Curative Treatment After Surgery in Colorectal Cancer
Since 1990, nine randomized trials have been reported comparing different low-and high-intensity surveillance programs after curative surgery for adenocarcinoma of the colon or rectum. These trials are summarized in Table 2. Overall, the total number of recurrences was similar with the 'standard' protocol or with intensive surveillance. However, recurrences were discovered earlier or were more likely to be asymptomatic in the groups receiving intensive surveillance. Importantly, in five trials, surgery with curative intent was reported more frequently in the groups receiving intensive surveillance. The vast majority of surgery was for metastatic or recurrent disease as opposed to new primary cancers.
Regarding the most pertinent end point of overall survival, only two studies reported statistically significant improvement in overall survival in the group receiving intensive surveillance. The remaining trials failed to show a survival benefit from more intensive follow-up. However, many studies suffered from design flaws. Notably, many of the control groups underwent fairly intensive testing, making any small differences in outcome that may have been present less apparent. Several studies also included stage I patients who had very favorable outcomes, further diluting outcomes. Individual studies also had small sample sizes with insufficient statistical power.
Two early meta-analyses were published but were not informative as one was based almost completely on nonrandomized data and the other included a mixture of cohort studies and randomized data. Subsequently, four meta-analyses were performed based on the randomized trials above. These meta-analyses are summarized in Table 3. All of these meta-analyses showed that patients undergoing more intensive surveillance had an improved survival. The estimated benefit was a relative 20–33% reduction in all-cause mortality and absolute overall survival gain between 7 and 13%. On the basis of these data, intensive follow-up is now recommended as a standard of care by major oncology guidelines.
In these meta-analyses, the total number of recurrences in intensive and standard follow-up groups was similar although the incidence of asymptomatic recurrences was significantly more common in patients on intensive follow-up. This may be attributable to earlier detection of recurrences, and therefore the increased number of curative resections.
While intensive surveillance does appear to provide an overall survival advantage, it is not possible to determine which component contributes the most. Trials that include CEA monitoring and liver imaging demonstrated benefit, whereas trials not including these tests did not. However, all studies that utilized liver imaging also monitored blood CEA levels and the relative benefit of each modality alone is currently undetermined. Based on current data, intensive programs of surveillance do indeed improve survival. A combination of modalities is thought to increase the sensitivity of detection of recurrence. However, there is little data to define the optimal complement of tests that should be incorporated in the follow-up programs, nor specific data to recommend the frequency of these tests.
There are at least three ongoing clinical trials that will provide valuable data on survival outcomes, quality of life and cost–effectiveness of specific surveillance programs postcurative surgery: GILDA, FACS and COLOFOL (a pragmatic study to assess the frequency of surveillance tests after curative resection in patients with stage II and III colorectal cancer – a randomized study).
The GILDA trial involves more than 40 centers (predominantly in Italy) and randomizes patients to minimal or intensive follow-up. The trial began in 1998 and final results have not yet been reported. FACS is a UK-led study randomizing patients in a two by two design. FACS evaluates intensive imaging (CT chest/abdomen/pelvis 6 monthly for 2 years) and annually for 3 years versus a single CT at 12–18 months and CEA monitoring 3–6 monthly versus no CEA monitoring. Recruitment has been completed and interim analysis is expected in 2013. COLOFOL randomizes patients into intensive CEA and CT imaging at 6 monthly intervals for 3 years versus the same tests at 12 and 36 months postsurgery.
These trials will be informative on the optimal investigations and optimal frequency of investigations. They will also clarify the survival benefit of intensive surveillance. However, none of these trials address the duration of follow-up nor are they risk adapted to evaluate the intensity of follow-ups in subgroups of patients with varying risk profiles.
Beyond current imaging techniques, blood tests and endoscopy, the next few years could see the emergence of new investigation techniques as well as risk-adaptive surveillance that integrate clinicopathologic and biological information to individualize surveillance strategies. The authors will also touch on surveillance strategies postcurative treatment of metastatic disease.
