NanoString Technologies Inc (NSTG) 2022 Q3 法說會逐字稿

Good afternoon. Thank you for attending today's NanoString Third Quarter 2022 Operating Results Conference Call. My name is Megan, and I'll be your moderator for today's call. (Operator Instructions)

I would now like to pass the conference over to our host, Doug Farrell with NanoString. Doug, please go ahead.

Thank you, operator. Good afternoon, everyone. Joining me on the call today is Brad Gray, our President and CEO; as well as Tom Bailey, our CFO. Earlier today, we released our financial results for the third quarter of 2022. During this call, we may make statements that are forward-looking, including statements about financial and operating projections, future business growth, trends and related factors, expectations regarding future operating results, cash flows, current and future orders, prospects for expanding and penetrating our addressable markets, our strategic focus and objectives and the development status and anticipated success of recently planned product launches as well as the impact of macroeconomic factors.

Looking forward, these statements are subject to risks and uncertainties, including those described in our SEC filings. Our results may differ materially from these projected and we undertake no obligation to update those. Later in the call, Tom will be discussing our financial results and the 2022 guidance. We have prepared as a supplement to GAAP financial measures, selected non-GAAP adjusted measures calculation of which are described in detail in our press release. Throughout the call, all financial measures will be GAAP, unless otherwise noted. You can also find a reconciliation of GAAP to non-GAAP measures as well as the description, limitations and rationale for using these in this afternoon's press release.

To aid analysts and investors in building their models, we have posted exhibits under the Financial Information tab of our Investor Relations home page that include a presentation of non-GAAP or adjusted measures or other selected financial data. I'd like to remind everyone that next week, we'll be participating in the Stifel Healthcare Conference in New York as well as the Jefferies Healthcare Conference in London. We look forward to having the chance to speak with many of you then.

Now I'd like to turn the call over to Brad.

Good afternoon, and thank you for joining us. Spatial biology is a dynamic field, yielding powerful scientific insights. While expansion will last a decade or more, the next few years will determine the mix of technologies in which key discoveries will be made. While there are other emerging spatial biology providers, we believe that our early mover advantage, best-in-class technology, ecosystem and synergistic platforms and Blue Chip partners position us to lead. For this reason, our top priority is to maximize the number of our spatial biology instruments placed in leading research labs worldwide.

During the third quarter, we made great progress on this top priority, generating orders for approximately 60 spatial biology instruments, an increase of approximately 70% compared to the same quarter a year ago. We have now sold more than 100 CosMx Spatial Molecular Imagers and accumulated our revenue backlog valued at approximately $23 million. We expect this backlog to increase in Q4 and have set us up for robust growth for 2023. During Q3, we also advanced our scientific leadership, images generated by our customers using the GeoMx Digital Spatial Profiler were featured on the cover of not 1 but 2 of the world's leading scientific journals. The seminal description of CosMx was published in another leading journal, and we strengthened our informatic ecosystem through a partnership with Visiopharm, whose mission is to transform pathology through AI-based image analysis and workflow standardization.

The degree to which customer interest in spatial biology is trending towards platforms that provide spatial imaging at single cell resolution has been even greater than we previously expected. The intense customer focus on single cell imaging resulted in fewer orders for the GeoMx instrument. And in Q3, we captured 20 GeoMx orders compared to the 25 to 30 orders we had expected for the quarter. Fortunately, our CosMx Spatial Molecular Imager is a leading product in the single cell imager category, and we are capturing demand at a significant pace. We generated approximately 40 CosMx orders during Q3, substantially exceeding the 25 to 30 CosMx orders we had guided for the quarter. As a result, while we recognized less GeoMx revenue than previously expected, we exceeded our guidance for overall spatial instrument orders.

Our spatial biology instrument orders were strong. Q3 was a challenging quarter for consumable sales, which were the largest contributor to the shortfall in our Q3 revenue relative to guidance. We experienced a decline in annual consumable pull-through per system for both nCounter and GeoMx, which we believe was due to a combination of lumpy ordering patterns, customer life cycles and macroeconomic factors. For nCounter, the pull-through dropped across virtually all geographies and customer segments and seem to be driven by decreasing demand from the older systems in our installed base. For GeoMx, the weakness seems to be primarily a product of lumpy ordering patterns and the impact of a relatively large number of newly placed systems still ramping to their full consumable run rate. In both cases, weakness was exacerbated by broader market conditions, including continued clinical trial delays among large pharma customers, reduced spending in cash-strapped small biotechs and COVID-19 lockdowns in China.

