Medicare, Radiology, and the Strange Migration from Film to Digital
##
The CMS-RVU migration from film to digital radiology was not simply a technology upgrade; it became a revealing Medicare payment episode.
Around 2013–2017, CMS and the AMA/RUC translated the disappearance of film, processors, view boxes, and related supplies into the mechanics of practice expense RVUs. Film-era inputs were removed from hundreds of imaging codes, PACS workstations were added, and CMS estimated about $240 million in annual budget-neutral redistribution—small relative to the whole Physician Fee Schedule, but roughly 5% of radiology’s Medicare allowed charges.
The episode had a distinctive mood: inevitability, suspicion, and invoice-driven trench warfare. Radiology accepted that film was obsolete but argued that CMS was underpricing the digital replacement by treating PACS like a desktop computer.
The history is not a perfect parallel to digital pathology or AI, but it is a useful prologue. It shows how Medicare recognizes technology transitions: slowly, mechanically, and only when costs can be named, priced, and mapped to codes.
A previous blog summarized the 3-hour, annual Roche Diagnostics Day.
I asked Chat GPT to go back to the transcript and discuss the news as far as TROP2 CDx coming through FDA using the AZ QCS technology.
As always, take this as an example of the current state of "AI thinking and writing" - not as gospel truth.
###
Roche’s TROP2 / AstraZeneca CDx Strategy
Why Roche is treating this as more than “another IHC companion diagnostic”
1. The short version
Roche is positioning VENTANA TROP2 RxDx, developed with AstraZeneca and incorporating AstraZeneca’s QCS — quantitative continuous scoring — as the first major proof point for a new generation of computational pathology companion diagnostics.
In Roche’s framing, this is not simply a TROP2 IHC assay. It is a regulatory, commercial, and technical template for complex IHC-based CDx in which the final clinical result depends on a tightly integrated system: VENTANA staining, digital slide acquisition, image management, algorithmic scoring, pharma trial linkage, and FDA-reviewed RxDx labeling.
Roche says the TROP2 product is currently launched as RUO, with a CDx launch planned for Q1 2027, and describes it as the first computational pathology RxDx device.
The Deeper Message
The deeper message is that Roche wants to use TROP2 as the wedge product for a much larger AI-enabled CDx franchise. The transcript is explicit: Laura Apitz calls attention to the “TROP2 franchise,” says it will be the “very first computational pathology companion diagnostic” launching in lung cancer next year, and says the RUO TROP2 footprint is being grown globally to set up later companion diagnostic algorithms.
2. What Roche is actually claiming
Roche’s claims fall into four layers.
First, TROP2 is a specific AstraZeneca-linked CDx project. The deck says VENTANA TROP2 RxDx is developed in collaboration with AstraZeneca and incorporates AstraZeneca’s proprietary QCS, or quantitative continuous scoring. The footnote also cautions that PHCS launches depend on the success of the related pharmaceutical trials and drug launches, which is important: this is not a generic assay launch independent of a drug program. It is a drug-diagnostic co-development story.
Second, TROP2 is the first computational pathology RxDx device in Roche’s pipeline. Roche highlights TROP2 as the first computational pathology RxDx device, currently launched as RUO, with CDx launch planned for Q1 2027. It also says the QCS platform can support multiple indications and biomarkers and can fuel future pharma collaborations.
Third, TROP2 is part of a larger digital pathology buildout. Roche ties TROP2 to AISight, PathAI, scanner updates, and digital pathology adoption. AISight Dx is described as a cloud-based image-management solution with a pathologist-centric collaborative interface, designed to be interoperable with scanners, LIS, and PACS systems.
Fourth, TROP2 is a commercial proof point for Roche’s PHCS/CDx flywheel. Roche says it has more than 25 years of experience in personalized healthcare solutions, more than 85 pharma partners, more than 150 ongoing IVD programs, over 25 digital pathology IVD development projects, seven biomarkers, and 13 indications. That scale is central to the argument: TROP2 is not a one-off; it is the first visible example of a pipeline.
3. Why TROP2 is more complex than ordinary IHC
Traditional IHC CDx often reduces to a standardized stain, a pathologist-read score, and a cutoff embedded in a drug label. That model is already complicated — specimen handling, fixation, clone, platform, staining conditions, reader training, and cutoffs all matter. But Roche is implying that TROP2 moves beyond that classic model.
The phrase quantitative continuous scoring is the giveaway. Instead of a relatively simple binary or ordinal manual interpretation, QCS suggests a more granular quantitative or semi-quantitative readout, presumably linking TROP2 expression patterns to drug eligibility or treatment-response enrichment. Roche does not fully disclose the scoring architecture in the investor deck, but it makes clear that AstraZeneca’s QCS is proprietary and incorporated into the VENTANA TROP2 RxDx device.
That makes this a complex IHC CDx in several senses:
Complexity layer
Why it matters
Biology
TROP2 expression may not behave like a simple mutation-positive / mutation-negative marker.
Analytical pathology
IHC intensity, distribution, tumor heterogeneity, and compartment recognition may all matter.
Scoring
Quantitative continuous scoring implies algorithmic or computational assistance rather than a simple manual category.
Device integration
The final result may depend on stain, scanner, image management, and software.
Regulatory review
FDA must evaluate a drug-linked diagnostic system, not merely an antibody reagent.
Commercial deployment
Labs need validated digital workflows and possibly scanner/software compatibility.
This is why Roche repeatedly links TROP2 to digital pathology, not merely to advanced staining. In the transcript, Apitz discusses new multiplexing and computational pathology algorithms, then immediately pivots to the TROP2 franchise as the first computational pathology CDx in lung cancer.
4. Why AstraZeneca matters
AstraZeneca’s role is not incidental. Roche is not saying, “Here is a Roche TROP2 stain that might be useful for many drugs.” It is saying the VENTANA TROP2 RxDx device was developed with AstraZeneca and incorporates AstraZeneca’s proprietary QCS.
That matters because pharma companies increasingly need biomarkers that do more than identify a target’s presence. For antibody-drug conjugates, immuno-oncology combinations, and complex targeted therapies, a simple “marker present” assay may be inadequate. Pharma may want richer tissue information: expression intensity, distribution, tumor-cell localization, microenvironment context, heterogeneity, or a continuous score tied to response.
So the TROP2 / AstraZeneca project is a signal to other pharma companies: Roche can take your complex tissue biomarker and turn it into a regulated, globally deployable RxDx. Roche is not merely offering a stain. It is offering a development partner that can combine VENTANA IHC, digital pathology, algorithmic scoring, global lab reach, and FDA-facing CDx experience.
This is the “partner pharma” message. Roche is marketing itself as the place where pharma can bring complex tissue biomarkers that are too hard for simple manual IHC and need computational pathology to become clinically usable.
5. Why Roche calls this a franchise
The phrase “TROP2 franchise” is significant. It implies that Roche sees TROP2 not just as one assay for one drug, but as a platformable class of work.
In the transcript, Apitz says Roche is growing the RUO TROP2 footprint globally now, which “sets us up really nicely” for the companion diagnostic algorithms. This will allow experience and help prepare the laboratories before the regulated CDx launch. RUO deployment can support familiarity, pharma development, workflow testing, institutional readiness, and potentially evidence generation, while the formal CDx path proceeds.
