DOI: https://doi.org/10.21203/rs.3.rs-1502216/v1
To study the pattern and treatment outcome of rectal cancer (RC) with concurrent locoregional recurrence (LR) and distant metastasis (DM) after total mesorectal excision (TME) and to identify patient-, disease-, and treatment-related factors associated with differences in prognosis after concurrent LR and DM.
RC patients who were diagnosed with concurrent LR and DM after TME from May 2015 to June 2019 were included in our study. All patients received single or multiple treatment modalities under the guidance of multidisciplinary team (MDT) of colorectal cancer in Fudan University Shanghai Cancer Center.The prognostic of various clinicopathological factors for survival calculated by Kaplan-Meier curves and Cox regression analysis.
A total of 74 RC patients with conurrent LR and DM after TME were eligible for analysis with median follow-up of 27 months. Median survival of patients included was 34 months and 30 patients (41%) died. 59 patients (80%) underwent comprehensive treatments. Patients with oligometastatic disease (OMD) achieved no evidence of disease (NED) status more frequently than those with multiple metastases (P = 0.003). In the univariate analysis, patients achieving NED, diagnosed with OMD and five or less peritoneal metastases tended to have longer survival after LR and DM diagnosis (P < 0.05). In the multivariate analysis, attaining NED status was the only independent factor for survival (hazard ratio (HR), 2.419; P = 0.032). Survival after concurrent LR and DM in the non-NED group was significantly shorter than in the NED group (median survival, 32 vs 46 months; HR, 2.7; P = 0.014).
The pattern and treatment outcome of RC with concurrent LR and DM after TME has changed with development of multiple treatment modalities. Although the prognosis is still poor, pursuing NED status through comprehensive treatments may improve the survival of RC patients with concurrent LR and DM after TME.
Colorectal cancer (CRC) is the third most common cancer worldwide, and its overall 5-year survival rate is approximately 65% (1, 2). Approximately 50% of patients with CRC develop distant metastasis (DM) after curative resection, the most common of which is liver metastases (3). Rectal cancer (RC) accounts for 29% of all CRCs (2). Total mesorectal excision (TME) combined with pre-operative or postoperative radiotherapy (RT) or chemoradiotherapy (CRT) significantly reduce the locoregional recurrence (LR) rate in patients with RC to less than 10%, even 5% in some clinical center. There are about 3% RC patients diagnosed with concurrent LR and DM after TME, which may cause severe disabling symptoms and usually have fatal outcomes (4–7).
For early and locally advanced RC, normative guidelines can be adopted for standardized treatment, while no consensus of treatment has been reached for concurrent LR and DM after TME. With the development of different treatment modalities, perioperative chemotherapy, palliative chemotherapy, targeted therapy, RT, radiofrequency ablation (RFA) and surgical resection can be applied singly or multiply.Several key problems have not been solved, including the sequence of local intervention and systemic treatment, the selection of surgical resection or RFA for local treatment, and the evaluation of tumors’ sensitivity to chemotherpy or RT. Thus, the individualized and comprehensive treatment of concurrent LR and DM still needs to be intensively studied.
Here, our study was designed to study the pattern and treatment outcome of RC with concurrent LR and DM after TME and to identify patient-, disease-, and treatment-related factors associated with differences in prognosis after concurrent LR and DM.
Study Design and Patients
RC patients who were diagnosed with concurrent LR and DM after TME from May 2015 to June 2019 and fulfilled the following criteria were eligible for study entry: (i) aged from 18 to 80 years old at the time of diagnosis of concurrent LR and DM; (ii) diagnosed with resectable RC (histologically proven rectal adenocarcinoma) and received TME surgery with or without preoperative CRT; (iii) received treatment for LR and DM at Fudan University Shanghai Cancer Center (FUSCC); (iv) had complete medical records since RC diagnosis. After being diagnosed with concurrent LR and DM, all patients received single or multiple treatment modalities, including palliative chemotherapy, RT, RFA and surgical resection with or without preoperative chemotherapy under the guidance of multidisciplinary team (MDT) of CRC at FUSCC.