While we are clearly disappointed by the shortfall, I remain confident that we are set up for a strong '23 and beyond. Importantly, we have recently reevaluated our key priorities to ensure that we invest in the most promising aspects of spatial biology while streamlining our cost structure. As a result of this review, we are eliminating approximately 95 positions and reducing spend in other nonpersonnel areas. While these decisions are difficult, especially as we consider the impact on our people, these actions will ensure that a more substantial portion of expected 2023 revenue growth falls to our bottom line and that we are positioned to deliver on our commitment to achieve breakeven without the need for additional financing.

Tom will provide more details on our financial outlook during his prepared remarks. Before I hand the call over to him, I'd like to provide an overview of our progress towards our strategic objectives for the year. Given the tremendous customer interest, [I will] start with an update on our objective to launch CosMx as the industry-leading molecular imaging platform. Customer interest in CosMx has exceeded our expectations throughout the year as we prepare to ship our first commercial systems in the weeks ahead. We are more confident than ever that CosMx SMI has the best performance specs and the ecosystem to be a platform of choice. CosMx leads with best-in-class performance metrics, including highest plex RNA assays, the capability to image proteins and robust data quality, all done and challenging FFPE tissue.

CosMx is helping us reach new customers and show strong synergy with the whole transcriptome capabilities of GeoMx boding well for our long-term leadership in spatial biology. In Q3, 2/3 of CosMx instrument orders came from new to NanoString customers, showing how CosMx is extending our customer base into discovery research. 1/3 of these new customers chose to adopt both CosMx and GeoMx at the same time by purchasing a bundle. Meanwhile, 1/3 of all Q3 CosMx instrument orders came from labs who had previously adopted GeoMx, primarily within translational research. To date, CosMx has penetrated only 20% of the GeoMx installed base, suggesting many more opportunities for cross-selling going forward.

In the near term, we expect our spatial instrument order mix to remain CosMx heavy. CosMx represents about 60% of opportunities added to our spatial instrument follow in Q3, and we expect CosMx to continue to account for about 2/3 of spatial instrument orders. Our beta program is progressing well with CosMx instruments installed and integrating of AtoMx at 3 beta customer sites. Customer feedback has been overwhelmingly positive. And through these interactions, we've identified several training and software updates that we are incorporating into the final commercial rollout. We're concluding the validation of CosMx and AtoMx is an integrated solution and are preparing our first commercial systems for shipments in the weeks ahead. We expect to ship between 5 and 10 CosMx systems before the end of the year, and we can't wait to see the great science these systems will generate.

Our second objective is to drive GeoMx DSP further into mainstream research, broadening adoption across multiple areas of research. From a scientific perspective, our CosMx customers had a wildly successful third quarter. Papers and images generated using GeoMx raised the covers of 2 prestigious scientific journals, Nature Genetics and Clinical Cancer Research. In total, our customers published more than 30 new papers, bringing our total loan to more than 160 peer-reviewed publications as of September 30. This productivity underscores the utility of GeoMx, particularly in translational research for human oncology and immunology.

Demand for GeoMx instruments from translational researchers remain strong even our discovery researchers opt for single cell resolution of CosMx, and translational researchers drove the vast majority of the approximately 20 GeoMx instruments ordered during the third quarter. We remain focused on ramping the consumable pull-through of our growing installed base of GeoMx systems. We've observed that customers take time to build their pipeline of spatial projects, reaching steady state utilization 12 or 18 months after instrument purchase. While GeoMx consumable revenue has grown, the large number of instruments sold in the second half of 2021 and activated in the first half of 2022, have weighed on the consumable pull-through per system.

We are bolstering our support for customer brand and consumable pull-through going forward by refocusing our PhD field application specialists, primarily on spatial biology experiments [full] design, helping customers accelerate their pipeline of projects on our platforms. We expect this will benefit both GeoMx and CosMx pull-through in the long term. Meanwhile, we continue to invest in partnerships that enhance the GeoMx workflow using Visiopharm's AI-driven digital pathology software, onco-topics discovery, researchers will be able to analyze [4] color fluorescent images generated on GeoMx and combine these images with those using traditional H&E staining to better understand the number and types of cells that are present within regions of interest. We expect the combination of these technologies will accelerate biomarker discovery and validation for both whole transcriptome RNA and HiFlex protein analysis.