The deck reinforces this by saying QCS is a platform for multiple indications and biomarkers and that TROP2 can fuel a pipeline of future pharma collaborations.
That is the franchise logic:
TROP2 RUO footprint → pharma trial use → computational scoring validation → FDA RxDx launch → broader QCS / algorithmic CDx platform → more pharma collaborations.
6. Why FDA significance is high
The regulatory significance is that Roche is describing this as the first computational pathology RxDx device and elsewhere as an FDA Breakthrough Device Designation for the first AI-based CDx in oncology. The deck’s AI timeline specifically notes “VENTANA TROP2: FDA BDD for 1st AI-based CDx in oncology.”
That positions TROP2 as a test case for FDA review of a new CDx category: not just IHC, not just image analysis, not just software, and not just a digital pathology workflow, but an integrated AI-enabled or computational pathology companion diagnostic linked to a therapy.
For FDA, the hard questions likely include:
FDA issue
Why TROP2 may matter
Algorithm performance
How accurate and reproducible is the QCS output?
Clinical validity
Does the score identify patients who benefit from the AstraZeneca therapy?
Analytical reproducibility
Does the result hold across sites, tissue handling, instruments, scanners, and operators?
Human factors
How does the pathologist interact with the algorithmic output?
Device boundaries
What exactly is the regulated device: antibody, stain, scanner, software, algorithm, IMS, or all of the above?
Change management
How are future software, scanner, algorithm, and scoring updates handled?
Labeling
How will the drug label and diagnostic label specify use, cutoff, and eligible population?
This is why TROP2 is a bellwether. If Roche can get FDA comfortable with this integrated model, it creates a precedent for additional computational IHC CDx devices.
If it struggles, the entire AI-enabled CDx category may move more slowly.
7. How PathAI changes the story
PathAI is strategically important because it gives Roche more of the missing software and AI layer. Roche already had VENTANA staining, scanners, and pharma CDx relationships. PathAI adds image management, AI-driven analysis, workflow capabilities, clinical-trial services, translational research, and biomarker discovery. Roche describes PathAI as complementary to Diagnostics and synergistic with Pharma, including workflow enhancement, regulated AI-enabled CDx, translational research, biomarker discovery, and drug-target discovery.
In the transcript, Andy Beck goes further. He says AI-powered CDx is expected to accompany most new approvals by 2035, that AI may be run on every slide, and that massive real-world datasets from millions of slides could drive R&D, regulatory, and policy decisions. He also describes AISight as a best-in-class digital pathology image-management system that received CE/IVDR and FDA clearance and a predetermined change control plan to help expand the label and add new scanners and monitors.
That matters for TROP2 because the product is not just an antibody. Roche is trying to create the end-to-end environment in which algorithmic CDx can be developed, deployed, monitored, and scaled.
8. The PHCS/CDx flywheel
Roche’s pathology strategy is built around a flywheel. The deck says Roche has a pathology menu with more than 250 ready-to-use IHC/ISH assays and personalized healthcare services involving more than 85 pharma partnerships and more than 150 ongoing IVD programs. It then describes pharma CDx partnerships as a flywheel: partnerships accelerate menu growth, menu growth drives instrument placements, placements increase access to Roche’s diagnostic menu, and increased access makes Roche more attractive for CDx development.
TROP2 fits perfectly into that flywheel. A high-profile AstraZeneca-linked computational CDx can do several things at once:
Strengthen Roche’s pharma-partner credibility. Other pharma companies can see Roche as a partner for complex tissue biomarkers.
Increase demand for VENTANA digital pathology infrastructure. If the CDx requires digital scoring, labs need compatible scanners, image management, software, and validated workflows.
Drive IHC/ISH menu differentiation. Roche can argue its pathology menu is moving beyond conventional stains toward AI-enabled therapeutic decision tools.
Create pull-through for PathAI / AISight. TROP2 gives a concrete reason for labs and pharma partners to adopt the computational layer.
Create a repeatable FDA pathway. If TROP2 succeeds, it becomes the model for future AI-enabled RxDx products.
9. Why this could be the first of many “complex IHC” CDx products
Roche also discusses translucent chromogens and multiplexing, including P40/TTF1 in lung cancer and SOX10/Ki-67 in melanoma / lymph node assessment. Apitz describes these as the first of many multiplexes Roche plans to bring out in companion diagnostic, companion digital pathology, and non-gynecologic cytology spaces.
This is important. TROP2 is not isolated from the multiplexing roadmap. The general direction is toward IHC that becomes more information-rich:
single marker → multiplex marker → quantitative score → digital pathology algorithm → companion diagnostic → pharma-linked treatment selection.
That path is a major shift for IHC. The historical strength of IHC is that it is cheap, familiar, visual, and installed everywhere. The historical weakness is that interpretation can be subjective, semi-quantitative, and hard to standardize across sites. Roche is trying to preserve the installed-base advantage of IHC while adding algorithmic reproducibility and pharma-grade scoring.
10. An interpretation
Roche is using TROP2 as a category-creating product. The immediate commercial story is an AstraZeneca-linked lung cancer companion diagnostic. The larger strategic story is that Roche wants to own the regulated infrastructure for computational IHC companion diagnostics.
That infrastructure includes:
VENTANA stain + BenchMark instruments + DP scanners + AISight image management + PathAI algorithms + pharma CDx development + FDA regulatory pathway + global pathology installed base.
If successful, TROP2 becomes the proof that Roche can take a complex tissue biomarker, convert it into a quantitative computational pathology result, support a drug approval, and deploy the diagnostic globally. That would be a major competitive advantage versus companies that own only the antibody, only the scanner, only the image-management software, or only the algorithm.
The risk is that this is hard. FDA will need to be comfortable with the device boundaries, scoring method, clinical cutoff, software updates, scanner dependencies, and pathologist role. Labs will need digital pathology infrastructure. Pharma trials must succeed. The clinical label must be meaningful. And the algorithm must be robust across real-world tissue variability.
But Roche’s message is clear: TROP2 is the opening act for AI-enabled CDx IHC. If TROP2 lands, it will not just be a successful TROP2 assay. It will be the first credible example of Roche’s next-generation pathology strategy: complex IHC made quantitative, digital, algorithmic, pharma-linked, and FDA-reviewable.
I gave the materials separately to Chat GPT and Opus 4.7, which provided 13p and 24p reports. I asked Chat GPT to produce a single combined report with a cover page (15pp).
Remember, this isn't an example of gospel-truth, it's an example of the current state of AI reading and writing.
To Fix Software and AI Reimbursement - Do We Need to Understand the RUC?
Maybe.
For at least five years, CMS has had difficulty pricing software-intensive services like AI. In 2025, CMS included an RFI for "public comment" on these problems in both the July 2025 Physician (Part B) rulemaking and the Hospital Outpatient rulemaking.
One Damn Thing After Another
One case study was the great difficulty CMS had when in pricing "Heartflow" software in both the OPPS and Part B settings.
Another case study was the difficulty in pricing a photographic device that gives retinal diagnostics. Processed through the RUC, the AMA valuation committee gave 92229 a few dollars for technician time ($9), seventy cents for capital equipment ($14,000 device x 13 minutes), and a software fee.
