Background:

Clinical laboratories began offering whole-exome sequencing in 2011 at a cost between $4,500 and $9,000. Reported detection rates for deleterious mutations range from 25 to 50%. Based on the experience of our clinical genetics service, actual success rates may be lower than estimated rates. We report results from our own experience along with a survey of clinical geneticists to ascertain (i) current success rates for causal gene detection in a clinical setting; (ii) if there are insurance authorization issues; and (iii) if turnaround times quoted by the clinical laboratories are accurate; we also gauge provider opinions toward clinical whole-exome sequencing.

Methods:

We reviewed our results and the results of a survey that was electronically distributed to 47 clinical genetics centers.

Results:

A total of 35 exome reports were available. If all positive results are collated, we observe a success rate of 22.8%. One result incorrectly identified a known benign variant as pathogenic. Some insurers covered all testing, whereas others denied any insurance coverage. Only three (23.1%) of our reports were available within the laboratory’s quoted turnaround times. More than 50% of clinicians queried in our survey had not ordered whole-exome sequencing at the current time, many stating concerns regarding interpretation, insurance coverage, and cost.

Conclusion:

Clinical whole-exome sequencing has proven diagnostic utility; however, currently many clinicians have concerns regarding interpretation of results, insurance coverage, and cost.

Genet Med 16 9, 717–719.

Clinical laboratories began offering whole-exome sequencing (WES) in 2011 with a cost range between $4,500 and $9,000. Yields quoted by some clinical laboratories have been as high as 50% for causal gene identification, whereas others have been more conservative in their estimates, quoting a 25% causative deleterious mutation detection rate.¹ Reports in the published literature have shown success rates as high as 50% for detection of causal gene mutations.² However, based on the experience of our clinical genetics service, actual success rates may be lower than estimates. To our knowledge, the use of WES in a typical clinical setting has not yet been well defined. We report results from our own experience along with results from a survey of clinical geneticists performed in December 2012 to ascertain (i) current success rates for causal gene detection in a clinical setting; (ii) if there are insurance authorization issues; and (iii) if turnaround times (TATs) quoted by the clinical laboratories are accurate; and we also gauge provider opinions toward clinical WES.

The study cohort consists of patients who have been seen in our medical genetics clinic and by medical geneticists in North America. Our clinic is a regional referral center in northern California. We see patients ranging in age from newborns to adults, and the most common presentations include developmental delay and congenital malformations.

We undertook interview of genetics professionals in our service. We asked whether they had ordered clinical WES, whether they had received results, and whether these results showed the causal gene mutation for the patient’s condition. An eight-item survey was electronically distributed to 47 clinical genetics centers across North America. The survey questions were (i) Have you ordered WES on a clinical basis? (ii) If yes, how many samples have you sent? (iii) How many results have you received and how many are likely the true answer for your patient’s clinical condition? (iv) If you had any answers that identify the likely causal mutation, retrospectively, how many do you feel could have been found using alternative methods (free-text explanation if needed)? (v) Have you been successful in getting insurance to pay for clinical WES? (vi) What clinical contexts have insurance companies authorized testing (options: multiple congenital anomalies, intellectual disability alone, intellectual disability plus congenital anomalies, have not authorized testing, and other (please specify))? (vii) Have you had pressure from other physicians or patients’ families to order clinical WES? and (viii) Please feel free to provide your comments on clinical WES.

As of March 2013, we have sent 27 WES samples in total from our service and have received 14 results. Two of these results (15.4%) identified the likely causal genetic abnormalities; both were considered in the differential. One result (7.69%) incorrectly identified a known benign variant as pathogenic. The other 11 reports (78.6%) did not identify a causal gene mutation; one reported a known benign variant in SHANK3 as pathogenic and likely disease causing despite recent literature describing this variant as a polymorphism.

From our survey sent to the other centers, we received 30 responses (63.8%). Thirteen of the 30 responses (43.3%) indicated that clinical WES had been ordered from their service. From these, 41 samples had been sent, and 21 results were available. Six of these results (28.6%) identified the causal mutations; however, two (9.52%) diagnoses could have been made using alternative methods such as single-gene testing. There were four unexpected results (19%) that were considered to be causative. Fifteen results (71.4%) did not provide a causal genetic etiology for the patient’s condition.

If all positive results are collated, we observe a success rate of 22.8%. If results for which the diagnosis could be made using alternative methods are excluded, we observe a 17.1% success rate (Figure 1).

Whole-exome sequencing (WES) results. *Wrong diagnosis: SHANK3 mutation. ^†Turnaround time (TAT): of 13 results available, only 3 (23.1%) were available within the laboratory’s quoted TAT. Average TAT was 123 days (18 weeks), and range was from 101 to 146 days (14.4–20.9 weeks).