We're committed to delivering informatic solutions that enhance research productivity, bringing us to our third strategic objective, which is launching our AtoMx spatial informatics platform. Spatial experiments require unprecedented bioinformatics capabilities to support image analysis, data visualization and global collaboration. We believe that the highly scalable compute and sports capacity of our cloud-based AtoMx platform is essential to drive broader adoption of spatial biology and that AtoMx is an underappreciated competitive differentiator.

We expect to achieve our goal of launching AtoMx in the weeks ahead as AtoMx becomes available to the first customers receiving their commercial CosMx systems. We expect to roll AtoMx out to our GeoMx users during the first half of 2023. AtoMx employs a flexible data structure, so it's ready to be leveraged using artificial intelligence and machine learning from day 1. Together, NanoString and Visiopharm are exploring opportunities for further integration by connecting our AtoMx to Visiopharm software so that researchers can leverage artificial intelligence and machine learning to provide spatial biology for both of our spatial biology platforms.

Our fourth objective for 2022 is to sustain our nCounter franchise. Q3 was a challenging quarter for nCounter as new instrument placements and consumable pull-through both fell short of expectations. Instrument placements slowed among academic researchers, especially in Europe and were hindered by a year-on-year decrease in the number of GeoMx plus nCounter bundles that are being sold. Consumable pull-through was challenged by an increasing age of our nCounter installed base, which first commercially launched back in 2009. At this stage in the nCounter life cycle, we believe that some older systems are becoming inactive as researchers who originally purchased them transition to new roles or take their research in new directions.

While the annualized consumable pull-through on each active nCounter system has remained relatively stable over the last 24 months, growth in the number of active nCounter systems has flattened as the number of new encounters being placed approximately [equals] the number of older encounter systems being inactivated. While we do not view nCounter as a future growth driver, it remains an important foundational business that continues to generate great science and provide substantial cash flow to support our global commercial channel and innovation.

I'd now like to turn the call over to Tom to review the details of our financial results and outlook for the balance of the year.

Thanks, Brad, and thanks all for joining us today. Revenue for the third quarter was $29.5 million, reflecting a CosMx heavy spatial instrument mix, lower GeoMx and nCounter consumable sales that are forecast syndicated and about a $1 million negative foreign currency impact. For CosMx, we generated orders for about 40 new systems in Q3, adding approximately $9 million of revenue backlog to be recognized in future periods. As of September 30, 2022, our cumulative CosMx orders stated over 100 systems, translating to a total revenue backlog of about $23 million.

Q3 GeoMx revenue was $9.3 million. GeoMx instrument revenue was $4.7 million, reflecting approximately 20 new system shipments and consumables revenue was $4.6 million. Q3 annualized GeoMx consumables pull-through was about $58,000 per installed system. At the end of Q3, our GeoMx installed base was approximately 330 instruments with about 15 new instruments installed during the quarter. For our nCounter business, which includes all service, Q3 revenue was $20.2 million. nCounter instrument revenue was $3.3 million, consumables revenue was $12 million and Q3 annualized nCounter consumable pull-through was approximately $44,000 for installed systems. At the end of Q3, our nCounter installed base was approximately 1,105 instruments with about 20 new instruments installed during the quarter.

Turning to margins and expenses. I'll provide results on a non-GAAP or adjusted basis, which removes the impact of stock-based compensation, depreciation and certain other items. Please refer to our press release as well as the exhibits we have posted to our Investor Relations web page for detailed information on how our non-GAAP or adjusted measures are prepared. Q3 adjusted gross margin was 57%, an improvement of 100 basis points as compared to Q3 of last year and an improvement sequentially, reflecting management of manufacturing personnel and other expenses to align with current product sales volumes and efficiency improvements. These impacts were partially offset by investments we are making in manufacturing capacity, primarily to support commencement of CosMx instrument shipments and consumables.

