Study: Hydro-stomach CT inherently poor at cancer detection
CT fails to reliably depict early gastric cancers in patients ingesting water as a distending contrast agent, as four-fifths of cancers went undetected by gastrointestinal radiologists who experienced no significant improvements in detection even after being made were aware of patients’ cancers, reported a study published Feb. 28 in the World Journal of Gastroenterology.
As screening for gastric cancer has become feasible and available, the volume of cases of the disease has soared—by 40 percent in South Korea—yielding favorable outcomes for a cancer whose five-year post-resection survival is greater than 90 percent.
“Traditionally, both air and tap water have been used as oral contrast agents to achieve adequate gastric distension. As compared to the use of CT gastrography with air-distension, stomach CT with water-distension (hydro-stomach CT) is less hindered by artifacts caused by air in the lumen,” explained Ki Jeong Park, MD, and colleagues from the department of radiology and the Center for Imaging Science at Samsung Medical Center, Sungkyunkwan University School of Medicine in Seoul, South Korea.
In the study, 110 patients with known early gastric cancer drank 500 to 1,000 mL of tap water, fasted for six hours and were injected with 2 mg/kg of nonionic contrast material iopromide prior to CT. Park and colleagues sought to evaluate the diagnostic performance of hydro-stomach CT in the detection of early gastric cancers, while also comparing detection rates for all patients when both interpreting radiologists were blinded and unblinded (aware) of patients’ surgical-histological cancer findings.
The first reviewer delivered a sensitivity of 20 percent and a specificity of 98 percent when blinded; reviewer two showed a 19 percent sensitivity and a 98 percent specificity when blinded. These results did not significantly improve when the gastrointestinal radiologists were unblinded, with reviewer one demonstrating a 27 percent sensitivity and reviewer two a 25 percent mark.
Reviewer one showed a specificity of 100 percent unblinded, while reviewer two’s rate was unchanged (98 percent). Both gastrointestinal radiologists missed 71 of 110 cancers, with one or both reviewers observing 39 cancers.
For mucosal cancer (in a total of 71 patients), blinded sensitivity was 17 percent for reviewer one and 10 percent for reviewer two, while unblinded sensitivities were 21 percent and 14 percent, respectively. Intraobserver agreements for all cancers were good (weighted kappa = 0.677 and 0.666 for reviewers 1 and 2, respectively); interobserver agreements were fair (0.371) when blinded and moderate (0.558) when unblinded.
The authors called blinded false positives “frequent” for both reviewers one and two (five and four, respectively).
“Our study demonstrated that EGCs [early gastric cancers] are poorly visible on hydro-stomach CT images even though both axial and MPR [multiplanar reconstruction] (sagittal and coronal) images were evaluated. The detection rate was not significantly different between blinded and unblinded analysis for both reviewers,” Park and colleagues explained. The authors said their findings confirmed previous studies demonstrating still worse results for hydro-stomach CT, wherein 43 of 44 cancers were missed.
Park and co-authors found that the depth of invasion and the size of tumors were significantly associated with visibility as the degree of each grew. On the other hand, “Gross morphology was not a factor that affected EGC visibility in our study,” the authors found.
Park and colleagues acknowledged several limitations to their study, including the limits to generalizability inherent in a retrospective, single institution study involving the experience of only two radiologists. Moreover, the authors did not assess CT gastrography with air-distension, nor did they evaluate the visibility of cancers according to gastric distension, which influence the cancers’ visibilities still further.
“In conclusion, hydro-stomach CT imaging was not a reliable tool for the detection of EGC,” Park and colleagues summed. They said that the fact that visibility did not improve when reviewers were unblinded “indicated that EGC detection on hydro-stomach CT imaging is intrinsically limited due to poor detection performance.”
As screening for gastric cancer has become feasible and available, the volume of cases of the disease has soared—by 40 percent in South Korea—yielding favorable outcomes for a cancer whose five-year post-resection survival is greater than 90 percent.
“Traditionally, both air and tap water have been used as oral contrast agents to achieve adequate gastric distension. As compared to the use of CT gastrography with air-distension, stomach CT with water-distension (hydro-stomach CT) is less hindered by artifacts caused by air in the lumen,” explained Ki Jeong Park, MD, and colleagues from the department of radiology and the Center for Imaging Science at Samsung Medical Center, Sungkyunkwan University School of Medicine in Seoul, South Korea.
In the study, 110 patients with known early gastric cancer drank 500 to 1,000 mL of tap water, fasted for six hours and were injected with 2 mg/kg of nonionic contrast material iopromide prior to CT. Park and colleagues sought to evaluate the diagnostic performance of hydro-stomach CT in the detection of early gastric cancers, while also comparing detection rates for all patients when both interpreting radiologists were blinded and unblinded (aware) of patients’ surgical-histological cancer findings.
The first reviewer delivered a sensitivity of 20 percent and a specificity of 98 percent when blinded; reviewer two showed a 19 percent sensitivity and a 98 percent specificity when blinded. These results did not significantly improve when the gastrointestinal radiologists were unblinded, with reviewer one demonstrating a 27 percent sensitivity and reviewer two a 25 percent mark.
Reviewer one showed a specificity of 100 percent unblinded, while reviewer two’s rate was unchanged (98 percent). Both gastrointestinal radiologists missed 71 of 110 cancers, with one or both reviewers observing 39 cancers.
For mucosal cancer (in a total of 71 patients), blinded sensitivity was 17 percent for reviewer one and 10 percent for reviewer two, while unblinded sensitivities were 21 percent and 14 percent, respectively. Intraobserver agreements for all cancers were good (weighted kappa = 0.677 and 0.666 for reviewers 1 and 2, respectively); interobserver agreements were fair (0.371) when blinded and moderate (0.558) when unblinded.
The authors called blinded false positives “frequent” for both reviewers one and two (five and four, respectively).
“Our study demonstrated that EGCs [early gastric cancers] are poorly visible on hydro-stomach CT images even though both axial and MPR [multiplanar reconstruction] (sagittal and coronal) images were evaluated. The detection rate was not significantly different between blinded and unblinded analysis for both reviewers,” Park and colleagues explained. The authors said their findings confirmed previous studies demonstrating still worse results for hydro-stomach CT, wherein 43 of 44 cancers were missed.
Park and co-authors found that the depth of invasion and the size of tumors were significantly associated with visibility as the degree of each grew. On the other hand, “Gross morphology was not a factor that affected EGC visibility in our study,” the authors found.
Park and colleagues acknowledged several limitations to their study, including the limits to generalizability inherent in a retrospective, single institution study involving the experience of only two radiologists. Moreover, the authors did not assess CT gastrography with air-distension, nor did they evaluate the visibility of cancers according to gastric distension, which influence the cancers’ visibilities still further.
“In conclusion, hydro-stomach CT imaging was not a reliable tool for the detection of EGC,” Park and colleagues summed. They said that the fact that visibility did not improve when reviewers were unblinded “indicated that EGC detection on hydro-stomach CT imaging is intrinsically limited due to poor detection performance.”