Denise Napoli

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

Compared with oncologic surgery of T1 colorectal carcinomas, endoscopic resection had statistically similar outcomes, even in the presence of factors thought to confer increased risk for cancer recurrence, reported Dr. Alexander Meining and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

The findings suggest that endoscopic resection and surveillance may be a reasonable course, regardless of the histologic character of the tumor involved, considering the mortality rate of the alternative – oncologic colorectal resection – the authors wrote (Clin. Gastroenterol. Hepatol. 2011[doi:10.1016/j.cgh.2011.02.002]).

Dr. Meining, of the department of gastroenterology at the Technischen Universität München in Munich, and his colleagues studied 474 patients who underwent endoscopic resection for T1 colorectal cancers between 1974 and 2002 at Neuperlach Hospital in Munich.

Cancers originating in patients with known inflammatory bowel disease, hereditary cancer syndromes, metachronous advanced colorectal cancers, or known cancers of other origin were excluded.

All the patients had T1 cancers, which are defined by infiltration of the tumor into the submucosal layer. Surgery with oncologic resection was performed in 161 of the patients, and the remaining 313 had their cancers removed by endoscopic means alone, either via snare polypectomy, submucosal saline injection, or – in the case of larger lesions – piecemeal resection.

The decision to perform surgery was based on conventional high-risk characteristics of the tumors, including poorer tumor grading (G3 or G4) or infiltration of lymphatic vessels, as well as individual patients’ preferences and physical ability to withstand surgery.

At the time of resection, tumor grading, resection margins, and determination of venous or lymphatic vessel infiltration were noted, as were polyp characteristics. The depth of submucosal infiltration was not routinely noted, according to the authors.

All patients with complete follow-up data for at least 2 years were then reassessed between January 2003 and September 2004, and the researchers classified each patient as having either good or poor outcomes. Poor outcomes were defined as locoregional cancer relapse, distant metastasis, lymph node metastasis, or death related to advanced colorectal cancer.

Follow-up data were available for 390 patients, or 83% of the original cohort, including 141 surgical patients and 249 patients who were solely treated with the initial endoscopic resection. The mean age of these patients was 63.8 years, 54% were male, and the mean follow-up period was 87 months.

Overall, the authors found that 39 patients (10% of the total) had a poor outcome during either surgery or follow-up, including 17 nonsurgical patients who had endoscopic resection only.

Among these 17 nonsurgical patients, the authors found that lymphatic vessel invasion (stratified as either L0 or L1) carried the highest odds ratio for subsequent poor outcome on follow-up, at 7.8 (P less than .001).

Higher tumor grading (G3/G4 versus G1/G2) also carried a substantial 3.4 odds ratio for poor outcome on follow-up (P = .009). And incomplete resection margins (R0 versus R1/RX) carried an odds ratio of 2.6 (P = .027).

Other presumed risk factors, including tumor configuration (sessile versus pedunculated), localization (right vs. left sided), and venous vessel infiltration (V0 versus V1) did not carry a significant odds ratio for poor outcome during follow-up – a surprising finding, wrote Dr. Meining and his colleagues.

However, while the negative predictive values were high for the risk factors with significant odds ratios (94.6% for lymphatic vessel infiltration, 94.2% for tumor stage, and 96.5% for incomplete endoscopic resection), the corresponding positive predictive values were low: 44.4% for lymphatic vessel infiltration, 42.9% for tumor stage, and 19.6% for incomplete endoscopic resection.

According to Dr. Meining, while the study confirms that lymphatic vessel infiltration is the strongest predictor of metastasis and other poor outcomes, the poor positive predictive values nevertheless confirm not only that "precise assessment of risk remains difficult," but that even in cases of infiltration, surgery may not be warranted."

And regarding the finding that incomplete endoscopic resection also carried an elevated risk, Dr. Meining noted that "endoscopic retreatment of areas suspicious for residual neoplasia should be ... performed if resection was incomplete to the lateral margins."

Nevertheless, Dr. Meining conceded that while "the overwhelming majority of patients can be regarded cured from their cancers in the absence of histological risk factors," surgery may sometimes be the best course for an individual patient, despite the risks.

The researchers wrote that they received no outside funding for this study, and stated that they had no individual interests to disclose.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

A significant, nonlinear relationship exists between fruit and vegetable intake and colorectal cancer risk, based on a meta-analysis of prospective studies, reported Dagfinn Aune and his colleagues in the July issue of Gastroenterology.

Moreover, the greatest risk reduction occurs among patients in the lowest stratum of baseline consumption who subsequently raise their intake – a prime target for preventive public health initiatives, the authors said.

Photo credit: James Reinaker/Elsevier Global Medical News

According to Mr. Aune, a researcher in the department of epidemiology and biostatistics at the Imperial College London, and his colleagues, the possible link between fruit and vegetable consumption and colorectal cancer risk has been the topic of numerous studies, although the findings have rarely been conclusive.

Indeed, two major meta-analyses from the World Cancer Research Fund/American Institute for Cancer Research, conducted 10 years apart, came to different conclusions.

The first, in 1997, found "convincing evidence" that vegetable intake, but not fruit intake, was associated with reduced rates of colon cancer. But the second analysis, in 2007, concluded there was "limited suggestive evidence for risk reduction by fruits and nonstarchy vegetables ... thus a downgrading of the judgment of the evidence compared with the previous report," wrote Mr. Aune and his associates.

In the present meta-analysis, Mr. Aune and his coauthors searched the PubMed database up until May 2010 for cohort studies of fruit and vegetable intake and colorectal cancer risk. To be included, studies had to have a prospective cohort, case-cohort, or nested case-control design.

First, the authors assessed the associations between colorectal cancer risk and the highest and lowest intakes of both fruits and vegetables. To accomplish this, they analyzed 11 of the included studies, comprising 11,853 cases of cancer among 1,523,860 participants.

They found that the summary relative risk (RR) for the highest intake, versus the lowest, was 0.92 (95% confidence interval, 0.86-0.99), a "small, but statistically significant reduction," the authors wrote (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.04.013]).

The finding was similar when they focused on fruit alone, an analysis that included 14 studies, 14,876 cases, and 1,558,147 participants. For highest versus lowest fruit intake, the relative risk for cancer was 0.90 (95% CI, 0.83-0.98). And for vegetables alone, the summary RR was 0.91 (95% CI, 0.86-0.96) across 16 studies with 16,057 cases and 1,694,236 participants. Next, the authors looked for a dose-response relationship in each of the three intake categories. For fruits and vegetables combined, the summary RR per 100 g/day was 0.99 (95% CI, 0.98-1.00).

For fruit alone, the summary RR per 100 g/day was 0.98 (95% CI, 0.94-1.01), but the authors also found evidence of a nonlinear association between fruit intake and colorectal cancer risk (with P for nonlinearity less than .001), "with most of the risk reduction observed when increasing intake up to about 100 g/day," reported Mr. Aune and his colleagues. "Higher intakes were associated with a further, but more modest, decrease in risk."

The finding was similar for vegetables alone, with a summary RR per 100 g/day of 0.98 (95% CI, 0.97-0.99), and a significant, nonlinear association between vegetable consumption and cancer risk, "with the greatest reduction for an intake between 100 and 200 g/day, but little evidence of a further reduction with higher intakes," wrote the authors.

