Description

Intro

Early- to mid-stage surgical device companies can exhibit a number of compelling investment characteristics.  Surgeons are very habitual and dislike switching away from an instrument they have gotten comfortable with, no matter the cost – the risk of a complication or dissatisfied patient makes the switching costs very high.  Surgeons are often trained on specific instruments during residency and loathe to change.  Since generally neither the surgeon nor the patient bear the cost, they are quite price insensitive, resulting in very healthy gross margins.  Additionally, while sell-side analysts like to model a linear, decaying YoY growth rate, the adoption of surgical procedures can often follow something more similar to the “S-curve”, resulting in accelerating revenue growth and dramatic upside to consensus forecasts.  A study of history shows a number of ~$100mm revenue med tech companies inflecting to $500mm+ in 3-5 years.  Lastly, with very high gross margins and duplicative costs of a sales force/public company expenses, small medical device companies are often incredibly compelling M&A targets for larger companies.  As adoption takes hold and the product is proven out, revenue growth accelerates, estimates go up, and then ultimately get acquired at a big multiple of gross profit. 

I believe SI-BONE (SIBN, ~$1.1bn mkt cap, $1bn EV) sits at a compelling inflection point in terms of adoption further buoyed by a re-opening/pent-up surgical backlog demand.  The company made a number of key advances (improved insurance coverage and reimbursement) over the last 12-18 months that have been masked by COVID.  I believe you can make a compelling case for a 2-3x in the next 3-4 years with downside of only ~30% if things go wrong.  The simple upside math we’ll discuss in more detail is revenues inflecting from ~$100mm to ~$435mm in 2025, at 10x gross profit, you have an ~$3bn company.  The downside is revenue growth, which had accelerated to +27% pre-pandemic, rapidly petering out around 5%, combined with some share loss, and a sale at ~10x gross profit for ~$800mm.  

SIBN’s iFuse is used in a minimally invasive, outpatient surgical procedure addressing as much as 15-25% of lower back pain, often replacing complex open procedures or addictive pain meds.  Before we get into the details of the product and business, I would note that SIBN’s product (“iFuse”) was invented by orthopedist Dr. Mark Reiley who also founded co-founded Kyphon (and invented the kyphoplasty procedure).  Medtronic acquired Kyphon for $3.9bn in 2007. I would also note that Tony Recupero, newly promoted President of Commercial Operations is also an alum of Kyphon, where he drive sales from $260k to ~$200mm over ~5 years.  

Background

Si-Bone was founded in 2008 and pioneered a minimally invasive implant system called “iFuse” to treat sacroiliac joint (“SI joint”) dysfunction.  The sacroiliac joint connects the sacrum (sits at base of spine) to the ilac bone (major bone of the pelvis).  It’s the largest joint in the human body and supports the weight of the entire upper body which over time wears down, resulting in severe lower back pain.  The iFuse procedure is 3 rigid, triangular titanium implants inserted through a quick, minimally invasive procedure (~1.5 inch incision laterally in the buttock) to stabilize the joint, alleviate the pain, and over time allows the bones to grow into the implant and permanently fuse the joint.  ~80% of procedures are done in outpatient/ASC settings.  iFuse’s triangular titanium implants have 3x the strength and 6x the rotational resistance of a traditional screw.  The latest generation of the product (“iFuse 3D”) is porous, designed to look like cancellous bone, which essentially means it’s honeycombed and allows bone to grow into it to make it permanent.  Historically, surgeons have not performed many SI joint fusions because they were 1) long, risky, open/invasive in-patient surgeries and 2) SI joint dysfunction is challenging to diagnose. 

From a high level, in 2021, SIBN (~$1bn EV) will do ~$93mm of revenue with ~87% gross margins.  This is ~9,500 procedures at ~$10k each.  SIBN contends they have ~65% market share, implying ~13.5k procedures done annually.  The key investment debate (and the co’s contention) is that SI joint dysfunction is heavily underdiagnosed and that there are ~280k eligible cases for SI joint fusion per year.  SIBN is essentially building a new market, supported by significant clinical data – but this is still heavy lifting and the major risk to our investment thesis.  With ~280k procedures per year and maintaining their 70% market share, at the same $10k cost, the company would do $2bn of revenue or $1.7bn of gross profit.  Now that we’ve set the goal posts (revenue somewhere between $100mm and $2bn…!), we’ll begin with some historical context and then delve into two key debates: 1) Is the market opportunity closer to 10k procedures or 280k procedures and 2) can SIBN maintain their 70% market share, $10k ARPU, and ~90% gross margins for what is essentially three titanium screws? 

