Evaluating Veracyte's Medium And Long-Term Prospects

About: Veracyte, Inc. (VCYT)
by: BioTrench

Afirma has been making large gains in the thyroid space over the past year with additional runway for growth over the next few quarters.

VCYT has added Percepta and Envisia to their line-up and looks to ramp up sales as they continue acquiring coverage and hiring sales reps.

Given current growth trends, VCYT should be able to break-even in under 5 years.

We evaluate VCYT's long-term potential but note a number of variables that could significantly alter VCYT's valuation.

Veracyte (VCYT) is a diagnostics company that combines machine learning with gene sequencing to facilitate the diagnostic process. VCYT currently offers three products: Afirma, Percepta and Envisia. Afirma has been on the market since 2011; in contrast, Percepta and Envisia are both (relatively) new products. Percepta has been commercially available for a year while Envisia is going commercial in 2019. VCYT has had a good run in recent quarters, going from $6.60 at the beginning of the year to $12.50 at the time of writing.

Chart VCYT data by YCharts

This is largely due to the high level of growth that they have been delivering over the past year.

Source: November 2018 Canaccord Presentation

With a market cap around $500M, VCYT may still have room to grow as it tries to break-even within the next few years.

Afirma - A raising standard in thyroid nodule diagnostics:

Afirma is being readily adopted in the thyroid nodule market with additional room for growth. Nodules are simply lumps of cells. They can be harmless or harmful, depending on a number of factors. Thyroid nodules have an estimated incidence of ~0.1%, which suggests that over 320,000 people require assessment and follow-up each year. As nodules can develop due to a variety of reasons, there is a need to identify which ones are actually malignant and require treatment.

Diagnosis of thyroid nodules, along with follow-up surveillance, is achieved through use of fine needle aspiration (FNA). As such, one estimate puts the number of FNA performed per year in the US at 600,000. Results from FNA can be broadly classified as benign, indeterminate, malignant or non-diagnostic. A number of studies found indeterminate rates to be in the 10-15% range1,2,3,4,5. This suggests that 60,000 to 90,000 tests come out as indeterminate each year.

The Afirma system is used to determine if an indeterminate nodule is not malignant. In other words, Afirma has a high negative predictive value (NPV), and will tell you if the nodule is likely harmless. This ultimately helps to avoid unnecessary surgeries. The original system used a microarray and was labeled as Afirma Genomic Expression Classifier (GEC). This system has since been updated to use a deep RNA-sequencing platform and renamed to Genomic Sequencing Classifier (GSC). The transition from GEC (93% NPV) to GSC (96% NPV) was completed mid-2018, which has led to improved diagnostic rates. Afirma's performance is on par with CBLPath/Sonic Healthcare's ThyroSeq V3 (97% NPV) and ahead of Interpace's testing solution (94% NPV).

Afirma has performed 22,556 tests in the first nine months of 2018, and will likely hit ~30,000 by EoY (vs. 26,000 in 2017). VCYT has estimated that their market penetration may be above 35% by the end of Q4, which is inline with our market size estimate. With high YoY growth over the past few years, it looks like there is still a decent runway for the next few quarters. Moreover, they have recently launched their Afirma Xpression Atlas solution, which can be used to sequence malignant nodules and create more targeted therapies.

Percepta - Improving lung nodule screening:

Percepta looks ready for a ramp up, but its market size may be smaller than expected. Percepta is meant to complement bronchoscopies and has a high NPV in low- to intermediate-risk patients (100% and 93% respectively). In a multi-center prospective study, Percepta was able to reclassify 34% of patients with lung nodules to a lower level of risk. As a result, only 14% of the sub-group had an invasive procedure, as compared to 34% that would have had a procedure in the absence of a genomic test. By 12 months, 83% remained procedure free, meaning an additional 3% underwent a procedure. This suggested that Percepta could assist in diagnosis and identify benign nodules that could be managed with basic surveillance in lieu of something more invasive.

