德康醫療 (DXCM) 2006 Q3 法說會逐字稿

Welcome to DexCom's third-quarter results conference call. Today's call is being recorded. At this time, all lines are on a listen-only mode. There will be a Q&A session following today's Q&A session (sic). Instructions will be given at that time for your questions.

Now I'd like to turn the conference over to the Chief Executive Officer, Andy Rasdal. Mr. Rasdal, please go ahead.

Thank you all for joining us this afternoon for our third-quarter results call. Just real briefly here to get started, I'd like to just review the agenda of prepared remarks that we have. We will go over a quick financial review, give some color to the STS three-day sensor launch, provide some background on reimbursement, the R&D pipeline, current litigation, and then sum up and as always open it up for a few questions.

So with that, I'd like to turn it over to our Chief Financial Officer, Steve Kemper, to begin with the financial review.

Good afternoon, everyone. We will be making forward-looking statements today about various aspects of our business. So please refer to our 10-Qs, 10-Ks, S-1s and all of our SEC documents for a full discussion of our risk factors.

DexCom today reported a net loss of $13.4 million or $0.48 per share for the third quarter of 2006, compared to a loss of $6.3 million or $0.25 per share for the third quarter of 2005. For the nine-month period ending September 30, we reported a loss of $35.5 million or $1.32 per share, compared to a loss of $18.8 million or $1.12 per share for the same nine months in 2005.

Today we reported revenues of $841,000 for the third quarter of 2006, our second full quarter post launch. Cost of goods sold was $3.4 million for the third quarter and $7.6 million for the nine-month period ending September 2006. For prior 2005 period, these manufacturing costs were included in R&D as we were still in the development stage.

Comparing our second quarter of this year and third quarter of this year sequentially, we saw our revenues increase to $841,000 for the third quarter, up 76% compared to the $479,000 we recorded for the second quarter of this year.

Our cost of sales for the third quarter of $3.4 million compares to an adjusted cost of sales for the second quarter of $3.2 million. If you recall, our second-quarter cost of sales was lowered by $1.1 million due to a onetime credit for materials purchased and expensed during 2005. So, consequently, our reported cost of sales was $2.1 million versus a 3.2 if you excluded that credit.

Our adjusted cost of sales then increased approximately $200,000 from the second quarter on $362,000 in higher revenues, as we are just starting to see some operating leverage from our recurring sensor sales.

Research and development expense including stock-based comp decreased $409,000 to $4.6 million for the third quarter of 2006, compared to $5 million for the third quarter of 2005. Changes in the major elements of R&D expense included $172,000 increase in development expenses and $566,000 increase in clinical and regulatory expenses as we continued to invest in our next generation of monitors.

These increases were offset by a $1.1 million decrease in manufacturing expenses that are now classified in cost of sales. Major items in research and development expense include about $0.5 million in lower outside design services, partially offset by 312,000 in increased clinical expenses and $279,000 in higher stock-based compensation expenses.

Comparisons between 2006 and 2005 are more difficult for R&D, as manufacturing was included in R&D during 2005. So we think it is helpful to briefly describe the changes from the second quarter of '06.

When we compare the last two quarters sequentially, we see that we drove down our costs for R&D from $5.4 million in the second quarter of this year to $4.6 million in the third quarter of this year. The sequential decrease was due to lower clinical study costs, as well as lower tooling and fixturing costs, as we completed the initial setup of our production lines.

Selling, general and administrative expense, including stock-based comp, increased $5.3 million to $7.1 million for the third quarter of 2006 compared to $1.8 million for the third quarter of 2005. The increase was primarily due to $4.2 million in higher sales and marketing costs as we continue to launch our STS system in the United States. Changes in selling, general and administrative expense from the 2005 period include $2 million in higher compensation costs, $1.6 million for increased promotions and advertising, and $729,000 in higher stock-based compensation.

Sequentially from the second quarter to third quarter, our SG&A increased approximately $2.1 million, primarily due to higher salary recruiting associated with our new certified diabetes educators that we hired during the quarter, as well as higher costs for advertising and promotional programs. We also had higher trade show expenses related to our AADE attendance, including our very successful education program for the CDEs. Other G&A departments, such as IT, accounting, executive, were flat quarter-to-quarter.

Our results show that we are investing heavily in sales and marketing and endeavoring to hold everything else flat as we set in place systems and people today to drive our future top-line growth.

For the third quarter, in total we reported approximately $1.4 million in stock-based compensation costs. Our interest and other income net increased $287,000 to $818,000 for the third quarter of 2006 compared to $531,000 for the third quarter of 2005. The increase was due to higher cash and cash equivalents due to our April 2006 follow-on offering, partially offset by $29,000 in interest expense on our capital equipment bank line.

We ended the third quarter with a strong balance sheet, including $64.7 million in cash and securities. Comparing our cash balance at the end of June to our balance at the end of September, the cash balance declined only $7.7 million, and only $8.9 million if you add back the $1.2 million we accessed on our bank line to offset our capital expenditure equipment purchases.