CMS declined the software fee as a payable item, leaving the $40 service valued at about $10. CMS arbitrarily assigned an RVU value paying $46. This dates back to 2019/2020.
On the Clin Lab Fee Schedule side, things used to work, then broke. AMA issued and CMS accepted around 10 codes for complex whole slide imaging (WSI) tests, and generally priced at $700. Then, AMA stopped issued WSI PLA codes, so nothing happened for a couple years.
Amidst many AMA initiatives around AI policy, AMA is now revising its Appendix S, and has begun putting new WSI tests into "Category III." But we don't know if CMS will include these Category III lab test codes in its summer lab crosswalk/gapfill process. I think they can and should, since CMS has classed WSI tests as CLFS tests, even when they are closely reviewed as ADLTs. And FDA and CLIA and, say, New York State, classifies computatioinal pathology tests as CLIA tests. So they should go into the summer CLFS crosswalk/gapfill process. File that under "track this."
All About the RUC
In 2016, Miriam Laugesen of Columbia published a wonderful book about the AMA RUC process, "Fixing Medical Prices," (Harvard University Press.)
It turns out there are a lot of RUC documents, some intricately detailed, on the AMA website.
For example, there is an encyclopedic 27 page article about practice expense components - which will be relevant, if only for contrast, to future software valuation. Find it here:
At the bottom of the long sidebar, you can get files for RUC decisions and meeting minutes going back to 1993 (!!!). The AMA RBRVS subscription database (about $300) lists the last RUC review by date for every CPT code, and with that date, you can get the original files at AMA. And you can also track the most recent debates for how the RUC is handling new Category I codes with software. Currently as recent as...May 2026!
We often hear there are 1000 FDA software/AI devices or SaMD for radiology, and only a few for pathology.
Let's look at pathology, and see what the FDA regulatory publications are. We also look at the sometimes-confusing delays and timelines in the appearance of documents.
Classification Letter and "Safety and Effectiveness" Review
FDA issues a clearance or classification letter, reflecting the date of clearance and documenting the date of submission. FDA lists the regulatory category (e.g. 864.3755).
The above letter enables the company to issue a press release about its new FDA product.
At a delay of 2-12 months, the FDA publishes a "Safety and Effectiveness" review, a technology assessment, often circa 20 pages long.
Regulatory Category and Product Code Quick; Federal Register "Someday"
For a new de novo category (which will be a future producdt 510(k) category), FDA assigns a name of the product category and a CFR category citation (such as 21 CFR 864.3750 and 864.3755).
However, that CFR is a placeholder. It make take 1-3 years for FDA to publish a several-page announcement in the Federal Register that the regulatory category is officially created.
At that point, the regulatory category will appear in the eCFR. Some regulatory categories in the eCFR are relatively short and some run several detailed paragraphs long.
At the time of the classification letter, FDA will also have assigned a Product Category (a 3-letter code, such as QPN).
Remember; The regulatory code (864.3755) and the product code (QPN) will be issued right away. It's the Federal Register publication, which actually puts that number into the eCFR, that can lag a couple years.
Software algorithm device to assist users in digital pathology
A software algorithm device to assist users in digital pathology is an in vitro diagnostic device intended to evaluate acquired scanned pathology whole slide images. The device uses software algorithms to provide information to the user about presence, location, and characteristics of areas of the image with clinical implications. Information from this device is intended to assist the user in determining a pathology diagnosis.
SFH - Artera Prostate
Pathology software algorithm device analyzing digital images for cancer prognosis
A pathology software algorithm device analyzing digital images for cancer prognosis is a software intended to analyze scanned whole slide images (WSIs) from prostate biopsies prepared from formalin fixed paraffin-embedded (FFPE) tissue and stained using Hematoxylin & Eosin (H&E) stains. The device provides prognostic risk estimates which are intended to assist physicians with prognostic risk-based decisions along with other clinicopathological factors. The device is not intended to determine a clinical diagnosis.
SHW - Artera Breast
Pathology software algorithm device analyzing digital images for breast cancer prognosis
A pathology software algorithm device analyzing digital images for cancer prognosis is a software intended to analyze scanned whole slide images (WSIs) from breast cancer specimens prepared from formalin fixed paraffin-embedded (FFPE) tissue and stained using Hematoxylin & Eosin (H&E) stains. The device provides prognostic risk estimates which are intended to assist physicians with prognostic risk-based decisions along with other clinicopathological factors. The device is not intended to determine a clinical diagnosis.
21 CFR 864.3750
Paige Prostate falls under 864.3750. Artera Prognostic Prostate and Prognostic Breast fall under 3755, which is a reserved number with no text.
864.3750 is lengthy.
21 CFR 864.3750 Software algorithm device to assist users in digital pathology.
(a)Identification. A software algorithm device to assist users in digital pathology is an in vitro diagnostic device intended to evaluate acquired scanned pathology whole slide images. The device uses software algorithms to provide information to the user about presence, location, and characteristics of areas of the image with clinical implications. Information from this device is intended to assist the user in determining a pathology diagnosis.
(b)Classification. Class II (special controls). The special controls for this device are:
(1) The intended use on the device's label and labeling required under § 809.10 of this chapter must include:
(i) Specimen type;
(ii)Information on the device input(s) (e.g., scanned whole slide images (WSI), etc.);
(iii)Information on the device output(s) (e.g., format of the information provided by the device to the user that can be used to evaluate the WSI, etc.);
(iv)Intended users;
(v) Necessary input/output devices (e.g., WSI scanners, viewing software, etc.);
(vi)A limiting statement that addresses use of the device as an adjunct; and
(vii)A limiting statement that users should use the device in conjunction with complete standard of care evaluation of the WSI.
(i) A detailed description of the device, including the following:
(A) Detailed descriptions of the software device, including the detection/analysis algorithm, software design architecture, interaction with input/output devices, and necessary third-party software;
(B) Detailed descriptions of the intended user(s) and recommended training for safe use of the device; and
(C) Clear instructions about how to resolve device-related issues (e.g., cybersecurity or device malfunction issues).
(ii)A detailed summary of the performance testing, including test methods, dataset characteristics, results, and a summary of sub-analyses on case distributions stratified by relevant confounders, such as anatomical characteristics, patient demographics, medical history, user experience, and scanning equipment, as applicable.
(iii) Limiting statements that indicate:
(A) A description of situations in which the device may fail or may not operate at its expected performance level (e.g., poor image quality or for certain subpopulations), including any limitations in the dataset used to train, test, and tune the algorithm during device development;
(B) The data acquired using the device should only be interpreted by the types of users indicated in the intended use statement; and
(C) Qualified users should employ appropriate procedures and safeguards (e.g., quality control measures, etc.) to assure the validity of the interpretation of images obtained using this device.
