Castle Biosciences (CSTL) Q4 2025 earnings review
Core Growth Masked by Medicare Cliff as Profitability Reverses
Castle Biosciences closed 2025 with mixed financial signals. While total revenue was stable, inching up 1% YoY to $87.0 million, this headline number obscures a massive structural shift. Excluding the discontinued IDgenetix and the Medicare-defunded DecisionDx-SCC tests, core revenue actually surged 43%. However, the loss of high-margin SCC revenue has taken a severe toll on the bottom line. The company reversed from a $9.6 million net profit in 24Q4 to a $2.3 million net loss in 25Q4, driven by gross margin compression and soaring SG&A expenses. Management guided for flat FY26 revenue of $340-$350 million, reflecting the ongoing absorption of the SCC revenue crater against the hyper-growth of its newer TissueCypher and AdvanceAD-Tx franchises.
๐ Bull Case
The gastroenterology franchise is in hyper-growth. TissueCypher test volume accelerated dramatically, up 77% YoY in Q4 to 11,803 reports, surpassing the legacy DecisionDx-Melanoma volume for the second consecutive quarter.
The limited launch of AdvanceAD-Tx opens a $33 billion total addressable market. This test addresses a critical need in moderate-to-severe atopic dermatitis, diversifying Castle away from oncology.
๐ป Bear Case
The Novitas Medicare non-coverage decision for DecisionDx-SCC removed a major, high-margin revenue stream. Consequently, the company reversed into unprofitability, posting a $24.2M net loss for FY25 after generating $18.2M in net income in FY24.
Adjusted Gross Margin decelerated from 81.1% in 24Q4 to 77.6% in 25Q4. The mix shift toward the rapidly growing but structurally lower-margin TissueCypher test is permanently altering the company's margin profile.
โ๏ธ Verdict: โช
Neutral. Castle is successfully executing a difficult pivot. The underlying adoption of TissueCypher is spectacular, but the financial realities of losing SCC Medicare coverage mean investors must tolerate compressed margins and a return to net losses while the new pipeline scales.
Key Themes
TissueCypher Surpasses Legacy Flagship
TissueCypher is now the undisputed volume engine of the company. It delivered 11,803 test reports in Q4 (up 77% YoY), firmly overtaking DecisionDx-Melanoma (10,022 reports). This accelerating adoption is driven by a recently expanded ~65-rep GI sales force and strong clinical validation identifying high-risk Barrett's esophagus patients that traditional pathology misses.
AdvanceAD-Tx Commercial Launch
Castle officially initiated the limited access launch of AdvanceAD-Tx in Q4, targeting a massive $33 billion U.S. TAM (approx. 10 million patients aged 12+). The test predicts systemic therapy response (JAKi vs. Th2 biologics). While management previously indicated revenue contribution will be 'immaterial in 2026' as they build reimbursement from ground zero, this is the company's most significant long-term catalyst.
DecisionDx-Melanoma Core Stability
Despite a slight seasonal sequential dip in Q4, DecisionDx-Melanoma maintained a stable growth trajectory, delivering 10,022 reports (+16% YoY). Management previously shifted the dermatology sales force to focus 95-100% on this test after the SCC coverage loss, and it appears to be successfully sustaining mid-to-high single-digit volume growth.
Gross Margin Deterioration
The mix shift from high-margin dermatology GEP tests to gastroenterology is dragging down corporate profitability. Adjusted Gross Margin dropped to 77.6% in 25Q4 from 81.1% a year prior. With TissueCypher acting as the primary growth engine and SCC revenue gone, investors should view the mid-70s as the new normal for gross margins.
Operating Leverage Reversing
While total revenue grew only 1% in Q4 due to the SCC cliff, SG&A expenses surged 13.5% YoY to $56.7 million. The company is funding an expanded GI sales force and a new launch (AdvanceAD-Tx) without the cash-cow benefit of Medicare SCC reimbursement. This caused Adjusted EBITDA to decelerate violently, dropping 46% YoY to $11.5M.
DecisionDx-SCC Overhang Persists
Test volumes for DecisionDx-SCC are decelerating, falling to 3,971 in Q4 from 4,299 a year ago. The company is processing these tests without Medicare coverage while awaiting a reconsideration request from Novitas. The continued processing incurs cost-of-goods without corresponding revenue, directly penalizing the P&L.
Other KPIs
Stable. Despite the GAAP net loss of $2.3M, cash generation remains robust, largely due to $11.4M in stock-based compensation and working capital timing. This proves the core business model remains highly cash-generative even amidst the SCC reimbursement shock.
Accelerating slightly from $287.5M at the end of Q3. The balance sheet is a fortress, providing ample runway to fund the reimbursement slog for AdvanceAD-Tx and potential further tuck-in M&A without needing dilutive capital raises.
Guidance
Decelerating headline growth. The midpoint ($345M) implies just 0.2% YoY growth compared to FY25's $344.2M. However, this includes the base effect of losing a full year of DecisionDx-SCC and IDgenetix revenue in FY26. Therefore, hitting this guidance actually requires substantial acceleration in TissueCypher and DecisionDx-Melanoma to fill the gap.
Key Questions
AdvanceAD-Tx Reimbursement Roadmap
With the limited launch underway, what specific milestones or timelines should investors track regarding commercial and Medicare coverage for AdvanceAD-Tx throughout 2026?
Margin Floor for TissueCypher
As TissueCypher becomes the dominant volume driver, what is the long-term structural margin profile of this test at scale, and when will the blended corporate gross margin bottom out?
DecisionDx-SCC Reconsideration Update
With the Novitas LCD reconsideration requests submitted in mid-2025, has there been any constructive dialogue, and at what point does management stop processing unfunded tests to protect gross margins?
SG&A Trajectory
Given the 13.5% YoY spike in SG&A against flat total revenue in Q4, how is management thinking about operating expense controls for FY26 to prevent further deterioration in Adjusted EBITDA?
