In This Issue...

Challenges Remain in Clinical Interpretation of Whole Genome Sequencing

An exploratory study of 12 adult volunteers without evident inherited disease found whole genome sequencing (WGS) was associated with incomplete coverage of inherited disease genes, low reproducibility of genetic variations with the highest potential clinical effects, and uncertainty about clinically reportable WGS findings (JAMA 2014;311:1035–44). In some cases, however, WGS identified clinically actionable genetic variants that warranted further medical intervention.

The findings suggest that while the analytical validity of WGS is improving, technical challenges remain for sensitive and accurate analysis of individual genetic variations.

The authors performed the analysis because WGS and whole exome sequencing are beginning to be applied in clinical settings. The American College of Medical Genetics (ACMG) recently published recommendations regarding a minimum list of inherited disease genes subject to discovery, reporting, and clinical follow-up, regardless of the primary reason for genetic sequencing.

The patients received genetic counseling at the time they gave blood samples for the study; all samples underwent genomic analysis at Illumina, and nine underwent confirmatory sequencing at Complete Genomics. Five physicians proposed initial clinical follow-up based on the study findings.

The investigators found that depending on the sequencing platform, 10–19% of inherited disease genes were not covered to accepted standards for single nucleotide variant discovery, including 9–17% recommended by ACMG as reportable. While there was high overall genotype concordance between the two platforms and higher than previously reported concordance between insertion/deletion variants, in this study the authors found that the second sequencing platform confirmed less than one-third of insertion/deletion variants in inherited disease genes. This suggests that genetic variants "quite likely to be pathogenic" are identified inconsistently.

Even after stringent bioinformatic filtering, the investigators still had to curate about 100 findings from each participant, which took about 60 minutes per finding.

VEGF-A Accurate in Distinguishing Non-Malignant Pancreatic Cysts

Vascular endothelial growth factor (VEGF)-A is specifically elevated in cyst fluid and tissue samples from pancreatic serous cystic neoplasms (SCN), and could, as part of a diagnostic panel of other cyst fluid or DNA-based tests, positively identify benign pancreatic lesions (J Am Coll Surg 2014;

The authors conducted their study because unlike other pancreatic lesions, SCN does not progress to pancreatic cancer, so being able to distinguish it from malignant cysts would reassure patients and spare them from undergoing aggressive treatments to prevent pancreatic cancer. Two other biomarkers, carcinoembryonic antigen (CEA) and KRAS mutation status, have proven to be elevated in mucinous cysts which are at risk for becoming malignant, but not in SCN.

The study involved collecting pancreatic fluid at the time of endoscopy or surgery. The investigators used an enzyme-linked immunosorbent assay to test for VEGF-A or VEGF-C. They also used immunohistochemistry, western blot analysis, and multiplex polymerase chain reaction testing on tissue samples to detect VEGF-A and VEGF receptor 2 expression and genetic alterations in the von Hippel-Lindau gene.

With respective cutoffs of 8,500 pg/mL and 200 pg/mL, VEGF-A had 100% sensitivity and 87% specificity and VEGF-C had 100% sensitivity and 90% specificity in distinguishing SCN. The researchers proposed that patients with levels of VEGF-A >8,500 pg/mL but <16,000 pg/mL might undergo additional testing for CEA, KRAS, von Hippel-Lindau gene mutation status, and VEGF-C to enable 100% accuracy in identifying SCN.

Pretest Clinical Probability, Age-Adjusted D-dimer Cutoff Rule Out Pulmonary Embolism

In comparison to a fixed D-dimer cutoff of 500 µg/L, the combination of pretest clinical probability along with an age-adjusted D-dimer cutoff ruled out pulmonary embolism (PE) in more patients, and had a low likelihood of subsequent venous thromboembolism (VTE) (JAMA 2014;311:1117–24).

The authors conducted the study because D-dimer measurement has proven to be important in the work-up of suspected PE, but its clinical utility is limited in elderly patients. They previously had derived and validated a progressive age-adjusted D-dimer cutoff defined as the patient's age multiplied by 10 in patients who were at least 50 years old. Their current study was a prospective validation of this cutoff to see if it could be implemented in clinical practice.

Participating clinicians used either Geneva score or Wells score to assess clinical probability of PE. Patients with high or likely scores went directly to computed tomography pulmonary angiography. Those with low, intermediate, or unlikely clinical probability had D-dimer testing performed with one of six different quantitative high-sensitivity D-dimer assays.

Overall, 19% of patients had PE. Among the 2,898 who had low, intermediate, or unlikely clinical probability of PE, 28% had D-dimer results <500 µg/L, while 11.6% had a result >500 µg/L and their age-adjusted cutoff. In patients who were at least 75 years old, using their age-adjusted cutoff instead of a fixed cutoff of 500 µg/L increased from 6.4% to 29.7% those in whom PE could be excluded without any additional false-negative findings.