"In contrast to previous meta-analyses, which have assumed a linear association between fruit and vegetables and colorectal cancer risk, we found evidence of a nonlinear inverse association, with the greatest risk reduction when increasing intake from low levels," the investigators wrote, adding that "the lack of significance of the results in the previous meta-analyses and for fruit in the present linear dose-response analysis might be because the linear model does not fit well with the data."

The meta-analysis does have potential limitations, including the fact that fruit and vegetable intake may be associated with a healthier lifestyle in general; however, most of the studies controlled for potential confounders related to lifestyle. Also, the large sample size and low heterogeneity (for example, the lack of difference between men and women) constitute important strengths, the investigators said. There was also no sign of publication bias affecting the analyses.

The researchers disclosed that the study was funded by the World Cancer Research Fund. They stated that they had no relevant financial disclosures.

Enteral nutrition is superior to total parenteral nutrition for the treatment of grade B postoperative pancreatic fistulas, reported Dr. Stanislaw Klek and his colleagues in the July issue of Gastroenterology.

Indeed, not only does enteral nutrition facilitate fistula closure, but closure occurs more quickly and the enteral approach is less costly, wrote the researchers (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.03.040]).

Dr. Klek of the Jagiellonian University Medical College in Krakow, Poland, and his associates studied 78 adult patients with postoperative pancreatic fistula treated at a single academic center in Poland. All patients had a grade B fistula, according to the criteria of the International Study Group on Pancreatic Fistula. "Postoperative pancreatic fistula is the most common and potentially life-threatening complication after pancreatic surgery," they noted.

Patients were excluded if they had any cardiocirculatory, pulmonary, renal, or liver failure; if their fistula required surgical intervention; or if nutritional status was so depleted that it required both parenteral and enteral feeding.

Patients were randomized into two demographically similar groups; for 30 days, one group received enteral feeding while the other received total parenteral nutrition (TPN). All patients were treated conservatively without surgery, somatostatin analogues, or proton pump inhibitors.

Enteral feedings were started 2-4 hours after placement of a nasointestinal tube, located 20 cm below the Treitz’s ligament or, in the case of pancreaticoduodenectomy, 20 cm below the last jejeunal anastomosis. The formula, called Peptisorb (manufactured by Nutricia), contains 1 kcal/mL, is low in fat, and is peptide based, with an initial flow rate of 10 mL/hr, progressing up to 125 mL/hr.

Parenteral nutrition infusions also commenced 2-4 hours post insertion of a central venous catheter, and were prepared by the hospital pharmacy.

The mean age in both groups was 57 years, and almost half (45%) of each group was female.

Fistula closure was defined as an output of less than 10 mL during a 48-hour period, with no recurrence in the subsequent 30 days and no evidence of fluid collection on ultrasound or computed tomography scan.

At 30 days, the rate of fistula closure was 60% in the enteral group (24 of 40 patients) and 37% in the parenteral group (14 of 38 patients).

That amounted to an odds ratio of 2.57 for the probability to be a responder with enteral feedings, compared with TPN (95% confidence interval, 1.03-6.41; P = .043).

The authors analyzed the data to identify other variables that might predict fistula closure, including demographic information, diagnosis (pancreatic cancer, ampullary cancer, or other); concomitant disease; body mass index; baseline lymphocyte and albumin levels; weight loss; type of surgery; and fistula output (less than 200 mL/day versus 200 mL or greater).

In this stepwise logistic regression analysis, "only two factors were significantly associated with fistula closure," wrote the authors. They were enteral nutrition (OR, 6.1; 95% CI, 1.2-41.6, P = .043) and initial fistula output of 200 mL/day or less (OR, 12.7; 95% CI, 9.1-47.2, P less than 0.001).

The researchers also looked at median time to closure. In the enteral nutrition cohort, it was 27 days (95% CI, 21-33 days). No median time was reached in the parenteral group (P = .047).

Cost was significantly lower for the enteral patients, with the median 30-day price tag for nutritional intervention being $532 for the enteral group and $968 for the parenteral cohort (P less than .001).

Fistula recurrence was found in two TPN patients, at 8 and 31 days after TPN withdrawal. No recurrence was seen in the enteral group. Both treatments were without major complications or safety issues, except for two cases of bacteremia in the TPN cohort, plus one case of diarrhea in the enteral group.

Although evidence-based recommendations for nutritional support in the setting of a postoperative pancreatic fistula are currently limited, "fasting with nutritional support, along with adequate drainage and skin protection, are currently the basic aspects of treatment," wrote the authors.

However, prior to this study, the choice between TPN and enteral feedings has been "essentially arbitrary, because the effects of both diets on closure rates of postoperative pancreatic fistula have not been compared in a randomized clinical trial."

Dr. Klek and his coauthors postulated that the benefits of enteral feeding are likely related to its stimulation of the release of "specific gut peptides, forming a negative feedback control system, and thus inhibiting pancreatic secretion." They added, however, that the mechanisms of this negative feedback loop are poorly defined.

Dr. Klek reported receiving educational and research grants from Nutricia, maker of the enteral formula used in this study, as well as Fresenius Kabi and B Braun, which make enteral and parenteral products, respectively. The study was also funded by Nutricia.

Enteral nutrition is superior to total parenteral nutrition for the treatment of grade B postoperative pancreatic fistulas, reported Dr. Stanislaw Klek and his colleagues in the July issue of Gastroenterology.

Indeed, not only does enteral nutrition facilitate fistula closure, but closure occurs more quickly and the enteral approach is less costly, wrote the researchers (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.03.040]).

Dr. Klek of the Jagiellonian University Medical College in Krakow, Poland, and his associates studied 78 adult patients with postoperative pancreatic fistula treated at a single academic center in Poland. All patients had a grade B fistula, according to the criteria of the International Study Group on Pancreatic Fistula. "Postoperative pancreatic fistula is the most common and potentially life-threatening complication after pancreatic surgery," they noted.

Patients were excluded if they had any cardiocirculatory, pulmonary, renal, or liver failure; if their fistula required surgical intervention; or if nutritional status was so depleted that it required both parenteral and enteral feeding.

Patients were randomized into two demographically similar groups; for 30 days, one group received enteral feeding while the other received total parenteral nutrition (TPN). All patients were treated conservatively without surgery, somatostatin analogues, or proton pump inhibitors.

Enteral feedings were started 2-4 hours after placement of a nasointestinal tube, located 20 cm below the Treitz’s ligament or, in the case of pancreaticoduodenectomy, 20 cm below the last jejeunal anastomosis. The formula, called Peptisorb (manufactured by Nutricia), contains 1 kcal/mL, is low in fat, and is peptide based, with an initial flow rate of 10 mL/hr, progressing up to 125 mL/hr.

Parenteral nutrition infusions also commenced 2-4 hours post insertion of a central venous catheter, and were prepared by the hospital pharmacy.

The mean age in both groups was 57 years, and almost half (45%) of each group was female.

Fistula closure was defined as an output of less than 10 mL during a 48-hour period, with no recurrence in the subsequent 30 days and no evidence of fluid collection on ultrasound or computed tomography scan.

At 30 days, the rate of fistula closure was 60% in the enteral group (24 of 40 patients) and 37% in the parenteral group (14 of 38 patients).

That amounted to an odds ratio of 2.57 for the probability to be a responder with enteral feedings, compared with TPN (95% confidence interval, 1.03-6.41; P = .043).

The authors analyzed the data to identify other variables that might predict fistula closure, including demographic information, diagnosis (pancreatic cancer, ampullary cancer, or other); concomitant disease; body mass index; baseline lymphocyte and albumin levels; weight loss; type of surgery; and fistula output (less than 200 mL/day versus 200 mL or greater).