Patients were followed up regularly according to Chinese guidelines for CRC and ended at date of death or on December 31, 2019. Physical examination and CEA were performed every 3-6 months for the first 2 years, every 6 months within the third to fifth year, and then annually. Chest/abdominal/pelvis computed tomography was performed annually for up to 5 years, and colonoscopy was performed for proper patients the first year after treatment and repeated in the third year if no advanced adenoma was found and then every 5 years. Clinical and pathological data were collected from electronic medical record system. Data on treatments and follow-up were gathered from surgeons, medical oncologists and radiologists. Patient data were collected prospectively using a standard form.The study was reviewed and approved by Institutional Ethics Committees of Fudan University Shanghai Cancer Center.
Evaluations of LR, DM and NED
LR was defined as radiologic and/or histologic evidence of a tumor within the lesser pelvis or the perineal wound after a macroscopically complete resection. LR location was categorized according to an adapted version of the subdivision proposed by Philipsen et al (8) into recurrences located at the level of the anastomosis, regional lymph node and pelvic recurrences. DM was defined as radiologic and/or histologic evidence of a tumor in any other area. In this study, oligometastatic disease (OMD) was defined as DM in up to 2 organs or structures including liver, lung and localized lymph node, absence of ascites and peritoneal, bone and central nervous system metastasis.No evidence of disease (NED) status was defined as all LR and DM being grossly resected or ablated and with no sign of remnant disease at one month after surgery. Clear circumferential margin of local recurrence was not mandatory for determination of NED. Two fixed senior radiologists checked all images reported LR and DM.
Statistical Analyses
Chi-square tests were used to compare proportions, and Mann-Whitney U tests were used to compare continuous variables. Kaplan-Meier analyses were used to compare overall survival in patients between different groups. Cox regression was used for univariate and multivariate analyses with hazard ratios (HRs) and 95% confidence intervals (CI). Factors that were statistically significant in the univariate analysis were included in the multivariate analysis. P < 0.05 was considered as significant. Data on patients who were alive were censored at date of last contact. Because the aim of the study was to document the pattern and treatment outcome of RC patients with concurrent LR and DM after TME, the starting point for all survival analyses was the date of LR and DM diagnosis. All analyses were performed with SPSS statistical software (version 19.0 for Windows; SPSS Inc, Chicago, IL).
Among 8,376 patient with RC received TME surgery at FUSCC from May 2015 to June 2019, a total of 74 patients diagnosed with concurrent LR and DM were included in our study. The diagnostic rate of concurrent LR and DM in RC patients was 0.88%. Median time between date of LR and DM and date of primary tumor diagnosis was 16 months (range, 1 to 60 months) (Fig. 1).
The clinicopathological characteristics of 74 eligible patients were summarized in Table 1. Among all patients, male patients were more likely to be diagnosed with concurrent LR and DM (male 60.8% v. female 39.2%, P = 0.081, by binomial test). Half were aged over 60 years old. The primary tumor of 62.2% (46/74) patients were located over 5cm from the anal verge. In terms of characteristics for the primary tumor, 91.9% (68/74) were diagnosed as T3-4, 60.8% as positive lymph nodes involved and 23.0% as G3 tumors. 74.3% (55/74) had anterior resection and 31.1% (23/74) had preoperative CRT or RT. For the type of local recurrence, 47.3% (35/74) of patients were diagnosed as regional lymph node recurrence while 31.1% (23/74) as anastomotic recurrence and 21.6% (16/74) as undetermined pelvic recurrence. For the type of distant metastasis, 71.6% (53/76) of patients were diagnosed with OMD and 28.4% (21/74) of patients were diagnosed to have metastases in 3 or more organs or in peritoneal. 73.0% and 85.1% of observed LR and DM had occurred within 2 and 3 years.