(3)Design verification and validation must include:
(i) A detailed description of the device software, including its algorithm and its development, that includes a description of any datasets used to train, tune, or test the software algorithm. This detailed description of the device software must include:
(A) A detailed description of the technical performance assessment study protocols (e.g., regions of interest (ROI) localization study) and results used to assess the device output(s) (e.g., image overlays, image heatmaps, etc.);
(B)The training dataset must include cases representing different pre-analytical variables representative of the conditions likely to be encountered when used as intended (e.g., fixation type and time, histology slide processing techniques, challenging diagnostic cases, multiple sites, patient demographics, etc.);
(C)The number of WSI in an independent validation dataset must be appropriate to demonstrate device accuracy in detecting and localizing ROIs on scanned WSI, and must include subsets clinically relevant to the intended use of the device;
(D)Emergency recovery/backup functions, which must be included in the device design;
(E) System level architecture diagram with a matrix to depict the communication endpoints, communication protocols, and security protections for the device and its supportive systems, including any products or services that are included in the communication pathway; and
(F)A risk management plan, including a justification of how the cybersecurity vulnerabilities of third-party software and services are reduced by the device's risk management mitigations in order to address cybersecurity risks associated with key device functionality (such as loss of image, altered metadata, corrupted image data, degraded image quality, etc.). The risk management plan must also include how the device will be maintained on its intended platform (e.g. a general purpose computing platform, virtual machine, middleware, cloud-based computing services, medical device hardware, etc.), which includes how the software integrity will be maintained, how the software will be authenticated on the platform, how any reliance on the platform will be managed in order to facilitate implementation of cybersecurity controls (such as user authentication, communication encryption and authentication, etc.), and how the device will be protected when the underlying platform is not updated, such that the specific risks of the device are addressed (such as loss of image, altered metadata, corrupted image data, degraded image quality, etc.).
(ii) Data demonstrating acceptable, as determined by FDA, analytical device performance, by conducting analytical studies. For each analytical study, relevant details must be documented (e.g., the origin of the study slides and images, reader/annotator qualifications, method of annotation, location of the study site(s), challenging diagnoses, etc.). The analytical studies must include:
(A)Bench testing or technical testing to assess device output, such as localization of ROIs within a pre-specified threshold. Samples must be representative of the entire spectrum of challenging cases likely to be encountered when the device is used as intended; and
(B) Data from a precision study that demonstrates device performance when used with multiple input devices (e.g., WSI scanners) to assess total variability across operators, within-scanner, between-scanner and between-site, using clinical specimens with defined, clinically relevant, and challenging characteristics likely to be encountered when the device is used as intended. Samples must be representative of the entire spectrum of challenging cases likely to be encountered when the device is used as intended. Precision, including performance of the device and reproducibility, must be assessed by agreement between replicates.
(iii) Data demonstrating acceptable, as determined by FDA, clinical validation must be demonstrated by conducting studies with clinical specimens. For each clinical study, relevant details must be documented (e.g., the origin of the study slides and images, reader/annotator qualifications, method of annotation, location of the study site(s) (on-site/remote), challenging diagnoses, etc.). The studies must include:
(A) A study demonstrating the performance by the intended users with and without the software device (e.g., unassisted and device-assisted reading of scanned WSI of pathology slides). The study dataset must contain sufficient numbers of cases from relevant cohorts that are representative of the scope of patients likely to be encountered given the intended use of the device (e.g., subsets defined by clinically relevant confounders, challenging diagnoses, subsets with potential biopsy appearance modifiers, concomitant diseases, and subsets defined by image scanning characteristics, etc.) such that the performance estimates and confidence intervals for these individual subsets can be characterized. The performance assessment must be based on appropriate diagnostic accuracy measures (e.g., sensitivity, specificity, predictive value, diagnostic likelihood ratio, etc.).
###
864.3755: For Cancer Prognostics; Pending since Paige Prostate in 2025...
Announced last summer, CMS has now instituted the WISeR program to force prior authorization into Medicare Part B using outside venders and their software.
There's a new piece about it in JAMA Internal Medicine currently. Kannarkat et al. They note,
CMS puts forth a thoughtful effort through WISeR to reduce health care waste....With WISeR, AI tools prescreen PA documentation for completeness and match against Medicare coverage rules before human review,
Well, wait a minute. Does anyone really read what CMS has done? The procedures include some with no CMS rules (for erection therapy, an old NCD says, basically, "Impotence therapy is covered when it is medically necessary.") That AI Prior Auth software may well implement ten or twenty rules and rejection triggers, but they're not CMS rules, 'cause there aren't any.
Even more alarming is the inclusion of deep brain stimulation in Parkinson's. CMS's NCD for DBS is approaching 25 years old - 2003 - and is grossly outdated. See a Chat GPT assessment of this policy mess, below. (See also an Opus 4.7 assessment).
Nobody seems to have even noticed these obvious problems in the months of large teams working on WISeR.
(Admittedly, CMS has delayed implementation of DBS in WISeR, but it took nine months to do so, and the reason doesn't seem to have anything to do with the outdated NCD.)
Chat GPT on DBS Outdated Rules
It's a mess. The CMS NCD is strikingly frozen in 2003 language. It still covers the basic durable points — idiopathic PD, levodopa responsiveness, STN/GPi targets, movement-disorders involvement, and exclusion of atypical parkinsonism — but a modern DBS center would think about DBS much differently than the NCD reads. CMS itself still lists NCD 160.24 as Version 1, effective April 1, 2003, with no ending effective date. (Centers for Medicare & Medicaid Services)
Header: For years, one of the Grand Questions in genomics has been "in house versus send out." I asked Claude Opus 4.7 to write a book chapter on how the pro's and con's currently align.
As always with "AI Guest Column," the point is not the "answer" but to show us a checkpoint on how AI currently researches data and assembles an argument.
####
Will new landmark events, like the approval of the 500-gene ILMN TSO PMA platform at FDA, shape the insource-outsource questions? I asked Claude Opus 4.7, which came back with a 20 page computer-generated book chapter on the question.
Precision oncology has advanced rapidly because genomic
testing has made cancer treatment increasingly targetable, measurable, and
data-driven. Yet the business landscape remains hard to predict: FDA approval,
strong reimbursement, and technical performance do not automatically translate
into broad adoption.
This AI-generated report (Opus 4.7) uses Illumina’s TruSight Oncology Comprehensive
assay as a test case for a larger industry question: whether distributed,
platform-based IVD kits — now reaching the scale of 500-gene comprehensive
genomic profiling — can shift oncology testing away from national reference
laboratories and back toward hospital and academic medical centers.
The analysis reviews the current market, the strategic
appeal of in-house CGP, and the barriers that may constrain adoption, including
local validation, uneven case volumes, oncologist preferences, EHR integration,
and the broader data ecosystems built by Foundation Medicine, Tempus, Caris,
and similar firms.
This report is intended to aid both investors and
policy-makers.It provides the reader
with both a market-facing review and a demonstration of how AI-generated
analysis can help frame the present state of a fast-moving field and explore
plausible paths for its evolution.
Header: Digital pathology at warp speed - Tempus acquires Paige, Roche aquires PathAI, ArteraAI gets a true FDA application (Breast cancer prognostics). What should ArteraAI do next? I asked Claude Opus 4.7 to write a guest column.
