Good morning. My name is Dan, I will be your conference operator today. At this time, I would like to welcome everyone to the Biogen second quarter 2018 financial results and business update conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question at that time, simply press star followed by the number 1 on your telephone keypad. If you would like to withdraw your question, you may press the pound key. Thank you. I would now like to turn the conference over to Mr. Matt Calistri, Vice President, Investor Relations. Please go ahead.
Thank you, welcome to Biogen's second quarter 2018 earnings conference call. Before we begin, I encourage everyone to go to the investor section of biogen.com to find the press release and related financial tables, including a reconciliation of the GAAP to non-GAAP financial measures that we will discuss today. Our GAAP financials are provided in tables 1 and 2. Table 3 includes a reconciliation of our GAAP to non-GAAP financial results. We believe non-GAAP financial results better represent the ongoing economics of our business and reflect how we manage the business internally. We have also posted slides on our website that follow the discussions related to this call. I would like to point out that we will be making forward-looking statements which are based on our current expectations and beliefs. These statements are subject to certain risks and uncertainties, our actual results may differ materially.
I encourage you to consult the risk factors discussed in our SEC filings for additional detail. On today's call, I'm joined by our Chief Executive Officer, Michel Vounatsos, Dr. Michael Ehlers, EVP of Research and Development, our CFO, Jeffrey Capello. We will also be joined for the Q&A portion of the call by our Chief Medical Officer, Dr. Al Sandrock. Before I conclude, I would like to also point out that starting with this earnings release and call, we're posting press releases related to earnings calls and investor events on the investor section of Biogen's website, www.biogen.com, will issue a statement on X when they become available. We will do this instead of publishing press releases related to future earnings calls, earnings releases, and investor events via Newswire services. Our Twitter handle is @Biogen. I'll turn the call over to Michel.
Good morning, everyone, thank you for joining us. First, let me begin with some financial highlights. Compared to the same period a year ago, Biogen's second quarter revenues grew 9% to a record $3.4 billion. Second quarter GAAP earnings per share grew 3% to $4.18, non-GAAP EPS grew 15% to $5.80. Our performance for the second quarter was strong, we are raising our full-year revenue guidance. Let me review the accomplishments we delivered in the second quarter. First, our MS core business, including Ocrevus royalties, delivered revenues of $2.3 billion, a decrease of 2% versus prior year. Adjusting for inventory dynamics, global MS revenues were stable versus prior year. The number of patients on our MS products globally remains stable versus the prior year as patient growth outside of the U.S. offset increased competition within the U.S.
We were pleased to see that new patient starts for TECFIDERA in the U.S. in the second quarter were at the highest level since the launch of Ocrevus just over a year ago. Outside the U.S., we continued our strong progress in several international markets, such as Japan, where close to half of MS patients are now treated with a Biogen product. Second, SPINRAZA. Global revenues grew to $423 million, driven by quarter-over-quarter and year-over-year revenue growth in the U.S., and even greater revenue growth outside the U.S. The number of commercial patients on SPINRAZA increased by approximately 28% from last quarter, and we now have over 5,000 patients on SPINRAZA, including the expanded access program and clinical trials. In the U.S., we are encouraged by the progress we are making with adults as the number of adult patients on SPINRAZA grew by over 20% compared to last quarter.
We remain focused on increasing our reach with adults by supporting the establishment of adult treatment sites, working to broaden our reimbursement coverage, and communicating the potential value of SPINRAZA for this patient population. Outside the U.S., the pace of reimbursement for SPINRAZA across multiple geographies led to meaningful revenue growth, with significant revenue contribution, not just from Europe, but also from Asia Pac and Latin America. We expect the launch of SPINRAZA to continue expanding globally throughout 2018 and 2019. We have filed for regulatory approval in seven additional countries, including China, and we expect to file in up to four more countries by the end of the year. We also secured formal reimbursement in seven additional countries and now have formal reimbursement in 24 markets. Third, we made significant progress advancing our differentiated neuroscience pipeline, a critical source of potential future value generation for Biogen.
We entered into an exclusive option agreement to acquire TMS-007, a phase II compound being developed for acute ischemic stroke, which we discussed in detail during our webcast on June 28th. Today, we also announce we acquired an early-stage program from AliveGen for muscle enhancement, which will initially be studied in SMA and has the potential to benefit patients across a broad range of neuromuscular diseases, including ALS. In line with our expectation, this month, we completed enrollment of both ENGAGE and EMERGE, our two phase III trials studying our investigational treatment, aducanumab, in early Alzheimer's disease. In May, we exercised our option with Neurimmune to further reduce the royalty rates payable on potential future sales of aducanumab. Along with our collaboration partner, Eisai, we announced positive top-line results of the final analysis of the phase II study of BAN2401.
We believe these results provide evidence to further support amyloid beta as a therapeutic target for Alzheimer's disease. Importantly, our cash generation remained very strong and continue to provide us with significant optionality and flexibility to allocate capital. In addition to the two BD deals I discussed earlier, during the second quarter, we repurchased $2.75 billion worth of shares. We exercised our option to increase our ownership of Samsung Bioepis to approximately 49.9% and will pay approximately $700 million for these option shares. We believe our share of Samsung Bioepis is worth significantly more, and we believe this investment is aligned with our goal of creating meaningful value for our shareholders. We invested $1 billion in a new collaboration agreement with Ionis, which capitalizes on Biogen's expertise in neuroscience R&D and Ionis's leadership in antisense oligonucleotides with the goal of meaningfully expanding our neuroscience pipeline.