In this stepwise logistic regression analysis, "only two factors were significantly associated with fistula closure," wrote the authors. They were enteral nutrition (OR, 6.1; 95% CI, 1.2-41.6, P = .043) and initial fistula output of 200 mL/day or less (OR, 12.7; 95% CI, 9.1-47.2, P less than 0.001).

The researchers also looked at median time to closure. In the enteral nutrition cohort, it was 27 days (95% CI, 21-33 days). No median time was reached in the parenteral group (P = .047).

Cost was significantly lower for the enteral patients, with the median 30-day price tag for nutritional intervention being $532 for the enteral group and $968 for the parenteral cohort (P less than .001).

Fistula recurrence was found in two TPN patients, at 8 and 31 days after TPN withdrawal. No recurrence was seen in the enteral group. Both treatments were without major complications or safety issues, except for two cases of bacteremia in the TPN cohort, plus one case of diarrhea in the enteral group.

Although evidence-based recommendations for nutritional support in the setting of a postoperative pancreatic fistula are currently limited, "fasting with nutritional support, along with adequate drainage and skin protection, are currently the basic aspects of treatment," wrote the authors.

However, prior to this study, the choice between TPN and enteral feedings has been "essentially arbitrary, because the effects of both diets on closure rates of postoperative pancreatic fistula have not been compared in a randomized clinical trial."

Dr. Klek and his coauthors postulated that the benefits of enteral feeding are likely related to its stimulation of the release of "specific gut peptides, forming a negative feedback control system, and thus inhibiting pancreatic secretion." They added, however, that the mechanisms of this negative feedback loop are poorly defined.

Dr. Klek reported receiving educational and research grants from Nutricia, maker of the enteral formula used in this study, as well as Fresenius Kabi and B Braun, which make enteral and parenteral products, respectively. The study was also funded by Nutricia.

Enteral nutrition is superior to total parenteral nutrition for the treatment of grade B postoperative pancreatic fistulas, reported Dr. Stanislaw Klek and his colleagues in the July issue of Gastroenterology.

Indeed, not only does enteral nutrition facilitate fistula closure, but closure occurs more quickly and the enteral approach is less costly, wrote the researchers (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.03.040]).

Dr. Klek of the Jagiellonian University Medical College in Krakow, Poland, and his associates studied 78 adult patients with postoperative pancreatic fistula treated at a single academic center in Poland. All patients had a grade B fistula, according to the criteria of the International Study Group on Pancreatic Fistula. "Postoperative pancreatic fistula is the most common and potentially life-threatening complication after pancreatic surgery," they noted.

Patients were excluded if they had any cardiocirculatory, pulmonary, renal, or liver failure; if their fistula required surgical intervention; or if nutritional status was so depleted that it required both parenteral and enteral feeding.

Patients were randomized into two demographically similar groups; for 30 days, one group received enteral feeding while the other received total parenteral nutrition (TPN). All patients were treated conservatively without surgery, somatostatin analogues, or proton pump inhibitors.

Enteral feedings were started 2-4 hours after placement of a nasointestinal tube, located 20 cm below the Treitz’s ligament or, in the case of pancreaticoduodenectomy, 20 cm below the last jejeunal anastomosis. The formula, called Peptisorb (manufactured by Nutricia), contains 1 kcal/mL, is low in fat, and is peptide based, with an initial flow rate of 10 mL/hr, progressing up to 125 mL/hr.

Parenteral nutrition infusions also commenced 2-4 hours post insertion of a central venous catheter, and were prepared by the hospital pharmacy.

The mean age in both groups was 57 years, and almost half (45%) of each group was female.

Fistula closure was defined as an output of less than 10 mL during a 48-hour period, with no recurrence in the subsequent 30 days and no evidence of fluid collection on ultrasound or computed tomography scan.

At 30 days, the rate of fistula closure was 60% in the enteral group (24 of 40 patients) and 37% in the parenteral group (14 of 38 patients).

That amounted to an odds ratio of 2.57 for the probability to be a responder with enteral feedings, compared with TPN (95% confidence interval, 1.03-6.41; P = .043).

The authors analyzed the data to identify other variables that might predict fistula closure, including demographic information, diagnosis (pancreatic cancer, ampullary cancer, or other); concomitant disease; body mass index; baseline lymphocyte and albumin levels; weight loss; type of surgery; and fistula output (less than 200 mL/day versus 200 mL or greater).

In this stepwise logistic regression analysis, "only two factors were significantly associated with fistula closure," wrote the authors. They were enteral nutrition (OR, 6.1; 95% CI, 1.2-41.6, P = .043) and initial fistula output of 200 mL/day or less (OR, 12.7; 95% CI, 9.1-47.2, P less than 0.001).

The researchers also looked at median time to closure. In the enteral nutrition cohort, it was 27 days (95% CI, 21-33 days). No median time was reached in the parenteral group (P = .047).

Cost was significantly lower for the enteral patients, with the median 30-day price tag for nutritional intervention being $532 for the enteral group and $968 for the parenteral cohort (P less than .001).

Fistula recurrence was found in two TPN patients, at 8 and 31 days after TPN withdrawal. No recurrence was seen in the enteral group. Both treatments were without major complications or safety issues, except for two cases of bacteremia in the TPN cohort, plus one case of diarrhea in the enteral group.

Although evidence-based recommendations for nutritional support in the setting of a postoperative pancreatic fistula are currently limited, "fasting with nutritional support, along with adequate drainage and skin protection, are currently the basic aspects of treatment," wrote the authors.

However, prior to this study, the choice between TPN and enteral feedings has been "essentially arbitrary, because the effects of both diets on closure rates of postoperative pancreatic fistula have not been compared in a randomized clinical trial."

Dr. Klek and his coauthors postulated that the benefits of enteral feeding are likely related to its stimulation of the release of "specific gut peptides, forming a negative feedback control system, and thus inhibiting pancreatic secretion." They added, however, that the mechanisms of this negative feedback loop are poorly defined.

Dr. Klek reported receiving educational and research grants from Nutricia, maker of the enteral formula used in this study, as well as Fresenius Kabi and B Braun, which make enteral and parenteral products, respectively. The study was also funded by Nutricia.

Enteral nutrition is superior to total parenteral nutrition for the treatment of grade B postoperative pancreatic fistulas, reported Dr. Stanislaw Klek and his colleagues in the July issue of Gastroenterology.

Indeed, not only does enteral nutrition facilitate fistula closure, but closure occurs more quickly and the enteral approach is less costly, wrote the researchers (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.03.040]).

Dr. Klek of the Jagiellonian University Medical College in Krakow, Poland, and his associates studied 78 adult patients with postoperative pancreatic fistula treated at a single academic center in Poland. All patients had a grade B fistula, according to the criteria of the International Study Group on Pancreatic Fistula. "Postoperative pancreatic fistula is the most common and potentially life-threatening complication after pancreatic surgery," they noted.

Patients were excluded if they had any cardiocirculatory, pulmonary, renal, or liver failure; if their fistula required surgical intervention; or if nutritional status was so depleted that it required both parenteral and enteral feeding.

Patients were randomized into two demographically similar groups; for 30 days, one group received enteral feeding while the other received total parenteral nutrition (TPN). All patients were treated conservatively without surgery, somatostatin analogues, or proton pump inhibitors.