###
1. Overview
2. ArteraAI Options Map
3. What a CEO Thinks at Night
4. Using High-Low Scenario Modeling
####
1. Overview
Three AI Pathology Companies — Acquisitions, ArteraAI's Story, and Exit Landscape
Recent AI pathology acquisitions
Tempus → Paige (Aug 2025). Tempus AI acquired Paige, an AI digital pathology company, for $81.25 million, paid predominantly in Tempus common stock plus assumption of Paige's remaining Microsoft Azure commitment. Paige brought a dataset of nearly 7 million digitized pathology slides licensed from Memorial Sloan Kettering, plus a foundation model trained on more than 2.3 million whole-slide images. Tempus then launched Paige Predict in January 2026, an H&E-based biomarker prediction suite covering 123 biomarkers across 16 cancer types. Tempus AI + 2
Roche → PathAI (May 7, 2026). Roche agreed to acquire PathAI for up to $1.05 billion: $750M upfront plus up to $300M in milestone payments, with closing expected in the second half of 2026. The deal builds on a partnership established in 2021 and expanded in 2024 for AI-enabled companion diagnostic algorithms; PathAI will fold into Roche's Diagnostics division. Roche specifically wanted PathAI's Image Management System (IMS) and intends to scale it globally, pairing it with Roche's companion-diagnostics franchise. MedTech Dive + 2
The price gap is striking — roughly 13x — and reflects what each buyer was actually paying for. Tempus got a data set and a team that complement an existing oncology platform; Roche got a productized lab IT layer (AISight IMS) plus five years of co-developed companion diagnostics for its drug pipeline.
ArteraAI — origins, work, goals
(ArteraAI is a Los Altos prostate cancer company. There's a separate, unrelated Santa Barbara company called Artera.IO doing patient communications that recently raised $65M.)
Origins. ArteraAI was founded in 2021 with the goal of using AI to globally personalize medical decisions and improve outcomes for cancer patients, built on the belief that histopathology images contain signals that traditional gene-expression tests miss. CEO and co-founder Andre Esteva is well known in medical AI (lead author of the 2017 Nature dermatology-AI paper at Stanford). The company emerged from stealth in March 2023 with $90 million in funding from Coatue, Johnson & Johnson Innovation, Marc Benioff and others. A further $20M followed in February 2024, bringing total raised to roughly $110M. Amazon Web Services + 2
The work. Artera's multimodal AI (MMAI) platform combines digital biopsy images with clinical variables; the algorithm was developed using thousands of patients and tens of thousands of pathology slides, clinically validated across multiple Phase 3 randomized trials. The flagship ArteraAI Prostate Test estimates 10-year risk of distant metastasis and prostate-cancer-specific mortality and predicts hormone-therapy benefit. The SaMD version of their prostate test received FDA De Novo authorization on July 31, 2025. ArteraAI See 24 page FDA review.
Goals. Beyond prostate, the long-term goal is a pan-tumor foundation model that can assess risk and therapy benefit across any cancer sample, starting with breast cancer. Amazon Web Services
Breast cancer — SABCS 2025 abstract data
Artera presented three abstracts at the San Antonio Breast Cancer Symposium (SABCS), December 9–12, 2025. The studies leveraged data from four independent Phase III trials across Germany, Austria, and North America, validating the MMAI model across more than 7,000 patients with HR+ early breast cancer. 01net
The headline finding was from the NSABP B-20 analysis: among patients aged 50 and older, MMAI high-risk individuals experienced a 52% relative reduction in 10-year distant metastasis with chemotherapy, while MMAI low-risk patients derived no additional benefit. This is the predictive (not just prognostic) claim — the same kind of "who actually benefits from treatment" question that made Oncotype DX a multi-billion-dollar franchise in breast cancer. Morningstar
A separate development abstract used data from over 12,000 patients enrolled in six Phase III trials in the US, Germany, and Austria to build the MMAI model for predicting distant metastasis in HR+ early-stage invasive breast cancer. BioSpace
The breast test currently started a laboratory-developed test (the AMA Cat III code application for "AI prognosis of HR+/HER2- breast cancer" was pending as of Apr 30–May 2). The SABCS data package looks deliberately constructed to support both an FDA filing and the NCCN/payer evidence bar for commercial coverage.
###
###
2. Capital and Exit Options.
ArteraAI: Fundraising, Development, and Exit Options
Where Artera stands today
ArteraAI is now a meaningfully different company than it was a year ago. As of May 2026, it has:
Two FDA authorizations: ArteraAI Prostate (De Novo, July 2025) and ArteraAI Breast (510(k), May 6, 2026) — the first and only FDA-cleared digital pathology-based risk stratification tools in either indication.
European regulatory access: CE marks under EU IVDR for both the prostate biopsy assay and breast cancer assay (April 2026).
A Predetermined Change Control Plan (PCCP) allowing iterative model updates without new submissions.
NCCN guideline inclusion for prostate (since early 2024), with breast inclusion in NCCN as the logical next push.
Medicare payment [pricing] on claims for the prostate LDT, with CPT code 0376U effective since April 2023.
Commercial coverage by 73 health plans covering 70.1M lives (as of March 2026), including Anthem and Concert Genetics.
Approximately $110M raised across seed and follow-on rounds from Coatue, Johnson & Johnson Innovation, Marc Benioff, Prosperity7 Ventures and others, with the last disclosed round being a $20M extension in February 2024.
The strategic picture: Artera now has a regulated multi-cancer AI pathology platform, not just a single product. That distinction drives valuation.
Why the timing is unusually favorable
Three external developments have re-set the market for an asset like Artera:
Roche / PathAI (May 2026) — up to $1.05B ($750M upfront + $300M milestones) for a company that did not have an FDA-authorized SaMD prognostic device or NCCN-recommended status. This is the most directly relevant comp.
Tempus / Paige (August 2025) — $81.25M, mostly stock, for a data-and-team asset. Smaller, but it removed one of Artera's most visible competitors and demonstrated public-market acquirers are active.
Consolidation pressure — with two of the four most-recognized AI pathology names now inside large strategics, the remaining standalone universe is small, which makes scarcity work in Artera's favor.
Path 1: Late-stage growth round
The most straightforward near-term option. Every diagnostics-investor underwriting checkbox has now been ticked: FDA authorization (twice), CE marking, guideline inclusion, Medicare payment, commercial coverage, PCCP, and a second indication validated in Phase 3 data.
Likely size: $75–150M Series C/D range, depending on dilution tolerance.
Likely investors: existing investors plus crossover funds (T. Rowe, Fidelity, RA Capital, Perceptive) that typically come in pre-IPO for diagnostics.
Valuation argument: PathAI comp + dual FDA + reimbursement traction supports a meaningful step-up from the last round.
Use of proceeds: breast commercial launch, CPT coding and payer coverage for breast, international expansion under CE mark, and continued platform expansion to additional tumor types.
Risk: a growth round preserves optionality but doesn't lock in value at today's elevated comp environment.
Path 2: Strategic acquisition
The exit pathway most directly supported by the Roche/PathAI comp. Logical acquirers fall into three buckets:
Diagnostics incumbents include
Exact Sciences — Oncotype DX is the direct target of ArteraAI Breast. Same-day, image-only, in-lab results competing with a send-out genomic test is exactly the disruption Exact would want to own rather than fight.
Veracyte — already competes with Artera SaMD in prostate guidelines via Decipher; offensive or defensive logic both apply.