As we have demonstrated in the past, we are committed to maximizing returns for our shareholders while continuing to bring innovative therapies to patients. Something that demands a thoughtful approach towards all our investment over both the short and the long term. As we aim to continue allocating capital to expand our pipeline, this quarter, we hired Daniel Karp as EVP, Corporate Development. We look forward to the contribution Daniel will make as we continue to refine our corporate strategy and evaluate new opportunities for potential business development and M&A. In summary, Biogen executed on its strategy and continued to deliver noticeable progress. Our core MS business demonstrated resilience. We continue to believe there is significant growth potential for SPINRAZA, and we are encouraged by our recent progress in reaching adults and expanding market access globally.
We enhanced our pipeline with the addition of two more assets as we aim to diversify our neuroscience R&D portfolio and build depth in our core and emerging growth areas. We completed phase III enrollment for aducanumab, which we believe is one of the most promising investigational therapy for Alzheimer's disease. Together with Eisai, we announced positive top-line results for BAN2401, which we believe provides further support for the amyloid hypothesis. We have made progress implementing a leaner and simpler operating model designed for the future. We strategically allocate capital as we invest to develop and expand our neuroscience portfolio, again, with the goal of maximizing shareholder return and bringing innovative therapies to patients. I will now turn the call over to Mike for a more detailed update on our recent progress in R&D.
Thank you, Michel, and good morning to everyone on the phone. This is a very exciting time for Biogen R&D as we have made meaningful advances across our industry-leading Alzheimer's portfolio, added two new clinical-stage assets, and continue to progress our broader neuroscience pipeline. Let me review the progress we made across our core and emerging growth areas. I'll start with the significant advances we made in Alzheimer's disease and dementia. Earlier this month, along with our collaboration partner, Eisai, we announced positive top-line results from the final 18-month analysis of the phase II study of BAN2401, a beta amyloid antibody being studied in early Alzheimer's disease, including subjects with mild cognitive impairment due to AD and a subset of mild AD. Based on data from 856 patients, this study demonstrated both the potential disease-modifying effect on clinical function and beta amyloid lowering in the brain.
Specifically, the study demonstrated a dose-dependent reduction in beta amyloid in the brain, as measured using amyloid PET, as well as a concomitant dose-dependent slowing of the rate of clinical progression as measured by the Alzheimer's Disease Composite Score, or ADCOMS, a novel endpoint which incorporates elements of validated measures, including ADAS-Cog, CDR Sum of Boxes, and the Mini-Mental State Examination. While the study did not achieve its primary outcome measure at 12 months based on a Bayesian statistical analysis, using conventional frequentist statistics, the highest treatment dose of BAN2401, 10 milligrams per kilograms biweekly, showed a statistically significant effect on ADCOMS as early as six months and also at 12 months and 18 months. These new data provide compelling evidence for further support blocking or clearing beta amyloid as a therapeutic approach for Alzheimer's disease. In this study, BAN2401 demonstrated an acceptable tolerability profile through 18 months.
The most common treatment-emergent adverse effects were infusion reactions and amyloid-related imaging abnormalities, or ARIA. These results will be presented by Eisai in an oral session at the 2018 Alzheimer's Association International Conference tomorrow, July 25th, at 3:30 P.M. Central Time in Chicago. A live webcast of the presentation is expected to be available. In addition to BAN2401, we continue to advance aducanumab, a distinct antibody that binds to both soluble and insoluble aggregated forms of beta amyloid, including oligomers, protofibrils, and fibrils. Aducanumab demonstrated significant plaque removal, as well as slowing of cognitive decline in the phase I-B PRIME study. Both phase III studies for aducanumab, ENGAGE and EMERGE, are now fully enrolled. Eisai also reported phase II data for elenbecestat, a BACE inhibitor designed to reduce the production of beta amyloid.
Elenbecestat demonstrated an acceptable safety and tolerability profile in this 70-patient study. The results demonstrated a statistically significant difference in beta amyloid levels in the brain as measured by amyloid PET. A numerical slowing of decline in functional clinical scales was also observed, although this effect was not statistically significant. Elenbecestat is currently being evaluated in two phase III studies. Beyond beta amyloid, we are advancing a number of assets targeting tau, which we believe plays an important and complementary role in Alzheimer's disease pathogenesis. Specifically, in the second quarter, we dosed the first patient in the phase II Alzheimer's disease trial of BIIB092, the anti-tau antibody we licensed from BMS last year. Turning to MS and neuroimmunology. We previously communicated data regarding the potential impact of extended-interval dosing on reducing the risk of PML for TYSABRI.
We have now finalized plans for a worldwide prospective study to evaluate whether extended-interval dosing provides a similar level of effectiveness to standard dosing. We believe this study will provide sufficient data to inform clinical practice, and we plan to initiate the study by the end of the year. In collaboration with Alkermes, we continue to advance BIIB098, or diroximel fumarate, as a novel oral fumarate in phase III development for the treatment of relapsing forms of multiple sclerosis, where we aim to provide a differentiated GI tolerability profile. This quarter, we received data from the ongoing open label long-term safety study that will be included as part of the new drug application to support registration. The safety data were consistent with what has been presented at prior scientific meetings and includes results from approximately 700 patients, of which roughly 500 were treated for at least one year.
Alkermes expects to file with the FDA by the end of the year. We continue to enroll the EVOLVE-MS-2 head-to-head tolerability study versus TECFIDERA, with data expected in the first half of 2019. This study is employing an adaptive trial design in which results from part A are used to inform part B. We are committed to expanding our leadership position in multiple sclerosis, and we believe this new-generation fumarate is an important part of our strategy. Turning to our progress in movement disorders. We are actively recruiting a phase II study of our anti-tau antibody BIIB092 in Progressive Supranuclear Palsy, or PSP. BIIB092 is a potent selective anti-tau antibody that has demonstrated reductions in CSF free tau levels of over 90%. We expect to complete enrollment in this 52-week study in the coming months.
