By David Douglas
NEW YORK (Reuters Health) - 7/2/2019
In patients with early gastric cancer the collagen signature in the tumor microenvironment is an independent indicator of lymph-node metastasis, according to Chinese researchers.
"A prediction model based on the collagen signature is useful in treatment decision-making for patients with early gastric cancer," Dr. Jun Yan of Southern Medical University, in Guangzhou, told Reuters Health by email.
In a paper online January 30 in JAMA Surgery, Dr. Yan and colleagues note that the arrangement and orientation of collagen have proven to be indicators of tumor metastasis in breast and other cancers, but their role in early gastric cancer is unclear.
To investigate further, the researchers retrospectively studied 232 patients with histologically confirmed gastric cancer. Between 2008 and 2012, the patients underwent radical gastrectomy and received a diagnosis of T1 gastric cancer. None of the patients had had neoadjuvant radiotherapy, chemotherapy, or chemoradiotherapy.
A further 143 patients who had received the same diagnosis between 2011 and 2013 provided a validation cohort.
The researchers extracted collagen features in specimens using multiphoton imaging. They then developed a prediction model based on the collagen signature.
The signature was significantly associated with lymph-node metastasis (odds ratio, 5.47). Multivariate analysis showed that the depth of tumor invasion, tumor differentiation and the collagen signature were all independent predictors of metastasis.
The team developed a nomogram using these three independent predictors. To quantify the discrimination of the nomogram, they calculated the area under the receiver operating characteristic curve (AUROC).
This model showed good discrimination in both the primary cohort (AUROC, 0.955) and in the validation cohort (AUROC, 0.938).
The optimum cutoff point in the primary cohort gave a sensitivity of 86.8%, a specificity of 93.3% and an accuracy of 92.2%.
In all 375 patients, the corresponding proportions were 87.3%, 92.1% and 91.2%. The positive predictive value was 72.1% and the negative predictive value was 96.9%.
Overall, say the researchers, "Our findings suggest that lymph node metastasis is more likely to appear in patients with an undifferentiated histologic result, submucosal invasion, and a high collagen signature."
The prediction model based on this collagen signature, they conclude, "appears to be useful in decision making associated with tailored surgical interventions."
In an accompanying editorial, Drs. Jashodeep Datta and Vivian E. Strong of Memorial Sloan Kettering Cancer Center, in New York City, congratulate the authors "for their novel and important contribution to the field."
However, Dr. Strong told Reuters health by email that there are many questions still to be answered. Among the most important, given our evolving understanding of differences in gastric cancer biology between those from Asia and the West, "is whether the findings associated with the collagen signature will hold up when they are validated in patients from Western countries."