We ended the quarter with about $119,000 in Accounts Receivable. Our Accounts Receivable days outstanding have been fluctuating between 9 and 12, so we are very obviously pleased with that type of performance. We also decreased inventories 1.3 million during the quarter, and ended [retired] payables, also helping to conserve our cash. We will continue to focus on cash conservation as we move forward.

Year-to-date, we have used approximately $34 million in cash and operating activities and raised about $48 million from the sale of equity securities. So our cash balance has increased over $14 million during the first nine months of the year.

With that, I will turn back to Andy Rasdal, our CEO.

Great. Thanks, Steve. I'd like to just start off with reviewing where we see we are in the STS launch here, in the second full quarter of launch. Just a couple of metrics, some of which Steve hit, but I'll quickly review and then build on that.

Overall revenue grew 76% from Q2 to Q3. Breaking that out just a little bit further, starter kit unit sales increased 8% quarter-to-quarter, although revenue for starter kits actually increased 21%. Average selling price for the durables increased 12%. So contrary to rumors, we're not giving away durables to patients free of charge, as the data indicates.

I think most encouraging for long-term viability of the business model was a 205% increase in sensor revenues quarter-to-quarter; and sensor ASPs remain unchanged from Q2 to Q3.

And then finally, I think the organization was very cash efficient in their Q3 efforts, with only $7.7 million in that cash used, so we are pleased with that.

Qualitatively now maybe a few thoughts and comments. We do believe that new patient and overall revenue growth in Q3 was limited by several issues. First, as we talked about on the last call at the end of Q2, the unexpected surge of new patients at the end of Q2 in June created a backorder situation until late July, thereby limiting the number of new patients we could recruit and address at the beginning of Q3. And also delaying sensor usage and thus the reorder from existing patients.

We also experienced an increased out-of-box sensor failure in both August and September, which obviously disappointed potential new patients and prescribers in their initial experiences with our system, again, thereby slowing new patient and account conversion through those months. We replaced these sensors at no charge for existing customers, also thereby impacting sensor reorder revenue.

The particular failure mode is acute; it essentially doesn't allow the sensor to start working initially. And thereby it doesn't create any safety concerns; only inconvenience for the user. We believe this failure is a result of the issues related to rapidly scaling our sensor manufacturing again here in the beginning of the third quarter, and we've taken appropriate actions to reduce and eliminate the problem. However, as a result of implementing many of these changes related to the issue, along with the increase in patient base during Q3, we have been supply constrained throughout October and are currently backordered on sensors.

We've also, like many of you, read the blogs and now here anecdotes of patients defeating the three-day shutoff and using the product beyond three days. We survey this and pay attention carefully. We are not aware of any performance or safety issues as a result of this off-label usage.

Obviously in the near-term, this has a potential to hurt short-term sensor revenue. However, strategically over the longer-term, especially prior to widespread reimbursement and the introduction of better-performing next-generation products, patients; daily use of CGM versus more sporadic use may ultimately be beneficial for the category. And ask we've said before, we believe our seven-day STS will have a much more robust end-of-life shutoff.

We also had a minor issue in our receiver -- this is the hand-held portion, which displays the data to the patient -- related to a component that we source outside of DexCom, and also believe we've resolved that issue. Again, although there have been no safety issues reported related to this failure, this may have also disappointed patients and prescribers in their initial experiences with the product, potentially slowing new patient account conversions. And most certainly, patients who experienced both a receiver and a sensor issue together would have been disappointed, and we had to work harder and longer to provide them with a positive set of experiences to enable them to enable them to benefit from CGM.

We have made progress in our manufacturing operations, a couple of which were supported by recent regulatory approvals. We've received approval from the FDA for a new method of applying the membrane systems to our sensor, which increases our manufacturing efficiencies and at the same time, based upon our data, reduces the variability.

We also now have the ability to use a much less expensive battery in our transmitter, thereby reducing the cost of our durables in the future. We expect that this lower-cost component will be first utilized in conjunction with our STS seven-day system launch, although we could incorporate it into the current three-day STS as needed.

In the category, we continue to see increasing excitement and clinical interest in the category of CGM, whether it be at the scientific medical meetings or in daily interactions with patients, physicians and educators. We continue to be surprised by the extent of our first-to-launch leadership position that is enabling us to do a number of things early. First is to establish an initial installed base of users early and first.

We're also, because we have a product out there, setting many of the category expectations for how continuous glucose monitoring products function, and the opportunity to build habit around those differentiable features. Perhaps most importantly, we're the first to obtain critical real field experiences that can't be obtained in clinical trials, which better enables us to introduce meaningful improvements to our product and services as desired by physicians and patients moving forward.

Looking to the future, we believe the key is accelerating the category growth for CGM. I think we're focused on three key aspects that are well proven tactics for category growth. First is technology and product improvement. We are intending to rapidly introduce new product generations with meaningful improvements we believe will appeal to broader segments of the target populations.

Second, we have to increase the clinical adoption and utilization with healthcare providers. We need to provide the evidence and experiences that establish CGM as the best standard of care for diabetes management. And third is reimbursement on a more widespread basis.