Agendia (MammaPrint) — smaller, threatened, less likely as buyer than as competitor.
Myriad Genetics, Natera — adjacent diagnostics platforms might consider for pathology AI exposure.
Pharma/diagnostics with companion-diagnostic strategy
Roche — just spent $1.05B on PathAI, but Artera's FDA-cleared prognostic SaMD with guideline inclusion is a different asset class. Possible, though probably not in the next 12 months.
Scanner and stack companies (Fujitsu, Danaher, others) — scanner and lab-IT footprint, no flagship AI pathology asset.
Tempus — already moved with Paige, but an FDA-authorized SaMD with NCCN inclusion is the kind of tuck-in they buy.
Likely deal structure: upfront + milestones, similar to PathAI. A reasonable range — using PathAI as the upper bound and adjusting for revenue scale — would be $600M–$1.2B total deal value, with substantial milestone weighting.
Path 3: IPO
Plausible but probably not the first move.
The Tempus 2024 IPO and the broader diagnostics IPO window have improved, but Artera's revenue scale is likely still below where public investors comfortably underwrite a diagnostics name.
Better positioned 18–24 months out, after breast claims revenue ramps and coverage broadens.
IPO valuation discipline tends to undervalue platform stories versus strategic buyers who can quantify cross-sell and synergy.
Path 4: Stay private and build
The "do nothing" option, which is rarely actually nothing.
Each subsequent milestone — breast CPT code, breast NCCN inclusion, breast payer coverage, third indication clearance — adds enterprise value.
The risk: pathology AI is consolidating in real time around 2–3 large platforms. A standalone with great technology can still get squeezed on distribution, hospital purchasing committees, and pharma partnership access.
Best suited if leadership has a specific milestone (e.g., $50M+ ARR or a third FDA clearance) at which they believe a markedly better exit becomes available.
What to watch over the next 6–12 months
Breast reimbursement milestones: AMA Cat III code decision for HR+/HER2-, CMS pricing, first commercial payer policies.
NCCN breast guideline inclusion: the single biggest catalyst for breast adoption and payer coverage.
A third indication: bladder, kidney, or another tumor type would convert "two-product company" to "true platform."
Strategic process signals: banker hires, named CFO with public-company experience, IPO-track audit firm engagement.
Pharma partnership announcements: companion-diagnostic deals with large oncology pharma would both generate non-dilutive revenue and signal to acquirers.
Recommended framing
The strongest argument Artera can make right now — to investors or acquirers — is that it is the only company in AI pathology with FDA authorizations in two indications, a regulated change-control mechanism, an established reimbursement footprint in one indication, and a clinically-validated pipeline behind it. That combination did not exist anywhere in the market six months ago, and the recent strategic activity has made it scarce and valuable simultaneously. Whichever path is chosen, the negotiating leverage is at a high-water mark.
###
###
3. Night Thoughts of a CEO
CEO/Chair Memo — Strategic Reflection, Mid-2026
Note: Artera has an actual CEO, but this is a FICTIONAL CEO created by Opus 4.7.
Personal working notes. Not for distribution.
What I'm thinking
We are in the best position this company has ever been in, and probably the best position we will be in for the next 18 months unless we execute nearly perfectly. That asymmetry is the thing I keep coming back to.
Six months ago we were a one-product company with a great FDA story and a real but narrow commercial footprint in prostate. Today we are — credibly, not aspirationally — a regulated multi-cancer AI pathology platform. The market just told us what that's worth: Roche paid up to $1.05B for a company that didn't have what we now have. That comp will not stay fresh forever. Strategic memory in M&A is about 12–18 months; after that, boards anchor to whatever the next deal prints, and the next deal might be smaller.
At the same time, I am acutely aware that "best position ever" is exactly the moment founders most often misread. The temptation is to believe the milestones will keep compounding linearly. They won't. Breast reimbursement is going to be a slog. The Cat III pathway is narrowing for AI [with new rules like AMA Appendix S] and possible enw AMA codes like "CMAA" - tbd.)
Payers have raised the bar. Distribution into hospital pathology labs is harder than my investor SaMD slide deck makes it look. And the consolidation around Tempus/Paige and Roche/PathAI changes the competitive dynamics in ways I don't fully understand yet — particularly around hospital purchasing committees and pharma partnership access.
So the real question isn't "what's the optimal path." It's: what decision do I need to make in the next 90 days that preserves the most optionality if I'm wrong about which path is optimal?
The core dilemma
It comes down to one trade-off, framed three ways:
Financial framing: Raise now at a high mark and dilute, or run a strategic process now and crystallize value, or push through to more milestones and risk the window closing.
Mission framing: I started this company to build a pan-tumor foundation model that personalizes cancer therapy globally. That mission survives — and arguably accelerates — inside Roche or Exact. It also survives independently with more capital. It does not survive a botched solo run that ends in a distressed sale in 2028.
Personal framing: If I sell now, I almost certainly stay on for 2–3 years inside the acquirer, then what? If I raise money and run it, I'm signing up for another 5–7 years minimum with materially higher execution risk. Neither answer is obviously wrong. Both answers have a version where I look back and wish I'd done the other.
Given my bias to an independent Artera, I should discount my own instincts here by maybe 20%.
The pivot points I'm actually weighing
1. Do we run a process now, or wait one more milestone cycle?
Arguments for now: PathAI comp is fresh, scarcity is real, we have the leverage. Arguments for waiting: breast NCCN inclusion plus first commercial payer wins plus a third FDA filing would, in theory, push us from "platform with two indications" to "the AI pathology platform" — and the multiple expansion from that re-rating could be larger than the multiple compression risk from waiting.
The honest answer is I don't know which is bigger. I think they're roughly the same magnitude, which means the tiebreaker is execution risk and team energy, not financial modeling.
2. If we raise, do we take strategic money?
A large pharma or diagnostics player taking a minority stake plus a commercial or co-development deal is the highest-information option I haven't fully explored. It lets us test partnership chemistry with a likely future acquirer without committing.
But it also signals to other potential buyers that we're "spoken for" and can suppress the open-auction dynamic later.
3. What's the right second-product priority?
We've been opportunistic — prostate, then breast because the data was there. The next one matters more strategically. Bladder is the easiest scientifically (we have the Cat III code already approved). Kidney is bigger commercially. A pan-tumor foundation model demo is more impressive to acquirers but takes longer. I don't have a clean answer and I should.
4. Commercial scale-up vs. regulatory/clinical momentum
We've been a regulatory-and-evidence machine. We have not been a commercial machine. Our hospital lab sales motion for the SaMD product is still nascent. If we're staying independent, this is the gap I lose sleep over. If we're selling, the acquirer's salesforce solves it overnight — which is itself an argument for selling, and I should be careful not to use that as a rationalization. Assuming the acquirer's sales force welcomes us and has the right incentives and skills.
5. The team question
I have people who joined to build a generational company. Some of them will be excited by an exit; some will feel sold-out. I have other people who joined for the IPO arc. Some of them will leave if we sell early.
Retention math on an acquisition has to account for who actually stays through the earnout — and a platform's value erodes fast if the AI/ML team scatters. This is not a hypothetical concern; it's the thing that determined whether Paige and PathAI were good buys for their acquirers, and we won't know the answer on those for another year.
