Prioritized Gene Testing vs Testing for All Genes as a Cost-Reduction Strategy

Jacquelyn K. Beals, PhD

October 29, 2009

October 29, 2009 (Honolulu, Hawaii) — A project aimed at reducing genetic testing costs has developed a model based on pheochromocytoma. A new study validates the efficacy of using clinical parameters to predict individuals who carry the mutations. The study also shows that prioritizing specific genetic sequences in testing for diseases is an effective way to reduce analytical costs in genetic diseases.

Zoran Erlic, MD, from the Department of Nephrology, Section of Preventive Medicine, University Medical Center, Albert-Ludwigs University, in Freiburg, Germany, presented the cost-reduction strategy here during the American Society of Human Genetics 59th Annual Meeting.

Pheochromocytoma is one of a larger group of paraganglioma (PGL) tumors (head and neck paragangliomas [HNPs]). Malignant or nonmalignant tumors of the adrenals, as well as of the neck and thorax, can be sporadic or part of a hereditary syndrome. Genetic screening has shown that pheochromocytomas typically involve changes in at least 5 susceptibility genes:

  • SDHB and SDHD are succinate dehydrogenase subunit genes B and D; changes in these genes cause PGL syndrome type 1 and type 4.

  • von Hippel-Lindau disease, occurring in about 20% of pheochromocytoma patients, is caused by germline mutations in the VHL tumor suppressor gene.

  • Changes in RET cause multiple endocrine neoplasias (MEN2).

  • Neurofibromatosis 1 is associated with mutations in the NF1 gene; approximately 1% of neurofibromatosis patients have pheochromocytomas.

The cost of diagnostic scanning for all 5 genes is close to $8000. Scanning for NF1 is the most expensive ($4800), and the remaining scans each cost more than $3000.

The current study scanned 1149 index cases, 30% of which had a mutation in 1 of the 5 susceptibility genes. About 70% of the cases were mutation-negative. It was clear that patients carrying the NF1 mutation had clinical features of the NF1 syndrome. NF1 screening would not be required in other patients.

Among VHL carriers, 54% had "clinical anamnestical evidence" of von Hippel-Lindau syndrome — that is, they demonstrated a secondary response to an immunogenic substance after the antibodies were no longer detectable in the blood. Similarly, 61% of RET mutation carriers showed clinical anamnestical evidence of MEN. Dr. Erlic asked in his presentation whether such clinical observations could predict testing results, including the presence of mutations.

To determine the order in which genes should be tested, investigators "looked for the frequencies of these patients based on different patient characteristics . . . and made the statistical approach in testing order," Dr. Erlic told Medscape Pathology & Lab Medicine. "Then we validated our flow of gene testing orders . . . and it was confirmed that it was really working."

The prioritized order of genetic tests for patients with head and neck paragangliomas [HNP] was SDHD, SDHB, RET, and VHL. Patients with no HNPs and with a single tumor outside the adrenals would be better served by a different testing order: SDHD, VHL, SDHB, and RET. The algorithm developed by this group had a sensitivity of 95.7%.

The average cost of testing for each patient would be approximately $1876, rather than $3116 — a savings of about 40%. Eight carriers were missed by applying this algorithm, but Dr. Erlic believes that such failures would decrease in frequency. Asked about the 8 missed cases, he noted that there were no clinical predictors prior to mutation screening. Two cases had medullary thyroid carcinoma at ages 47 and 52 years, respectively; a tumor had been found in the other 6, but with no further manifestation years later. One case was lost to follow-up.

"What we really wanted to create was a model [that is] really easily obtainable with simple data like gender, tumor number, tumor location, without looking for a . . . chemical profile, which can be treated somehow, depending on the lab that is performing it," explained Dr. Erlic. "The clinical cost . . .  is going to be cheaper and cheaper because we believe that more and more genes will be involved, modifier genes will be identified. So we really want to introduce it to clinical practice, the genetics clinic," he said, "and we need to learn to consider these economic reasons."

Session comoderator David Malkin, MD, associate chief of research (clinical), senior staff oncologist, senior scientist, and codirector of the Cancer Genetics Program at The Hospital for Sick Children, Department of Pediatrics, Division of Oncology, in Toronto, Ontario, expressed concern about not incorporating biochemical assays into the model.

"I think that's very important and should be incorporated," he told Medscape Pathology & Lab Medicine. "I mean, [assays] have a cost, but they're not going to be as costly as a [genetic] test. So that might make it easier to select out the group that the test will be beneficial for, and also cost," Dr. Malkin observed.

Taking questions from the audience, Dr. Erlic agreed that chemical testing "would help us improve our model, and also involves the order to test in." However, "this is the first study published so far, and only 2 centers can do this chemistry. We want it to be easily available to each pathologist in the world. We want an algorithm based on simple data to get the price down," he concluded.

Dr. Erlic and Dr. Malkin have disclosed no relevant financial relationships.

American Society of Human Genetics (ASHG) 59th Annual Meeting: Abstract 116. Presented October 23, 2009.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.