PET scans provide functional/biological information. PET uses FDG-glucose to highlight areas of increased glucose metabolism and indirectly reflects the increased metabolic rate in proliferative tissue. However, PET alone is limited by poor anatomic delineation. The hybrid PET/CT scan fuses PET and CT images. A recent systematic review demonstrated that in diagnosing colorectal metastases, PET/CT was more sensitive than CT for extrahepatic disease (75–89 vs 58–64%) although specificities were similar (95–96 vs 87–97%). For hepatic disease, PET/CT had higher sensitivity (91–100 vs 78–94%) and specificity (75–100 vs 25–98%). For local recurrence, PET/CT also had higher sensitivity (93–100 vs 0–100%) but similar specificity (97–98% for both modalities).
There are limitations of using PET/CT for colorectal metastases. In particular, PET/CT cannot detect tumors less than 5 mm and has diminished discriminatory capability for lesions 5–10 mm. False-negative results may occur with recent chemotherapy (less than 1 month before the scan) although this is of less concern in surveillance.
CT alone may underdiagnose widespread disease (e.g., extrahepatic disease). Presently, PET/CT's role is to improve preoperative identification of resectable metastases. PET/CT is therefore currently deployed after CT detection of metastases to clarify disease burden and possibly avoid unnecessary surgery. While PET/CT is clearly an emerging modality, its role in routine surveillance remains to be ascertained.
To date, a variety of detection technologies have been developed to enrich and detect circulating tumor cells (CTCs) from the peripheral blood. Studies have shown that the presence of CTCs in the peripheral blood of metastatic colorectal cancer patients has promising prognostic value. In patients with resected nonmetastatic cancer, data on the significance of CTCs are more limited. A recent meta-analysis reported that the detection rates of CTC in peripheral blood of patients with resected non-metastatic colorectal cancer varied from 4 to 57%. In addition, several studies and a meta-analysis have found that presence of CTC postoperatively is a prognostic marker of poor disease-free survival.
Early in tumor development, apoptotic and necrotic cells of the primary tumor release DNA into the bloodstream. Teams are beginning to develop assays that detect cell-free tumor DNA in the peripheral blood of cancer patients. These studies have thus far focused on metastatic patients but are likely to be extended to nonmetastatic patients.
The use of serial CTC or circulating tumor DNA assays in surveillance has not been systematically studied as yet in randomized trials or multicenter trials. Since circulating material are supposed to be specifically identified 'tumor' cells or secreted tumor DNA detectable in peripheral blood, they may represent the earliest sign of cancer dissemination. Thus, serial CTC and serial circulating DNA assays could have great potential as surveillance tools to enable early diagnosis of recurrence with a lead time to either radiological or symptomatic disease.
Currently, surveillance guidelines recommend a uniform intensive surveillance approach for patients with stage 3 and high-risk stage 2 disease. Aside from clinical staging, several clinicopathologic parameters and multigene assays have been developed to provide additional discriminatory prognostic information regarding relapse risk. It is conceivable that in patients with a higher risk of early relapse, further intensification of surveillance may yield greater benefits. Thus, beyond prognostication, these biological and pathological parameters could serve to individualize surveillance strategies and focus resources upon higher-risk patients. Nevertheless, currently, such biomarkers or 'signatures' are highly population dependent, thus more validation studies are required before they can be translated into clinics.
While guidelines recommend surveillance strategies to be employed postcurative resection of primary early-stage cancer, the surveillance strategies postresection of metastatic disease remains undefined. The duration of follow-up of these patients is also not determined. In a series of 309 consecutive patients with complete 10-year follow-up data, 11% of the patients who were disease-free at 5 years developed later recurrence. Given that more patients are now undergoing surgery for metastatic colorectal cancer, many of whom will develop recurrence that is amenable to further surgery, it is also important to clarify the optimal surveillance strategy postmetastectomy.