Enteral feedings were started 2-4 hours after placement of a nasointestinal tube, located 20 cm below the Treitz’s ligament or, in the case of pancreaticoduodenectomy, 20 cm below the last jejeunal anastomosis. The formula, called Peptisorb (manufactured by Nutricia), contains 1 kcal/mL, is low in fat, and is peptide based, with an initial flow rate of 10 mL/hr, progressing up to 125 mL/hr.

Parenteral nutrition infusions also commenced 2-4 hours post insertion of a central venous catheter, and were prepared by the hospital pharmacy.

The mean age in both groups was 57 years, and almost half (45%) of each group was female.

Fistula closure was defined as an output of less than 10 mL during a 48-hour period, with no recurrence in the subsequent 30 days and no evidence of fluid collection on ultrasound or computed tomography scan.

At 30 days, the rate of fistula closure was 60% in the enteral group (24 of 40 patients) and 37% in the parenteral group (14 of 38 patients).

That amounted to an odds ratio of 2.57 for the probability to be a responder with enteral feedings, compared with TPN (95% confidence interval, 1.03-6.41; P = .043).

The authors analyzed the data to identify other variables that might predict fistula closure, including demographic information, diagnosis (pancreatic cancer, ampullary cancer, or other); concomitant disease; body mass index; baseline lymphocyte and albumin levels; weight loss; type of surgery; and fistula output (less than 200 mL/day versus 200 mL or greater).

In this stepwise logistic regression analysis, "only two factors were significantly associated with fistula closure," wrote the authors. They were enteral nutrition (OR, 6.1; 95% CI, 1.2-41.6, P = .043) and initial fistula output of 200 mL/day or less (OR, 12.7; 95% CI, 9.1-47.2, P less than 0.001).

The researchers also looked at median time to closure. In the enteral nutrition cohort, it was 27 days (95% CI, 21-33 days). No median time was reached in the parenteral group (P = .047).

Cost was significantly lower for the enteral patients, with the median 30-day price tag for nutritional intervention being $532 for the enteral group and $968 for the parenteral cohort (P less than .001).

Fistula recurrence was found in two TPN patients, at 8 and 31 days after TPN withdrawal. No recurrence was seen in the enteral group. Both treatments were without major complications or safety issues, except for two cases of bacteremia in the TPN cohort, plus one case of diarrhea in the enteral group.

Although evidence-based recommendations for nutritional support in the setting of a postoperative pancreatic fistula are currently limited, "fasting with nutritional support, along with adequate drainage and skin protection, are currently the basic aspects of treatment," wrote the authors.

However, prior to this study, the choice between TPN and enteral feedings has been "essentially arbitrary, because the effects of both diets on closure rates of postoperative pancreatic fistula have not been compared in a randomized clinical trial."

Dr. Klek and his coauthors postulated that the benefits of enteral feeding are likely related to its stimulation of the release of "specific gut peptides, forming a negative feedback control system, and thus inhibiting pancreatic secretion." They added, however, that the mechanisms of this negative feedback loop are poorly defined.

Dr. Klek reported receiving educational and research grants from Nutricia, maker of the enteral formula used in this study, as well as Fresenius Kabi and B Braun, which make enteral and parenteral products, respectively. The study was also funded by Nutricia.

Enteral nutrition is superior to total parenteral nutrition for the treatment of grade B postoperative pancreatic fistulas, reported Dr. Stanislaw Klek and his colleagues in the July issue of Gastroenterology.

Indeed, not only does enteral nutrition facilitate fistula closure, but closure occurs more quickly and the enteral approach is less costly, wrote the researchers (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.03.040]).

Dr. Klek of the Jagiellonian University Medical College in Krakow, Poland, and his associates studied 78 adult patients with postoperative pancreatic fistula treated at a single academic center in Poland. All patients had a grade B fistula, according to the criteria of the International Study Group on Pancreatic Fistula. "Postoperative pancreatic fistula is the most common and potentially life-threatening complication after pancreatic surgery," they noted.

Patients were excluded if they had any cardiocirculatory, pulmonary, renal, or liver failure; if their fistula required surgical intervention; or if nutritional status was so depleted that it required both parenteral and enteral feeding.

Patients were randomized into two demographically similar groups; for 30 days, one group received enteral feeding while the other received total parenteral nutrition (TPN). All patients were treated conservatively without surgery, somatostatin analogues, or proton pump inhibitors.

Enteral feedings were started 2-4 hours after placement of a nasointestinal tube, located 20 cm below the Treitz’s ligament or, in the case of pancreaticoduodenectomy, 20 cm below the last jejeunal anastomosis. The formula, called Peptisorb (manufactured by Nutricia), contains 1 kcal/mL, is low in fat, and is peptide based, with an initial flow rate of 10 mL/hr, progressing up to 125 mL/hr.

Parenteral nutrition infusions also commenced 2-4 hours post insertion of a central venous catheter, and were prepared by the hospital pharmacy.

The mean age in both groups was 57 years, and almost half (45%) of each group was female.

Fistula closure was defined as an output of less than 10 mL during a 48-hour period, with no recurrence in the subsequent 30 days and no evidence of fluid collection on ultrasound or computed tomography scan.

At 30 days, the rate of fistula closure was 60% in the enteral group (24 of 40 patients) and 37% in the parenteral group (14 of 38 patients).

That amounted to an odds ratio of 2.57 for the probability to be a responder with enteral feedings, compared with TPN (95% confidence interval, 1.03-6.41; P = .043).

The authors analyzed the data to identify other variables that might predict fistula closure, including demographic information, diagnosis (pancreatic cancer, ampullary cancer, or other); concomitant disease; body mass index; baseline lymphocyte and albumin levels; weight loss; type of surgery; and fistula output (less than 200 mL/day versus 200 mL or greater).

In this stepwise logistic regression analysis, "only two factors were significantly associated with fistula closure," wrote the authors. They were enteral nutrition (OR, 6.1; 95% CI, 1.2-41.6, P = .043) and initial fistula output of 200 mL/day or less (OR, 12.7; 95% CI, 9.1-47.2, P less than 0.001).

The researchers also looked at median time to closure. In the enteral nutrition cohort, it was 27 days (95% CI, 21-33 days). No median time was reached in the parenteral group (P = .047).

Cost was significantly lower for the enteral patients, with the median 30-day price tag for nutritional intervention being $532 for the enteral group and $968 for the parenteral cohort (P less than .001).

Fistula recurrence was found in two TPN patients, at 8 and 31 days after TPN withdrawal. No recurrence was seen in the enteral group. Both treatments were without major complications or safety issues, except for two cases of bacteremia in the TPN cohort, plus one case of diarrhea in the enteral group.

Although evidence-based recommendations for nutritional support in the setting of a postoperative pancreatic fistula are currently limited, "fasting with nutritional support, along with adequate drainage and skin protection, are currently the basic aspects of treatment," wrote the authors.

However, prior to this study, the choice between TPN and enteral feedings has been "essentially arbitrary, because the effects of both diets on closure rates of postoperative pancreatic fistula have not been compared in a randomized clinical trial."

Dr. Klek and his coauthors postulated that the benefits of enteral feeding are likely related to its stimulation of the release of "specific gut peptides, forming a negative feedback control system, and thus inhibiting pancreatic secretion." They added, however, that the mechanisms of this negative feedback loop are poorly defined.