So I'd like to just delve a little bit deeper into each of those three and talk a little bit about some of our efforts, starting with the last, which is reimbursement. There has been continued support for CGM from Congress. Three additional letters have now come from the House and the Senate diabetes caucuses to Health & Human Services and CMS, supporting CGM, making a total of five letters that we are aware of from Congress on CGM.

A Senate hearing was held on September 27th on CGM and the artificial pancreas concept, largely initiated by Juvenile Diabetes Research Foundation efforts, all of which is providing, I think, an awful big impetus and an awful lot of political legislative support for the category and for diabetes in general.

In the third quarter we also had private meetings with key members of CMS and personally with the house diabetes caucus on the pending decision for establishing a HCPCS for CGM. We expect a decision on these HCPCS codes based upon Medtronics' application to come later this fall. If code is not created this year based on Medtronics' application, we'd expect to submit our own application for HCPCS code decision in 2007.

As many of you are aware, a Medicare Recovery Advisory Committee, or MCAC, meeting was held in August. We believe it's very positive that coverage issues for CGM are being considered formally so soon. The discussion amongst panel members at these types of meetings is very hard to interpret, even for people experienced in attending these types of independent and often eclectic panels. And so I'd like to sort of provide our take on this, having sat through a number of these and other FDA panels.

The meeting was predominately focused on the appropriateness of fingerstick monitoring, we believe, in people with Type II diabetes over 64 years of age. There appeared to be agreement about benefits of tight glucose control in people with type I diabetes. There's less agreement and less conclusive data on the benefits of fingerstick monitoring for older Type IIs, especially those not using insulin.

Their discussion was interesting, suggesting that the outcomes in landmark randomized clinical trials, such as the DCCT, would have been even better if lower, more normal A1C levels could have been achieved, and an inference that CGM may facilitate such results in the future.

There was then a small session which was devoted to CGM, and I think a number of the panel comments were overall encouraging. The panel seemed to understand the concepts, the advantages of CGM over finger sticks. They seemed to understand the concept that CGM studies could be conducted with endpoints based on reducing glucose variability, such as time spent, high time spent low, that we and others have published. But I also wanted to then validate it to A1C measures, some of which we believe we already have our investigators have published such data.

As expected, the panel would like to see more outcome as evidence; that is always desired. But I think importantly in that context, the panel members suggested that it should be relatively straightforward to demonstrate such outcomes for specific populations in relatively small studies. So I think all in all, we are pleased with the MCAC and the progress that was made there so early in the category.

On the private side, which openly has more patients involved, we continue to get anecdotal feedback from physicians and patients informing us of favorable coverage decisions by their insurance companies, including Blue Shield, Blue Cross, United, Prudential and Aetna, after they had already purchased their initial starter kit system. For now, patients seem to be making the decision to purchase and then seek independent reimbursement coverage from their payers later on. We are helping patients and physicians facilitate this process when and as appropriate.

We've also had a few instances where patients reported coverage had been declined, although we have few specifics. At the same time, however, we're also aware that some initial negative decisions have been overturned on appeal with physician support.

We would expect to begin having formal meetings with private payers and increase our activity in that field related to formalizing coverage decisions with different payers as the support and evidence for CGM builds.

We finally were very pleased and appreciative of JDRF's leadership and sponsorship for a large, randomized clinical trial that they have announced, which we believe should provide very credible third-party outcomes data to support reimbursement efforts. And we intend to both support and participate in this trial.

From a category development standpoint, as I've said, I think we continue to be surprised by the extent of our first-to-market and first-to-launch position. And with our leadership role in the category now, we are beginning to invest more in some of the activities to grow the category by increasing the acceptance of CGM as the best standard of care for diabetes management.

We continue to conduct clinical studies and publish results providing evidence of improved outcomes from the use of CGM. We just found out the results from our 86-patient seven-day STS pivotal trial, demonstrating statistically improved glucose control, have now been accepted for peer review application publication in the December issue of Diabetes Care. And so gives us another credible data source over a longer period of time for the short-term sensor platform. And we'd expect to continue providing such clinical evidence for CGM.

As planned, we've added approximately 15 clinical education specialists in the third quarter to help with the fundamental training and education of staff at key diabetes centers. Our direct field personnel exceeds 40 going into this fourth quarter here.

We've sponsored several educational sessions on continuous glucose monitoring, offering continuing medical education units for medical professionals in conjunction with diabetes meetings, such as the AADE. The interest and response for these events have been very enthusiastic, almost always selling out and being space-limited. We are leveraging these programs into an Internet accessible program to facilitate broader access for healthcare professionals and to, again, grow the category.

Finally and perhaps most importantly is the R&D pipeline. We believe very strongly that the rapid evolution of continuous technology and the introduction of next-generation products is critical to establish a category into building the franchise for DexCom. While we believe our first generation STS is helping to build the continuous category and helping patients better manage their diabetes today, we still remain focused on and committed to rapidly introducing these future generations.

I think we've continued to execute in a timely manner on our product development efforts, starting first with our seven-day STS. Our PMA supplement was filed during Q2, as we said, on schedule and as previously committed. We know the filing has been under active review by the FDA and we just received a list of questions at the end of last week.