Incidental lung nodules are common, with one study estimating that 1.57M people per year have nodules appear during their computed tomography (CT) scans. Evaluation of nodules may involve a number of procedures, one of much is a bronchoscopy, wherein a tube is inserted into the airway for observation and sample collection purposes. According to the National Lung Screening Trial (NLST) research team, which screened 56,000 people on a yearly basis over 3 years, around 4% of people with nodules received a bronchoscopy during follow-up.

There are two likely contributors behind the low number of bronchoscopies. Firstly, there is a high number of false positives (less than 4% of nodules were cancerous). Secondly, a large number of alternative diagnostic/invasive procedures were also employed. Whether this is due to shortcomings with bronchoscopies or not is largely unknown. While bronchoscopies are generally safe, the problem is that its sensitivity (ability to detect disease) can vary quite significantly. This is particularly true for peripheral nodules (in far/hard to reach places), where sensitivity can range from 34-63% depending on nodule size. Hence, additional tools are required to assist in diagnosis. While advances have improved diagnostic rates, it is still less than ideal.

For treatment purposes, patients can be categorized as high-, intermediate- or low-risk. High-risk patients are likely to receive surgery in order to remove nodules, so Percepta provides minimal benefit. In the original AEGIS studies, 43% of patients had non-diagnostic results, of which 57% were classified as low- or intermediate-risk. Hence, 24% of the population had an inconclusive bronchoscopy and either a low or intermediate pre-test malignancy risk. In contrast, a 2016 abstract for VCYT's Percepta registry showed that 53% of patients had an inconclusive bronchoscopy and a low- or intermediate-risk profile.

Due to a lack of information, we will assume that 40% of patients receiving a bronchoscopy will receive a non-diagnostic result and have a low- or intermediate-risk profile. If there are 1.6M-2M patients with nodules, then 64,000-80,000 would receive bronchoscopies (average of 72,000). Of those, roughly 30,000 would benefit from Percepta. One thing to note is that VCYT suggests that 350,000 bronchoscopies are performed per year to diagnose nodules, which is significantly higher than our value of 72,000.

If their value is correct, then our estimate will significantly underestimate the market size. However, as a reminder, a significant number of lung cancer cases every year (~70%) are diagnosed in the advanced/metastatic stage (i.e. stage 3/4), so even if bronchoscopies are utilized, it's unlikely Percepta would be required.

Percepta has had 1 year of sales and had 374 tests performed in Q3 2018. Currently, it only has Medicare coverage, but the team is working on getting private coverage, which will help it significantly. Moreover, there is runway for additional growth. It has been estimated that of the 7-8M smokers eligible for CT screening, only 2-4% take the time to get one done1,2. In addition, if Percepta is truly beneficial, then more physicians may opt to use bronchoscopies over time instead of other more invasive techniques. As an aside, if VCYT can complete their nasal swab lung cancer screen, then it could significantly expand their revenue as well.

Envisia - Improving lung disease diagnostic rates:

Envisia is opening up a new market opportunity in interstitial lung disease and sales will start picking up in 2019. Interstitial lung disease (ILD) is a broad family of lung disorders. Idiopathic pulmonary fibrosis (IPF) is one of the most common ILD and accounts for 31% of cases in the US. IPF has a poor clinical outcome, with half the patients living only 2-3 years, although disease progression can vary significantly. Due to a mix of data insufficiency and regional differences, there are no accurate figures regarding IPF incidence.

Incidence in North America is estimated to be between 5-10 per 100,000 people if using a narrow case definition of IPF and goes up to 17 per 100,000 people if using a broad case definition. The difference between the two is that narrow case definition uses a stricter set of requirements for confirming diagnosis. Assuming an incidence of 10 per 100,000 people, there would be 33,000 new cases of IPF per year in the US out of 106,000 cases of ILD.