What I'd want to know before deciding
In rough priority order:
About the market
What would ABC actually pay, and would they engage? They're the most natural defensive acquirer for the breast asset and we have no read on their appetite. A quiet, banker-led temperature check would tell us a lot without committing us.
Is Roche really done after PathAI, or are they buyers of a complementary asset within 12 months? My instinct says they're done for now, but I'd want to test it.
What did Tempus actually do with Paige post-close? If integration is going well, they're a more credible acquirer for us than the consensus thinks. If it's struggling, they're out for 2+ years.
Where is DEF strategy heading? They're the one player I can imagine making a competing bid that meaningfully changes our negotiating posture.
About our own business
What does the realistic breast revenue ramp look like over 24 months, base / bull / bear, given no CPT code yet? I have rough numbers but I want them stress-tested by someone who has actually launched a breast diagnostic before.
What's our actual hospital lab sales conversion data from the prostate SaMD early access program? Not the happy pitch deck version — the real funnel math. This tells me whether the SaMD commercial motion is working or whether we're still effectively an LDT company with FDA stickers.
What's the cash runway under each path? I think we have 18–24 months but I want to know what changes if breast launch costs more than planned.
What's the regulatory cost and timeline for a third indication, and which one maximizes platform narrative versus revenue?
About the buyers and partners
Who would do a strategic minority investment plus partnership without a path-to-acquisition lockup? This is the highest-leverage structure available to us and I don't know who's actually open to it.
What's the current pharma appetite for companion diagnostic partnerships specifically built on AI pathology? Post-PathAI, this market should be hotter, but pharma BD cycles are slow.
For each plausible acquirer: who is the actual decision-maker, what's his/her current strategic agenda, and what's their integration track record?
Generic "Roche might buy us" is not helpful to me.
"The Roche Diagnostics SVP, John Smith, who championed PathAI, is now looking for X"… I can move with that.
About the team and myself
Who on the leadership team has genuinely processed the trade-offs and can give me honest counsel versus telling me what they think I want to hear? The board can do this, but the board doesn't carry the operational weight. I need 2–3 internal voices I trust to push back.
What do I actually want? This should align but doesn't always, and I'd rather notice the gap now than discover it during a deal.
What do my co-founders think? Don't assume.
What I'd do in the next 90 days regardless of path
These are no-regret moves:
Hire a banker on a quiet retainer. Not to run a process — to give us real market intelligence and be ready if we need to move fast. Goldman, Centerview, or Jefferies for the diagnostics specialization.
Take one or two exploratory meetings. Through the banker, not direct. With ABC and one pharma. Listen, don't pitch. The information value is enormous and the commitment cost is zero.
Pressure-test the breast launch plan. Bring in 1–2 senior advisors who have personally launched competing diagnostics. Their assessment of our commercial readiness will dominate any board discussion of timing.
Lock in the AI/ML team. Refreshed equity grants, retention packages, whatever it takes. Whether we sell or scale, the model team is the asset. If they walk, every path gets worse.
Get the third-indication decision made. Not the work done — the decision. Bladder, kidney, or foundation model demo. Pick one and commit, because indecision here costs us 6 months of narrative.
Have the conversation with my co-founders and with the board chair separately. Where each of them actually is on the spectrum from "build for a decade" to "take the right offer." I should know this before any banker conversation moves past exploratory.
What I'm not going to do
I'm not going to start telling people we're "exploring strategic alternatives." The minute that phrase leaks, valuation gets harder and the team starts updating LinkedIn.
I'm not going to chase a pre-emptive offer. Pre-empts are almost always under-priced; if someone wants us badly enough to pre-empt, they'll pay more in a process.
I'm not going to over-index on the PathAI comp publicly. It's our anchor internally, but waving it around with bankers and acquirers makes us look like we've already decided to sell.
I'm not going to let the breast launch get under-resourced because we're distracted by strategic optionality. Whatever path we choose, breast commercial execution is the proof point that justifies our value. A weak launch tanks every other option and outcome.
The honest summary
The right answer is probably: raise a modest strategic round in the next 6 months with optionality built in, run a quiet market check in parallel, and make a binding decision on path by Q1 2027 when we'll have the first 6 months of breast commercial data.
That's not the bold answer. It's not the simple answer. But it's the answer that preserves the most value across the widest range of futures, which is what I'm paid to do.
I should sleep on it. And I should talk to my co-founder before I do anything that locks anything in.
###
###
4. High-Low Scenario Map
Two ArteraAI Scenarios: Mid-2026 to End-2028
Both scenarios assume the same starting point — May 2026, two FDA clearances, ~$110M raised, prostate generating real revenue, breast just cleared. Both are realistic. The point isn't to predict; it's to make the shape of each future concrete enough to plan against.
Upside Scenario: "The Platform Thesis Validates"
Mid-to-late 2026
The breast launch goes better than the board's base case. The hospital lab sales motion built for prostate transfers cleanly — same buyers, same workflow, incremental sell. Within 90 days of launch, three top-20 academic medical centers commit to integrating ArteraAI Breast alongside prostate, and one of them publishes early implementation data showing same-day reporting versus the 5–10 day send-out for Oncotype DX. That single data point becomes the entire commercial pitch.
A late-2026 growth round closes at materially higher valuation than the last mark. The round is led by a crossover fund (T. Rowe or Fidelity), with one strategic — a large pharma, not a diagnostics player — taking a minority position alongside a multi-year companion diagnostics partnership. The strategic stake is structured deliberately without right-of-first-refusal or any acquisition lockup. Total raised crosses $200M. Runway extends to 36 months.
The AMA Cat III code for breast moves through the editorial panel on the first submission, helped by the FDA clearance and the SABCS evidence package. CMS pricing follows the prostate playbook. We start billing Medicare on breast claims in Q2 2027.
Through 2027
NCCN inclusion for ArteraAI Breast in HR+/HER2- early-stage breast cancer arrives in the v2.2027 guidelines. This is the single most important commercial event in the company's history — bigger than either FDA clearance — because it forces payer coverage conversations and unlocks the academic medical center buyer who needs guideline cover to deploy.
A third indication clears FDA via 510(k) by mid-2027. The PCCP infrastructure means the submission takes 8 months instead of 18. The "platform" narrative is no longer marketing — it's three indications, two pathways, and a regulatory mechanism for continuous improvement. This is the moment Artera stops being compared to single-product diagnostics companies and starts being compared to AI platforms.
The pharma partnership generates its first milestone payment in late 2027, validating the companion diagnostic thesis. A second pharma signs a similar deal in Q4. Non-dilutive revenue becomes a real line item.
Commercial revenue hits roughly $60–80M ARR by end of 2027, growing 100%+ year-over-year, with healthy gross margins on the SaMD product and improving margins on the LDT business as volume scales.
Through 2028
Multiple things happen in parallel that compound:
A fourth indication enters FDA review.
The foundation model work matures into a demonstrable pan-tumor capability — not commercial yet, but credible enough to anchor investor conversations.
The hospital lab installed base crosses 200 sites, creating a network effect where new indications can be cross-sold into existing accounts at near-zero CAC.
A second large pharma partnership expands into a co-development arrangement.