The primary endpoints are efficacy as assessed by the PSP Rating Scale, a 0-100 measure of disability across six domains, as well as safety. If the data from this phase II study are positive, we believe there is a potential it could be registrational and could thereby possibly support a regulatory filing. Moving to neuromuscular disorders, including spinal muscular atrophy. We continue to build on the growing body of clinical evidence for SPINRAZA as the standard of care in SMA. Earlier this month at the International Congress on Neuromuscular Diseases, we presented a new analysis of Type 3 SMA patients initially enrolled in the CS2 study between the ages of 11 and 14, followed by its open label extension study.
Of the 11 patients included in this analysis, 10 demonstrated stabilization or clinically meaningful improvement in motor function as compared to the decline over time expected based on natural history. As we aim to bolster our leadership position in SMA, we are pursuing new therapies across multiple modalities. To that end, as Michel mentioned, today we were pleased to announce the acquisition of two programs from AliveGen targeting the myostatin pathway for potential muscle enhancement across a range of neuromuscular diseases, including SMA. Myostatin is a widely studied negative regulator of skeletal muscle mass that signals to ActRIIA and ActRIIB receptors. ALG-801, now known as BIIB110, and its backup compound, ALG-802, are pseudo-hybrids of the ActRIIA and ActRIIB receptors and act as ligand traps to inhibit myostatin as well as other muscle-suppressing ActRIIA/IIB ligands, including activin A and activin B.
Recent research has indicated that inhibiting both myostatin and activins may be necessary to maximize clinical efficacy. However, previous approaches to simultaneously inhibit these ligands have resulted in inhibition of BMP9, a related growth factor involved in both vascular and bone homeostasis, whose inhibition increases the risk for vascular-related toxicity. AliveGen has carefully designed BIIB110 with the aim of sparing BMP9 while sequestering the muscle-inhibiting factors myostatin and activins A/B, conferring the potential for both greater efficacy and improved safety compared to other compounds in the class. Preclinical data for BIIB110 have demonstrated compelling increases in muscle mass, as well as improvements in bone density, which we believe may further differentiate the asset. BIIB110 is currently in a phase I-A study in healthy volunteers.
We believe there is a broad opportunity for BIIB110 with an initial focus on development as an adjunct therapy for SMA, but with potential applicability for other rare diseases such as ALS and Charcot-Marie-Tooth disease, or more prevalent conditions such as sarcopenia. Development in SMA, in particular, has the potential for complementary benefit when used in combination with a disease-modifying therapy such as SPINRAZA. As previously communicated, we have been working to advance our SMA gene therapy program in collaboration with the University of Pennsylvania. We were recently informed that the FDA has placed our IND on clinical hold. We believe we have a path forward, and we are working to lift the hold as quickly as possible. We are not able to provide details or updated timing, but we intend to bring the program to the clinic as soon as possible. Moving to acute neurology.
We have now begun initiating sites in the phase III CHARM study of BIIB093, our first-in-class IV glibenclamide therapeutic targeting brain edema and large hemispheric infarction. As Michel mentioned, during the second quarter, we also entered into an option agreement for TMS-007, which is a potential best-in-class thrombolytic drug candidate currently in phase II in Japan. Due to its highly targeted mechanism of action, we believe TMS-007 may have several important benefits compared to the current standard of care of tissue plasminogen activator, or tPA, including a potentially longer treatment window and lower bleeding risk. I encourage you to review our recent webcast from June 28th for more details on BIIB093, TMS-007, and our overall strategy to target stroke.
Overall, we have made significant progress advancing our pipeline, including encouraging new data in Alzheimer's disease, continued new additions to our pipeline, and strong execution across our clinical programs. Since the beginning of 2017, we have meaningfully expanded our pipeline by adding or advancing 14 clinical stage programs. We are building our pipeline with the goal of being the leader in neuroscience. We see compelling evidence that this is just the beginning of a period of transformative breakthroughs in the treatment of neurological diseases. We are positioning Biogen to take full advantage of the innovation and progress that we anticipate will continue to occur in neuroscience. With that, I will now pass the call to Jeff.
Thanks, Mike. Good morning, everyone. I will now review our financial performance for the second quarter of 2018, starting with revenues. Total revenues for the second quarter were $3.4 billion, growing 9% year-over-year. Let me now provide more detail on our MS franchise revenues. Overall, our MS business delivered revenues of $2.3 billion in Q2 2018, including Ocrevus royalties of approximately $113 million. MS revenues in Q2 '18 were down 6% versus prior year without Ocrevus royalties or down 2% including Ocrevus royalties. Total ex-U.S. MS product revenues benefited by approximately $42 million versus the prior year due to changes in foreign exchange rates net of hedging. Global second quarter Tecfidera revenues were $1.1 billion, a 2% decrease versus the prior year.
This included revenues of $826 million in the U.S., a decrease of 6% versus Q2 '17, and $261 million outside the U.S., an increase of 11% versus the second quarter 2017. In the U.S., we saw an inventory drawdown of approximately $35 million. We also believe we experienced the final quarter of difficult comparisons due to the timing of the Ocrevus launch. Both of these factors weighed on our U.S. Tecfidera performance for the second quarter. We continue to see stabilization in share of total prescriptions and growth in Tecfidera share of new prescriptions, building on the favorable trends we saw in the first quarter of 2018. New patient starts for Tecfidera in U.S. in Q2 '18 were at the highest level since the launch of Ocrevus just over a year ago.
