FIT and Colonoscopy in Familial Colorectal Cancer Screening
FIT and Colonoscopy in Familial Colorectal Cancer Screening
In this prospective randomized trial involving first-degree relatives of sporadic CRC index cases, repeated FIT screening detected all CRC and 61% of advanced adenomas, proving equivalent to one-time colonoscopy screening in terms of diagnostic yield and tumor staging. However, colonoscopy was superior to the FIT strategy for the detection of any neoplasm. The study also revealed that the number of subjects requiring colonoscopy to detect one advanced neoplasm was 4 times less in first-degree relatives screened by FIT than in those screened by colonoscopy. This finding indicates that FIT screening may save a large amount of unnecessary colonoscopies in this population.
Our study has a number of strengths. First, it is the first randomized trial to compare colonoscopy with FIT screening in familial CRC. Second, considering that the noninvasive nature of the FIT screening strategy can compensate for its lower capacity for detecting advanced neoplasia with respect to colonoscopy by increasing the screening uptake, we designed an equivalence analysis including consecutive screening-naïve first-degree relatives. Third, the trial was carried out in accordance with the strict requirements of an equivalence study. Equivalence was an a priori research aim and assessed on the basis of intention-to-screen and per-protocol analyses. First-degree relatives were randomly assigned to either screening strategy; an equivalence boundary (±3%) was established for the detection of advanced neoplasia and sample size was calculated for equivalence. Finally, follow-up was completed in >95% of individuals in both arms of the study.
The main finding of our study was that cumulative FIT screening yielded an equivalent advanced neoplasia detection rate to one-time colonoscopy. Equivalence between the 2 strategies was supported by the fact that the 95% CI of the advanced neoplasia detection rate difference was within the predefined equivalence boundary (±3%) at different participation rates. The error margin of ±3% was empirically established, considering that it was clinically acceptable for a 3-year rescreening period with FIT. In fact, the results obtained in the current study were better than expected. For an estimated advanced neoplasia global prevalence of 7.7%, with detection rates of 56.5% for FIT and 96.5% for colonoscopy, we expected to find an advanced neoplasia detection rate of 4.4% and 7.4% for FIT and colonoscopy groups, respectively. However, in the intention-to-screen analysis, we found an advanced neoplasia detection rate of 4.2% and 5.6% for the FIT and colonoscopy strategies, respectively, which was less than expected. The most plausible explanation for this finding is that the global prevalence of advanced neoplasia in this population was lower than that expected a priori, which could be directly related to the fact that >50% of participants were younger than 50 years of age. Equivalence was achieved despite suboptimal compliance with the FIT strategy in our study, as only 72.6% of subjects underwent at least 2 tests. This finding was in concordance with a longitudinal study performed in an average-risk population showing similar compliance with FIT screening.
The yield of FIT to detect advanced neoplasia in familial CRC varied according to the cutoff value established for fecal Hb detection. A recent cross-sectional study testing a qualitative FIT at a high cutoff level (50 μg Hb/g of feces, equivalent to 250 ng Hb/mL buffer) showed low sensitivity for CRC (25%) and for advanced adenoma (35%) in this population. However, the elevated advanced neoplasia miss rate using this cutoff value makes it unacceptable in screening programs. By contrast, when the cutoff level is decreased to 10 μg Hb/g feces (equivalent to 50 ng Hb/mL buffer), as used in the current study, sensitivity for CRC reached 100%, and for advanced adenoma it ranged from 37% to 85%. The false-negative rate for advanced adenoma detection in these studies varied from 16.6% to 37.5%. In our longitudinal study, a colonoscopy performed in 86.3% of patients who tested negative for FIT confirmed that repeated FIT screening detected all CRC and 25 out of 41 advanced adenomas, that is, a false-negative rate of 39% for advanced adenoma. Advocates of colonoscopy might argue that it is unacceptable to miss 39% of advanced adenomas, and others might consider that a 61% sensitivity with only 3 years of screening is a very good result. Considering that large polyps (>1 cm in diameter) become CRC at a rate of roughly 1% per year, a screening program such as annual FIT, allows the detection of most advanced adenomas on subsequent screens before they become malignant. All together, these findings indicate that repeated FIT screening at a cutoff value of 10 μg Hb/g feces has an acceptable advanced neoplasia yield in this moderate-risk population.