Characteristics | No. (%) |
---|---|
Gender | |
Male | 45 (60.8) |
Female | 29 (39.2) |
Age (years) | |
< 60 | 37 (50.0) |
≥ 60 | 37 (50.0) |
Primary tumor location: distance from the anal verge (cm) | |
> 5 | 46 (62.2) |
≤ 5 | 28 (37.8) |
Type of local recurrence | |
Anastomotic recurrence | 23 (31.1) |
Regional lymph node recurrence | 35 (47.3) |
Undetermined pelvic recurrence | 16 (21.6) |
Distant metastasis | |
Liver/lung/localized lymph node | 53 (71.6) |
3 or more organs/structures involved or peritoneal metastases | 21 (28.4) |
Type of resection of primary tumor | |
Anterior resection | 55 (74.3) |
Abdominoperineal resection | 19 (17.6) |
Others | 6 (8.1) |
T stage of primary tumor | |
T1-2 | 6 (8.1) |
T3 | 41 (55.4) |
T4 | 27 (36.5) |
N stage of primary tumor | |
N0 | 29 (39.2) |
N1 | 36 (48.6) |
N2 | 9 (12.2) |
Tumor grade of primary tumor | |
G1-2 | 57 (77.0) |
G3 | 17 (23.0) |
Preoperative CRT or RT of primary tumor | |
Yes | 23 (31.1) |
No | 51 (68.9) |
Time to recurrence | |
< 24 months | 52 (70.3) |
24–36 months | 10 (13.5) |
> 36 months | 12 (16.2) |
Treatment modalities for 74 RC patients with concurrent LR and DM were listed in Table 2. 70 patients (94.6%) underwent at least one of the local treatments including surgical resection, RT and RFA. 48 patients (64.9%) received systemic treatments such as perioperative chemotherapy and palliative chemotherapy. 59 patients (79.7%) underwent multiple treatments. The results suggested that the vast majority of patients with LR and DM received comprehensive treatment no matter aggressively or palliatively.
Treatment modality | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
Surgical resection | √ | √ | √ | √ | √ | |||||
Perioperative chemotherapy | √ | √ | ||||||||
Palliative chemotherapy | √ | √ | √ | √ | ||||||
Radiotherapy | √ | √ | √ | √ | ||||||
Radiofrequency ablation | √ | √ | √ | √ | ||||||
Cases | 8 | 12 | 5 | 16 | 3 | 10 | 9 | 4 | 3 | 4 |
Percentage (%) | 10.8 | 16.2 | 6.8 | 21.6 | 4.1 | 13.5 | 12.2 | 5.4 | 4.1 | 5.4 |
During follow-up, 30 patients (40.5%) died and for the remaining 44 patients, median time between date of LR and DM diagnosis and date of last contact was 27 months (range, 17 to 48 months). Median survival after LR and DM diagnosis was 34 months (95% CI, 28.6 to 39.4 months) and three-year survival after LR and DM was estimated at 49.3% (Fig. 2).
In univariate analysis, type of distant metastasis (HR, 2.464; 95%CI, 1.132–5.362;P = 0.023), number of peritoneal metastases (HR, 2.637; 95%CI, 1.140–2.229; P = 0.023) and NED status (HR, 2.727; 95%CI, 1.229–6.049; P = 0.014) were associated with survival (Table 3). Kaplan-Meier analysis showed that patients achieving NED (P = 0.009), diagnosed with OMD (P = 0.017) and five or less peritoneal metastases (P = 0.017) tended to have longer survival after LR and DM diagnosis (Fig. 3).