In addition, we were very pleased with TECFIDERA's performance outside the U.S., driven by strong year-over-year patient growth across each large European market, solid emerging market growth, and particularly strong performance in Japan, where TECFIDERA has now reached over 25% market share. For the balance of the year, we expect continued patient growth for TECFIDERA outside the U.S., though we are closely monitoring some evolving pricing pressure in certain European markets. Ex-U.S. TECFIDERA revenues benefited by approximately $17 million versus the prior year due to changes in foreign exchange rates net of hedging. TYSABRI worldwide revenues were $467 million this quarter, a decrease of 6% versus the second quarter of 2017. This included $266 million in the U.S. and $202 million outside the U.S. In the U.S., revenues declined 8% versus the prior year, primarily due to the launch of Ocrevus.
We are encouraged that this was the second straight quarter of stable U.S. volumes for TYSABRI as we saw stability in TYSABRI share of both new prescriptions and total prescriptions. Outside the U.S., TYSABRI revenues decreased 2% versus the prior year. TYSABRI patients increased in all major European markets versus the prior year, along with strong double-digit patient growth in the emerging markets. Ex-U.S. TYSABRI revenues this quarter benefited by approximately $12 million versus the prior year due to changes in foreign exchange rates net of hedging. interferon revenues, including both AVONEX and PLEGRIDY, were $626 million during the second quarter, a decrease of 9% versus Q2 2017. This included $445 million in the U.S. and $181 million in sales outside the U.S.
In the U.S., we saw a decrease of interferon inventory of approximately $10 million in Q2 2018. Ex-U.S. interferon revenues benefited by approximately $12 million versus the prior year due to changes in foreign exchange rates net of hedging. Overall, U.S. MS performance versus the prior year was impacted by both the launch of Ocrevus and channel dynamics. We expect both factors to be less significant on a year-over-year basis as we move throughout the year, with a potential benefit if there is channel build again in the second half of this year. Given the trends we see, combined with the anniversary of the Ocrevus launch, we expect second half 2018 worldwide MS product revenues to be stable when compared on a year-over-year basis. Let me now move to SPINRAZA. Global second quarter SPINRAZA revenues were $423 million.
This included revenues of $206 million in the U.S., representing 10% growth compared to the first quarter, and $217 million outside the U.S., representing 23% growth compared to Q1. This is the first quarter where revenues outside the U.S. exceeded revenues in the U.S. The number of patients on therapy in the U.S. increased by over 10% as compared to the end of the first quarter, including a greater than 20% increase in the number of adults on therapy, as we continue to make progress in reaching adults in the U.S. We saw a continued increase in the revenue contributions for maintenance doses this quarter. In the U.S., approximately 55% of SPINRAZA revenues in the second quarter were attributed to maintenance doses as compared to approximately 40% in the first quarter. In the second quarter, the average doses per patient was approximately 1.1, roughly the same as in Q1.
In the second quarter, approximately 15% of U.S. SPINRAZA units were dispensed through our free drug program, a decrease from approximately 20% in Q1. As some patients enter their second year on therapy, we have benefited from improved insurance coverage we have today versus a year ago, allowing some patients to transition from free to commercial product. We believe both inventory levels and discounts and allowances for SPINRAZA were relatively flat in the second quarter versus Q1. We continue to believe there is a significant opportunity in adults in the U.S., and we are very encouraged by the progress we're making. We estimate we now have reached approximately 50% of all infants, 50% of pediatric patients, and 10% of adults.
Of note, our data indicate that approximately 5% of the prevalent SMA population are infants, 35% are pediatric patients, and 60% are adults, highlighting the large number of untreated adult patients that we believe could benefit from SPINRAZA. Assuming a similar free drug percentage and continued progress in reaching adults, we expect to show continued stability in U.S. SPINRAZA revenues compared to the second quarter of 2018. Outside the U.S., the number of commercial SPINRAZA patients increased almost 50% versus the prior quarter, and there are still approximately 300 patients active in the expanded access program. We recorded revenue from over 25 international markets with approximately 75% of ex-U.S. SPINRAZA revenues in the second quarter coming from Germany, Italy, Japan, Turkey, Brazil, Spain, and Australia.
We continue to believe that the international opportunity for SPINRAZA is even greater than in the U.S. as we continue the momentum of new country launches, and we believe there is significant ex-U.S. opportunity, not just in Europe, but also in Asia Pacific and Latin American markets. Let me now move on to our biosimilars business, which generated $127 million in revenue this quarter, a 40% increase versus the prior year. We estimate that more than 90,000 patients in Europe are currently receiving Biogen biosimilars. BENEPALI continues to be the market leader in countries such as the U.K., Denmark, and Norway, and we estimate BENEPALI market share of approximately 40% across all major launched markets. Flixabi units grew 48% versus the prior quarter, led by Germany, Italy, and France, where we estimate market share to now exceed 10%.
As more competitors enter the anti-TNF biosimilar market, we believe the continuity of our global supply chain may prove to be a competitive advantage. We aim to maintain 100% uninterrupted supply, and we have delivered nearly 5 million doses since launch without an interruption. We expect growth to moderate in the second half of this year as volume growth is offset by pricing pressure. We expect to be well-positioned in 2019 with the expected launch of Imraldi in October this year. We estimate that the originator product, HUMIRA, has annual revenues of approximately $4 billion in Europe, representing a large potential market for biosimilars. Last month, we exercised our option to purchase additional shares of Samsung Bioepis, our joint venture with Samsung Biologics. We will pay Samsung Biologics approximately $700 million to increase our ownership share in Samsung Bioepis to approximately 49.9%.