Dr. Klek reported receiving educational and research grants from Nutricia, maker of the enteral formula used in this study, as well as Fresenius Kabi and B Braun, which make enteral and parenteral products, respectively. The study was also funded by Nutricia.

Enteral nutrition is superior to total parenteral nutrition for the treatment of grade B postoperative pancreatic fistulas, reported Dr. Stanislaw Klek and his colleagues in the July issue of Gastroenterology.

Indeed, not only does enteral nutrition facilitate fistula closure, but closure occurs more quickly and the enteral approach is less costly, wrote the researchers (Gastroenterology 2011 July [doi:10.1053/j.gastro.2011.03.040]).

Dr. Klek of the Jagiellonian University Medical College in Krakow, Poland, and his associates studied 78 adult patients with postoperative pancreatic fistula treated at a single academic center in Poland. All patients had a grade B fistula, according to the criteria of the International Study Group on Pancreatic Fistula. "Postoperative pancreatic fistula is the most common and potentially life-threatening complication after pancreatic surgery," they noted.

Patients were excluded if they had any cardiocirculatory, pulmonary, renal, or liver failure; if their fistula required surgical intervention; or if nutritional status was so depleted that it required both parenteral and enteral feeding.

Patients were randomized into two demographically similar groups; for 30 days, one group received enteral feeding while the other received total parenteral nutrition (TPN). All patients were treated conservatively without surgery, somatostatin analogues, or proton pump inhibitors.

Enteral feedings were started 2-4 hours after placement of a nasointestinal tube, located 20 cm below the Treitz’s ligament or, in the case of pancreaticoduodenectomy, 20 cm below the last jejeunal anastomosis. The formula, called Peptisorb (manufactured by Nutricia), contains 1 kcal/mL, is low in fat, and is peptide based, with an initial flow rate of 10 mL/hr, progressing up to 125 mL/hr.

Parenteral nutrition infusions also commenced 2-4 hours post insertion of a central venous catheter, and were prepared by the hospital pharmacy.

The mean age in both groups was 57 years, and almost half (45%) of each group was female.

Fistula closure was defined as an output of less than 10 mL during a 48-hour period, with no recurrence in the subsequent 30 days and no evidence of fluid collection on ultrasound or computed tomography scan.

At 30 days, the rate of fistula closure was 60% in the enteral group (24 of 40 patients) and 37% in the parenteral group (14 of 38 patients).

That amounted to an odds ratio of 2.57 for the probability to be a responder with enteral feedings, compared with TPN (95% confidence interval, 1.03-6.41; P = .043).

The authors analyzed the data to identify other variables that might predict fistula closure, including demographic information, diagnosis (pancreatic cancer, ampullary cancer, or other); concomitant disease; body mass index; baseline lymphocyte and albumin levels; weight loss; type of surgery; and fistula output (less than 200 mL/day versus 200 mL or greater).

In this stepwise logistic regression analysis, "only two factors were significantly associated with fistula closure," wrote the authors. They were enteral nutrition (OR, 6.1; 95% CI, 1.2-41.6, P = .043) and initial fistula output of 200 mL/day or less (OR, 12.7; 95% CI, 9.1-47.2, P less than 0.001).

The researchers also looked at median time to closure. In the enteral nutrition cohort, it was 27 days (95% CI, 21-33 days). No median time was reached in the parenteral group (P = .047).

Cost was significantly lower for the enteral patients, with the median 30-day price tag for nutritional intervention being $532 for the enteral group and $968 for the parenteral cohort (P less than .001).

Fistula recurrence was found in two TPN patients, at 8 and 31 days after TPN withdrawal. No recurrence was seen in the enteral group. Both treatments were without major complications or safety issues, except for two cases of bacteremia in the TPN cohort, plus one case of diarrhea in the enteral group.

Although evidence-based recommendations for nutritional support in the setting of a postoperative pancreatic fistula are currently limited, "fasting with nutritional support, along with adequate drainage and skin protection, are currently the basic aspects of treatment," wrote the authors.

However, prior to this study, the choice between TPN and enteral feedings has been "essentially arbitrary, because the effects of both diets on closure rates of postoperative pancreatic fistula have not been compared in a randomized clinical trial."

Dr. Klek and his coauthors postulated that the benefits of enteral feeding are likely related to its stimulation of the release of "specific gut peptides, forming a negative feedback control system, and thus inhibiting pancreatic secretion." They added, however, that the mechanisms of this negative feedback loop are poorly defined.

Dr. Klek reported receiving educational and research grants from Nutricia, maker of the enteral formula used in this study, as well as Fresenius Kabi and B Braun, which make enteral and parenteral products, respectively. The study was also funded by Nutricia.

A large percentage of patients with cirrhosis and variceal bleeding will mount an inadequate adrenal response with respect to their condition, and may be classified as having critical illness–related corticosteroid insufficiency, reported Dr. Christos Triantos and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

Dr. Triantos, of the department of gastroenterology at the University Hospital of Patras, Greece, and his colleagues evaluated 20 patients with cirrhosis and bleeding varices and compared them with 74 controls: 14 healthy volunteers and 60 patients who had stable cirrhosis and no bleeding (Clin. Gastroenterol. Hepatol. 2011 [doi:10.1016/j.cgh.2011.03.033]).

The participants were divided into five groups. Group 1 included 10 with variceal bleeding, who underwent a conventional-dose short synacthen test (SST), administered by injecting a 250-mcg dose of the synthetic adrenocorticotropin hormone synacthen. Serum cortisol concentration was measured at baseline, 30 minutes, and 60 minutes.

Group 2 comprised the remaining 10 patients with variceal bleeding, who underwent the low-dose short synacthen test (LDSST), in which a 1-mcg dose of synacthen was administered. Again, serum cortisol levels were measured at baseline, 30 minutes, and 60 minutes. Two hours later, these patients underwent the SST as well.

Group 3 consisted of the 14 healthy volunteers, who also underwent both tests. Group 4 comprised 50 stable (nonbleeding) cirrhosis patients who underwent the LDSST, and the fifth group comprised the remaining 10 stable cirrhotics, who underwent the SST.

Among nonstressed (nonbleeding) patients who underwent the LDSST, adrenal insufficiency was diagnosed when a peak cortisol level was 18 mcg/dL or less. In stressed (bleeding) patients, the definition of adrenal insufficiency was a peak cortisol level of less than 25 mcg/dL, or a maximum change of less than 9 mcg/dL from baseline.

On the SST, adrenal insufficiency was defined as a peak cortisol level of 18 mcg/dL or less in nonstressed patients. In patients with bleeding, it was a maximum difference between peak and basal levels of less than 9 mcg/dL, or a random total cortisol level of less than 10 mcg/dL.

The authors found that patients with variceal bleeding had higher median basal cortisol levels (15.4 mcg/dL) compared with stable cirrhotic patients (8.7 mcg/dL, P = .001) or healthy volunteers (10.1 mcg/dL, P = .01).

Looking at peak cortisol, patients with bleeding varices also had higher median levels than did nonbleeding cirrhotic patients, both on the SST (32.7 mcg/dL vs. 21 mcg/dL, P = .0001) and on the LDSST (28 mcg/dL vs. 18.5 mcg/dL), although the difference on the latter test did not reach statistical significance.

However, the change between peak and basal values was not significantly different between patients with and without bleeding, either on the SST (15.4 mcg/dL vs. 12.7 mcg/dL, respectively) or the LDSST (9.3 mcg/dL vs. 8.1 mcg/dL).