Related factors | Univariate | Multivariate | |||||
---|---|---|---|---|---|---|---|
HR | 95% CI | P value | HR | 95% CI | P value | ||
Gender | 0.296 | ||||||
Male | 1.000 | ||||||
Female | 0.658 | 0.300-1.433 | |||||
Age (years) | 0.333 | ||||||
< 60 | 1.000 | ||||||
≥ 60 | 0.691 | 0.327–1.460 | |||||
Primary tumor location: distance from the anal verge (cm) | 0.856 | ||||||
> 5 | 1.000 | ||||||
≤ 5 | 0.933 | 0.443–1.965 | |||||
Type of local recurrence | 0.159 | ||||||
Anastomotic recurrence | 1.000 | ||||||
Regional lymph node metastasis | 0.461 | 0.205–1.038 | |||||
Undetermined pelvic recurrence | 0.565 | 0.182–1.751 | |||||
Distant metastasis | 0.023 | 0.068 | |||||
Liver/lung/localized lymph node | 1.000 | 1.000 | |||||
3 or more organs/structures involved or peritoneal metastases | 2.464 | 1.132–5.362 | 2.106 | 0.947–4.684 | |||
Localized abdominal recurrence | 0.060 | ||||||
0 | 1.000 | ||||||
1–3 | 0.695 | 0.279–1.734 | |||||
> 3 | 2.011 | 0.696–5.806 | |||||
Peritoneal metastases | 0.023 | 0.513 | |||||
0–5 | 1.000 | 1.000 | |||||
> 5 | 2.637 | 1.140–6.099 | 1.380 | 0.526–3.623 | |||
Type of surgery of primary tumor | 0.268 | ||||||
Anterior resection | 1.000 | ||||||
Abdominoperineal resection | 2.064 | 0.857–4.971 | |||||
Others | 1.111 | 0.257–4.815 | |||||
T stage of primary tumor | 0.362 | ||||||
T1-2 | 1.000 | ||||||
T3 | 1.668 | 0.220-12.624 | |||||
T4 | 0.965 | 0.121–7.668 | |||||
N stage of primary tumor | 0.424 | ||||||
N0 | 1.000 | ||||||
N1 | 0.979 | 0.441–2.174 | |||||
N2 | 1.904 | 0.652–5.565 | |||||
Tumor grade of primary tumor | 0.199 | ||||||
G1-2 | 1.000 | ||||||
G3 | 0.551 | 0.222–1.368 | |||||
Preoperative treatment of primary tumor | 0.242 | ||||||
Chemoradiotherapy or radiotherapy | 1.000 | ||||||
None | 1.801 | 0.673–4.822 | |||||
Time to recurrence | 0.233 | ||||||
< 24 months | 1.000 | ||||||
24–36 months | 3.074 | 0.835–11.315 | |||||
> 36 months | 1.779 | 0.599–5.290 | |||||
No evidence of disease | 0.014 | 0.032 | |||||
NED | 1.000 | 1.000 | |||||
Non-NED | 2.727 | 1.229–6.049 | 2.419 | 1.078–5.427 |
After multiple variables adjustment in the Cox proportional hazards regression model, number of peritoneal metastases lost its statistically significance (HR, 1.380; 95%CI, 0.526–3.623;P = 0.513) and type of distant metastasist was marginal statistically significant for predicting survival (HR, 2.106; 95%CI, 0.947–4.684; P = 0.068) (Table 3). NED status was the only independent factor for survival after LR and DM diagnosis (HR, 2.419; 95%CI, 1.078–5.427; P = 0.032) (Table 3).
The relationship between clinicopathological features and NED status was then analyzed (Table 4). The type of distant metastasis (P = 0.003), number of localized abdominal recurrence (P = 0.005), number of peritoneal metastases were all significantly related with NED status (P = 0.001). Thus, patients with OMD can achieve NED status more frequently.