Catching IPF early can help slow down disease progression, but misdiagnoses occur frequently, resulting in delays. IPF is typically diagnosed using high resolution CT (HRCT) and/or surgical biopsy. First, usual interstitial pneumonia (UIP) is identified. When the cause of UIP cannot be found, IPF is diagnosed. One study found that 70% of HRCT readings lead to a "possible" or "inconsistent with UIP" result.

Matched surgical biopsies indicated that at least half of those possible and inconsistent with UIP results were likely to be UIP. Unsurprisingly, predicting UIP is difficult in this population. When HRCT results are inconsistent with UIP, negative and positive predictive values can be lower than 40%.

Envisia helps bring clarity to possible and inconsistent with UIP HRCT results. It can be performed alongside a bronchoscopy and doesn't involve surgery. A recent abstract showed that Envisia had a PPV of 81% and an NPV of 85% in a retroactive study. With 100,000 ILD cases a year, the market size for Envisia is ~70,000 patients in the US.

Currently, Envisia has received a draft Medicare coverage and is waiting for the outcome early 2019. If successful, VCYT will have a third revenue stream up and running. In the meantime, they have an early access program running since May 2018, so a handful of tests have been performed to date. Although Envisia's test price has not been set yet, management has suggested that it would be in-line with Percepta and Afirma, meaning it would be likely in the $1,000-3,000 range.

Gauging the likelihood of breaking even by 2021:

Given the current trends, there is a high likelihood that VCYT will be breaking even within the next 5 years, if not sooner. We can estimate this roughly using Afirma sells, as we believe it will be the biggest driver towards breaking even.

Afirma tests were ordered 22,556 times in the first nine months of 2018 and will likely hit ~30,000 by EoY. Afirma volumes have shown decent YoY growth for the past 3 years (2016: 19.6%; 2017: 11.7%, 2018: est. 14.6%). The increase in sales this year likely stems from VCYT's sales force expansion, as they look to close the year with 110-115 sales representatives, as compared to 71 at the beginning of the year.

VCYT has estimated that their market share will likely be higher than 35% by the end of 2018. From this, we will assume one of three possible scenarios (increased, maintained, or reduced growth rates) for 2019, followed by a decline afterwards to reflect gradual market saturation, as shown below.

Afirma growtt rate prediction

Source: Author's projections.

For estimating revenue, we have the following:

And finally on average, we accrued between $2700 and $2800 for the Afirma Genomic Classifier test including variants meeting our revenue recognition standard which was between 90% and 95% of the reported Afirma Classifier test volume.

Source: 2018 Q2 conference call

Based off of VCYT's update, we will assume a cost of $2,750 per test, with 92.5% of tests being accrued. This translates to overall Afirma revenues of $95-111M in 2021, depending on which of the three scenarios occur. VCYT also offers a cytopathology service, but it seems to be slow growing ($8.5M in 2017 and est $9-9.5M for 2018). Assuming that the pathology service will be ~$10M for the coming years, FNA-related revenues will be ~$105-120M.

Currently, VCYT’s gross margins are ~65% and operating costs were $19.5M for Q3 2018. Accounting for the Q4 sales force expansion (which we assume to be close to completion), operating costs will likely be $21-22.5M/qtr or $90M/year. Based off of the above assumptions, an annual revenue of $140M would be required to break even. This would mean that an additional $20-35M in Percepta and Envisia revenue would be required.

Percepta costs $1,000 per test currently, although this may increase in the future. Likewise, Envisia is going to be priced similar to the other two tests (~$1,000-3,000). Using a minimum price point of $1,000 for Percepta and Envisia, a total of 20,000-30,000 tests would be required in order to break even. This level of volume is on the higher end and may be difficult to achieve in 2021, although it may be possible by 2022. As mentioned previously, VCYT has been increasing their sales force rapidly over the course of the year. Moreover, these sales representatives carry all three product lines. From their conference call:

Samples for Percepta and Envisia are now being collected from the same bronchoscopic suite and even more evidence that this approach is working, 35% of our customers adopting Percepta are also using Afirma.