We believe the intrinsic value of this additional equity is significantly greater than our expected payment. This transaction is subject to certain regulatory conditions and is expected to close in the second half of 2018, at which point we would anticipate that we would begin to record our share of the net profits and losses of the JV below the operating line. Turning to our anti-CD20 revenues, we reported $490 million in Q2, an increase of 23% versus the prior year, primarily driven by Ocrevus royalties. This includes our estimated Ocrevus royalties of $113 million for the second quarter. Total other revenues were $109 million in the second quarter, more than doubling versus Q2 2017 as we continue to benefit from greater contract manufacturing.
Although the timing can be difficult to predict for the third and fourth quarter, we expect roughly similar contribution from other revenues compared to the second quarter. Q2 GAAP and non-GAAP gross margins were 87%, a slight improvement versus Q1 due to lower contract manufacturing. Q2 GAAP R&D expense was 29% of revenue or $981 million. Q2 non-GAAP R&D expense was 24% of revenue or $819 million. Both GAAP and non-GAAP R&D expense include $324 million out of the $375 million upfront payment to Ionis, with the remaining $51 million capitalized as prepaid R&D, which will be amortized into the P&L over the life of the agreement.
GAAP R&D expense also included $162 million, comprised of the equity premium of approximately $94 million, as well as a discount of approximately $68 million to reflect the effect of certain holding period restrictions, which will be amortized into the P&L for the next couple of years. The remaining $463 million, which is the difference between the fair market value of our investment in Ionis upon deal closing and the discount, was recorded as an investment and will be marked to market each quarter as a GAAP-only expense. Q2 GAAP SG&A was 15% of revenue, or $516 million. Q2 non-GAAP SG&A was also 15% of revenue at $512 million. GAAP other net expense, which includes interest, was $35 million in the second quarter versus $69 million in the second quarter of last year. Non-GAAP other net expense was $40 million in Q2 versus $69 million in Q2 of last year.
In Q2, our GAAP tax rate was approximately 22%, and our non-GAAP tax rate was approximately 21%. Our weighted average diluted share count for Q2 was approximately 207 million shares, and we finished the quarter with 201 million shares outstanding. We repurchased approximately 9.6 million shares in Q2 at an average price of $286. For a total value of $2.75 billion, completing our previously announced $5 billion share repurchase authorization, which now brings us to our diluted earnings per share. In the second quarter, we booked GAAP earnings per share of $4.18, an increase of 3% versus last year, and non-GAAP earnings of $5.80 per share, an increase of 15% versus last year. We generated approximately $1.1 billion in net cash flows from operations in Q2, which was impacted by approximately $537 million due to the Ionis collaboration agreement.
We ended the quarter with approximately $4.4 billion in cash, marketable securities, and $5.9 billion in debt. We believe we have, and we will continue to have ample capacity to execute meaningful future business development M&A activity, as well as return capital to shareholders. We will continue to be disciplined in our approach and focused on value creation. Let me now turn to our updated full-year guidance for 2018. We expect revenues of approximately $13 billion-$13.2 billion, which would represent year-over-year growth of approximately 6%-7.5%. The low end of our guidance reflects emerging potential risks to our MS business in Europe, including pricing pressure as well as uncertainty in channel inventory levels in the U.S. We expect gross margin as a percentage of sales to be consistent with Q2 2018.
As a result of business development transactions in the first half of the year, we now anticipate GAAP R&D expense between 19% and 20% of sales and non-GAAP R&D expense between 18% and 19% of sales. Of note, guidance does not include any impact from potential acquisitions or large business development transactions, as both are hard to predict. We continue to expect SG&A expense to be approximately 15%-16% of revenues. We anticipate our GAAP tax rate for 2018 to be between 21.5%-22.5%, and our non-GAAP tax rate to be between 20.5%-21.5%. This reflects both the anticipated benefit of the U.S. corporate tax reform as well as the tax impact of the business development activity that's occurred so far in 2018.
We anticipate full-year GAAP diluted EPS of $21.80-$22.40 for 2018, representing growth of 83%-87% versus 2017, and non-GAAP diluted EPS to be $24.90-$25.50, representing growth of 14%-17%. I'll now turn the call back over to Michel for his closing comments.
Thank you, Jeff. We closed the second quarter with record quarterly revenues, double-digit non-GAAP earnings growth, an increase to our full-year revenue guidance, and exciting progress across our pipeline. Consistent with our strategic priorities, we allocated capital to both invest for future growth and opportunistically return capital to shareholders. We do not intend to pause or slow down here. We believe there will be plenty more exciting opportunities for Biogen as we move into the second half of the year. Looking forward, within the next 12 months, we expect further progress across our neuroscience pipeline, including Eisai's presentation of the BAN2401 data tomorrow and assessing next steps. Dosing the first patient in our phase III studies in stroke and neuropathic pain. Data readouts across MS, Alzheimer's, neuropathic pain, ophthalmology, and ALS. Filing for regulatory approval in the U.S. for BIIB098 in MS.
As Michael noted earlier, we continue to see compelling evidence across the neuroscience landscape that leads us to believe we are at the beginning of a period of transformative breakthroughs in the treatment of neurological diseases. We believe we are uniquely positioned to benefit from these potential breakthroughs, and Biogen's goal is to be the long-term leader in neuroscience, as we believe it will become one of the most promising thematic areas of investment for biopharma and growth investors. To deliver on our aspiration, we remain focused on executing well on our strategic priorities to fortify our core business in MS and SMA and allocate capital to expand and progress our neuroscience pipeline.