We anticipated receiving questions, of course, on our seven-day supplements from the FDA. Given that we have just received the questions, we can't precisely forecast how long it will take to respond at this time, but where targeting the submission within three or four weeks should answer all of their questions.

We believe the seven-day STS offers many improvements, again, beyond just functioning for a full seven days, including improved accuracy and stability through the full time period, a waterproof claim, and another full-size reduction in needles size for the patient.

We also intend for the seven-day STS to be commercialized, as we said, with a much more robust and sophisticated end-of-use shutoff than our current three-day STS. And the seven-day obviously represents an attractive business model opportunity for us, but also for our patients and payers. We expect to price the seven-day higher than the three-day, but in a manner that lowers the cost per day to patients, but will still provide a higher margin percenter to DexCom.

On the pediatric approval side, we still feel strong commitment, both ethically and businesswise, to that population. We had a very productive in-person Pre-IDE meeting with the FDA in September on pediatric approval requirements. The meeting, we believe, was cooperative and collaborative. and I think many good things came out of it.

We've conducted a series of pediatric feasibility trials, and based upon our experience, we believe there are performance differences between pediatrics and adults, and we intend to commercialize a product optimized for the pediatric population. Based on our meeting with the FDA, we intend to submit an IDE prior to initiating pivotal trial, and we believe that IDE approval will be required for all pediatric CGM studies going forward.

We also had much talk about the replacement claim. We continue to be pleased with our initial labeling versus fingersticks and do not see current labeling as a major impediment with physicians or patients. We also had a productive in-person Pre-IDE meeting with the FDA in September on replacement claim trial design requirements and specific protocols.

Our replacement feasibility studies, published and presented at the ADA by our investigators, demonstrated statistically that diabetes management decisions made using solely the STS were more appropriate in achieving glycemic control than decisions using fingersticks alone. The data said 71% of decisions using solely the STS CGM system were considered appropriate versus about 53% for fingerstick alone.

The accumulated effect of these appropriate decisions statistically decreased the time spent high and low, and increased the time spent in the target zone when compared to the fingerstick decisions alone. So we presented this data, of course, to the FDA in the meeting, and we believe there is compelling evidence to continue moving forward in that direction.

However, we continue to believe that actual replacement performance is vastly more important than just labeling claims, and are still highly focused on evolving our technology to provide such performance. We're currently conducting human feasibility trials with a third generation STS, and we're in animals with a fourth generation technology aimed at more replacement like performance.

Our professional PC software to allow healthcare professionals to download, analyze and print patient data from the STS has been available since May, which obviously helped in our efforts with the larger regional diabetes centers. We received approval for our patient use PC download software in late August and began making it available to patients in September. Having the ability to download data from our receiver to a PC was one of our most frequent requests or complaints in the first months of launch. We're also developing next generation software products which would provide additional features.

On the long-term sensor program, we got a chance to review several months of data from our fully implantable G3 [OUS] human feasibility studies. And while we've learned a lot more about the technology, the performance at this time does not warrant investing in a large pivotal trial to seek approval at this time for the G3.

Currently, we're utilizing some of the team to leverage the most recent biomaterials from our LTS into our short-term sensor product platform in the future generations with very encouraging results, and currently assessing how to best translate the convergence we are seeing between our long-term and short-term technologies into the most meaningful product and technology platform.

Finally, on the product platform is the in-hospital application for technology. We believe the hospital application for CGM provides a very large and attractive opportunity complementary to our current STS, and we intend to aggressively pursue this opportunity. Each year in the U.S. there are more than 4 million people with diabetes admitted to the hospital.

Increasingly, there's a greater emphasis put on frequent monitoring and tighter glucose control for these patients, including from some hospital associations. There's also compelling published data supporting the need for frequent glucose monitoring and tight glucose control. This is specifically in the critical care areas of the hospitals.

We have now conducted our human feasibility trials with the STS and we believe we've initially characterized the in-hospital performance requirement for a CGM system and are currently developing two specific in-hospital product platforms -- one specifically for the critical care use, ICU, CCU type of usage; and another in the noncritical care uses, on the floors and wards, as a means to more easily, conveniently and effectively monitor glucose.

Relative to our litigation and the Abbott litigation, the court has granted DexCom's motion to stay the case. And one significant claim in that case was dismissed. These decisions effectively suspend litigation activity until late '07 or when the reexamination of the patents involved in the dispute are resolved.

The court also disallowed Abbott's attempt to amend the complaint with three additional patents. Abbott subsequently filed a separate action on those three patents. We've been aware of those patents and believe the complaint is without merit and will contest it.

Based upon examination of the prosecution history of those patents, we've also filed a request to have these three patents reexamined, as we did with the other four. Again, 74% of reexaminations result in reduced or invalidated claims, and the process takes about 21 months on average.

On the intellectual property side for our own, we had an additional four patents issued in the third quarter and two are allowed. So we currently have 11 issued patents, three allowed patents and 97 pending applications, making what we think is one of the strongest intellectual property portfolios in continuous glucose monitoring.