In agreement with previous studies carried out in the average-risk and familial CRC risk populations, we found that FIT screening was associated with a significantly lower detection rate of nonadvanced adenomas than colonoscopy. This finding could be considered an advantage of FIT, as most of these lesions are probably low risk and less likely to progress to CRC than advanced adenomas. If, as in average-risk individuals, the presence of <3 nonadvanced adenomas implies a risk similar to that of individuals without adenomas, this finding could help save unnecessary colonoscopy in the familial-risk population. However, the real clinical significance of nonadvanced lesions in this moderate-risk group remains to be clarified.
The usefulness of FIT screening as an alternative to colonoscopy in the familial risk population will finally depend on the capacity of FIT to improve screening uptake. This question cannot be answered by the current study, as the uptake of first-degree relatives assigned to the FIT group was influenced by the fact that most were referred to our high-risk CRC clinic to undergo colonoscopy screening, and also because the study design included colonoscopy after 3 years of FIT screening, if so desired.
We are also aware of the study limitations. First, as stated, we could not evaluate the efficacy of annual FIT screening for a period longer than 3 years due to ethical concerns, a fact that could have had improved FIT performance. Second, a potential bias for increased detection of advanced neoplasia in the FIT group was possible, as participants knew that they could opt out of their assigned screening strategy before signing informed consent. However, this limitation was minimized by the per-protocol analysis, which corroborated there was no significant difference in the detection rate of advanced neoplasia between the 2 arms of the study. Third, we cannot provide accurate information as to whether repeated FIT is able to detect sessile serrated polyps or traditional serrated polyps, because at the time the study was designed (2006), these polyps were classified as hyperplastic and considered to have no malignant potential. In fact, clinical guidelines did not recommend colonoscopy surveillance for individuals with these polyps until recently. Fourth, the current study was not powered to analyze the different risk groups separately. However, the higher-risk subgroup of first-degree relatives having at least 2 family members with CRC constituted a small minority (7.6%) of the whole study sample. To minimize the effect of this potential flaw, the analysis was adjusted for confounding variables, including index-case age and having siblings or more than one family member with CRC. In addition, both arms of the study were homogeneous regarding demographic data and family history.
Finally, the potential advantage of FIT screening to spare endoscopic resources in this moderate-risk population also warrants comment. First, the number of first-degree relatives that needed to undergo colonoscopy to detect one advanced neoplasia was 4 in the FIT group as compared with 18 in the colonoscopy group. Second, up to 71% of colonoscopies in the colonoscopy arm had a normal or nonsignificant result. Overall, repeated screening with FIT needed about 86% fewer colonoscopies to detect 100% CRC and 61% of advanced adenomas, which is consistent with the results of previous studies. Therefore, a potential benefit of FIT over primary colonoscopy in familial CRC screening is that it may save a substantial number of unnecessary colonoscopies, thus preventing harm and lowering costs. Given its equivalence in terms of advanced neoplasia detection, FIT screening should be considered when colonoscopy capacity is limited.
In conclusion, the results of this randomized trial demonstrate that in asymptomatic first-degree relatives of patients with CRC, screening with FIT is equivalent to one-time colonoscopy for the detection of advanced neoplasia. In addition, it also provides evidence for the benefit of repeated FIT screening in terms of colonoscopy resources. This strategy should be taken into account when colonoscopy capacity is limited and particularly in populations where FIT is better accepted than colonoscopy as a CRC screening strategy. Additional studies are needed to determine acceptance of this strategy and its efficacy in reducing the incidence and mortality associated with familial CRC.