Overall, "among patients analyzed by SST, 6/20 with variceal bleeding had adrenal insufficiency (AI) (30%), compared with 3/10 controls with [stable] cirrhosis (30%)," wrote the authors.

And on the more sensitive LDSST, "6/10 patients with variceal bleeding had AI (60%) vs. 24/50 controls with cirrhosis (48%)," they added.

The lack of a significant difference between patients with and without bleeding, with regard to the change between peak and basal levels, shows that many of those with bleeding meet the criteria for "critical illness–related corticosteroid insufficiency" with respect to their severity of illness, wrote the authors. None of the healthy volunteers was found to have adrenal insufficiency.

According to the researchers, a large proportion of those with variceal bleeding showed a dysfunctional response on both the SST assay and the more sensitive LDSST test, thought to be a better mimic of physiologic adrenocorticotropic hormone secretion.

"Although better methods of assessing adrenal function in cirrhosis are needed, the clinical significance of the abnormal ACTH tests and [critical illness–related corticosteroid insufficiency] may only be resolved by a clinical trial of corticosteroid supplementation in variceal bleeding as is used in sepsis."

The authors declared no conflicts of interest related to this study.

A large percentage of patients with cirrhosis and variceal bleeding will mount an inadequate adrenal response with respect to their condition, and may be classified as having critical illness–related corticosteroid insufficiency, reported Dr. Christos Triantos and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

Dr. Triantos, of the department of gastroenterology at the University Hospital of Patras, Greece, and his colleagues evaluated 20 patients with cirrhosis and bleeding varices and compared them with 74 controls: 14 healthy volunteers and 60 patients who had stable cirrhosis and no bleeding (Clin. Gastroenterol. Hepatol. 2011 [doi:10.1016/j.cgh.2011.03.033]).

The participants were divided into five groups. Group 1 included 10 with variceal bleeding, who underwent a conventional-dose short synacthen test (SST), administered by injecting a 250-mcg dose of the synthetic adrenocorticotropin hormone synacthen. Serum cortisol concentration was measured at baseline, 30 minutes, and 60 minutes.

Group 2 comprised the remaining 10 patients with variceal bleeding, who underwent the low-dose short synacthen test (LDSST), in which a 1-mcg dose of synacthen was administered. Again, serum cortisol levels were measured at baseline, 30 minutes, and 60 minutes. Two hours later, these patients underwent the SST as well.

Group 3 consisted of the 14 healthy volunteers, who also underwent both tests. Group 4 comprised 50 stable (nonbleeding) cirrhosis patients who underwent the LDSST, and the fifth group comprised the remaining 10 stable cirrhotics, who underwent the SST.

Among nonstressed (nonbleeding) patients who underwent the LDSST, adrenal insufficiency was diagnosed when a peak cortisol level was 18 mcg/dL or less. In stressed (bleeding) patients, the definition of adrenal insufficiency was a peak cortisol level of less than 25 mcg/dL, or a maximum change of less than 9 mcg/dL from baseline.

On the SST, adrenal insufficiency was defined as a peak cortisol level of 18 mcg/dL or less in nonstressed patients. In patients with bleeding, it was a maximum difference between peak and basal levels of less than 9 mcg/dL, or a random total cortisol level of less than 10 mcg/dL.

The authors found that patients with variceal bleeding had higher median basal cortisol levels (15.4 mcg/dL) compared with stable cirrhotic patients (8.7 mcg/dL, P = .001) or healthy volunteers (10.1 mcg/dL, P = .01).

Looking at peak cortisol, patients with bleeding varices also had higher median levels than did nonbleeding cirrhotic patients, both on the SST (32.7 mcg/dL vs. 21 mcg/dL, P = .0001) and on the LDSST (28 mcg/dL vs. 18.5 mcg/dL), although the difference on the latter test did not reach statistical significance.

However, the change between peak and basal values was not significantly different between patients with and without bleeding, either on the SST (15.4 mcg/dL vs. 12.7 mcg/dL, respectively) or the LDSST (9.3 mcg/dL vs. 8.1 mcg/dL).

Overall, "among patients analyzed by SST, 6/20 with variceal bleeding had adrenal insufficiency (AI) (30%), compared with 3/10 controls with [stable] cirrhosis (30%)," wrote the authors.

And on the more sensitive LDSST, "6/10 patients with variceal bleeding had AI (60%) vs. 24/50 controls with cirrhosis (48%)," they added.

The lack of a significant difference between patients with and without bleeding, with regard to the change between peak and basal levels, shows that many of those with bleeding meet the criteria for "critical illness–related corticosteroid insufficiency" with respect to their severity of illness, wrote the authors. None of the healthy volunteers was found to have adrenal insufficiency.

According to the researchers, a large proportion of those with variceal bleeding showed a dysfunctional response on both the SST assay and the more sensitive LDSST test, thought to be a better mimic of physiologic adrenocorticotropic hormone secretion.

"Although better methods of assessing adrenal function in cirrhosis are needed, the clinical significance of the abnormal ACTH tests and [critical illness–related corticosteroid insufficiency] may only be resolved by a clinical trial of corticosteroid supplementation in variceal bleeding as is used in sepsis."

The authors declared no conflicts of interest related to this study.

A large percentage of patients with cirrhosis and variceal bleeding will mount an inadequate adrenal response with respect to their condition, and may be classified as having critical illness–related corticosteroid insufficiency, reported Dr. Christos Triantos and his colleagues in the July issue of Clinical Gastroenterology and Hepatology.

Dr. Triantos, of the department of gastroenterology at the University Hospital of Patras, Greece, and his colleagues evaluated 20 patients with cirrhosis and bleeding varices and compared them with 74 controls: 14 healthy volunteers and 60 patients who had stable cirrhosis and no bleeding (Clin. Gastroenterol. Hepatol. 2011 [doi:10.1016/j.cgh.2011.03.033]).

The participants were divided into five groups. Group 1 included 10 with variceal bleeding, who underwent a conventional-dose short synacthen test (SST), administered by injecting a 250-mcg dose of the synthetic adrenocorticotropin hormone synacthen. Serum cortisol concentration was measured at baseline, 30 minutes, and 60 minutes.

Group 2 comprised the remaining 10 patients with variceal bleeding, who underwent the low-dose short synacthen test (LDSST), in which a 1-mcg dose of synacthen was administered. Again, serum cortisol levels were measured at baseline, 30 minutes, and 60 minutes. Two hours later, these patients underwent the SST as well.

Group 3 consisted of the 14 healthy volunteers, who also underwent both tests. Group 4 comprised 50 stable (nonbleeding) cirrhosis patients who underwent the LDSST, and the fifth group comprised the remaining 10 stable cirrhotics, who underwent the SST.

Among nonstressed (nonbleeding) patients who underwent the LDSST, adrenal insufficiency was diagnosed when a peak cortisol level was 18 mcg/dL or less. In stressed (bleeding) patients, the definition of adrenal insufficiency was a peak cortisol level of less than 25 mcg/dL, or a maximum change of less than 9 mcg/dL from baseline.

On the SST, adrenal insufficiency was defined as a peak cortisol level of 18 mcg/dL or less in nonstressed patients. In patients with bleeding, it was a maximum difference between peak and basal levels of less than 9 mcg/dL, or a random total cortisol level of less than 10 mcg/dL.

The authors found that patients with variceal bleeding had higher median basal cortisol levels (15.4 mcg/dL) compared with stable cirrhotic patients (8.7 mcg/dL, P = .001) or healthy volunteers (10.1 mcg/dL, P = .01).