By mid-2028, the company is in a position where three outcomes are simultaneously available:
A premium strategic acquisition at $2.5–4B. Most likely Roche (returning for a complementary asset to PathAI) or ABC (defensive, expensive). Possibly /PharmaX/ or a non-obvious pharma buyer through the partnership relationship.
An IPO in a window that opens for diagnostics names with $100M+ ARR, growth above 80%, and a platform story. Underwriters compete for the mandate.
A continued private path with another late-stage round at a $3B+ valuation, building toward a 2029–2030 outcome.
The luxury of having all three options is the actual prize. The CEO gets to choose based on what serves the mission, not what's available.
Key features of this scenario: Compounding, not breakthrough. No single miracle. Every individual step is something the company has already done at least once — clear FDA, build commercial, secure reimbursement, sign pharma. The success comes from doing all of them in parallel and on time.
Probability assessment: Plausible. Maybe 25–30%. It requires execution at roughly the 70th percentile across six independent workstreams. Each individual workstream is achievable; the conjunction is what makes it hard.
Downside Scenario: "The Slow Squeeze"
Mid-to-late 2026
The breast launch starts slower than projected. Hospital pathology lab procurement cycles are longer than the prostate experience suggested — those early prostate accounts had champions who were already digital pathology believers, and the next cohort doesn't. The first three "easy" accounts take six months to close instead of three.
The growth round happens, but it's smaller than hoped — $75–100M — and the valuation step-up is modest, around 1.5x the last mark rather than the 3x the PathAI comp suggested. The reason isn't the asset; it's the diagnostics financing environment, which softens in late 2026 when a couple of public comps miss earnings. Crossover funds pass. The round is led by existing investors with one new healthcare specialist, which is a flat-to-down signal that the strategic acquirers register.
The AMA Cat III code for breast hits a snag — not a rejection, but a request for revisions that pushes the effective date to 2028. CMS pricing follows the code, so Medicare billing on breast doesn't start when planned.
Through 2027
NCCN inclusion for breast doesn't come in the v2.2027 guidelines. The panel wants longer follow-up on the predictive (chemotherapy benefit) claim, not just the prognostic data. This is a defensible scientific position but a commercial disaster — it pushes inclusion to v2.2028 at earliest, which means most commercial payers won't write coverage policies until 2028, which means hospital lab buyers won't commit volume until coverage exists. The chicken-and-egg problem the slides described is real and we're now living inside it.
Meanwhile, two competitive dynamics get worse:
Tempus + Paige launches a breast biomarker prediction product as part of Paige Predict, leveraging their existing oncology install base and reference lab relationships. It's not as good clinically but it's bundled into a workflow customers already have. Sales cycles get harder.
Exact uses glass slide archives, which are plentiful, to develop a test that's additive with Oncotype.
Announces an AI overlay to Oncotype DX — an "Oncotype DX AI" product that adds image analysis to their existing genomic test.
It's a defensive product, scientifically weaker than ours, but it carries the brand and the payer contracts. We start hearing it in customer conversations.
A pharma partnership conversation that looked promising in early 2027 stalls. The pharma partner does a similar deal with PathAI/Roche instead, partly because Roche can offer the full integrated companion diagnostic stack and we can offer the AI piece only.
The third indication clears FDA, but later than planned — Q4 2027 instead of mid-year — and the data package was harder to assemble than expected because the right validation cohort wasn't readily available.
Revenue hits ~$30–40M ARR by end of 2027. Growth is real but is below plan, and the gross margin story is muddier than the board deck suggested because hospital lab sales have higher implementation costs than the model assumed.
Through 2028
This is the "downside" forecast, so the squeeze tightens. Three forces compound in the wrong direction:
Cash burn vs. revenue ramp. With a smaller-than-hoped 2026 round and slower revenue, runway becomes a 2028 problem. A bridge round in mid-2028 happens at flat-to-down valuation, and existing investors have to choose between supporting it and accepting dilution. The signal to outside acquirers is unmistakable.
Strategic acquirer interest cools. The Roche/PathAI integration is going well enough that Roche is uninterested in a second pathology asset. Tempus is focused on absorbing Paige. Exact's defensive AI product is in market and they no longer feel they need to buy us. The natural buyer set has effectively closed, at least for the next 18 months.
Talent risk crystallizes. Two senior AI/ML leaders leave for better-funded competitors or to start their own companies. Replacements are findable but the institutional knowledge in the foundation model work degrades. The "platform" narrative gets harder to tell honestly.
By mid-2028, the company faces a harder choice set:
A strategic acquisition at $400–700M — real money, but a fraction of the PathAI comp and well below what was achievable in 2026. The likely acquirer is a tier-2 diagnostics player or a private equity rollup, not a premium strategic. Founders and early investors do well; later investors are roughly flat.
A down round at a $600–900M post-money, with new investor protections that materially dilute the founders and reset the cap table.
A continued grind that buys time but doesn't change the trajectory — eventually leading to option 1 or 2 in 2029 on worse terms.
The IPO option is gone. It doesn't come back until 2029–2030 at earliest, and only with a meaningful inflection that's hard to engineer from this position.
Key features of this scenario: No catastrophe. No FDA rescission, no clinical scandal, no fraud. Just a series of small misses that compound. Each individual miss is recoverable; the conjunction is what creates the squeeze. The company never fails in any dramatic sense. It just becomes worth less than it should have been.
Probability assessment: Equally plausible. Maybe 25–30%. The reimbursement timeline risk is the single biggest swing factor and it's largely outside management's control.
What both scenarios share
Looking at these side by side, three observations matter more than the outcomes themselves:
1. The high-success scenario doesn't require luck — it requires reimbursement to work. The single biggest variable between the two scenarios is whether breast CPT coding, CMS pricing, NCCN inclusion, and commercial payer policy all move on schedule. That entire workstream is partially outside management's control and is the right place to over-invest attention and resources right now.
2. The downside scenario is largely about competitive dynamics, not internal execution. In the bad version, Artera doesn't fail at anything — it gets slowly squeezed by larger, better-distributed competitors. This argues for moving faster on strategic optionality, not slower, because the window where Artera looks scarce and special is open now and may not be open in 18 months.
3. The optionality value of acting in 2026 is asymmetric. A strategic conversation initiated in 2026 from a position of strength can be walked away from if the high-success path materializes. A strategic conversation initiated in 2028 from a position of weakness happens on the acquirer's terms. The cost of optionality is low; the cost of not having it is potentially enormous.
What to plan against
The right planning posture isn't to bet on the high-success scenario or to insure against the downside — it's to set up the 2026 decisions so that:
If the high-success scenario materializes, the company is positioned to choose among three good outcomes.
If the downside scenario materializes, the company has already established the relationships and information base needed to execute a strategic sale before the squeeze tightens.
The actions look similar in both cases for the next 6 months: hire the banker, take exploratory meetings, lock in the team, over-resource reimbursement work, get the third-indication decision made. The divergence in tactics happens around Q1 2027, when the first breast commercial data and the NCCN decision will tell management which world they're actually living in.
That's the practical value of running both scenarios — not to predict, but to identify the actions that look right in both, and execute those first.
###
The original output sometimes included mutiple citations widgeted together, the blog here captures the 1st one.