Discussion
In this prospective randomized trial involving first-degree relatives of sporadic CRC index cases, repeated FIT screening detected all CRC and 61% of advanced adenomas, proving equivalent to one-time colonoscopy screening in terms of diagnostic yield and tumor staging. However, colonoscopy was superior to the FIT strategy for the detection of any neoplasm. The study also revealed that the number of subjects requiring colonoscopy to detect one advanced neoplasm was 4 times less in first-degree relatives screened by FIT than in those screened by colonoscopy. This finding indicates that FIT screening may save a large amount of unnecessary colonoscopies in this population.
Our study has a number of strengths. First, it is the first randomized trial to compare colonoscopy with FIT screening in familial CRC. Second, considering that the noninvasive nature of the FIT screening strategy can compensate for its lower capacity for detecting advanced neoplasia with respect to colonoscopy by increasing the screening uptake, we designed an equivalence analysis including consecutive screening-naïve first-degree relatives. Third, the trial was carried out in accordance with the strict requirements of an equivalence study. Equivalence was an a priori research aim and assessed on the basis of intention-to-screen and per-protocol analyses. First-degree relatives were randomly assigned to either screening strategy; an equivalence boundary (±3%) was established for the detection of advanced neoplasia and sample size was calculated for equivalence. Finally, follow-up was completed in >95% of individuals in both arms of the study.
The main finding of our study was that cumulative FIT screening yielded an equivalent advanced neoplasia detection rate to one-time colonoscopy. Equivalence between the 2 strategies was supported by the fact that the 95% CI of the advanced neoplasia detection rate difference was within the predefined equivalence boundary (±3%) at different participation rates. The error margin of ±3% was empirically established, considering that it was clinically acceptable for a 3-year rescreening period with FIT. In fact, the results obtained in the current study were better than expected. For an estimated advanced neoplasia global prevalence of 7.7%, with detection rates of 56.5% for FIT and 96.5% for colonoscopy, we expected to find an advanced neoplasia detection rate of 4.4% and 7.4% for FIT and colonoscopy groups, respectively. However, in the intention-to-screen analysis, we found an advanced neoplasia detection rate of 4.2% and 5.6% for the FIT and colonoscopy strategies, respectively, which was less than expected. The most plausible explanation for this finding is that the global prevalence of advanced neoplasia in this population was lower than that expected a priori, which could be directly related to the fact that >50% of participants were younger than 50 years of age. Equivalence was achieved despite suboptimal compliance with the FIT strategy in our study, as only 72.6% of subjects underwent at least 2 tests. This finding was in concordance with a longitudinal study performed in an average-risk population showing similar compliance with FIT screening.
The yield of FIT to detect advanced neoplasia in familial CRC varied according to the cutoff value established for fecal Hb detection. A recent cross-sectional study testing a qualitative FIT at a high cutoff level (50 μg Hb/g of feces, equivalent to 250 ng Hb/mL buffer) showed low sensitivity for CRC (25%) and for advanced adenoma (35%) in this population. However, the elevated advanced neoplasia miss rate using this cutoff value makes it unacceptable in screening programs. By contrast, when the cutoff level is decreased to 10 μg Hb/g feces (equivalent to 50 ng Hb/mL buffer), as used in the current study, sensitivity for CRC reached 100%, and for advanced adenoma it ranged from 37% to 85%. The false-negative rate for advanced adenoma detection in these studies varied from 16.6% to 37.5%. In our longitudinal study, a colonoscopy performed in 86.3% of patients who tested negative for FIT confirmed that repeated FIT screening detected all CRC and 25 out of 41 advanced adenomas, that is, a false-negative rate of 39% for advanced adenoma. Advocates of colonoscopy might argue that it is unacceptable to miss 39% of advanced adenomas, and others might consider that a 61% sensitivity with only 3 years of screening is a very good result. Considering that large polyps (>1 cm in diameter) become CRC at a rate of roughly 1% per year, a screening program such as annual FIT, allows the detection of most advanced adenomas on subsequent screens before they become malignant. All together, these findings indicate that repeated FIT screening at a cutoff value of 10 μg Hb/g feces has an acceptable advanced neoplasia yield in this moderate-risk population.
