Radiotherapy and Oncology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Original article
Diagnostic accuracy and prognostic impact of restaging by magnetic resonance imaging after preoperative chemoradiotherapy in patients with rectal cancer Jung Wook Huh a, Hee Cheol Kim a,⇑, Soon Jin Lee b, Seong Hyeon Yun a, Woo Yong Lee a, Yoon Ah Park a, Yong Beom Cho a, Ho-Kyung Chun a a
Department of Surgery, Sungkyunkwan University School of Medicine; and b Department of Radiology, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea
a r t i c l e
i n f o
Article history: Received 1 April 2014 Received in revised form 6 August 2014 Accepted 12 August 2014 Available online xxxx Keywords: Rectal cancer Preoperative chemoradiation Restaging MRI
a b s t r a c t Background: The prognostic role of restaging rectal magnetic resonance imaging (MRI) in patients with preoperative CRT has not been established. The goal of this study was to evaluate the diagnostic accuracy and prognostic role of radiological staging by rectal MRI after preoperative chemoradiation (CRT) in patients with rectal cancer. Methods: A total of 231 consecutive patients with rectal cancer who underwent preoperative CRT and radical resection from January 2008 to December 2009 were prospectively enrolled. The diagnostic accuracy and prognostic significance of post-CRT radiological staging by MRI was evaluated. Results: The sensitivity, specificity, positive predictive value, and negative predictive value of radiological diagnosis of good responders (ypTNM stage 0–I) were 32%, 90%, 65%, and 69%, respectively. The overall accuracy of MRI restating for good responders was 68%. The 5-year disease-free survival rates of patients with radiological and pathological TNM stage 0, stage I, and stage II–III were 100%, 94%, and 76%, respectively (P = 0.037), and 97%, 87%, and 73%, respectively (P = 0.007). On multivariate analysis, post-CRT radiological staging by MRI was an independent prognostic factor for disease-free survival. Conclusion: Radiological staging by MRI after preoperative CRT may be an independent predictor of survival in patients with rectal cancer. Ó 2014 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2014) xxx–xxx
Multidisciplinary treatment including preoperative chemoradiation (CRT) followed by radical resection for locally-advanced rectal cancer is currently a treatment standard that is associated with improved local control and disease-free survival [1,2]. Several nonrandomized studies suggest that good responses (ypT0-2N0) after preoperative CRT and curative surgery are associated with improved oncologic outcomes [3,4]. Many oncologists have suggested that preoperative identification of good and poor responses to preoperative therapy using imaging modalities enables further tailoring of treatment. Achieving a complete clinical response to CRT may allow for an organ preservation strategy such as local transanal excision or a wait-and-see approach in select patients with rectal cancer [5,6]. In addition, patients with poor response may benefit from intensified strategies to improve resectability and treatment completion [7,8]. Although the accuracy and validity of rectal cancer restaging by current imaging modalities remain ⇑ Corresponding author. Address: Department of Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 135-710, Republic of Korea. E-mail address: [email protected] (H.C. Kim).
uncertain, a recent promising study indicated that magnetic resonance imaging (MRI) assessment of tumor regression grade after preoperative CRT correlated with survival in patients with rectal cancer [9]. High-resolution MRI has been established as the most accurate method for preoperative staging of rectal cancer [10]; however, its role in the restaging of patients who receive preoperative CRT has not yet been determined [9–12]. The goal of this study was to evaluate the diagnostic accuracy and prognostic relevance of post-CRT MRI assessment of response to preoperative therapy in patients with rectal cancer.