Historical Context

While we’ll delve into those, I think it’s helpful to begin with some history.  While SIBN was founded in 2008, a brief discussion of the history of the procedure tells me why I think now is an extremely interesting time to study this business.  When the iFuse was originally launched in 2013, it very quickly ramped to ~$60mm of revenue (nearly today’s levels) and became profitable on only ~$40mm of funding.  Surgeons were using the only existing relevant billing code at the time that dated back to 1985.  The AMA quickly stepped in and said that since this is a minimally invasive procedure, it has to get a new billing code.  With a new code, they suddenly lost insurance coverage and had to begin rebuilding reimbursement. 

Over the next 5 years, the company compiled (and supported) significant clinical evidence in 90 peer-reviewed papers including the only (two) randomized control trials supporting an SI joint fusion device as well as the only prospective clinical studies demonstrating 5-year prospective improved pain, patient function, and quality of life.  You can review them all here.  (https://si-bone.com/providers/clinical-evidence/data).  They are published in prominent, top tier journals and many of the authors advocating for SIBN at major conferences (see North American Spine Society presentation: https://investor.si-bone.com/static-files/d7a06515-36c7-4ba3-b82d-874a169c2eee ) are well regarded in the field (Yale med faculty, hospital CMO, Spinal Surgery division head at Scripps). 

So after 5 years of compiling clinical evidence and working with insurance for reimbursement, BCBS not only began covering the procedure, but covering iFuse (SIBN’s product) exclusively because of the significant body of clinical evidence.  Today, the iFuse is covered by nearly all payers (including Medicare) representing ~312mm covered lives in the US, nearly double from ~160mm in 2017.  37 of these payors, covering 80mm lives, have exclusive coverage for iFuse, which is relatively unique among surgical devices.  This body of evidence would be extremely difficult and time consuming for a competitor to build.  One helpful note of context for generalists like myself – it typically takes 4-5 years to do a 2-year clinical study including writing the protocol, recruiting surgeons, enrolling patients, then following patients for 2 years, then 6-9mo to publish the paper. 

I begin with this history lesson to cover the “why now?” and illustrate how much of the heavy lifting (building clinical evidence, gaining insurance coverage) has been accomplished already, potentially putting us at a compelling inflection point.  In the 5 quarters between SIBN’s IPO and COVID, revenue growth accelerated from +13% to +19% to +21% to +23% to +27%.  Interestingly, since that last period of accelerating growth, SIBN has gained improved reimbursement levels for surgeons and increased coverage from CIGNA, Aetna, and Humana, the benefits of which have been masked by COVID driving deferrals of procedures. 

 Key Debate: Market Size/Adoption  

We can safely agree that lower back pain is a very common problem in the US and very challenging and frustrating to solve.  ~30mm Americans suffer from lower back pain.  Academics and clinicians estimate anywhere from 15% to 30% of back pain is related to the SI joint.  Of those, ~30% are eligible for surgery.  On average, a patient will suffer through 5 years of pain before doing a surgery, resulting in an estimate of ~280k eligible cases per year.  A sanity check against this number is that ~280k patients received multiple steroidal injections into the SI joint (and this is back in 2017) and today there are ~1.2mm therapeutic injections into the SI joint annually.  Another helpful piece of context is that there are ~300k lumbar spinal fusions performed per year and they have only an ~60% success rate and an ~12% revision rate. 

We assume the low end of the 15-30% estimate of back pain due to SI joint to get to the company’s estimated market opportunity.  I fully admit there’s a lot of wiggle room there, so we heavily haircut this number in our valuation sensitivities.  

The lack of adoption stems from a few challenges.  I believe these challenges are the reason this opportunity exists today.  For the most part, spine surgeons did not learn SI joint procedures in their training as historically there was no minimally invasive procedure and the risks of an open procedure were viewed as too high compared to the benefits.  While that is remedied with minimally invasive procedures, the same slow-to-change habits of surgeons that make these devices compelling businesses works against their adoption.  As more and more surgeons (and more patients) see successful pain relief from minimally invasive SI joint fusions, the body of knowledge and awareness will continue to grow.  

The second challenge is diagnosis.  Spine surgeons typically diagnose lower back pain with X-rays, CT scans, MRI’s where they can see a bulge, compression, or a fracture making it clear what needs to be done.  Unfortunately, SI joint issues are not visible on a scan and require a physical exam to diagnose the issue.  This historically has been a big sticking point - spine surgeons are not used to diagnosing via physical exam.  Over time, they have perfected a 2-3 minute exam across 5 tests can typically identify SI joint issues.  If the SI joint is suspected, from there, the surgeon injects ~1-1.5cc of anesthesia into the joint (and ~1.2mm of these injections are currently done annually), and if it relieves pain, it confirms the patient is a good candidate for an SIJ fusion.  

