Autologous Fat Grafting into Rectovaginal Space Reduces Radiation Dose and Acute Injury to Rectum for Uterine Cervix-Targeted Radiation Therapy in Swine

Abstract

Brain metastasis is a commonly encountered clinical problem amongst patients diagnosed with cancers of the breast, lung, colorectum and melanoma, but there are currently no clinical parameters that can predict this outcome. Identifying prognostic indicators may allow for prophylactic cranial irradiation (PCI) of patients with these cancers, which may reduce the incidence of brain metastasis and improve median survival rates. Altered gene expression patterns of transcription factor 7-like 2 (TCF7L2) have been implicated in various cancers including breast cancer formation and clear cell renal cell carcinoma metastasis. We postulated that single nucleotide polymorphisms (SNPs) within TCF7L2 (rs4506565 and rs7901695) and interleukin-6, IL-6 (rs1800795) increase the risk of developing brain metastases from these 4 primary cancers. Materials/Methods: Patients from de-identified electronic medical records (EMR) databases at 2 institutions were assigned to either cases (patients with brain metastasis) or controls (no metastasis) using ICD-9 and tumor registry code-based algorithms. To verify the algorithm integrity, we thoroughly reviewed primary diagnoses with dates and brain metastasis status with dates for all cases. We matched cases 1:1 with controls based on primary cancer, age (AE 10 years), and time until metastasis (AE 3 months), and pulled DNA samples for genotyping. Univariable conditional logistic regression models were fit to analyze association between each SNP and case or control status. Results: Our algorithms identified 1,512 cases and 6,387 controls from 250,219 subjects in both databases. We confirmed 1,044 algorithmassigned cases to be correct, reassigned 26 cases as no-metastasis controls, 52 cases as those with non-brain metastasis and 390 cases as invalid for lacking the primary cancers of interest. Among breast cancer subjects from one institution, we analyzed 174 cases with any-organ metastasis and 174 matched controls, and observed that rs7901695 and rs4506565 SNPs were not associated with any-organ metastasis (p Z 0.37 and 0.68, respectively). However, in a subset of 69 cases and 69 matched controls with non-missing rs1800795 SNP, rs1800795 was associated with anyorgan metastasis (pZ0.003): patients with "GC" had an odds 2.40 times greater than patients with "CC" (95% CI 0.75-7.64) and patients with "GG" had an odds 6.84 times higher than those with "CC" (95% CI 1.82-25.68). Conclusion: Careful manual review of algorithm-based patient assignments in EMR databases is essential for high quality data. A "GG" SNP at rs1800795 of TCF7L2 may indicate an increased risk of any-organ metastasis in breast cancer patients. We are currently verifying the controls to expand this analysis to primary lung, colorectal and melanoma cancers and test brain metastasis specifically.