Finally, I want to reiterate our commitment to maximizing returns to our shareholders and bringing innovative therapies to patients over the long term. This demands that we continue to allocate capital efficiently, effectively, and appropriately. As we have demonstrated in the past, we will always strive to have an optimal capital structure as well as aim for superior returns from the investment we make. Once again, I would like to thank our employees around the world who are dedicated to making a positive impact on patients' life, and all of the physicians, caregivers, and participants in our clinical development programs. Our past and future achievements could not be realized without their passion and commitment. With that, we'll open the call for questions.
And at this time, if you would like to ask a question, please press star followed by the number 1 on your telephone keypad. Your first question today comes from the line of Umer Raffat with Evercore. Please go ahead.
Hi. Thanks so much for taking our question. I wanted to focus on BAN2401, and I didn't want to ask any questions on the data. We'll find it all tomorrow, and we can discuss it. But instead, to help us prepare, my question was some clarity on trial design so we're all on the same page going into it. My question is this: So beyond the Bayesian stats, I'm going to focus on specifically on the traditional stats. Was there a stat hierarchy? If not, was there a multiplicity adjustment? If that's the case, what's the statistical significance threshold we should be pegging most of the results to? I think that's really what I'm trying to get clarity on. How many cuts of the data were there, and what's the P-value threshold we should be pegging all the results to? Thank you very much.
Hi, Umer. This is Al Sandrock. Look, we look forward to seeing the results tomorrow presented by Eisai, my colleague at Eisai, Dr. Kramer. There was a single primary endpoint, and the primary endpoint was a Bayesian analysis at 12 months, where they were looking for the threshold was an 80% probability of seeing a 25% treatment effect. That's the primary endpoint. There were a number of secondary endpoints that were pre-specified, and Dr. Kramer, I believe, has a plan to present those pre-specified secondary endpoints. I hope that answers your question.
Your next question comes from the line of Cory Kasimov with J.P. Morgan. Please go ahead.
Hey, good morning, guys, thanks for taking the question. I wanted to also ask a question on just the Alzheimer's kind of portfolio overall. With the understanding, as Umer said, that the 2401 data's tomorrow and you're limited in what you can say on the data, I'm curious on a broader level, what you see as the key elements of read-through from 2401 over to aducanumab. How much do these results impact your confidence on the ongoing studies there that you've just completed enrollment in? Thanks.
Yeah, Cory. Thanks. This is Al again. First, I want to point out some of the similarities between BAN2401 and Adu. They both bind tightly to soluble oligomers as well as insoluble fibrils of A-beta. Protofibrils are just a type of soluble oligomer, and thus both antibodies bind tightly to it. Neither antibody binds very well to monomers of A-beta, so they're both selected for aggregated forms of A-beta. Both antibodies bind to the N-terminal portion of the A-beta peptide, and we believe that the linear epitopes are very similar. They bind to very similar places on A-beta. But based on X-ray crystallography studies that we've been doing lately, we see that they bind in a different way because they're different antibodies.
In the human trials, as we've noted on the top line, both antibodies clear amyloid plaques as seen on PET scans in a dose-dependent manner, as Mike said this morning. In human trials, both antibodies seem to slow the progression of cognitive decline at the highest dose. You can see that there are a lot of similarities, we continue to believe that what we're doing right now in terms of these two large phase III trials of aducanumab, where we use an antibody that selectively targets aggregated forms of A-beta in the right patients, is the right approach. We think it's the most promising approach right now as we try to find the first disease-modifying therapy for Alzheimer's.
Cory, this is Mike. I think Al summarized it really well. The only thing I might add to that is we always thought going into this trial in terms of the read-through to aducanumab would be the relationship between brain amyloid lowering and slowing of cognitive decline. What we reported together with ACEI on the top line was that there was an observed both a reduction in brain amyloid by PET and in a slowing in cognitive decline. That's what we were looking to in terms of a potential read-through to aducanumab.
Your next question comes from the line of Ying Huang with Bank of America Merrill Lynch. Please go ahead.
Hi. Thanks for taking my question. I have one question on BAN2401 data as well. Given that the trial missed the primary endpoint but hit maybe the key secondary point, is it possible you could potentially file this data set with the FDA, or this trial could potentially be a pivotal trial for registration purpose? Secondly, maybe for Mike and Al, what's your thought on the Roche oral drug data in SMA? I know you guys just invested $1 billion in the Ionis collaboration, but would you consider developing an oral modality for SMA? Thank you.
Hi Ying, it's Al again on the BAN2401. Look, our next steps are to talk to FDA and other regulators. We'll see what they say. I think it's too early to speculate as to whether or not we can file with this.
Yeah, Ying, I'll take the question on the Roche PTC oral. First, the comment relative to our broader deal with Ionis. I would say, of course, this oral molecule really pertains to one specific target, SMN2 splicing, our broad collaboration with Ionis will be across many targets and many different mechanisms. It's not particularly related to the broader Ionis collaboration, I would say. In terms of SMA, we're very interested to continue seeing the data that Roche and PTC are reporting around their oral. It's still early days. We've always said from the beginning that we welcome additional modalities. We think that there's going to be a world in the future where there are multiple options for patients with SMA, whether it's SPINRAZA, whether potential gene therapy as we're seeing, whether oral splice modulators.
This kind of combinatorial effects or complementary effects to different mechanisms are things that we're actively exploring.
I just want to remind everyone, we have a lot of people in the queue, we'd really appreciate if you limit to one question each. Thank you.
Your next question comes from the line of Matthew Harrison with Morgan Stanley. Please go ahead.
Hey, great. Good morning. I will not ask a BAN2401 question. I was hoping maybe you could just comment. You mentioned a couple of times, I think when you talked about guidance for the rest of the year, you talked about 2H stability for MS, you also pointed to EU pricing pressure. Could you just expand upon what's happening in the EU, and what are the risks and the timelines for that playing out? Thanks.