Methods A total of 300 consecutive patients who underwent curative surgery after preoperative CRT for locally-advanced rectal cancer at our institution between January 2008 and December 2009 were prospectively enrolled in this study. Patients with metastatic disease, recurrent disease, local resection, or pretreatment TNM stage I, and those lacking pretreatment MRI assessment or follow-up data were excluded from the study cohort. Ultimately, 231 patients
http://dx.doi.org/10.1016/j.radonc.2014.08.023 0167-8140/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Huh JW et al. Diagnostic accuracy and prognostic impact of restaging by magnetic resonance imaging after preoperative chemoradiotherapy in patients with rectal cancer. Radiother Oncol (2014), http://dx.doi.org/10.1016/j.radonc.2014.08.023
2
Radiologic restaging in rectal cancer
were included in the present analysis. This study was reviewed and approved by the appropriate institutional review board. All patients underwent concurrent CRT consisting of preoperative 5-fluorouracil-based chemotherapy and pelvic radiation (4500–5040 cGy), which was followed by radical surgery 6– 8 weeks later [13,14]. Of 231 patients, 216 (93.5%) received postoperative chemotherapy. The mean interval between preoperative CRT and surgery was 55 days (median, 55 days; range, 26– 225 days). All patients underwent total mesorectal excision and the tumors were staged according to the 7th American Joint Committee on Cancer (AJCC) TNM classification. Responses to chemoradiation were evaluated using the tumor regression grade classification proposed by Dworak et al. [15], as follows: grade 0, no regression; grade 1, minimal regression; grade 2, moderate regression; grade 3, near-complete regression; and grade 4, complete regression. All patients received rectal MR imaging within a mean of 40 days (median, 41 days; range, 12–204 days) after completion of preoperative CRT. MRI was performed using a 3.0-T whole body system (Intera Achieva 3T, Philips Medical Systems, Best, Netherlands) with a cardiac phased-array coil in a supine position [16,17]. First, a sagittal localizing image was obtained for the selection of axial and coronal images with a T2-weighted turbo spinecho sequence (3-mm slice thickness, 800 785 matrix, 360-mm field of view, and four signals acquired). The oblique axial and coronal T2-weighted turbo spin-echo images were obtained orthogonal and parallel to the long axis of the rectal tumor. An axial T1-weighted turbo-field-echo sequence (3-mm slice thickness, 400 400 matrix, 360-mm field of view, and two signals acquired) was also performed. All sequences were performed without fat saturation. All MR images were independently interpreted by an expert radiologist (S.J.L.) who was blinded to the results of other imaging examinations and histopathological evaluations. For rectal wall (mrT) staging, we used the following criteria: mrT1 (the tumor signal intensity was confined to the submucosal layer), mrT2 (the tumor signal intensity extended into the muscle layer leading to an irregular or thickened muscle layer), mrT3 (the tumor signal intensity extended through the muscular layer into the perirectal fat), and mrT4 (the tumor signal intensity extended into adjacent organs). Lymph nodes were classified as malignant if they exhibited an indistinct border of any size, an irregular margin, or mixed signal intensity, instead of using size criteria [18]. The patients were followed at 3-month intervals for 2 years, at 6-month intervals for the next 3 years, and annually thereafter. For patients who did not return for observation after 1 year, information was obtained by mail or telephone. On a semi-annual basis or when there was a suspicion of recurrence, the follow-up examination included a clinical history, physical examination, serum carcinoembryonic antigen (CEA) level, chest X-ray, abdominopelvic computed tomography (CT) or MRI, colonoscopy, and positron emission tomography (PET) scanning where available. Recurrence was determined by clinical and radiological examination or using histological confirmation. The main pattern of recurrence was recorded as the first site of detectable failure during the followup period. Statistical evaluation was carried out using the statistical package SPSS for Windows (Version 14.0; SPSS Inc., Chicago, IL, USA). Differences between the two groups were tested using Student’s t-test and the chi-squared test as appropriate. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated from a 2 2 contingency table. Disease-free survival curves were calculated using the Kaplan–Meier method and differences between curves were evaluated using the log-rank test. A Cox model multivariate analysis was generated by a forward stepwise selection of variables, and a P value of 0.1 was adopted as the limit for inclusion of a covariant. A P value 60.05 was considered statistically significant.