Adjusted R&D expense was $14.5 million, a decrease of 12% year-over-year, primarily due to the capitalization of approximately $3.5 million of software-related product development costs that will be expensed in future periods. Inclusive of capitalized amounts, our Q3 R&D spend reflects continued investment in our spatial biology platforms, including in hardware, consumables and software development in CosMx and AtoMx in advance of the expected shipment of our first commercial CosMx instruments in Q4. Adjusted SG&A expense was $28.4 million, an increase of 20% year-over-year and driven primarily by investments made in spatial biology related commercial initiatives.

Q3 adjusted EBITDA loss was $26.1 million, and our cash and cash equivalents at September 30 were $230.5 million. We also announced today that we have taken steps to streamline our cost structure by eliminating selected positions and activities while maintaining key investments in spatial biology. As our business mix has evolved, we decided to take these steps to prioritize our portfolio of technologies, operational and commercial initiatives. These steps will allow us to better leverage our operating expenses while investing in our spatial biology business and will support our objective of reaching cash flow breakeven with our current balance sheet resources. We expect to record a charge of approximately $3.5 million in the upcoming fourth quarter related to these changes.

Turning to guidance. Our Q4 outlook reflects a revised mix of expected spatial biology system orders and consumables pull-through rates for GeoMx and nCounter based on our Q3 experience with the upper end of our guidance ranges, assuming modest Q4 driven seasonal improvement. For the fourth quarter, we expect to receive orders for over 60 spatial biology systems with an approximately 66%, 33% mix between CosMx and GeoMx, implying a cumulative total of over 200 spatial systems sold in 2022. Spatial biology revenue of $12 million to $13 million, which we expect will derive from sales of GeoMx systems, the shipment of our first CosMx systems and from consumables.

We would expect approximately $7 million to come from instrument sales and approximately $5 million to $6 million to come from consumable sales. We expect Q4 nCounter revenue of $21 million to $22 million in total Q4 revenue of $33 million to $35 million. We have also updated our 2022 full year outlook, which is detailed in today's press release to reflect our year-to-date actual results and our Q4 outlook. Our updated full year outlook also incorporates approximately 1 month impact of our announced cost reduction initiatives.

Looking ahead to 2023, the CosMx Systems orders we've already secured in 2021 and 2022 provides the foundation for expected revenue growth of 40% to 50% in 2023. We also expect improvement in our EBITDA loss next year, reflecting expected revenue growth combined with the full year's impact of our cost reduction initiatives and anticipate ending 2023 with approximately $140 million to $150 million of cash. We look forward to offering more details when we provide our 2023 annual guidance for our usual process in February.

Now I'll turn the call back over to Brad for our closing comments.

Thanks, Tom. We are focused on the dual objectives of spatial biology and market leadership and achieving cash flow breakeven on our current resources. We believe our leadership in spatial biology is best measured through the pace of new instrument orders, and by that measure, we are having a solid year. Our successful launch of CosMx sets the stage for strong revenue growth in 2023. We're addressing trends in our nCounter and GeoMx business with the sites of action and are streamlining our cost structure to maintain balance. With our unique spatial biology portfolio and strong balance sheet, we believe we are poised to deliver both market leadership and future profitability.

With that, we'll now open the line for questions.

(Operator Instructions) Our first question comes from the line of Kyle Mikson with Canaccord Genuity.

I guess, Brad, just starting with the quarter, could you just provide some more detail on like why consumables for GeoMx [was lumpy, by ordering was lumpy]? Like how do you kind of prevent that going forward? And you didn't mention the realigned commercial team much. I mean, was any of the shortfall here due to that realignment earlier in '22?

Yes. I'll start with the second one first. We do not believe that the shortfall in Q3 was related to our sales force realignment in the first quarter. We're really past the issues that realignment created many of the differences we saw between territories that hadn't been [realigned are] now raised. The junior consumable reps have caught up to the senior consumable reps in terms of performance, et cetera. We think now that we have those execution issues behind us when we're focused on macro and market trends and issues that are very specific to our installed base in terms of the life cycle of the customer.

So looking at the Q3 GeoMx consumable pull-through, I'd say it's impacted by 3 different factors, which are sort of tricky to parse apart. The first is won't be ordering patterns. And we've come to appreciate there that new sites for GeoMx very often build inventory early on and then burn it off slowly over time. We know that we saw some stocking orders in the third quarter of 2021 that did not repeat in the third quarter of this year. And as a result, that's part of the explanation for the year-on-year drop. The best evidence for lumpiness is the very strong Q4 seasonality that we have seen in both 2020 and 2021, where our pull-through reached over 110,000. That wasn't really pull through on a run rate basis in terms of people's actual experimentation, but rather they're stocking behavior.