Thanks, Matthew. It's Jeff. Let me two parts of that question. One is the stability. One of the things we're looking forward to as we enter the back half of this year is we now have a period in the U.S. from an MS perspective, where we're now up against a comparable period where we had Ocrevus in the market. The stability in terms of volumes were directed at that equation. We feel good about that in terms of our potential to be stable in the U.S. Outside the U.S., the volumes have been growing pretty well outside interferons. We believe that will continue. The challenge is each of the European markets is different, and there are pricing pressures. We'll have to work our way through those.
Certain countries are a little bit more aggressive than others, we'll have to work through that as we work our way through the back half of the year.
If I may add to Jeff's comment, there is nothing new in the EU. We've seen and we've been facing year after year this type of pressure on price. The good news is that the volume is going strong and according to what we did plan. In some major countries, we have conventions, and we have some timeline where we need to meet back with authorities, regulators based on drug utilization, and price is being negotiated. It's too early to speculate, but nothing new from the European setting. It's a constant, I would say, engagement with regulators. The important element is that the underlying demand is doing well, mostly for the key products.
Your next question comes from the line of Geoffrey Porges with Leerink Partners. Please go ahead.
Thank you very much, and congratulations on all the progress. Unfortunately, back to BAN2401, perhaps Michel, you could give us a sense of Biogen's corporate stance. At this stage, are you willing to commit to another Alzheimer's pivotal trial program? Perhaps you could give us some of your insights so far on whether another pivotal would have to be of the same size and scale as the aducanumab pivotal trials. If so, when we might hear about such a development plan. Thanks.
Thanks for the great question. That might be a bit too early for us to answer. As you can anticipate, we are encouraged by the news. We look forward to the presentation tomorrow. We are engaging very closely with our partners at Eisai. This latest data lands well on the amyloid hypothesis. This reinforces aducanumab. We believe so. We need to reconvene. We are well positioned with our portfolio that is already large. We believe, as many of you, that it's not one single treatment that will treat all the patients at all the stages. We need to envision new studies, including potentially combination therapies or sequencing therapies to best answer the unmet medical need. Again, too early to be precise on what we have to do. Clearly, there is an investment case.
Your next question comes from the line of Michael Yee with Jefferies. Please go ahead.
Thanks for the question. Congrats again on the BAN data. I guess just to clarify some earlier comments about the next steps, is it fair to say that you would want to seek to aggressively file, or that the goal is clearly to start more trials in different patient populations? I guess what are the scenarios. More importantly, I guess, what's the timing of all this to get some more clarity to the street? Thanks so much.
Hi, Michael. It's Al again. Well, look, we can get a meeting with FDA pretty quickly, I believe. I think we'll have to see what they say. A whole variety of options are potentially available to us. We'll do the best thing that we can for AD patients who need a disease-modifying therapy approved as soon as possible. In terms of the size and scale of what we'll have to do next, that will all be based on what these conversations say with respect to regulatory authorities, not just in the U.S., but also across the world.
Your next question comes from the line of Geoff Meacham with Barclays. Please go ahead.
Hey, guys. Thanks for the question. Another one on 2401. Obviously, we'll learn a lot more on the data tomorrow, can you speak out to the ADCOMS a bit more? What's the regulatory view of it? What validation has been done? What animal studies have already been completed? I'm just thinking preclinical toxicity, things like that. Thank you.
Well, obviously, preclinical studies, extensive ones were done in order to allow both companies to enroll patients into fairly large trials. Even this phase II trial with BAN2401 had more than 800 patients. Clearly, a full and adequate preclinical package was done. Sorry, I forgot the first part of your question, Jeff.
AdCom.
Yeah, the ADCOMS. The ADCOMS is a composite of three already well-validated endpoints. The CDR Sum of Boxes, the ADAS-Cog, and the MMSE. It well, actually uses all of the CDR Sum of Boxes, which is about 60% of the composite endpoint, and chooses the portions of the ADAS-Cog and the MMSE that increase the sensitivity of the outcome measure to change in this early population, MCI and mild patients. They've published on this. How the regulators will view it, we'll have to see what they say when we go to talk to them.
Your next question comes from the line of Terence Flynn with Goldman Sachs. Please go ahead.
Hi. Thanks for taking the question. Was just wondering if you'd be able to expand at all on your comments on your SMA gene therapy program in terms of maybe timeframe for getting off clinical hold. Based on the orphan drug designation, it looks like you guys are using the same promoter that Novartis is. Are you willing to share what vector you're using at this point? Thanks.
Yeah. This is Mike. All I can really say at this point is we believe we've got a path forward. We're working hard to get this off clinical hold, and we look forward to initiating the trial and getting this to patients as rapidly as possible. Once we're able to do that, we will have more to say about the precise nature of our gene therapy agent.
Your next question comes from the line of Salim Syed with Mizuho. Please go ahead.
Hey, guys. Thanks for the question. One on BAN2401, if I may. When you're thinking about running additional trials versus filing, are those two decisions independent of each other? If you run additional trials, does that preclude you from filing off the data that you have thus far? Thank you.
Boy, that's a tough question to answer without the benefit of the conversations we're about to have with the regulators, I don't think I'll hazard a response.
Your next question comes from the line of Robyn Karnauskas with Citi. Please go ahead.
Hi. Thanks so much for taking my question, and congrats. A question on the aducanumab data that was presented yesterday for 110 weeks. It was confusing because it's versus placebo switchers, but it looked like the titration arm didn't do as well as the 10 mg per kg arm. Can you help us understand a little bit more what you think is going on at two years, and how to think about the drug, and what's going on in that arm at that time? Thanks.