Results Clinicopathological characteristics of patients are summarized in Table 1. The median age of participants was 56 years (range, 26–81 years) and 163 (70.6%) were male. The median distance from the anal verge was 4.0 cm (range, 0–10). The median number of resected lymph nodes was 11 (range, 1–44). Using the 7th AJCC TNM staging system, 24 and 207 patients had pretreatment stage II and III cancers, respectively, and 32, 55, 73, and 71 patients had pathologic stage 0 (complete response), I, II, and III cancers, respectively. We compared the results of post-CRT radiological TNM staging by MRI with the results of pathologic examination. The sensitivity, specificity, PPV, and NPV of radiological diagnosis of good responders (ypTNM stage 0–I) were 32%, 90%, 65%, and 65%, respectively. The overall accuracy of good responders by MRI was 68%. We also separately assessed the performance of MRI for establishing a complete response. The sensitivity, specificity, PPV, and NPV of radiological diagnosis of a complete response (ypTNM stage 0) were 3%, 99%, 33%, and 86%, respectively. The overall accuracy of complete response prediction by MRI was 86%. The sensitivity, specificity, PPV, NPV, and accuracy of post-treatment MRI diagnosis of regional nodal disease (ymrN+) for prediction of pathologically-diagnosed regional nodal metastases (ypN+) were 37%, 81%, 79%, 40%, and 52% respectively. We performed further analyses to address the question of whether it is safe to perform limited surgery in patients who are mrN+ on pretreatment MRI for pathologically-diagnosed ypN0 disease. The sensitivity, specificity, PPV, NPV, and accuracy of post-treatment MRI of ymrN0 disease in patients who were mrN+ on pre-treatment MRI (n = 207) for pathologically-diagnosed ypN0 disease were 33%, 84%, 81%, 37%, and 49% respectively. Moreover, we performed analysis to guide decision-making in terms of the resection of adjacent organs involved on pre-treatment MRI, but cancer-free on
Table 1 Patient characteristics (n = 231). No. (%) Median age, years (range) Gender Male Female Median distance from anal verge, cm (range) Pretreatment TNM stage II III Operation method Low anterior resection Abdominoperineal resection Histology Adenocarcinoma Mucinous carcinoma Median tumor diameter, cm (range) Positive lymphovascular invasion Positive perineural invasion Median no. of lymph nodes retrieved (range) Median preoperative CEA, ng/ml (range) ymrTNM stage 0 (complete response) I II III ypTNM stage 0 (complete response) I II III
56 (26–81) 163 (70.6) 68 (29.4) 4.0 (0–10) 24 (10.4) 207 (89.6) 211 (91.3) 20 (8.7) 221 (95.7) 10 (4.3) 2.3 (0–7.0) 45 (19.5) 12 (5.2) 11 (1–44) 2.8 (0.2–48.2) 3 (1.3) 40 (17.3) 43 (18.6) 145 (62.8) 32 55 73 71
(13.9) (23.8) (31.6) (30.7)
CEA, carcinoembryonic antigen; CRT, chemoradiotherapy; ymrT, MRI-assessed T staging post-treatment; ypTNM, pathologic TNM staging post-treatment.
Please cite this article in press as: Huh JW et al. Diagnostic accuracy and prognostic impact of restaging by magnetic resonance imaging after preoperative chemoradiotherapy in patients with rectal cancer. Radiother Oncol (2014), http://dx.doi.org/10.1016/j.radonc.2014.08.023
3
J.W. Huh et al. / Radiotherapy and Oncology xxx (2014) xxx–xxx Table 2 Multivariate Cox analyses of the prognostic factors for 5-year disease-free survival (DFS). Factor
Age, years
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Original article
Diagnostic accuracy and prognostic impact of restaging by magnetic resonance imaging after preoperative chemoradiotherapy in patients with rectal cancer Jung Wook Huh a, Hee Cheol Kim a,⇑, Soon Jin Lee b, Seong Hyeon Yun a, Woo Yong Lee a, Yoon Ah Park a, Yong Beom Cho a, Ho-Kyung Chun a a
Department of Surgery, Sungkyunkwan University School of Medicine; and b Department of Radiology, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea
a r t i c l e
i n f o
Article history: Received 1 April 2014 Received in revised form 6 August 2014 Accepted 12 August 2014 Available online xxxx Keywords: Rectal cancer Preoperative chemoradiation Restaging MRI
a b s t r a c t Background: The prognostic role of restaging rectal magnetic resonance imaging (MRI) in patients with preoperative CRT has not been established. The goal of this study was to evaluate the diagnostic accuracy and prognostic role of radiological staging by rectal MRI after preoperative chemoradiation (CRT) in patients with rectal cancer. Methods: A total of 231 consecutive patients with rectal cancer who underwent preoperative CRT and radical resection from January 2008 to December 2009 were prospectively enrolled. The diagnostic accuracy and prognostic significance of post-CRT radiological staging by MRI was evaluated. Results: The sensitivity, specificity, positive predictive value, and negative predictive value of radiological diagnosis of good responders (ypTNM stage 0–I) were 32%, 90%, 65%, and 69%, respectively. The overall accuracy of MRI restating for good responders was 68%. The 5-year disease-free survival rates of patients with radiological and pathological TNM stage 0, stage I, and stage II–III were 100%, 94%, and 76%, respectively (P = 0.037), and 97%, 87%, and 73%, respectively (P = 0.007). On multivariate analysis, post-CRT radiological staging by MRI was an independent prognostic factor for disease-free survival. Conclusion: Radiological staging by MRI after preoperative CRT may be an independent predictor of survival in patients with rectal cancer. Ó 2014 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2014) xxx–xxx
Multidisciplinary treatment including preoperative chemoradiation (CRT) followed by radical resection for locally-advanced rectal cancer is currently a treatment standard that is associated with improved local control and disease-free survival [1,2]. Several nonrandomized studies suggest that good responses (ypT0-2N0) after preoperative CRT and curative surgery are associated with improved oncologic outcomes [3,4]. Many oncologists have suggested that preoperative identification of good and poor responses to preoperative therapy using imaging modalities enables further tailoring of treatment. Achieving a complete clinical response to CRT may allow for an organ preservation strategy such as local transanal excision or a wait-and-see approach in select patients with rectal cancer [5,6]. In addition, patients with poor response may benefit from intensified strategies to improve resectability and treatment completion [7,8]. Although the accuracy and validity of rectal cancer restaging by current imaging modalities remain ⇑ Corresponding author. Address: Department of Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 135-710, Republic of Korea. E-mail address: [email protected] (H.C. Kim).
