 141 (47.3) 0 (0) 14 (4.7) 62 (20.8) 65 (21.8)
5.5 2.9 3.4 0 4 4.2 4.3 0 10.3 0 0 16.1 10.5
7.5 NA 7.5b 0 7.8 2.4 2.4 0 10.8 0 0 6.3 19.7
.199 .082 .117 .606 .576 .979 .178 .545
25.1 11.8 11.5 6.2 12.4 19 18.8 25 51.6 0 17.8 50.6 71.8
24.7 0 9b 0 9.4 8.5 8.5 0 42.3 0 36.5 34.1 52.5
.972 .809 .665 .117 .037 .042 .204 .219 .142 .093
73.3 88.2 85.3 93.8 83.9 80.1 80.3 75 47.4 100 82.2 46.9 25
73.2 86.6 0 86 90.3 90.3 56.2 63.5 65.9 43.7
.803 .897 .665 .978 .050 .056 .221 .219 .095 .126
Abbreviations: CRT, chemoradiotherapy; DFS, disease-free survival; NA, not applicable; Postop, postoperative; Preop, preoperative; UICC, International Union for Cancer Control. a Bold type indicates statistical significance. b Note that protocol-specified treatment for patients with pT1-2N0 disease after primary surgery was observation only.
modern methods,8,9 which in turn may improve the prognosis of patients with UICC stages I/II and III disease without any changes in individual outcomes or the prognosis of the entire group.10,11 Conversely, one could ask if “downstage migration,” as observed after preoperative 5FU-CRT, is associated with a worse outcome for corresponding ypTN categories and UICC stages without altering the individual outcomes or the outcome for the entire treatment arm? Or, in other words, is the outcome of a patient with, for example, ypT3N0 disease after neoadjuvant 5-FU-CRT different from that of a patient who received postoperative 5-FU-CRT for a pT3N0 tumor? In the CAO/AIO/ARO-94 phase 3 trial, patients with cT3-4 and/or cN1 rectal cancer were included after clinical staging with endorectal ultrasound and computed tomography. A total of 799 patients were randomized to receive preoperative 5-FU-CRT followed by total mesorectal excision surgery or immediate total mesorectal excision and postoperative 5-FU-CRT. Patients were well matched with regard to pretreatment cT and cN categories.3,4 To assess downstaging, we used the chi-square test to compare percentages (ie, the T, N and UICC stages of disease). Preoperative 5-FU-CRT resulted in significant downshifts of the ypT and ypN categories and, consequently, UICC downstaging compared with primary surgery: (y)pT0, (y)pT1, (y)pT2, (y)pT3, and (y)pT4: 10%, 6%, 29%, 51%, and 4% versus 0%, 2%, 25%, 68%, and 2
5%, respectively (40 patients, 23 patients, 117 patients, 203 patients, and15 patients vs 0 patients, 7 patients, 96 patients, 266 patients, and 19 patients, respectively; P