The final challenge is around physician incentives.  Historically, the procedure did not reimburse well for surgeons and was not a good use of time.  However, as of Jan 1, 2020, CMS increased the reimbursement by ~27%.  According to our field work (excerpts below), the new level of ~$1,000 for a 20-40 minute procedure is competitive with the $1500-2000 for a lumbar fusion which takes 1-2 hours.  Once again, I believe this inflection in surgeon incentives just as COVID was beginning has not fully shown up in the results yet.  We’ll include some excerpts on this topic below.  

Key Debate : Competitive/Patent Risks

This was originally my key concern, but I have become materially less concerned here over time.  The first thing to note is that the patents for the core iFuse product expire in November 2024.  However, 95% of the business is now iFuse 3D (which is porous, allowing rapid bone in-growth) with patents lasting through 2035.  

From a competitive perspective, the challenge is not that it’s difficult for someone else to make a triangular, titanium screw.  While none of SIBN’s competitors today are triangular (which drives 6x rotational resistance and 3x the strength), at some point a competitor will launch a triangular screw.  Further down the road, a competitor will launch a porous, triangular screw.  However, the challenge is in providing clinical evidence that it works and should be reimbursed.  To duplicate a 5-year study, would take ~7 years.  Further, surgeons are very hesitant to switch from what they are already used to using.  Surgeons do not bear the cost of the implant, the payers do, and they are now finally on board - often covering iFuse exclusively due to the level of clinical data provided by SIBN.  Surgeons care about revision and error rates - risk of things going wrong.  According to our field work, iFuse is viewed as best in class from an ease of use perspective as well as from success rates.  Again - using the only product supported by significant clinical evidence (even if in theory all of that evidence would also support the competitor products) is the type of “CYA” that surgeons want.  

SIBN competes against both larger companies with vaster resources but less focus on the SI joint as well as small companies that lack SIBN’s public equity funding.  I believe SIBN is well situated as both a public company with reasonable (but not infinite) resources to invest in physician training combined with the focus to innovate and not be distracted.  As mentioned above, it would take many years for any other company to develop the set of clinical evidence SIBN has amassed.  

Inflection: Drivers of Accelerating Growth

So far, we’ve discussed the improvements in payer coverage and physician reimbursement that should drive accelerating 25%+ annual revenue growth.  As a reminder, revenue growth was already accelerating every quarter to >25% between IPO and COVID hitting - this was before improved surgeon reimbursement and broader insurance coverage.  In 2020, SIBN saw the same opportunities and raised ~$70 in an equity offering specifically to invest further to accelerate revenue growth beyond that 25% annual level.  Specifically, they are targeting becoming a $300-500mm revenue company in the medium-term and believe they have a line of sight to it vs consensus of $140mm in 2023.  

They intend to increase their direct sales headcount by ~40% in 2021 to ~90 reps.  During COVID, SIBN developed an interesting “simulator” approach to training, reducing the need for travel and radiation exposure to train surgeons in a cadaver lab.  We have received fairly positive feedback on the simulator training from the surgeons we surveyed.  The combination of ~16 more simulators by the end of the year and the ramp-up of the larger sales force (takes ~1yr to mature), should drive accelerating revenue growth in 2022 and 2023, while consensus assumes a deceleration (27% and 21% respectively).  

One way to think about this ramp is that iFuse procedures have grown from 5,100 in 2017 to 7,500 in 2020.  This occurred as active surgeons grew from ~450 to ~588, indicating that not only did the number of surgeons performing the procedure grow but the number of procedures per active surgeon per year grew from ~11 to 13.  SIBN believes that they have trained ~2,300 of 7,500 relevant spine fellowship trained surgeons.  Of those, 1,600 have performed at least one procedure.  The difference between the 1,600 who have performed one, but only ~588 active for the most part has been around reimbursement, which as we have discussed is now materially improved.  

With 20 simulators now in the field, SIBN believe they could train ~3,000 surgeons per year, including refreshing surgeons who have been trained in the past but not recently active.  Further, according to our survey-work, the average surgeon we polled did ~21 iFuse procedures per year (not surprising it skews above the actual mean if they responded to our survey) but expects to do ~34 per year in 2021.  

Further, SIBN have increased spend on product development and will be launching at least two new products for different sacro-pelvic (deform and trauma) joint issues in 2021.  They have also been testing DTC ad campaigns to drive awareness of the SI joint as a potential explanation for unsolvable lower back pain.  