uncertain, a recent promising study indicated that magnetic resonance imaging (MRI) assessment of tumor regression grade after preoperative CRT correlated with survival in patients with rectal cancer [9]. High-resolution MRI has been established as the most accurate method for preoperative staging of rectal cancer [10]; however, its role in the restaging of patients who receive preoperative CRT has not yet been determined [9–12]. The goal of this study was to evaluate the diagnostic accuracy and prognostic relevance of post-CRT MRI assessment of response to preoperative therapy in patients with rectal cancer.
Methods A total of 300 consecutive patients who underwent curative surgery after preoperative CRT for locally-advanced rectal cancer at our institution between January 2008 and December 2009 were prospectively enrolled in this study. Patients with metastatic disease, recurrent disease, local resection, or pretreatment TNM stage I, and those lacking pretreatment MRI assessment or follow-up data were excluded from the study cohort. Ultimately, 231 patients
http://dx.doi.org/10.1016/j.radonc.2014.08.023 0167-8140/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Huh JW et al. Diagnostic accuracy and prognostic impact of restaging by magnetic resonance imaging after preoperative chemoradiotherapy in patients with rectal cancer. Radiother Oncol (2014), http://dx.doi.org/10.1016/j.radonc.2014.08.023
2
Radiologic restaging in rectal cancer
were included in the present analysis. This study was reviewed and approved by the appropriate institutional review board. All patients underwent concurrent CRT consisting of preoperative 5-fluorouracil-based chemotherapy and pelvic radiation (4500–5040 cGy), which was followed by radical surgery 6– 8 weeks later [13,14]. Of 231 patients, 216 (93.5%) received postoperative chemotherapy. The mean interval between preoperative CRT and surgery was 55 days (median, 55 days; range, 26– 225 days). All patients underwent total mesorectal excision and the tumors were staged according to the 7th American Joint Committee on Cancer (AJCC) TNM classification. Responses to chemoradiation were evaluated using the tumor regression grade classification proposed by Dworak et al. [15], as follows: grade 0, no regression; grade 1, minimal regression; grade 2, moderate regression; grade 3, near-complete regression; and grade 4, complete regression. All patients received rectal MR imaging within a mean of 40 days (median, 41 days; range, 12–204 days) after completion of preoperative CRT. MRI was performed using a 3.0-T whole body system (Intera Achieva 3T, Philips Medical Systems, Best, Netherlands) with a cardiac phased-array coil in a supine position [16,17]. First, a sagittal localizing image was obtained for the selection of axial and coronal images with a T2-weighted turbo spinecho sequence (3-mm slice thickness, 800 785 matrix, 360-mm field of view, and four signals acquired). The oblique axial and coronal T2-weighted turbo spin-echo images were obtained orthogonal and parallel to the long axis of the rectal tumor. An axial T1-weighted turbo-field-echo sequence (3-mm slice thickness, 400 400 matrix, 360-mm field of view, and two signals acquired) was also performed. All sequences were performed without fat saturation. All MR images were independently interpreted by an expert radiologist (S.J.L.) who was blinded to the results of other imaging examinations and histopathological evaluations. For rectal wall (mrT) staging, we used the following criteria: mrT1 (the tumor signal intensity was confined to the submucosal layer), mrT2 (the tumor signal intensity extended into the muscle layer leading to an irregular or thickened muscle layer), mrT3 (the tumor signal intensity extended through the muscular layer into the perirectal fat), and mrT4 (the tumor signal intensity extended into adjacent organs). Lymph nodes were classified as malignant if they exhibited an indistinct border of any size, an irregular margin, or mixed signal intensity, instead of using size criteria [18]. The patients were followed at 3-month intervals for 2 years, at 6-month intervals for the next 3 years, and annually thereafter. For patients who did not return for observation after 1 year, information was obtained by mail or telephone. On a semi-annual basis or when there was a suspicion of recurrence, the follow-up examination included a clinical history, physical examination, serum carcinoembryonic antigen (CEA) level, chest X-ray, abdominopelvic computed tomography (CT) or MRI, colonoscopy, and positron emission tomography (PET) scanning where available. Recurrence was determined by clinical and radiological examination or using histological confirmation. The main pattern of recurrence was recorded as the first site of detectable failure during the followup period. Statistical evaluation was carried out using the statistical package SPSS for Windows (Version 14.0; SPSS Inc., Chicago, IL, USA). Differences between the two groups were tested using Student’s t-test and the chi-squared test as appropriate. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated from a 2 2 contingency table. Disease-free survival curves were calculated using the Kaplan–Meier method and differences between curves were evaluated using the log-rank test. A Cox model multivariate analysis was generated by a forward stepwise selection of variables, and a P value of 0.1 was adopted as the limit for inclusion of a covariant. A P value 60.05 was considered statistically significant.