Intensive versus Less-intensive Surveillance Strategies
Randomized Trials
Since 1990, nine randomized trials have been reported comparing different low-and high-intensity surveillance programs after curative surgery for adenocarcinoma of the colon or rectum. These trials are summarized in Table 2. Overall, the total number of recurrences was similar with the 'standard' protocol or with intensive surveillance. However, recurrences were discovered earlier or were more likely to be asymptomatic in the groups receiving intensive surveillance. Importantly, in five trials, surgery with curative intent was reported more frequently in the groups receiving intensive surveillance. The vast majority of surgery was for metastatic or recurrent disease as opposed to new primary cancers.
Regarding the most pertinent end point of overall survival, only two studies reported statistically significant improvement in overall survival in the group receiving intensive surveillance. The remaining trials failed to show a survival benefit from more intensive follow-up. However, many studies suffered from design flaws. Notably, many of the control groups underwent fairly intensive testing, making any small differences in outcome that may have been present less apparent. Several studies also included stage I patients who had very favorable outcomes, further diluting outcomes. Individual studies also had small sample sizes with insufficient statistical power.
Two early meta-analyses were published but were not informative as one was based almost completely on nonrandomized data and the other included a mixture of cohort studies and randomized data. Subsequently, four meta-analyses were performed based on the randomized trials above. These meta-analyses are summarized in Table 3. All of these meta-analyses showed that patients undergoing more intensive surveillance had an improved survival. The estimated benefit was a relative 20–33% reduction in all-cause mortality and absolute overall survival gain between 7 and 13%. On the basis of these data, intensive follow-up is now recommended as a standard of care by major oncology guidelines.
In these meta-analyses, the total number of recurrences in intensive and standard follow-up groups was similar although the incidence of asymptomatic recurrences was significantly more common in patients on intensive follow-up. This may be attributable to earlier detection of recurrences, and therefore the increased number of curative resections.
While intensive surveillance does appear to provide an overall survival advantage, it is not possible to determine which component contributes the most. Trials that include CEA monitoring and liver imaging demonstrated benefit, whereas trials not including these tests did not. However, all studies that utilized liver imaging also monitored blood CEA levels and the relative benefit of each modality alone is currently undetermined. Based on current data, intensive programs of surveillance do indeed improve survival. A combination of modalities is thought to increase the sensitivity of detection of recurrence. However, there is little data to define the optimal complement of tests that should be incorporated in the follow-up programs, nor specific data to recommend the frequency of these tests.
Ongoing Clinical Trials
There are at least three ongoing clinical trials that will provide valuable data on survival outcomes, quality of life and cost–effectiveness of specific surveillance programs postcurative surgery: GILDA, FACS and COLOFOL (a pragmatic study to assess the frequency of surveillance tests after curative resection in patients with stage II and III colorectal cancer – a randomized study).
The GILDA trial involves more than 40 centers (predominantly in Italy) and randomizes patients to minimal or intensive follow-up. The trial began in 1998 and final results have not yet been reported. FACS is a UK-led study randomizing patients in a two by two design. FACS evaluates intensive imaging (CT chest/abdomen/pelvis 6 monthly for 2 years) and annually for 3 years versus a single CT at 12–18 months and CEA monitoring 3–6 monthly versus no CEA monitoring. Recruitment has been completed and interim analysis is expected in 2013. COLOFOL randomizes patients into intensive CEA and CT imaging at 6 monthly intervals for 3 years versus the same tests at 12 and 36 months postsurgery.
These trials will be informative on the optimal investigations and optimal frequency of investigations. They will also clarify the survival benefit of intensive surveillance. However, none of these trials address the duration of follow-up nor are they risk adapted to evaluate the intensity of follow-ups in subgroups of patients with varying risk profiles.
Emerging Paradigms
Beyond current imaging techniques, blood tests and endoscopy, the next few years could see the emergence of new investigation techniques as well as risk-adaptive surveillance that integrate clinicopathologic and biological information to individualize surveillance strategies. The authors will also touch on surveillance strategies postcurative treatment of metastatic disease.