The second issue is macro. And when we look at where our year-on-year pull-through dropped the most, we can see that it dropped the most amongst biopharma companies and within China. And we know that these groups have been under some macroeconomic pressure. The last one, and the trickiest one for us to model today is the life cycle impact of new sites. So it takes a lot for a GeoMx site to reach its full ramp of consumables. Sometimes there's logistical issues that needs to overcome such as the purchase of ancillary equipment for slide trap or data storage analysis solutions. Sometimes they're learning to plan new projects and taking on new concepts of spatial biology and intervention selection. And sometimes, Core Labs simply need to promote their new spatial biology capability to their local customers to build a book of business.

And overall, when we look back, a new site tends to do about 60% of what its long-term run rate for consumables will be during its first year and then ramp up to 80% in the second year and 100% thereafter. Because we had a very large number of GeoMx systems sold in the second half of last year and installed and activated in the first half of this year, we are experiencing a weight of relatively low pull-through systems that are still ramping up to their full run rate. So it's very tricky to assign the -- to allocate the pull-through drop across these 3 different factors. But it's something we're continuing to study and that we will provide updates on in future periods and when we guide for next year.

Okay. That was great. And then maybe just thinking about [GeoMx specifically just staying] on that. Going forward, I guess, spatial orders 1/3 being GeoMx, sounded like that was the expectation. Maybe just talk about your confidence in the health of that franchise and the kind of the visibility heading into '23 and beyond. And I was also curious if NGS readout and WTA as much as -- because that much of a headwind -- a tailwind that you guys thought that might have been last year [or a year earlier].

Yes. I'll take the second question first, again, Kyle. WTA and NGS readout now represent the vast majority of the GeoMx system. 85% of our new systems are placed in front of the sequencers and the Whole Transcriptome Atlas assays account for something like 75% of our consumable revenue. So that's been really a successful story. We're also long-term optimistic about the value of the GeoMx franchise. The science speaks for itself. It's no small thing to be on the cover of 2 different journals within just a few months' time. We now have 160 peer-reviewed papers, 30 a quarter being published. This for translational researchers who are doing immunology oncology, this is a very productive and powerful instrument.

I do think the idea of single selling a gene is sucking a lot of the oxygen out of the room right now with respect to people who are building their new spatial capabilities. But in the long term, we still believe GeoMx has an important place. And I think the best evidence we have today for that is the strong synergy between CosMx and GeoMx when it comes to who's purchasing the CosMx systems, a very large fraction of those who are buying CosMx are either doing so with the GeoMx in a bundle or our labs who've already purchased the GeoMx and want to have complementary capabilities.

All right. Perfect. And then just one more before I hop off. Just on the '23 outlook, I guess. And I know you're not going to provide much detail until it will be on the 4Q call. But Tom or Brad, just a lot of this hinges upon GeoMx next year, including pull-through? I mean, how should we sort of think about that as we kind of model going ahead here?

Yes, I think that the first answer is, obviously, we'll provide more detail in February, Kyle. But setting that aside, what we built our growth rate ranges, we didn't make heroic assumptions around GeoMx. I think that you can get to our numbers by assuming GeoMX repeats when it does, did this year for instruments and that pulled through repeats what it did this year as well. And you can get to those growth rate ranges for those conservative assumptions. So we'll comment more in February.

Okay. Got it. And just to repeat, the range was 40% to 50% growth, and I think it was the backlog for CosMx of $28 million. Is that correct?

Through $23 million currently, and we expect that to be about $30 million by the end of the year, dollar -- in dollar terms.

Our next question comes from the line of Dan Brennan with Cowen.

Great. Maybe just starting off with the updated guidance for the year. Fourth quarter arguably still looks a little steep. Just walk us through the visibility, particularly on genomics, I mean we're getting to like $12 million to $13 million in revenues versus $9 million, a little over $9 million this quarter. Just give us a sense, it looks like placements are expected to double sequentially. Is that about right? Just anything on what's implied for GeoMx in fourth quarter? And kind of what's the visibility there?