I guess finally just to sum up, we've continued to further establish and extend the first-to-launch leadership position in the category of continuous glucose monitoring. There continues to be growing interest in and excitement about CGM from patients, from physicians and from Congress.

We continue to receive very rewarding anecdotes from patients about the positive impact our STS system, even as a first-generation system, has had on their lives and their diabetes. Based on our experiences, we believe CGM will be increasingly important in the future of diabetes management.

Right now, we're focused on driving the CGM category growth through three well proven tactics, new product introductions, providing improvements that enable the device to field a broader section -- segments of the population; providing the education and clinical evidence to make CGM the standard of care for glucose monitoring with healthcare providers; and (indiscernible) widespread reimbursement.

We believe we have a full product pipeline from a new second-generation product waiting for FDA approval to a third-generation product already in human trials, to a technology platform for in-hospital application. We're pleased with the support from advocacy groups for reimbursement efforts and are increasing our own efforts with both public and private payers as momentum for CGM builds.

And perhaps lastly and most importantly, I'm pleased with the way the DexCom team has handled the issues expected with any first-generation product launch, and especially with the manner of which they've handled being thrust into category leadership responsibilities. We believe the experience we're obtaining in building a stronger organization and that new entrants in the categories will have to do their own learning while we have the ability to capitalize and execute on our early experiences.

So with that, I will open up to some questions.

(OPERATOR INSTRUCTIONS) Ben Andrew with William Blair.

Good afternoon, Andy. Just wanted to ask a few things. You mentioned some of the problems you've had with the sensors out in the field. Can you be more specific about what exactly the issues are as well? You said you are on back order. Is that constraining your ability to supply and grow the market at this point?

The answer to the first acute question, this is an acute failure that the sensor fails right out of the box when it first goes into the patient and it fails to essentially initialize and begin displaying to the patient. And we believe that that was a problem. The increased rate was introduced as we scaled manufacturing.

And yes, we enter October and have been backordered to support both existing customers and obviously to be able to drive new patients as quickly as we would like.

How severe is that backorder? When do you expect to clear it?

I can't really provide any estimates at this time. I mean it's a dynamic thing. The production guys start making good volumes, again a very good product, but demand is growing as well.

Okay. If I just look at the numbers you put up this quarter, it looks as though you sold about the same number of systems, about 900, as you did the previous quarter. And that would fit with an increase in sensor sales of about 200%. Is that in the ballpark?

Yes, I think we said up 8%, and 205% on sensor sales.

Right, okay. So from a standpoint of rep productivity, you're looking at something on the order of 300 or 400 in total for the organization per month at this point. How much do you think that was constrained during the quarter and what might we think about going forward?

I can't provide you any guidance. We're sort of starting October the way we started July, and that was sensor constrained going forward. And clearly, that impacts two ways. One is the ability to introduce new patients and sites to the product and for them to get it. But also, anytime you back order a patient, you ultimately ship that product, but you miss out on the usage period that they didn't have to use it as well.

And the patients that did experience what you call disappointments, whether it was the monitor or the sensor, what rate are they reordering versus dropping off the system completely?

I can't speak to that data at this time.

Okay. And then if we look at kind of the issues you mentioned on the pediatric side, you said that it's going to require an IDE because of differences in kind of product performance or relative to the patient. Are you seeing that in the field specifically at this point? Is that something that people have expressed concerned about relative to use in peds off-label?

I don't -- number one, I didn't certainly intend to provide any linkage between requirements for an IDE in pediatrics versus any type of performance. I think our understanding from the FDA is that need to have approved IDEs before conducting pediatrics studies is something that is really not [at their end], certainly not our end necessarily. Although I think we always thought we'd probably, as we always have, submit an IDE to try and get agreement on trial design and regulatory endpoints and forward.

And I think -- I can't speak to anything related to off-label usage or anything related to pediatric function now. I would say the data that we've observed in fairly extensive pediatric feasibility trials where we've carefully analyzed the product, says that there is a difference in performance between -- and probably physiology between young pediatric patients and full-grown adults. And we believe we've adjusted our sensor platform to be optimized for pediatrics.

You mean for the IDE study specifically?

Yes. Yes. We don't have pediatric approval today.

Right. And I guess I want to push you on the manufacturing at this point, because it is such a critical aspect of your ability to supply and grow the patient base. What is the major constraint -- is it the monitor or the sensor itself at this point?

Sensors.

Sensors. And so, I mean without making projections, seriously at what point do you think you can get off of backlog to the point where you can start to have the sales force be unconstrained? Because we're all sitting here trying to make some sort of projections relative to new patient adds.

Yes, I mean, I can't provide you exact guidance; it's a moving target. We continue to increase both the productivity and the throughput of the sensors, but we also, as we do that, we have increased end demand. And so we are trying to balance all of that.

But with any first launch, it is most important to scale appropriately, and the sensor failures that we had in August and September, we started out a little bit slow here coming into this quarter to make real sure that we have the best, highest performing, most robust product. Because nothing is more expensive and frustrating than to provide a disappointing experience initially.

Right. And then finally, I guess, just if you can, are we talking weeks, months, quarters, to get back off of backlog?

Something like that.

So it is not years?

It is definitely not years.