In agreement with previous studies carried out in the average-risk and familial CRC risk populations, we found that FIT screening was associated with a significantly lower detection rate of nonadvanced adenomas than colonoscopy. This finding could be considered an advantage of FIT, as most of these lesions are probably low risk and less likely to progress to CRC than advanced adenomas. If, as in average-risk individuals, the presence of <3 nonadvanced adenomas implies a risk similar to that of individuals without adenomas, this finding could help save unnecessary colonoscopy in the familial-risk population. However, the real clinical significance of nonadvanced lesions in this moderate-risk group remains to be clarified.
The usefulness of FIT screening as an alternative to colonoscopy in the familial risk population will finally depend on the capacity of FIT to improve screening uptake. This question cannot be answered by the current study, as the uptake of first-degree relatives assigned to the FIT group was influenced by the fact that most were referred to our high-risk CRC clinic to undergo colonoscopy screening, and also because the study design included colonoscopy after 3 years of FIT screening, if so desired.
We are also aware of the study limitations. First, as stated, we could not evaluate the efficacy of annual FIT screening for a period longer than 3 years due to ethical concerns, a fact that could have had improved FIT performance. Second, a potential bias for increased detection of advanced neoplasia in the FIT group was possible, as participants knew that they could opt out of their assigned screening strategy before signing informed consent. However, this limitation was minimized by the per-protocol analysis, which corroborated there was no significant difference in the detection rate of advanced neoplasia between the 2 arms of the study. Third, we cannot provide accurate information as to whether repeated FIT is able to detect sessile serrated polyps or traditional serrated polyps, because at the time the study was designed (2006), these polyps were classified as hyperplastic and considered to have no malignant potential. In fact, clinical guidelines did not recommend colonoscopy surveillance for individuals with these polyps until recently. Fourth, the current study was not powered to analyze the different risk groups separately. However, the higher-risk subgroup of first-degree relatives having at least 2 family members with CRC constituted a small minority (7.6%) of the whole study sample. To minimize the effect of this potential flaw, the analysis was adjusted for confounding variables, including index-case age and having siblings or more than one family member with CRC. In addition, both arms of the study were homogeneous regarding demographic data and family history.
Finally, the potential advantage of FIT screening to spare endoscopic resources in this moderate-risk population also warrants comment. First, the number of first-degree relatives that needed to undergo colonoscopy to detect one advanced neoplasia was 4 in the FIT group as compared with 18 in the colonoscopy group. Second, up to 71% of colonoscopies in the colonoscopy arm had a normal or nonsignificant result. Overall, repeated screening with FIT needed about 86% fewer colonoscopies to detect 100% CRC and 61% of advanced adenomas, which is consistent with the results of previous studies. Therefore, a potential benefit of FIT over primary colonoscopy in familial CRC screening is that it may save a substantial number of unnecessary colonoscopies, thus preventing harm and lowering costs. Given its equivalence in terms of advanced neoplasia detection, FIT screening should be considered when colonoscopy capacity is limited.
In conclusion, the results of this randomized trial demonstrate that in asymptomatic first-degree relatives of patients with CRC, screening with FIT is equivalent to one-time colonoscopy for the detection of advanced neoplasia. In addition, it also provides evidence for the benefit of repeated FIT screening in terms of colonoscopy resources. This strategy should be taken into account when colonoscopy capacity is limited and particularly in populations where FIT is better accepted than colonoscopy as a CRC screening strategy. Additional studies are needed to determine acceptance of this strategy and its efficacy in reducing the incidence and mortality associated with familial CRC.
Source...