Source: 2018 Q3 conference call

As long as Envisia finalizes coverage in 2019, then an aggregate of 20,000 to 30,000 in Envisia/Percepta volume could be reachable within 2-3 years, given cross sales potential. This can further be facilitated by test prices being higher than what has been assumed or if either tests receive private coverage. Moreover, revenue will also be generated from Afirma Xpression Atlas, which has largely been ignored so far. Up to 30% of indeterminate results can be malignant, which can be captured by the Atlas system. Likewise, there has been mention of biopharmaceutical relationships which can contribute additional revenue. If VCYT fails to break even in 2021, then the likelihood of it happening in 2022 is fairly high, given our current analysis.

Going for a long-term valuation:

While VCYT's medium-term potential looks decent, it may or may not translate to longer-term potential, depending on a number of factors. To assess long-term potential, we will first take a look at a base case. At the end of Q3 2018, VCYT had long-term debts of $25M, cash and equivalents of $77M, and a market cap slightly below $500M. Assuming:

  1. Collection on 92.5% of test volumes
  2. $2750/Afirma, $1,000/Percepta, $1,000/Envisia
  3. Minimal growth in cytopathology
  4. Eventual 50% market penetration in lung and ILD
  5. 30% of Afirma GSC tests are malignant, and half go for Atlas ($1,000/test)

We arrive at the following:

Revenue Stream Volume Gross Profit
Afirma GSC 45000 $114.5M
Afirma Xpression Atlas 6750 $6.2M
Percepta 15000 $13.9M
Envisia 35000 $32.4M
Cytopathology - $10.0M
Biopharma - unknown
Total - >$177.0M

We will increase gross margins from 65% to 68% for the following reason:

And I'd just remind because the script - this would go out to all investors that we do have about $9 million, $9.5 million of our revenues side of pathology services which is a low single-digit margin product. So as we do more and more volume on the genomic side our margins will approach 70%, so we do expect our margin expansion to continue.

Source: 2018 Q3 conference call

As such, gross profit will be $120.4M, which will fall to $20.4M after accounting for $100M in operating costs (assuming further expansion to accommodate sales increase). A look at other reference/diagnostics firms (MYGN, GHDX, QDEL, NEO) suggests EV/EBITDA multiples of 15x and above are not unusual. Using a 15-20x multiple would imply a valuation of $300-400M for VCYT. While this would suggest that VCYT is overvalued, there are several caveats of note.

The biggest factor is Percepta and Envisia's long-term pricing. For example, if Percepta/Envisia's average test cost were to eventually approach $2,000, then VCYT would be valued at $800M-1B, given our current assumptions. Secondly, as mentioned earlier, our estimate of Percepta's market size is significantly smaller than VCYT's estimate. A doubling of our Percepta market estimate, along with test volumes, would lead to a value of $450-600M.

Thirdly, the base case used a market penetration rate of 50%; however, if a 75% penetration rate was used instead, VCYT's value would come out to $540-720M. The above scenario changes would imply that VCYT is currently fairly valued, if not undervalued. Finally, biopharmaceutical collaborations could become a large variable in the long run, but it is currently difficult to assess its potential. Overall, while VCYT has the potential for continued upside in the long run, it cannot be considered de-risked. The commercial foundation being laid out in 2019 for its products will have a significant impact and attention will be required.

Concluding remarks

VCYT has been performing well in the thyroid arena since their initial launch in 2011. Since then, they have been working on adding additional sources of revenue, which look to begin ramping up in the next year or two. In 2017, they started commercial activities for Percepta, which has witnessed its first full year of sales. As of May 2018, they have started offering Afirma Xpression Atlas as a supplement to their GSC solution. Finally, they look to finalize Medicare coverage for Envisia and begin ramping up sales in 2019.

To this end, they have been increasing their sales force significantly over the past year. With the ability to cross-sell between the different products, volume should pick up quickly over the next year or two. As a result, there is a high probability that VCYT will break-even in under 5 years. Lastly, pending long-term pricing of the various products and additional factors, VCYT may offer an attractive proposition.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.