Okay, thanks a lot.

Alex Arrow with Lazard Capital Markets.

Thanks. Good afternoon, Andy. The seven-day STS, do you have a mechanism there to stop patients from using it longer, the way the three-day STS you mentioned was being used?

Well, the three-day STS has what I'll call software driven mechanism to do that. And the seven-day has a more hardware-facilitated shutoff. And so we believe it is much more robust and appropriate.

Okay, thanks. You mentioned the letters from the Congresspeople to get reimbursement. Can you give us your outlook for reimbursement? I think the last thing you publicly said, if I remember, was one year post approvals. So that would be approximately March of '07. Is that where your thinking is now and do the letters change that in any way? And are you talking more about CMS or private insurers?

Yes, I don't think we've ever given guidance of the year or anything specific. I think what we have before us is a HCPCS code decision at the Medicare label considering create national HCPCS codes for continuous glucose monitoring. Based upon Medtronics' application, we'd expect to have that decision made sometime later this fall. If successful, obviously, it moves things forward towards coverage decisions over the next couple of years. If it doesn't, then those things have too -- coverage and codes have to be either created simultaneously or sequentially coming back in '07, I think.

We didn't think that we would be in a position even to apply for a HCPCS code until kind of the calendar year after we had approval, which meant we always believed we'd be applying for HCPCS code at the beginning of '07 for a decision '07. We just have the opportunity to piggyback onto the Medtronic applications; the systems operate similarly.

On the private pay side, I think we're going to move prudently and appropriately in having the discussions at the local, moving to the national level, to try to ensure we give them the right information to make favorable and appropriate coverage decisions rather than trying to impose and force decisions too early.

Okay. So do the letters change anything? And do you know of any precedence of previous reimbursement decisions where Congressional letters seem to help?

I can only speak to my own experience. I've not seen this kind of legislative support for the reimbursement of a new category, especially so early. And I think it speaks very strongly of the understanding of the need to manage diabetes, both from a patient welfare and a national cost standpoint. And it's sort of a can't-wait disease state. We need to take early opportunity to see where we can impact the care of this disease.

And so I can't forecast how this will impact the timeliness or what the decisions will be. But certainly having the support of the House and the Senate and both the committees within and around those is a bit encouraging, unusual, and I think keeps us hopeful.

Okay, great. One last question, just on the run rate. I know in the month of June, you did about 500 systems. What is the most important factor that changed that run rate? Was it the sensor constrained back order, was it the out-of-box performance or what?

I don't know if I could categorize them. I think those are -- those are certainly both top ones right there. You have to have the product and product -- the most important thing in any product category is the product has to work appropriately and meet expectations. So I think both of those are high priorities.

Okay. Thanks, Andy.

Tom Gunderson with Piper Jaffray.

Good afternoon. The sensor issue, Andy, did I hear you correctly, the root cause is identified, it's been fixed, and now it's just a matter of implementing and getting capacity up. Is that right?

Well, you know, I say (indiscernible) and I believe we've identified the root cause and I believe we have fixed it, based upon our ability to re-create the problem and eliminate both in testing and in other trials.

And we're currently back up -- and we have been. We never had to stop shipping product; we'd just needed to slow down and make sure we could initiate the changes in an appropriate manner that ensured the product going out here was characterized appropriately to meet those standards. And so I think we're boosting volumes on a daily basis as we gain confidence in that.

And in the answer to previous questions, we've gotten a sense that you were in backorder in July and backorder now in October. Was that always a pressure for the other months, or was there a point there where you were unconstrained?

I don't know that we'd say that we'd necessarily be unconstrained; it's always been tight. But clearly, we didn't have the backlog in August and September that we had in July and we had going into October.

Okay. And then you mentioned that ASPs were stable for the sensor sequentially and up for the initial systems. Do you care to tell us what those ASPs are?

No, I'm afraid I'm going to have to decline at this time, Tom.

Okay. What about mix of revenues between sensors and systems?

I think other than to talk about the growth that we did, I don't want to provide specific balance of those things yet. Perhaps in the future.

Last quarter, you talked about the -- you had to do the one-offs in the early months. And then in June, you got to implement the strategic plan and actually do it, focusing on the top 45 centers, and gave us a percentage number. Was that carried through in Q3 and are you focusing on those 45 centers or has that number gone up?

I think strategically we'd still like to stay on the top 45 to 50 centers and we are. But we also need to help patients in some of the other areas. And so it's a balance; yes, we've expanded the number of centers that can be trained and know how to both prescribe and train patients on the product. But the bulk of our effort still is focused on those top 45 to 50, given that population of patients that represents.

Okay. And then last question. On the sales force add, mostly with the -- I think you said you were going to add 15 CDEs during the quarter and you said you did. Can you give us a little bit more color on the timing of that? Do you feel that as we're sitting now on October 23rd that those 15 are up to speed and you have a fully-trained direct field force?

Yes, I think certainly the original group of people, I think, had six months of experience. And in my experience, that is about the right time to really start to become productive.