One key note around the additional products in deform and trauma (Bedrock for deform is already in market) - these are much smaller markets, but they are essentially the strategic “trojan horse” to get in front of academic hospitals and talk about the SI joint.  Building relationships in academic/teaching hospitals is what ultimately cascades down through residency/fellowship training, building not only awareness, but ingraining SIBN’s products.  SIBN is now training in over 50 academic medical centers and have trained more than 200 fellows and residents.  

Surgeon Survey Results Supports iFuse Strengths As Well As Market Size

We polled ~30 spinal surgeons for their feedback on SIJ fusions and iFuse in particular.  This is obviously a small sample size, but we found it helpful.  While some physicians are certainly skeptical of the procedure, the majority were supportive, emphasizing ease of use, good training, acceptable levels of reimbursement, and expecting the procedure to grow within it’s practice.  

We also polled them on their estimate of the annual number of SI joint fusion procedures that should be done and while answers 10,000 to 100,000, the modal response was ~50,000, which is what we use below in our base case.  We also polled them on what percent of back pain they believed was caused by the SI joint with an average of ~17% (vs 15% used by the company in their TAM calculations).  Finally, we asked how many iFuse procedures the surgeons did per month in 2019, 2020, and how many they anticipated in 2021.  They anticipated ~60% growth from 2019 to 2021.  

Below, please see some excerpts we found useful on key positives of iFuse: 

“Porous coated titanium, laser printed lattice structure allowing ingrowth and ongrowth, MIS lateral approach, only system with multiple RCTs demonstrating efficacy in addition to the most peer reviewed publications demonstrating efficacy”

“Mechanical superiority due to shape and due to insertion across the joint”

“Proven clinical results, consistent fusion rates”

“Ease of use.  Reproducibility Safe.  Quick recovery  patients improve”

“Easy to use  Easy to remove  Anatomic  stabilization”

“Porous implant that allows for in-growth.”

“Reliable  Price points acceptable to institution  Good data”

“Ease of insertion and high fusion rates”

“Technology and data and surface technology”

“On contract  Good service  Data”

“Time tested, straight forward system”

“Ease of use   Good company support”

“Fast, easy to learn and perform”

Below, please see some excerpts we found useful on reimbursement: 

“It isn’t as bad as it used to be. Insurance has accepted the diagnosis and treatment by in large.”

“I think it is commensurate with the work being done.”

“Reimbursement levels have lagged regular spine procedures. We have recently petitioned CMS for an appropriate increase in reimbursement commensurate with the difficulty and work load associated with the appropriate evaluation, management and technical skill associated with surgical management is SI joint dysfunction and did see a slight correction, which is an improvement — but still not ideal.”

“Improved coverage but peer to peer still necessary in most case for private insurers. Documentation is laborious.”

“Overall fairly comparable to other orthopedic procedures of similar technical difficulty”

“Current reimbursement is within range for a procedure like this”

“Fair but low. Pay is similar to a micro discectomy and both procedures take approximately the same amount of time”

“Since performing in ASC, reimbursement is very good”

“Fair based on the amount of time it takes to perform procedure”

 “higher than expected for the time spent performing the procedure”

“Reimbursement seems reasonable”

Excerpts on barriers to doing more procedures - consistent with our discussion above around diagnostic difficulty - this is an educational issue, not a need issue

“Diagnosing people with SI joint pathology”

“Having an appropriate diagnosis.”

“Reliable diagnosis, reliable diagnostic injections”

“Referrals and educating referring physicians”

“Insurance authorization, Patient selection”

“Confirming the correct diagnosis and getting insurance carriers to approve the procedure.”

Valuation/Risk-Reward

Let’s preface that this is not an objectively cheap stock on near-term valuation metrics and that I can’t argue that you face limited downside - this has a wide range of outcomes, you can absolutely lose money here, and it should be sized accordingly.  In all of the math presented below, I ascribe no value to any of the new products the company will be launching this year, assume gross margins degrade from ~90% to 85%, market share flat at 65% (although they gained 5 pts of share in 2020) and use an ~10x gross profit exit multiple.  I also assume they burn ~$100mm in cash over the next 3 years before breaking even, which I think is conservative.  In the bear case I also assume market share declines.  

Let’s begin with downside (far right below) - if instead of accelerating 25%+ annual procedure growth, it instead immediately decays to ~5% per year, with procedure growth tapping out at 15,000 per year (5% of SIBN’s view of the TAM), with iFuse ceding some share, I basically come to revenue in 2024 right around their current 2021 guidance of $90mm.  They likely get taken out by a larger cap device company at ~10x gross profit (which would still be accretive given they could remove essentially all of the SG&A as well as try their hand at driving adoption with their sales force).  This implies an ~$24 share price for 30% downside.  