Full size image

Insurance authorization

Our experience of getting insurance to approve tests has been mixed, with some insurers covering all testing and others denying any insurance coverage. From our survey, of 19 free-text responses, concern was expressed over insurance coverage (n = 5; 26%) and cost (n = 2; 11%). Multiple respondents reported that all Medicaid testing in their state was denied, whereas we have had instances in which Medi-Cal (California state Medicaid) has covered testing. For preferred provider organization insurance, our experience has been mixed, ranging from requests for large copays and deductibles to no out-of-pocket expense for patients. Indeed, much of this is dependent on other factors such as the patient’s particular plan and if they have met yearly deductibles. In addition, there were respondents who reported lengthy insurance approval processes (more than 3 months in one instance) with multiple requests to provide additional paperwork. One respondent from Canada reported that their health-care system had chosen not to pay for clinical exome sequencing.

Of cases that were approved by insurance, the most common clinical context was intellectual disability plus congenital anomalies (60%; n = 9), followed by multiple congenital anomalies (33%; n = 5), and intellectual disability alone (6.6%, n = 1). Five respondents reported that insurance had not approved testing (35.7%). One respondent reported ordering WES for an autoinflammatory condition in which the family wanted testing regardless of insurance authorization; testing was negative.

Turnaround time

From our 13 reports, only 3 (23.1%) were available within the laboratory’s quoted TAT. The typical quoted TAT from the clinical laboratories is 11–16 weeks. The average TAT was 123 days (18 weeks), and the range was between 101 and 146 days (14.4–20.9 weeks).

Some geneticists (n = 10; 36%) reported interest from other specialists and families in ordering WES. More than 50% of clinicians queried in our survey had not yet ordered WES, many stating concerns regarding interpretation, insurance coverage, and cost (Figure 1).

WES is clinically available and has proven diagnostic utility. However, results from our own experience and that of other clinical genetics centers suggest that success rates are not as high as those reported by some of the clinical laboratories and those in the medical literature. Based on our experience and that of other clinical genetics services, we estimate the current diagnostic yield to be in the range of 15–25%. We note that this diagnostic yield is in line with a recent report by one of the clinical laboratories.¹ We feel that concerns over interpretation may be warranted based on the incorrect reporting of one of our samples with a known benign variant being interpreted as pathogenic. It is imperative that laboratories offering clinical WES conduct thorough literature reviews to ensure that the latest evidence is used when calling variants deleterious or benign. In addition, concern was expressed over interpretation of WES findings in 8 of 19 (42%) free-text comments in the survey. Insurance authorization is currently a challenge, including some cases of complete denial of coverage. The most commonly approved clinical indication from our survey is intellectual disability plus congenital anomalies, followed by multiple congenital anomalies or intellectual disability alone. From our clinic, we have also had insurance approval for the clinical context of developmental delay and seizures. Table 1 highlights two examples in which diagnoses were made. We also note that TATs are often longer than those quoted by the clinical laboratories.

From the reports we have received, overall, we feel that they are well written. While each laboratory has different labels for the different categories within the report, they all segregate mutations/variants into categories largely defined as “answers,” variant of uncertain significance relevant to clinical phenotype, and variant of uncertain significance not felt relevant to clinical phenotype. Overall and particularly for variants of uncertain significance that are relevant to clinical phenotype, we feel that laboratories should include parental testing (if available), confirmation by Sanger sequencing, in silico analysis and relevant references, along recommendations for further testing as needed. Information regarding conservation and whether amino acid substitutions are conservative/nonconservative is also useful.

Currently, the decision to order a multigene panel or WES as a first-tier test is an important question facing the community. Our approach has been to compare cost of testing, quoted yield (if available), and whether a negative multigene panel would lead to subsequent WES testing. We do not have objective data to compare multigene panels to WES, and this would certainly be an interesting follow-up study. Also of note is that the issues of insurance authorization, TAT, and yield are certainly not unique to WES, and indeed, multigene panels have similar issues.

Finally, the issue of performing WES, using commercially available kits—some of which lack complete coverage of known medically relevant genes, compared with a “medical exome” that would test and offer near-complete coverage of only known medically relevant genes (currently ~4,600)—is an important question at present. Moving forward, a sensible approach would be to make use of commercially available kits while concurrently enhancing analysis of known medically relevant genes, thereby optimizing clinical yield.

In summary, clinical WES has proven diagnostic utility; however, currently many clinicians have concerns regarding interpretation of results, insurance coverage, and cost. This study provides insight from the clinical “end-user” perspective. We hope that as our experience with WES increases, interpretation and diagnostic yield will improve and costs will fall, leading to a greater clinical utility.

The authors declare no conflict of interest.

Affiliations

1. Division of Medical Genetics, Stanford University Medical Center, Stanford, California, USA

   • Paldeep Singh Atwal
   • , Marie-Louise Brennan
   • , Rachel Cox
   • , Michael Niaki
   • , Julia Platt
   • , Margaret Homeyer
   • , Andrea Kwan
   • , Sylvie Parkin
   • , Susan Schelley
   • , Leah Slattery
   • , Yael Wilnai
   • , Jonathan Adam Bernstein
   • , Gregory M. Enns
   •  & Louanne Hudgins

Corresponding author

Correspondence to Paldeep Singh Atwal.

To obtain permission to re-use content from this article visit RightsLink.

Publication history

DOI

https://doi.org/10.1038/gim.2014.10