Looking at peak cortisol, patients with bleeding varices also had higher median levels than did nonbleeding cirrhotic patients, both on the SST (32.7 mcg/dL vs. 21 mcg/dL, P = .0001) and on the LDSST (28 mcg/dL vs. 18.5 mcg/dL), although the difference on the latter test did not reach statistical significance.

However, the change between peak and basal values was not significantly different between patients with and without bleeding, either on the SST (15.4 mcg/dL vs. 12.7 mcg/dL, respectively) or the LDSST (9.3 mcg/dL vs. 8.1 mcg/dL).

Overall, "among patients analyzed by SST, 6/20 with variceal bleeding had adrenal insufficiency (AI) (30%), compared with 3/10 controls with [stable] cirrhosis (30%)," wrote the authors.

And on the more sensitive LDSST, "6/10 patients with variceal bleeding had AI (60%) vs. 24/50 controls with cirrhosis (48%)," they added.

The lack of a significant difference between patients with and without bleeding, with regard to the change between peak and basal levels, shows that many of those with bleeding meet the criteria for "critical illness–related corticosteroid insufficiency" with respect to their severity of illness, wrote the authors. None of the healthy volunteers was found to have adrenal insufficiency.

According to the researchers, a large proportion of those with variceal bleeding showed a dysfunctional response on both the SST assay and the more sensitive LDSST test, thought to be a better mimic of physiologic adrenocorticotropic hormone secretion.

"Although better methods of assessing adrenal function in cirrhosis are needed, the clinical significance of the abnormal ACTH tests and [critical illness–related corticosteroid insufficiency] may only be resolved by a clinical trial of corticosteroid supplementation in variceal bleeding as is used in sepsis."

The authors declared no conflicts of interest related to this study.

Major Finding: Over 10 years of follow-up, 32.6% of patients with hypertension and diabetes recorded a brain infarct, compared with 15.1% of those without either condition.

Data Source: The Atherosclerosis Risk in Communities (ARIC) Study.

Disclosures: Lead author Dr. Knopman disclosed being a deputy editor of Neurology, the journal in which this study was published. He disclosed relationships with Eli Lilly, the Elan Corporation, Baxter International, and Forest Laboratories. Additionally, he and several other investigators stated they have received grants from the National Institutes of Health and the National Heart, Lung, and Blood Institute, which partially funded the ARIC study.

Diabetes and hypertension were strongly and independently associated with brain infarcts, as well as with atrophic changes such as increasing ventricular size and sulcal widening.

The finding, from one of the first longitudinal imaging studies to look at vascular risk factors and infarct, confirms that “control of blood sugar and blood pressure in midlife should reduce the likelihood of ischemic and atrophic changes in the brain in subsequent decades,” wrote Dr. David S. Knopman and his colleagues (Neurology 2011 May 4 [doi:10.1212/WNL.0b013e31821d753f]).

Dr. Knopman of the Mayo Clinic, Rochester, Minn., and his colleagues looked at an initially middle-aged cohort of patients from the Atherosclerosis Risk in Communities (ARIC) study, which in 1987 recruited nearly 16,000 adults aged 45-64 years from Forsyth County, N.C.; Jackson, Miss.; selected suburbs of Minneapolis; and Washington County, Md. The subset for the current study included 1,812 patients who underwent brain MRI in 1994-1995. They were 55 years and older at this time, and came from Forsyth County or Jackson.

Ten years later, between 2004 and 2006, the patients were invited to undergo a follow-up MRI and vascular health assessments. Overall, 1,112 of these follow-up images were of sufficient quality for inclusion in the present study (689 females; mean age 61.7 years).

“Compared with current participants, those who died, were ineligible, or refused to participate in the follow-up scan were older, had a much higher stroke rate, had a higher rate of diabetes and hypertension, and had worse imaging at the baseline scan,” wrote the authors.

At baseline, 50.3% of the included subjects had neither hypertension nor diabetes. (hese patients were classified as having “low vascular risk,” the researchers wrote. Patients with both conditions were referred to as “high vascular risk” and made up 9.2% of the total cohort studied, they added.

Among the high-risk group, incident infarcts were seen in 32.6%, compared with 15.1% in the low vascular risk group, and 20.1% in the overall cohort. The risk increased with disease severity, the authors found.

“Those in the highest tertile for both fasting blood sugar and systolic blood pressure had 3.68 higher risk (95% confidence interval, 1.89-7.19) of new infarcts compared with subjects in the lowest tertile for both conditions,” they added.

Diabetes alone was also associated with incident infarct, independent of hypertension. After adjustment for variables including age, sex, race, hypertension, and prevalent stroke, diabetes conferred a nearly two-fold risk of incident infarct, compared with those patients without the condition (odds ratio, 1.96; 95% CI, 1.23-3.10).

Similarly, hypertension alone was associated with an OR for incident infarct of 1.58, compared with those patients with normal blood pressure (95% CI, 1.08-2.30).

Looking at brain atrophic changes, Dr. Knopman found that most patients had a change in ventricular size, sulcal widening, and white matter hyperintensities over the 10-year period, and older age by itself accounted for worsening in these categories.

Vascular risk factors also played a role, as 84.7% of patients in the high risk group, versus 73.2% in the low risk group, had ventricular size progression of one grade or more over the study period.

Similarly, 76.5% of high risk patients versus 55.5% in the low-risk group showed white matter hyperintensity progression. And 80.0% of high-risk patients, versus 69.6% in the low-risk group, showed an increase in sulcal widening.

The authors found no race- or sex-specific interactions between changes in brain imaging and vascular risk factors.

The study's strengths include its large sample size, biracial composition, extensive risk factor assessment at baseline, and decade-long follow-up, the researchers said. Its weaknesses included the fact that many subjects were lost over the 10 years of follow-up. But “those persons who had follow-up scans were healthier in all respects including lower burdens of vascular risk factors, and less pathology on imaging,” they wrote.

Consequently, “our findings probably understate the links between diabetes and hypertension.”

In addition, at the time of the initial scans, volumetric MRI was not yet available, making measurement over time of that particular parameter impossible, they noted.

Major Finding: Over 10 years of follow-up, 32.6% of patients with hypertension and diabetes recorded a brain infarct, compared with 15.1% of those without either condition.

Data Source: The Atherosclerosis Risk in Communities (ARIC) Study.

Disclosures: Lead author Dr. Knopman disclosed being a deputy editor of Neurology, the journal in which this study was published. He disclosed relationships with Eli Lilly, the Elan Corporation, Baxter International, and Forest Laboratories. Additionally, he and several other investigators stated they have received grants from the National Institutes of Health and the National Heart, Lung, and Blood Institute, which partially funded the ARIC study.

Diabetes and hypertension were strongly and independently associated with brain infarcts, as well as with atrophic changes such as increasing ventricular size and sulcal widening.

The finding, from one of the first longitudinal imaging studies to look at vascular risk factors and infarct, confirms that “control of blood sugar and blood pressure in midlife should reduce the likelihood of ischemic and atrophic changes in the brain in subsequent decades,” wrote Dr. David S. Knopman and his colleagues (Neurology 2011 May 4 [doi:10.1212/WNL.0b013e31821d753f]).