I think a very conservative base case is that SIBN is only able to grow 20% per year through 2024, putting revenue at ~$150mm (vs ~$144mm 2023 consensus).  I assume some gross margin degradation to 85% and at 10x gross profit (similar logic as in the downside case) gets you to an ~$49 stock, for a boring 15% 3-year IRR when considering the risks here.  

Where it gets fun is if adoption actually takes off and does accelerate.  In my upside cases, I use 50,000 annual procedures (the modal response in our survey of spine surgeons for number of cases per year).  This is an ~34% CAGR from 2020 levels.  The company hits $300mm of revenue by 2025, not at all unprecedented given historical ramps from $100mm to $500mm in 3-7 years for peers.  Still at 10x gross profit, with many fast growing peers currently trading at >10x sales, and not accounting for any new product launches, the shares are worth $81 in 4 years for a 28% 4-year IRR.  

Lastly, just for fun and for the sake of really thinking through the true possible range of outcomes - let’s say they get to 140k procedures (50% penetration of TAM) by 2030.  This is an ~27% CAGR, and using all the same prior assumptions, you have an ~$8bn+ company, for a 9-year 24% IRR.  This is obviously a fairly speculative upside - but I want to highlight the skew.  

Sanity Check on the Math Above

We’ve discussed a few ways to model the opportunity here - working top down from the alleged 280k relevant cases per year as well as bottoms-up from the number of surgeons actively performing the procedure (588 as of 2020) compared to the number who have been trained and performed a procedure (1,600) or trained (2,300) or just the total relevant number of fellowship-trained spinal surgeons (7,500).  

To sanity check the moderate case, where I estimate ~20k annual iFuse procedures, that would imply 938 surgeons (vs ~588 in 2020, a 12% annual CAGR,) out of 7,500 relevant surgeons performing 20.8 procedures per year (60% higher, or a 12% CAGR from 2020).  This 60% increase in iFuse procedures per surgeon (~2 per month) foots with the mean response in our survey where we asked surgeons how many iFuse procedures they planned to do in 2021 vs 2019.  On an absolute basis, we likely sampled higher-volume surgeons who chose to respond about iFuse, but the average absolute response was that they intended to do ~34 iFuse procedures per year in 2021 - which makes me feel like our 21 estimate in 2024 is reasonable.  You can apply the same sanity check logic at the bottom of the risk-reward spreadsheet to the upside and downside cases - the downside case assumes active surgeons only grows from 588 to 692 by 2024 and no increase in procedures per surgeon per year.  That incremental ~100 active surgeons between 2020 and 2024 is less than the number added between 2018 and 2020 despite COVID.  

Historical Precedents: Medtech revenue ramps

You can see below that it is not unprecedented for products to inflect from $100mm to $500mm of revenue in 3-7 years.  My “upside” 27% IRR case has revenue going from ~$100mm in 2021 to ~$325mm by 2025.  

Risks

Hopefully I have highlighted a number of the risks along the way.  SIBN is attempting to build a market for a relatively new surgical procedure.  It’s a heavy lift and things can go wrong.  The company is unprofitable on a GAAP basis and burning cash.  

SIBN is building a new market – convincing surgeons that there is an under-diagnosed source of back pain.  Teaching them to diagnose it and do a new procedure.  If I had trained for 4 years of medical school, 6 years of ortho residency, and 2 years of fellowship, it’d be hard to convince me that there’s a joint causing back pain I never learned about too! 

Patent risk.  iFuse patent expires in 2024 as mentioned above.  90% of business is now iFuse 3D, covered through 2035 and surgeons seem unwilling to use an off-brand product without clinical evidence.  

Competitive risk.  iFuse is not a particularly complicated product to make.  SIBN has many deep-pocketed competitors.  As discussed above, I believe SIBN has the right combination of 100% focus on the SI joint and the resources of a public company to retain their dominant  65% (and increasing) market share.  SIBN has spent nearly 10 years building a “moat” based on clinical evidence and payer coverage.  

Management change.  CEO Jeff Dunn who has been CEO for ~13 years is becoming Exec Chairman and will work ~half time.  CFO Laura Francis will become CEO.  Laura seems highly capable - she was CFO of Bruker, an ~$10bn life sciences company as well as CFO of Auxogyn, which is now part of $4bn market cap public company Progyny (PGNY).  Management change always presents risk.  

Catalyst

Ramp of expanded sales force and increased surgeon training simulators driving...

Accelerating revenue growth resulting in upward consensus estimates (likely to also drive multiple expansion)

Launch of two new products in 2021, not included in estimates above 

Strategic actions