Hi, Robyn. It's Al again. Yeah, it is hard because we've lost the placebo control after one year. I think I do agree with when you look at the data, the 10 mg per kg group has always looked good since the very beginning and continues to look great. The titration arm, there's two questions related to that. One is that the dose was lower for a period of time in order to try to mitigate the effects of ARIA, which it does. Two things are happening. One is that you're starting with a lower dose, which we believe might be less effective initially. The time to get to the appropriate dose is a little bit later. When I look at the titration data, I have to say I'm pretty encouraged overall. I think it looks quite good.
It may not be as good as 10 mgs per kg, but I think the bottom line also, Robyn, is that we've lost the placebo control. We're down to very few patients at this point. It was not a large trial to begin with. Trying to look at differences between dose arms is pretty hard with this small number of patients.
Great. Thanks.
Your next question comes from the line of Phil Nadeau with Cowen and Company. Please go ahead.
Good morning. Thanks for taking my question. One commercial question on SPINRAZA. We saw a deceleration in new patient starts again in Q2. Is that simply the maturation of the product? Is there anything that can reverse that trend? On the adult patients specifically, you mentioned 60% of the population are adults. What proportion of those adults do you think are appropriate for therapy and could be on therapy over the long haul? Thanks.
Thanks for your question. I think we were encouraged by the progress we made from the first quarter to the second quarter. We mentioned it that 20% growth in adult patients being treated, and I think you have the statistic, right? The 60% of the patient pool is in adult. We think that still is a large underserved patient pool. The exact percentage of that's treatable, I think is still to be determined given some of the complexities, but we think it's a large percentage. We continue to be encouraged by that patient subsegment, and we've made good progress. We continue to believe as we work our way through the year, our efforts to focus in on those sites that treat adults, which we had not been focused on before, is going to pay dividends.
Our adult branded and targeted campaign to identify the patients is going to pay dividends, as well as the investment in the sales force. We continue to expect to see growth. We'll just have to see how much growth we get in the back half of the year versus next year.
Your next question comes from the line of Carter Gould with UBS. Please go ahead.
Congrats on the quarter, guys. Thanks for taking the question. For Michel and Jeff, on the BD front, we seem to be very active completing a number of tuck-ins and platform deals this year. Meanwhile, you've reiterated a number of times a preference for later-stage assets. Maybe you could just characterize where you sit now with your BD strategy and if you see sort of any evolution on that front, sort of after the first half of the year. Thank you.
I will get started and Jeff will continue. We always said that the early stage was the sweet spot for the organization. This is where we can add the most value based on the quality of the team, our ability to best develop the compound. We continue to be looking at opportunities that can meet and complement our current portfolio. It's very nice to see, like the last deal we've done on muscle enhancement, that this can bridge towards different priority areas that we have. This is very exciting. We look at all the opportunities in later stage 2, by remaining very disciplined, and I will let Jeff continue on the answer here.
Yeah. I think we're pleased by the progress we've made. I think since Michel has become the CEO, we've done 7 deals now. All smart deals, many which were later-stage deals that have really helped the pipeline if you look at the slides. I think we feel good about the pipeline advancement, but as we said from the beginning, we continue to say we've got ample capacity to both invest in early-stage deals, late-stage deals, and return capital to shareholders. That hasn't changed. The challenge is, I think as everybody can appreciate is, we have to find later-stage deals where there is a good scientific rationale, strong intellectual property, good synergies from a commercial and manufacturing perspective, and it makes sense financially. Those are a lot of wickets to get through to kind of get a deal through. There's no lack of attention and focus on it.
As Michel had said, we now have a strong head of business development, Daniel Karp. We'll be moving forward more aggressively, looking at the later-stage deals. We certainly hope to get something done, but it'll be something that'll be smart and appropriate and disciplined.
Your final question will come from the line of Brian Abrahams with RBC Capital Markets. Please go ahead.
Hi there. Thanks for taking my question and congrats on the progress in the quarter. On Alzheimer's, being that the ADCOMS endpoint I think is somewhat new to many of us. I was wondering if you could speak to your views as to the bar for clinical meaningfulness or the goal on that particular endpoint, maybe relative to CDR sum of boxes, and then maybe speak about how the 2401 results might affect your views of performing an interim analysis for aducanumab in the phase III program. Thanks so much.
Hi, Brian, it's Al. First, I'm going to start with the interim analysis. The answer that we give is that we're not commenting on it. In terms of the ADCOMS, it's a new endpoint, as you pointed out. I haven't seen any publications that relate to what the clinically meaningful difference is. I will say, though, that the BAN 2401 was powered on the Bayesian analysis, essentially to see a 25% treatment effect size, an 80% probability with the threshold. That answers that in terms of what the study was powered for. What the community believes is the clinically meaningful threshold has not yet been published or determined to my knowledge.
Hi, this is Matt. Thank you again for joining us for today's call. Before we conclude, I'd like to remind everyone that we plan to host a webcast tomorrow, July 25th, to discuss our Alzheimer's portfolio with Dr. Al Sandrock, our Chief Medical Officer, and Samantha Budd Haeberlein, Vice President, Alzheimer's Disease, Dementia, and Movement Disorders late-stage clinical development. This will be a webcast-only event, not in person, and will start at 5:00 P.M. Central Time, 6:00 P.M. Eastern Time. The link can be found on the investor section of Biogen's website at www.biogen.com. Thank you. We look forward to speaking with you in the future.
Thank you.
Thank you to everyone for attending today. This will conclude today's call, and you may now disconnect.