Results Clinicopathological characteristics of patients are summarized in Table 1. The median age of participants was 56 years (range, 26–81 years) and 163 (70.6%) were male. The median distance from the anal verge was 4.0 cm (range, 0–10). The median number of resected lymph nodes was 11 (range, 1–44). Using the 7th AJCC TNM staging system, 24 and 207 patients had pretreatment stage II and III cancers, respectively, and 32, 55, 73, and 71 patients had pathologic stage 0 (complete response), I, II, and III cancers, respectively. We compared the results of post-CRT radiological TNM staging by MRI with the results of pathologic examination. The sensitivity, specificity, PPV, and NPV of radiological diagnosis of good responders (ypTNM stage 0–I) were 32%, 90%, 65%, and 65%, respectively. The overall accuracy of good responders by MRI was 68%. We also separately assessed the performance of MRI for establishing a complete response. The sensitivity, specificity, PPV, and NPV of radiological diagnosis of a complete response (ypTNM stage 0) were 3%, 99%, 33%, and 86%, respectively. The overall accuracy of complete response prediction by MRI was 86%. The sensitivity, specificity, PPV, NPV, and accuracy of post-treatment MRI diagnosis of regional nodal disease (ymrN+) for prediction of pathologically-diagnosed regional nodal metastases (ypN+) were 37%, 81%, 79%, 40%, and 52% respectively. We performed further analyses to address the question of whether it is safe to perform limited surgery in patients who are mrN+ on pretreatment MRI for pathologically-diagnosed ypN0 disease. The sensitivity, specificity, PPV, NPV, and accuracy of post-treatment MRI of ymrN0 disease in patients who were mrN+ on pre-treatment MRI (n = 207) for pathologically-diagnosed ypN0 disease were 33%, 84%, 81%, 37%, and 49% respectively. Moreover, we performed analysis to guide decision-making in terms of the resection of adjacent organs involved on pre-treatment MRI, but cancer-free on
Table 1 Patient characteristics (n = 231). No. (%) Median age, years (range) Gender Male Female Median distance from anal verge, cm (range) Pretreatment TNM stage II III Operation method Low anterior resection Abdominoperineal resection Histology Adenocarcinoma Mucinous carcinoma Median tumor diameter, cm (range) Positive lymphovascular invasion Positive perineural invasion Median no. of lymph nodes retrieved (range) Median preoperative CEA, ng/ml (range) ymrTNM stage 0 (complete response) I II III ypTNM stage 0 (complete response) I II III
56 (26–81) 163 (70.6) 68 (29.4) 4.0 (0–10) 24 (10.4) 207 (89.6) 211 (91.3) 20 (8.7) 221 (95.7) 10 (4.3) 2.3 (0–7.0) 45 (19.5) 12 (5.2) 11 (1–44) 2.8 (0.2–48.2) 3 (1.3) 40 (17.3) 43 (18.6) 145 (62.8) 32 55 73 71
(13.9) (23.8) (31.6) (30.7)
CEA, carcinoembryonic antigen; CRT, chemoradiotherapy; ymrT, MRI-assessed T staging post-treatment; ypTNM, pathologic TNM staging post-treatment.
Please cite this article in press as: Huh JW et al. Diagnostic accuracy and prognostic impact of restaging by magnetic resonance imaging after preoperative chemoradiotherapy in patients with rectal cancer. Radiother Oncol (2014), http://dx.doi.org/10.1016/j.radonc.2014.08.023
3
J.W. Huh et al. / Radiotherapy and Oncology xxx (2014) xxx–xxx Table 2 Multivariate Cox analyses of the prognostic factors for 5-year disease-free survival (DFS). Factor
Age, years