Yes. So for fourth quarter, Dan, that $12 million to $13 million range also is inclusive of the 5 to 10 CosMx systems that we expect to ship that Brad mentioned in his script, so you have to consider that when looking at that range. And I think when you do that, you'll see that the range that we've established for GeoMx is a conservative range that assumes that at the bottom end of the range, that pull-through is about the same as we saw in Q3. And at the top end, is what you would characterize as a typical sequential improvement, which for GeoMx is about a 15% sequential improvement in pull-through. That would be the top end of our range, which is still lower than the old range that we had before. So I think that the new guide that we've set up incorporates those factors, and we feel are confident in the place [that we've planned at that] for Q4.

Got it. You guys talked about China and emerging biopharma in some of the remarks. Can you remind us like how did China do in the quarter? And how big is that emerging biopharma business for you since you're kind of citing that as a weak point?

Yes. So just China underperformed its plan during the quarter, Tom may have the exact year-on-year numbers. China represents about 5% of our overall revenue for the company. Small biopharma represents about 15% of our overall revenue for the company. But we also cited that large biopharma at least on the consumables side has also underperformed, and that makes up about another 10% of our business. So I think we -- this was not the only factor that impacts our performance, but we do have -- when you add those together, 25% to 30% of our business exposed to [China] markets that are facing their unique challenges.

Got it. And then when you think about CosMx going into '23, obviously, a really strong order book, a lot of momentum there. Like how do you see -- I guess, maybe since you threw out the 40% to 50%, how much does that assume additional CosMx orders in the first half of the year that could installed? Just wondering, you had a lot of momentum right now, which is great. And hopefully, that is sustained. But obviously, there's a lot of Zenium is coming out and you've got vision that's on the market and you have others that are coming. Just wondering kind of what you baked in for kind of additional order conversion as we look ahead into '23.

Yes, I got it. I'll repeat that we'll guide in February, but I think that we built that 40% to 50% range, Dan, that is a conservative assumption on what we would expect orders to be for CosMx next year. So you do not have to make a growth assumption on CosMx orders to get to the revenue range -- growth rate range that we articulated in 40% to 50%.

Got it. And then maybe a final one. Just if you think about -- you guys had the Investor Day, not that long ago now maybe no longer. When you look out, say, the next 3 to 5 years, I mean, you've got [3 elements] which is wobbling right now, but you guys still feel good longer term. It's got a real strong place and CosMx, you guys are excited. How do you think about the collective business say, growing, not over '23, when you look at, say, the next 3 to 5 years, 20% top line growth, 30% topic line growth, any range of outcomes that you would talk to about what the overall company should be able to do?

Yes, Dan, this is Brad. I don't think we're going to establish long-term growth rates here on this call. But I mean I think we believe we're going to be a market-leading company in a market -- in the spatial biology market, which is just getting started and will grow in a very meaningful way for a decade ahead. We obviously have some transient impacts this year with the inability to recognize revenue on CosMx and we're so learning the consumable pull-through dynamics on GeoMx. But the secular trend, heavy growth in the spatial biology industry should drive meaningful top line growth for the years ahead.

Our next question comes from the line of Dan Arias with Stifel.

Brad, why do you think biopharma spending is retracted? Historically speaking, the consistency there has been well above average in the R&D setting?

Well, we've been talking about for -- ever since the recovery from the pandemic about a slowdown in clinical trial approval that occurred during the COVID-19 pandemic. And because the translational biology that's done on our systems is downstream of that accrual, typically starting once all the clinical trial samples have been collected, we have seen a slower recovery in pull-through in biopharma than we had amongst the academics in the time since the 2020 pandemic. That has not improved. It has remained suppressed relative to pre-pandemic levels.

Okay. But if you were -- I remember you were talking about that earlier in the year when it didn't seem like that was impacting GeoMx pull-through at the time. Is this sort of like a follow-on effect or something that you're realizing later? I'm trying to understand the progression of the issue over the quarters.

Well, it's -- I mean it's an issue that has been impacting both the nCounter and GeoMx pull-through, ever since the post-pandemic period. I think we've talked about it most in the context of nCounter because the overall magnitude in absolute dollar terms for nCounter is bigger than it is for GeoMx at this stage of their life cycles. But this is not a new issue. It's probably one that we would have guessed would have recovered by now, that we have not seen the recovery in biopharma pull-through that we would have expected as those trials completed enrollment and translational biology began.