The additional clinical education specialists, most of them joined us late in the second half of the quarter. We've all done all the formal training, and they are out there. There is no shortage of activity, so obviously gaining valuable on-the-job training. And so I think we entered this quarter with a full complement of over 40 people.

Clearly, there is some of those people who are going to need to get another three to four months' of experience. But they all come with great track records and great backgrounds in diabetes. And so I'm pretty confident that they are making a big impact today.

Okay, thank you.

Wade King with Montgomery & Company.

A number of my questions have been asked already. I'll focus on a couple other areas. To go back to the sales force. Andy, you said 40 direct field sales force total, so that includes your reps and your clinical educators, is that correct?

Yes, we're in excess of 40 now, going into the quarter.

Can you give us a breakdown of the reps versus the education specialists?

I'm sorry -- I'm going to keep that as proprietary.

Okay. Secondly, you gave us an update on the PMA supplement for the seven-day sensor. If you respond in the month of November to the Agency's questions, once again, your understanding would be that their clock is on six months starting from the time of your response -- is that correct -- assuming they don't ask you further questions, to give you an answer?

Yes, they have six months.

Okay, so that would be May, is that correct?

I'm sorry?

That would be May, assuming you respond in the month of November -- is that right?

When we received the questions, obviously the clock stops for the FDA, as it did for our last regulatory approval. And then we will provide the response, and then yes, by letter of the law, they probably have up to 180 days to make their decision and respond.

Okay, very good. Can I ask you, in follow-up to some of the prior questions on the supply constraint, the sensor and receiver issues you described, you indicated that you had had some sensor failure issues in August and then September and then a minor issue in the receiver. Have you had any sensor failure reports from the field in the month of October?

Probably -- I would imagine that we do. But we are not failing sensors for the problem that we experienced that I referenced.

So once again, you may have had occasional issues that come up in the normal course of your expectations with this type of product. But you haven't had the sensor failure replications in the month of October, as far as you are aware?

Certainly the increased rate that we saw in August and September, we are not seeing currently.

Okay. And how about on the receiver side, Andy? You mentioned that as more of a minor issue, suggesting that that appears to be corrected and that the backorder situation is focused on the sensors as opposed to receivers. Have you had any replication of the minor issue with the receivers in the month of October?

Not to our knowledge. That was a specific hardware component issue, so that is a little more identifiable of a root cause when it is that simple.

Okay. And lastly, please, the decision on the Medtronics HCPCS application, what specifically do you expect that to provide you in terms of benefit? You've indicated clearly, consistent with prior commentary, you would expect to submit your own application in 2007.

But if Medtronic gets a favorable decision later this year in that regard, will there be specific additional things that you would expect to do in follow-up or are there specific benefits that you think it would provide you as a company in trying to develop and grow the market from here?

Sure, I think the HCPCS code, as it was constructed and applied for by Medtronic, essentially it created code for the receiver, the transmitter and the disposable sensor. So I think it would be applicable to all three systems that we see, either in commercialization or about to come to commercialization.

And this essentially establishes a national code, which we believe is warranted as diffusion of the category increases, especially with other folks coming forward, and would enable then the regional contractors for Medicare to utilize that code and make their own coverage decisions. And for the private insurers who are making decisions today, it would facilitate, streamline their ability to do so and to make those decisions. That is the primary benefit.

And then once the codes are created, we begin to focus more extensively on creating favorable coverage decisions.

All right, thank you. And actually, one last thing, if you don't mind. Last quarter, you provided a placements or shipments number for the last month of the quarter, then June, 500 systems. Would you be willing to provide the same for the month of September applied to the third quarter?

I think we've probably given the financial detail we are comfortable giving at this time.

Okay, gentlemen. Thank you.

Suraj Kalia with Rodman & Renshaw.

This is actually Charles (indiscernible) behalf of Suraj. Just one quick question. Just if we could get your installed base number as of quarter end.

No, we are not going to release actual patient numbers for this quarter for proprietary reasons right now.

Okay, understood. Thanks.

Sara Michelmore with Cowen & Company.

Hi, Andy. Hi, Steve. I think a lot of my questions had been answered also. Andy, if you could just talk about -- you mentioned it quickly -- about that you not only had dealt with some patient frustration with some of the bugs and supply issues, but also the clinicians. Can you just talk about what some of the feedback from the clinicians has been and what you guys have tried to do to work with the clinicians and trying to help them be patient with some of the issues that you've had to overcome here?

Well, I think in this field, the clinicians and the educators in diabetes are tremendously robust people who I think are as focused in on improving patient care as any specialty or subspecialty I've ever been involved with. And while we certainly want to make any initial experiences with the new technology as problem-free and seamless as possible, because that enables them to do their job easier and enables the patients to benefit more quickly, I think there are enough positive experiences that have been created even around those minor issues that people are starting to see the benefit, and for the most part, are committed to trying to get patients, as they deem appropriate, into continuous glucose monitoring as a way to better understand and better manage their disease.

I think the thing that was interesting this quarter was given that we put a number of patients in the second quarter on continuous glucose monitoring, this is really the first time when those patients have been on for three or for months, and they are coming back to their physicians for their quarterly Hba1c or A1C level checks, and the feedback we get that's not solicited but it's anecdotal is they are seeing --many patients are seeing significant reductions in their A1Cs for the first time ever.