Dr. Knopman of the Mayo Clinic, Rochester, Minn., and his colleagues looked at an initially middle-aged cohort of patients from the Atherosclerosis Risk in Communities (ARIC) study, which in 1987 recruited nearly 16,000 adults aged 45-64 years from Forsyth County, N.C.; Jackson, Miss.; selected suburbs of Minneapolis; and Washington County, Md. The subset for the current study included 1,812 patients who underwent brain MRI in 1994-1995. They were 55 years and older at this time, and came from Forsyth County or Jackson.

Ten years later, between 2004 and 2006, the patients were invited to undergo a follow-up MRI and vascular health assessments. Overall, 1,112 of these follow-up images were of sufficient quality for inclusion in the present study (689 females; mean age 61.7 years).

“Compared with current participants, those who died, were ineligible, or refused to participate in the follow-up scan were older, had a much higher stroke rate, had a higher rate of diabetes and hypertension, and had worse imaging at the baseline scan,” wrote the authors.

At baseline, 50.3% of the included subjects had neither hypertension nor diabetes. (hese patients were classified as having “low vascular risk,” the researchers wrote. Patients with both conditions were referred to as “high vascular risk” and made up 9.2% of the total cohort studied, they added.

Among the high-risk group, incident infarcts were seen in 32.6%, compared with 15.1% in the low vascular risk group, and 20.1% in the overall cohort. The risk increased with disease severity, the authors found.

“Those in the highest tertile for both fasting blood sugar and systolic blood pressure had 3.68 higher risk (95% confidence interval, 1.89-7.19) of new infarcts compared with subjects in the lowest tertile for both conditions,” they added.

Diabetes alone was also associated with incident infarct, independent of hypertension. After adjustment for variables including age, sex, race, hypertension, and prevalent stroke, diabetes conferred a nearly two-fold risk of incident infarct, compared with those patients without the condition (odds ratio, 1.96; 95% CI, 1.23-3.10).

Similarly, hypertension alone was associated with an OR for incident infarct of 1.58, compared with those patients with normal blood pressure (95% CI, 1.08-2.30).

Looking at brain atrophic changes, Dr. Knopman found that most patients had a change in ventricular size, sulcal widening, and white matter hyperintensities over the 10-year period, and older age by itself accounted for worsening in these categories.

Vascular risk factors also played a role, as 84.7% of patients in the high risk group, versus 73.2% in the low risk group, had ventricular size progression of one grade or more over the study period.

Similarly, 76.5% of high risk patients versus 55.5% in the low-risk group showed white matter hyperintensity progression. And 80.0% of high-risk patients, versus 69.6% in the low-risk group, showed an increase in sulcal widening.

The authors found no race- or sex-specific interactions between changes in brain imaging and vascular risk factors.

The study's strengths include its large sample size, biracial composition, extensive risk factor assessment at baseline, and decade-long follow-up, the researchers said. Its weaknesses included the fact that many subjects were lost over the 10 years of follow-up. But “those persons who had follow-up scans were healthier in all respects including lower burdens of vascular risk factors, and less pathology on imaging,” they wrote.

Consequently, “our findings probably understate the links between diabetes and hypertension.”

In addition, at the time of the initial scans, volumetric MRI was not yet available, making measurement over time of that particular parameter impossible, they noted.

Major Finding: Over 10 years of follow-up, 32.6% of patients with hypertension and diabetes recorded a brain infarct, compared with 15.1% of those without either condition.

Data Source: The Atherosclerosis Risk in Communities (ARIC) Study.

Disclosures: Lead author Dr. Knopman disclosed being a deputy editor of Neurology, the journal in which this study was published. He disclosed relationships with Eli Lilly, the Elan Corporation, Baxter International, and Forest Laboratories. Additionally, he and several other investigators stated they have received grants from the National Institutes of Health and the National Heart, Lung, and Blood Institute, which partially funded the ARIC study.

Diabetes and hypertension were strongly and independently associated with brain infarcts, as well as with atrophic changes such as increasing ventricular size and sulcal widening.

The finding, from one of the first longitudinal imaging studies to look at vascular risk factors and infarct, confirms that “control of blood sugar and blood pressure in midlife should reduce the likelihood of ischemic and atrophic changes in the brain in subsequent decades,” wrote Dr. David S. Knopman and his colleagues (Neurology 2011 May 4 [doi:10.1212/WNL.0b013e31821d753f]).

Dr. Knopman of the Mayo Clinic, Rochester, Minn., and his colleagues looked at an initially middle-aged cohort of patients from the Atherosclerosis Risk in Communities (ARIC) study, which in 1987 recruited nearly 16,000 adults aged 45-64 years from Forsyth County, N.C.; Jackson, Miss.; selected suburbs of Minneapolis; and Washington County, Md. The subset for the current study included 1,812 patients who underwent brain MRI in 1994-1995. They were 55 years and older at this time, and came from Forsyth County or Jackson.

Ten years later, between 2004 and 2006, the patients were invited to undergo a follow-up MRI and vascular health assessments. Overall, 1,112 of these follow-up images were of sufficient quality for inclusion in the present study (689 females; mean age 61.7 years).

“Compared with current participants, those who died, were ineligible, or refused to participate in the follow-up scan were older, had a much higher stroke rate, had a higher rate of diabetes and hypertension, and had worse imaging at the baseline scan,” wrote the authors.

At baseline, 50.3% of the included subjects had neither hypertension nor diabetes. (hese patients were classified as having “low vascular risk,” the researchers wrote. Patients with both conditions were referred to as “high vascular risk” and made up 9.2% of the total cohort studied, they added.

Among the high-risk group, incident infarcts were seen in 32.6%, compared with 15.1% in the low vascular risk group, and 20.1% in the overall cohort. The risk increased with disease severity, the authors found.

“Those in the highest tertile for both fasting blood sugar and systolic blood pressure had 3.68 higher risk (95% confidence interval, 1.89-7.19) of new infarcts compared with subjects in the lowest tertile for both conditions,” they added.

Diabetes alone was also associated with incident infarct, independent of hypertension. After adjustment for variables including age, sex, race, hypertension, and prevalent stroke, diabetes conferred a nearly two-fold risk of incident infarct, compared with those patients without the condition (odds ratio, 1.96; 95% CI, 1.23-3.10).

Similarly, hypertension alone was associated with an OR for incident infarct of 1.58, compared with those patients with normal blood pressure (95% CI, 1.08-2.30).

Looking at brain atrophic changes, Dr. Knopman found that most patients had a change in ventricular size, sulcal widening, and white matter hyperintensities over the 10-year period, and older age by itself accounted for worsening in these categories.

Vascular risk factors also played a role, as 84.7% of patients in the high risk group, versus 73.2% in the low risk group, had ventricular size progression of one grade or more over the study period.

Similarly, 76.5% of high risk patients versus 55.5% in the low-risk group showed white matter hyperintensity progression. And 80.0% of high-risk patients, versus 69.6% in the low-risk group, showed an increase in sulcal widening.

The authors found no race- or sex-specific interactions between changes in brain imaging and vascular risk factors.

The study's strengths include its large sample size, biracial composition, extensive risk factor assessment at baseline, and decade-long follow-up, the researchers said. Its weaknesses included the fact that many subjects were lost over the 10 years of follow-up. But “those persons who had follow-up scans were healthier in all respects including lower burdens of vascular risk factors, and less pathology on imaging,” they wrote.

Consequently, “our findings probably understate the links between diabetes and hypertension.”

In addition, at the time of the initial scans, volumetric MRI was not yet available, making measurement over time of that particular parameter impossible, they noted.