Characteristics | NED, n (%) (n = 41) | Non-NED, n (%) (n = 33) | χ2 | P value |
---|---|---|---|---|
Gender | ||||
Male | 25 (33.8) | 20 (27.0%) | 0.001 | 0.974 |
Female | 16 (21.6) | 13 (17.6%) | ||
Age (years) | ||||
< 60 | 17 (23.0) | 20 (27.0%) | 2.680 | 0.102 |
>/= 60 | 24 (32.4) | 13 (17.6%) | ||
Primary tumor location: distance from the anal verge (cm) | ||||
> 5 | 26 (35.1) | 20 (27.0%) | 0.061 | 0.804 |
≤ 5 | 15 (20.3) | 13 (17.6%) | ||
Type of local recurrence | ||||
Anastomotic recurrence | 10 (13.5) | 13 (17.6%) | 2.115 | 0.347 |
Regional lymph node metastasis | 22 (29.7) | 13 (17.6%) | ||
Undetermined pelvic recurrence | 9 (12.2) | 7 (9.5%) | ||
Distant metastasis | ||||
Liver/lung/localized lymph node | 35 (47.3) | 18 (24.3%) | 8.525 | 0.003 |
3 or more organs/structures involved or peritoneal metastases | 6 (8.1) | 15 (20.3%) | ||
Localized abdominal recurrence | ||||
None | 10 (13.5) | 4 (5.4%) | 10.636 | 0.005 |
</=3 | 29 (39.2) | 18 (24.3%) | ||
> 3 | 2 (2.7) | 11 (14.9%) | ||
Peritoneal metastases | ||||
None or </=5 | 39 (52.7) | 22 (29.7%) | 10.223 | 0.001 |
> 5 | 2 (2.7) | 11 (14.9%) | ||
T stage of primary tumor | ||||
T1-2 | 3 (4.1) | 3 (4.1%) | 1.161 | 0.560 |
T3 | 25 (33.8) | 16 (21.6%) | ||
T4 | 13 (17.6) | 14 (18.9%) | ||
N stage of primary tumor | ||||
N0 | 17 (23.0) | 12 (16.2%) | 2.920 | 0.232 |
N1 | 17 (23.0) | 19 (25.7%) | ||
N2 | 7 (9.5) | 2 (2.7%) | ||
Tumor grade of primary tumor | ||||
G1-2 | 30 (40.5) | 27 (36.5%) | 0.773 | 0.379 |
G3 | 11 (14.9) | 6 (8.1%) | ||
Preoperative treatment of primary tumor | ||||
Chemoradiotherapy or radiotherapy | 16 (21.6) | 7 (9.5%) | 2.708 | 0.100 |
None | 25 (33.8) | 26 (35.1%) | ||
Treatment modality | ||||
Single treatment | 8 (10.8) | 7 (9.5%) | 0.033 | 0.857 |
Multiple treatment | 33 (44.6) | 26 (35.1%) |
Further survival analysis showed that 11 patients (26.8%) in NED group and 19 patients (57.6%) in non-NED group died during follow-up (Table 5). Three-year survival after LR and DM was estimated to be 61.8% in NED group and 29.6% in non-NED group. Patients in NED group have longer median survival after LR and DM diagnosis of 46 months (95% CI, 37.5 to 54.5 months), compared with that of 32 months (95% CI, 24.2 to 39.8 months) in non-NED group. Consequently, RC patients with concurrent LR and DM after TME have a poor prognosis, but reaching NED status after treatments can improve patients’ survival.
Survival outcome | NED (n = 41) | Non-NED (n = 33) |
---|---|---|
Number of patients followed until death | 11 (26.8%) | 19 (57.6%) |
3-year survival rate | 61.8% | 29.6% |
Median survival time (month) | 46 | 32 |
Though the incidence of concurrent LR and DM after TME of rectal cancer is quite low, which is 0.88% in our study, the prognosis of this subset of patients is poor. Our study retrospectively collected the pattern and treatment outcome of 74 RC patients with oncurrent LR and DM after TME, to identify patient-, disease-, and treatment-related factors associated with differences in prognosis.