And whereas this is something that you can imagine being a person with diabetes, every quarter you have to go take a final exam and you have this A1C test, these people dread it. And I think for the first time after experiencing continuous, they're actually looking forward to come back and get there and see what kind of additional improvement they can and will make.

So I think it is encouraging. It doesn't give us an excuse not to provide robust product. It's not unusual to have hiccups and issues with first-generation products and I'm pleased with the way the organization has responded. And it is our goal to eliminate as many as quick as we can.

And I think clinicians just by their nature in this field are supportive of that.

Great. I appreciate the color on that. Can you talk a little bit too about resource-wise how you feel about the ratio you have right now of the clinical specialists versus salespeople, and how you think that your coverage is there and what your plans are in the next six months or so?

Yes. I think our coverage is sufficient, both upon the educators' side, where we need to take a stronger leadership role in driving continuous as a category, and then at the rep level, where we drive business and penetration within that. I'm pretty pleased, and I think that certainly given the current level of competitive efforts, we are fine. Many things can change. The category can begin to accelerate as critical mass is achieved, or competitive efforts can drive both increased demand and need for reps. But I think we are pretty comfortable right now with field force in excess of 40 going forward here.

Okay, that's great. And I just want to make sure I'm interpreting your comments on the implantable long-term sensor (indiscernible). You guys are not now pursuing that program, is that correct?

No, it's not -- we still have a program and we're still continuing to develop technology and we're trying to -- when you have a short-term sensor that keeps working longer and longer, I guess our view of the world three years ago was in order to achieve the stability and the length of life that we are now achieving in our next-generation short-term sensor partners, we thought we had to be fully implantable. That's what drove that need.

The technology is being a little bit more robust and we're pleased with that. And I think we are trying to figure out how to best converge those into something which either extends life convenience or stability beyond perhaps the current short-term product.

So by no means is just -- what I said specifically is based upon the performance and the convergence, it wouldn't make sense to conduct a lengthy and very expensive pivotal trial for the current G3 model.

For the G3? Okay. I appreciate the clarification there, Andy. And lastly, I don't know that you guys have talked about this publicly. I'm sure you have contemplated it. What should we expect in terms of the transition from a seven-day product to the three-day product? Will you stop selling the three-day product and is there going to be some sort of incentive program that you're going to be able to offer current patients on the three-day sensor to convert to the seven-day?

We haven't given any specific guidance on the timing and how that would go. Philosophically, one in answer to your last, is, yes, we would expect to treat existing patients very favorably. They are the early people with us and we value them.

And second, I think in general in terms of the transition, I think we would be hesitant to force something upon the marketplace; would rather let the marketplace tell us about what they prefer between three and seven days.

And are you guys picking up at all that there is people out there that are waiting to start the product for a seven-day sensor, or you just think that people that want to test the product are willing to test out the three-day?

I don't have anything that I could quantify for you on that at this time.

All right. Thank you, Andy.

Steve Ogilvie with ThinkEquity.

Hi, guys. First, on the sales force, just curious, obviously, with the manufacturing issues and the backorders and incentives there how you are managing your sales forces' expectations and how you are trying to retain them, as they might be a little frustrated at this point?

And then just more of a technical question, on the seven-day sensor. Are the durables that you are selling right now going to be completely compatible --

I don't see or sense any frustration within our field representatives. I couldn't be more impressed with the leadership each and every one of them is demonstrating within their own territories. And the senior leadership of Rod Kellogg, I think this is just unparalleled. None of these people are neophytes, none of these challenges are unexpected in launching a new technology into a brand new category by a brand new company.

And so we haven't had to do anything. They are busy every day. And I said, we didn't stop shipping product altogether. And so they are being more precise and more targeted.

And just on the seven-day sensor, is that completely compatible with the durables you are selling now?

I don't -- the answer is essentially yes. There will be ways for things to be compatible, although we think the advantages certainly in the new software and the receiver and even the transmitters, people want to upgrade those systems, but they won't lose the ability to use sensors.

Great, thank you.

They won't be able to benefit from all the advantages, but they won't lose it.

Steve, did you have any additional questions?

No, that is all. Thanks.

William Plovanic with First Albany.

Great, thank you. Good evening. A lot of the laundry list of questions have been asked. The one thing you did discuss briefly was the G3 in humans on your short-term sensor. First of all, is that in a pivotal -- are you in pivotal trials at this point? And can you give us a feel for the bells and whistles that you plan to incorporate or that are incorporated in that iteration of the product?

I'm not sure I understood the question. The G3 on the short-term platform?

Your third generation short-term that you mentioned --

Oh, the short-term. Okay. G3 is the full implantable; we are not in pivotal trials. In fact, we made the decision not to invest in pivotal trial for that.

On the third-generation STS, no, we have conducted a number of feasibility trials. And I don't want to give any insight into what the product features and benefits and improvements would be at this point for competitive reasons.

That is all I had. Thank you.

Great. I think that wraps up all of the Q&A. Thanks for joining us.

Thank you.