Okay. And then you mentioned this quarter and then last quarter as well, the pull-through is suffering from some of the newcomers that are just slow to ramp. What about the existing users those labs that bought an instrument in 2019, 2020, early 2021. I think there were about 200 devices or so that were installed by June of '21. So just curious how pull-through is trending amongst that crowd and how it compares to the new forks that you're referencing here.

Yes. When we look over a long period, like the trailing 12 months from the third quarter of 2022 back versus the previous trailing 12 months from the third quarter of 2021 back, they're very similar. That being said, if you just look at Q3, it was -- our pull-through was down sequentially from Q2 to Q3 across all vintages of our GeoMx installed base. So this particular quarter, even the more mature systems that have been out there a while, had lower ordering. And we attribute that to some of the other factors I mentioned in terms of lumpiness and the macro effects.

Okay. Last one for me. Just Tom, on CosMx. As orders continue to build here, I'm just curious what would you think about as far as a [placement max] for 2023. I know talking about '24 seems a little crazy right now, but just wondering what we need to sort of, in our minds, penicillin for backlog towards the end of the year there.

Yes. I think that in the guidance range that we have earned, it assumes that orders are similar to this year for CosMx that we deliver about half of our backlog. So that would suggest that you'd go into 2024 with about 60%, give or take in backlog headed into the -- that's based on the 40% to 50% range that we talked about. And obviously, we'll get more explicit commentary in February that could be to have some movement around that, but that's split that 40% to 50% would infer that number, Dan.

Our next question comes from the line of [Ruizhi Qin] with JPMorgan Chase.

This is (inaudible) on for Julia. So regarding the order dynamics between GeoMx and CosMx, specifically for your clients, could you provide more color into the customer preferences over which instrument to purchase perhaps by customer types? Or would you just say it's more or less equal throughout the various customers? And then how does that compare to your internal expectations?

I think your -- I have a little to hearing your question, the first part I know was talking about mapping the preference for CosMx and GeoMx to different customer types. So I think what we've observed so far this year, if we look -- well, going back to 2021, we sold quite a few GeoMx systems to both discovery customers and translational customers. So discovery customers remember when we are those who are doing very basic scientific research, not necessarily focused on curing disease and translational researchers or those who are doing work on human tissue samples for drug development or diagnostic development.

So our GeoMx sales last year were pretty nicely balanced against those 2 customer segments. This year, those discovery customers have shown a very strong preference for single cell imaging. And in fact, CosMx is predominantly going to discovery customers. Now interest in GeoMx remains strong from translational customers. But given the mix we had last year, which was balanced across the 2, we've seen a meaningful year-on-year drop in the total number of GeoMx systems we're placing.

Got it. Very helpful. And then the second question is, we're seeing a growing number of new entrants in the spatial market, especially with [GeoMx launching] their new product. Obviously, you have had start with CosMx, but as these continuing platforms launch, how do you think about those dynamics going forward? And what are [your most durable advantages for CosMx versus competitors]?

Yes. I heard the question [now, Tom]. So we feel really great about the position of CosMx and the durable advantage that it has. I mean, I think first, we picked the RNA plex as a key specification where we are currently market leading, and we expect to remain market leading. So we do 1,000 different RNAs at once most of the other competing imagers to about half that. The second area we distinguish ourselves is on protein imaging, where we'll launch with a 60-plex offering, and we've demonstrated already an ability to go to 100 flex. That is a feature that most of the other imagers are not able to offer their customers. So we feel very good about the competitive profile of CosMx. And I think so far, we have a very good win rate in those instances where customers are evaluating our offerings against others.

Our next question comes -- go ahead.

Operator, you broke up a little bit there.

Sorry about that. The next question comes from the line of Catherine Schulte with Baird.

First, I think you said the -- yes, can you hear me?

Yes, we can, Catherine.

Perfect. I think you said the number of active nCounters is remaining about flat as new systems are being offset by older systems that are being retired. Do you have any sense for why these systems are being retired? Is there a replacement cycle dynamic within accounts or users stopping nCounter-based research and if so, why?