Our study found that male patients (60.8%) were more likely to be diagnosed with concurrent LR and DM. This epidemiological feature was similar to the data of CRC patients underwent surgery in FUSCC from 2008 to 2017. We found that the vast majority of patients with LR and DM received comprehensive treatment no matter aggressively or palliatively. Although the prognosis is still poor, pursuing NED status through comprehensive treatments may improve the survival of RC patients with concurrent LR and DM after TME. There are several possible explanations for this finding.
The first explanation concerns the treatment modalities. With development of multiple treatment modalities and MDT, the pattern and treatment outcome of RC with concurrent LR and DM after TME has changed. Our results indeed showed that the majority of patients (79.7%) underwent multiple treatments no matter aggressively or palliatively. Compared with Dutch trial in 2004[参考文献], more drugs with better clinical applications, more options for local and systematic treatment and modified therapy with LR and DM can be reached at present. For example, short-term preoperative RT (a total dose of 25 Gy in five fractions over 5 to 7 days) was used at that time, while long-term preoperative RT (a total dose of 45 Gy in 25 fractions over about 5 weeks) is widely used in FUSCC at present. Meanwhile, treatment of metastasis is more agressive at present (9, 10). With these development, the median survival after LR and DM diagnosis was 34 months in our study and the median survival after LR 6.1 months in preoperative RT + TME group and 15.9 months in TME group in Dutch trial (4).
The second explanation concerns the survival. Although RC with concurrent LR and DM after TME has a poor prognosis, many studies have focused on attaining NED status after treatments to improve the overall survival which is also confirmed by our results. Furthermore, we found patients with OMD can achieve NED status after treatments more frequently. Consequently, patients with OMD after TME are the candidates to pursue NED status through upfront curative resection from the initial of the treatment, including CRT and RFA (11, 12).
The third explanation concerns surgical resection which is an important treatment modality to achieve NED status. However, if NED status not achieved, surgical resection of LR still plays an essential role. Due to the limited pelvic space, recurrent tumors are easy to compress other organs, such as ureter and blood vessels, leading to renal insufficiency and lower limb edema, which seriously affects the quality of life and subsequent treatment. Patients undergoing R0 resection have the greatest survival advantage following surgery for recurrent rectal cancer. Meanwhile, there is a survival advantage for R1 over R2 resection (13).
The present study has several limitations. First, the study design was a retrospective single-center trial. Second, we defined OMD as metastasis in up to 2 organs or structures including liver, lung and localized lymph node, without taking the number, size of tumors into account. In ASCO-GI 2020, OMD was defined as up to 5 metastasis, up to 3 metastasis in one organ, up to 3 affected organs, size ≤ 3 cm, absence of ascites and peritoneal, bone and central nervous system metastasis (14). Thus, it is possible that less patients were counted into OMD status.
In conclusion, our study showed that RC patients with concurrent LR and DM after TME have a poor prognosis. Patients with OMD are the candidates to pursue NED status through multiple treatments including curative resection which may improve the overall survival.
DATA AVAILABILITY STATEMENT
Data are available upon reasonable request.
ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Institutional Ethics Committees of Fudan University Shanghai Cancer Center. The authors are accountable for all aspects of the work.
FUNDING
The study was supported by grants from the National Natural Science Foundation of China (U1932145 to JP, 82002946 to YL), Science and Technology Commission of Shanghai Municipality (18401933402 to JP), Fudan University Shanghai Cancer Center Basic and Clinical Translational Research Seed Foundation (YJZZ201802 to SC), and Shanghai Sailing Program (19YF1409500 to YL).
AUTHOR CONTRIBUTIONS
Data collection: YC, XM. Case selection: YL, XM and XH. Evaluation: FL, SC, and JP. Statistical analysis: YL, SM. Paper writing: YC, YL, and XM. All authors contributed to the article and approved the submitted version.
CONFLICTS OF INTEREST
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
ACKNOWLEDGEMENTS
We would like to thank the patients and family members who gave their consent on presenting the data in this study, as well as the investigators and research staff involved in this study.