PET-CT Scans
PET scans provide functional/biological information. PET uses FDG-glucose to highlight areas of increased glucose metabolism and indirectly reflects the increased metabolic rate in proliferative tissue. However, PET alone is limited by poor anatomic delineation. The hybrid PET/CT scan fuses PET and CT images. A recent systematic review demonstrated that in diagnosing colorectal metastases, PET/CT was more sensitive than CT for extrahepatic disease (75–89 vs 58–64%) although specificities were similar (95–96 vs 87–97%). For hepatic disease, PET/CT had higher sensitivity (91–100 vs 78–94%) and specificity (75–100 vs 25–98%). For local recurrence, PET/CT also had higher sensitivity (93–100 vs 0–100%) but similar specificity (97–98% for both modalities).
There are limitations of using PET/CT for colorectal metastases. In particular, PET/CT cannot detect tumors less than 5 mm and has diminished discriminatory capability for lesions 5–10 mm. False-negative results may occur with recent chemotherapy (less than 1 month before the scan) although this is of less concern in surveillance.
CT alone may underdiagnose widespread disease (e.g., extrahepatic disease). Presently, PET/CT's role is to improve preoperative identification of resectable metastases. PET/CT is therefore currently deployed after CT detection of metastases to clarify disease burden and possibly avoid unnecessary surgery. While PET/CT is clearly an emerging modality, its role in routine surveillance remains to be ascertained.
Circulating Tumor Cells & Circulating DNA
To date, a variety of detection technologies have been developed to enrich and detect circulating tumor cells (CTCs) from the peripheral blood. Studies have shown that the presence of CTCs in the peripheral blood of metastatic colorectal cancer patients has promising prognostic value. In patients with resected nonmetastatic cancer, data on the significance of CTCs are more limited. A recent meta-analysis reported that the detection rates of CTC in peripheral blood of patients with resected non-metastatic colorectal cancer varied from 4 to 57%. In addition, several studies and a meta-analysis have found that presence of CTC postoperatively is a prognostic marker of poor disease-free survival.
Early in tumor development, apoptotic and necrotic cells of the primary tumor release DNA into the bloodstream. Teams are beginning to develop assays that detect cell-free tumor DNA in the peripheral blood of cancer patients. These studies have thus far focused on metastatic patients but are likely to be extended to nonmetastatic patients.
The use of serial CTC or circulating tumor DNA assays in surveillance has not been systematically studied as yet in randomized trials or multicenter trials. Since circulating material are supposed to be specifically identified 'tumor' cells or secreted tumor DNA detectable in peripheral blood, they may represent the earliest sign of cancer dissemination. Thus, serial CTC and serial circulating DNA assays could have great potential as surveillance tools to enable early diagnosis of recurrence with a lead time to either radiological or symptomatic disease.
Risk-Stratified Approach
Currently, surveillance guidelines recommend a uniform intensive surveillance approach for patients with stage 3 and high-risk stage 2 disease. Aside from clinical staging, several clinicopathologic parameters and multigene assays have been developed to provide additional discriminatory prognostic information regarding relapse risk. It is conceivable that in patients with a higher risk of early relapse, further intensification of surveillance may yield greater benefits. Thus, beyond prognostication, these biological and pathological parameters could serve to individualize surveillance strategies and focus resources upon higher-risk patients. Nevertheless, currently, such biomarkers or 'signatures' are highly population dependent, thus more validation studies are required before they can be translated into clinics.
Surveillance Postmetastectomy
While guidelines recommend surveillance strategies to be employed postcurative resection of primary early-stage cancer, the surveillance strategies postresection of metastatic disease remains undefined. The duration of follow-up of these patients is also not determined. In a series of 309 consecutive patients with complete 10-year follow-up data, 11% of the patients who were disease-free at 5 years developed later recurrence. Given that more patients are now undergoing surgery for metastatic colorectal cancer, many of whom will develop recurrence that is amenable to further surgery, it is also important to clarify the optimal surveillance strategy postmetastectomy.
SHARE
![Breast Cancer Screening (PDQ®): Screening - Health Professional Information [NCI]-Special Population](https://cdnus.oss-us-west-1.aliyuncs.com/libs/images/NewsJpg/cate_6/039_215x160.jpg)