Well, the 2 major reasons that we see people no longer use their nCounters are, one, the individuals who purchased that nCounter and who was responsible for experimentation on it moves from one job to another. And 2, the science moves in a different direction, perhaps away from bulk gene expression to some other area of science. And when we look back across our installed base and we look at where those systems are inactive, it's highly correlated with the age of the system because as you can imagine, with every cumulative year after an instrument has been placed, the cumulative probability of the employee to bought a leading goes up and the cumulative probability of the science going in a different direction goes up. So there is a life cycle to these systems that's totally independent of their usable life. I mean we have active systems that are around there from 2012 still running assays. But every year, the possibility that the researchers move in a different direction increases.

Got it. And then on pull-through, we've heard some of your peers talk about some weakness on the consumable side as well, just given maybe more pronounced seasonality and inventory work through. So can you just talk through how activity levels trended throughout the quarter? And maybe how October looked just to get a better sense for which of those were transitory and which of those dynamics are likely to stick around?

Well, we don't have perfect visibility into activity levels. Our systems are not connected with a telemetry that allows us to monitor their use remotely as some of our competitors are. So I can't really speak to that question, Catherine. I think our early October trends are consistent with the guidance that Tom gave, which is to say we are not expecting a substantial transitory recovery in the fourth quarter, and we have not built that to our guidance.

Our next question comes from the line of Tejas Savant with Morgan Stanley.

This is Edmund on for Tejas. The first question for me is what's embedded for your -- what's embedded for your nCounter franchise and your updated guidance? And given some of the dynamics that you observed in the quarter, we're thinking about nCounter next year as flat year-over-year would be the right way to think about it?

Yes, I think (inaudible) the fourth quarter, as I mentioned in the prep remarks and another comment, so don't built the guidance at the bottom end of the range for consumables at what the Q3 pull-through was the top end of the range, you could assume about a 10% sequential improvement in pull-through, which is typically what we would see on a seasonal basis in a normal year. So consistent with the comment that Brad just made about us not assuming any improvement in the macro factors, that's what that reflects. So that's where the overall guidance was built for Q4. And we'll give more commentary on next year as we get into the February time frame, but the growth rate ranges that we assumed for the 40% to 50% range assumes the typical flat instrument year-over-year that we typically assume in our nCounter guidance and pull through that was consistent with what we observed this year, so no improvements next year. So that still would suggest there would be some growth, but not any growth in the pull-through ranges that we experienced this year that were built into those preliminary assumptions.

Got it. And understanding you want to hold '23 guidance until later, but how are you thinking about academic budgets heading into next year in light of the upcoming midterms in the U.S. and other government priorities such as energy subsidies in the EU?

We have not, in our commentary about next year's growth rate factored any major changes into academic research funding in one direction or the other. And we'll keep an eye on it, but spatial biology is a relatively new field that I would hope would be funded pretty well. And you can see, I think that through some major recent initiatives, there's a huge new brain napping effort that's underway that's being run out of the Allen Institute up here in Seattle that represents a major application of spatial biology technology. It's just got, I think, $0.5 billion of research funding over multiple years ahead. So my hope is, regardless of what the budget does overall, the application of new technologies like spatial will be at the top of the priority list.

Got it. I appreciate that. And one final question for me. Appreciating the fact that China is only about 5% of your total business. You've also talked about the fact that you want to expand your exposure in the region. So I was wondering if you think back about all of the headwinds that you saw in COVID and the ongoing headwinds in the region, has your international expansion plan in China or your approach has changed in any way? And do you plan on further expanding from that 10 plus direct sales in the region in the next 12 months?

Yes. Our international expansion plans have not changed. Unfortunately, the investment we've made and the great team we've built in China haven't really yet had a chance to prove what they can do because of the rolling lockdowns that we've experienced over there. But I still have a high degree of confidence that they will succeed when provided with things kind of cool down with the virus. And we don't have any plans to reduce our base of investment there.

Thank you. There are currently no further questions registered. (Operator Instructions) There are no additional questions waiting at this time.

Very good. Thank you, everyone, for joining us today. A full transcript should be available as well as a replay in the next couple of hours. You can access that by dialing (866)-813-9403, the conference ID code 93-1170. Thank you again. This concludes our call.

That concludes the NanoString third Quarter 2022 operating results conference call. Thank you for your participation. You may now disconnect your lines.