|
|
|
|
||
Re: MS downgradeStepping aside ahead of APOLLO-B readout; Downgrading to Equal-weight Morgan Stanley April 25, 2022 With the stock suggesting ~55-60% probability of success (PoS) for the APPOLO-B readout in mid-2022, we believe the market is now pricing in the right expectations. Given our lack of conviction on a higher PoS compared to the market, we downgrade to EW on risk/reward ahead of the data readout. WHAT'S CHANGED? From: To: Alnylam Pharmaceuticals Inc (ALNY.O) Rating Overweight Equal-weight Price Target $210.00 $178.00 Shr price, close (Apr 22, 2022) $167.00 Mkt cap, curr (mm) $14,210 Downgrading to Equal-weight on risk-reward and lack of conviction on APOLLO-B readout. With investors debating the read through from BridgeBio's cardiomyopathy miss to Alnylam's APOLLO-B readout in mid-2022, we analyzed current hypotheses for the six-minute walk test failure. We see baseline heart failure severity and trial conduct quality as important variables in APOLLO-B. However, given the lack of information on these two factors, we're unable make a strong call on the trial outcome and are stepping aside ahead of the readout. We expect ALNY to trade up to $210-220 on positive data, and down to $100-120 on negative results suggesting slightly higher than 50/50 PoS being priced into the stock which we believe is right given the current information available. Multiple hypotheses attempt to explain why ATTRibute-CM did not meet the six-minute walk test efficacy endpoint: Based on contrasts in trial design and baseline characteristics in BridgeBio's ATTRibute-CM and Pfizer's ATTR-ACT trial, current hypotheses for the 6MWT miss include (1) false positive diagnosis of ATTR-CM with non-invasive technetium imaging; (2) disproportionate number of wild type ATTR-CM patients; (3) training effect from multiple baseline 6MWT screenings; (4) poor quality assurance (e.g., variation) of trial conduct; (5) small portion of patients with severe heart failure; and (6) geographic distribution of trial sites. We think baseline heart failure severity is an important variable in APOLLO-B: BridgeBio's ATTRibute-CM study enrolled less Class III heart failure placebo patients than Pfizer's ATTR-ACT trial (13.9% vs. 36%). Based on Yap et. al's literature review (here) suggesting patients with mild heart failure perform better on the 6MWT, we believe enrollment of a healthier patient population may have contributed to the robust placebo response in ATTRibute-CM. For Alnylam, we think enrollment of 25%+ patients with Class III heart failure would maximize the likelihood of success. We also think trial conduct quality control is a key success factor in APOLLO-B: The American Thoracic Society guidelines (here) identify numerous sources of variability within the 6MWT procedure itself, including but not limited to: shape of walking corridor, design of turnaround points, clothing, and time of day. Without visibility into the differences between how 6MWT is performed by BridgeBio and Alnylam, we use overlapping trial sites as a proxy for quality assurance risk in APOLLO-B. Our analysis estimates ~30% (21/66) of the APOLLO-B sites currently listed on clinicaltrials.gov are overlapping with ATTRIBUTE-CM. We note that clinicaltrials.gov does not yet list all 90 of the APOLLO-B trial sites. We do not see training effect, over-enrollment of wild type patients, or use of non-invasive diagnostics as major risks for APOLLO-B: We do not see training effect as a risk in APOLLO-B, although studies (here) have confirmed the phenomenon of learning effect after repeated same-day and next day tests on 6MWT. Our view is predicated on the fact that APOLLO-B is not conducting multiple baseline screening tests and there is no evidence to suggest learning effect is durable at 12 months. Additionally, we do not see over-enrollment of wild type patients as a risk for Alnylam, although wild type (non-hereditary) patients have a better prognosis than hereditary. Our view is informed by Pfizer's ATTR-ACT trial (here) and natural history studies (here) which suggest there is no difference between wild type and hereditary patients on 6MWT at 12 months. Finally, we do not see misdiagnosis of ATTR-CM with technetium imaging as a risk in APOLLO-B, as data suggests this approach is accurate when best practice guidelines are followed. 02Risk Reward - Alnylam Pharmaceuticals Inc (ALNY.O) Recently Approved Drugs & Advancing Pipeline to Drive Value EQUAL-WEIGHT THESISPRICE TARGET: $178.00 Alnylam's three RNA interference drugs provide the company with a solid base business and the stock with a valuation floor. These products include Onpattro in polyneuropathy from transthyretin amyloidosis, Givlaari in acute hepatic porphyria, and Oxlumo in primary hyperoxaluria type 1. Upside is primarily guided by clinical pipeline updates. Currently, the key pipeline driver is a Phase 3 readout from Onpattro (Patisiran) in cardiomyopathy with transthyretin amyloidosis in mid-2022. We believe the stock reflects a ~55-60% likelihood of success on the study and believe the market is pricing in the right expectations. 03ATTR Cardiomyopathy Trials ATTR-ACT | PFIZER ATTR-ACT was a Phase 3 study evaluating oral tafamidis (Vyndaqel / Vyndamax) compared to placebo in both hereditary and wild-type ATTR-CM. The primary analysis of the study compared a pooled Vyndaquel (80mg and 20mg) to placebo, which demonstrated a significant reduction (p=0.0006) in all-cause mortality (71% vs. 57%) and frequency of cardiovascular-related hospitalizations (0.297 vs. 0,455) compared to placebo over a 30-month period. Additionally, individual components of the primary analysis also demonstrated a relative reduction in the risk of all-cause mortality and frequency of cardiovascular-related hospitalization of 30% (p=0.026) and 32% (p<0.0001), respectively, ATTRIBUTE-CM | BRIDGEBIO ATTRibute-CM was a Phase 3 trial evaluating BridgeBio's acoramidis (transthyretin stabilizer) in cardiomyopathy from transthyretin amyloidosis. The company announced in 4Q21 that acoramidis failed to meet the six-minute walk test primary endpoint. Placebo patients experienced a less prominent decline from baseline on 6MWT at 12 months than acoramidis patients (-7.04m vs. -9.28m, p=0.76). Interestingly, the placebo arm performed in-line with healthy elderly patients (Exhibit 1), leaving investors to wonder (1) what could have caused the robust placebo response and (2) what the miss means for Alnylam's upcoming readout using the same primary endpoint in the Phase 3 APOLLO-B trial. APOLLO-B | ALNYLAM APOLLO-B is a Phase 3 study evaluating Alnylam's patisiran (transthyretin silencer) in cardiomyopathy from transthyretin amyloidosis. Results are expected in mid-2022. The primary endpoint will compare patisiran to placebo on six-minute walk test change from baseline at 12 months. Key secondary endpoints include cardiomyopathy symptoms, cardiac biomarkers, and death and hospitalization outcomes. 04Key Debates On APOLLO-B Below we outline, analyze, and provide our views on potential contributors to placebo effect in ATTRibute-CM and what this means for APOLLO-B. The current hypotheses for ATTRibute-CM's miss include: (1) False positive diagnosis of ATTR-CM with non-invasive technetium imaging; (2) Disproportionate number of wild type ATTR-CM patients; (3) Training effect; (4) Trial quality assurance; (5) Small portion of patients with severe heart failure; (6) Geographic distribution of trial sites. HEART FAILURE SEVERITY The Concern: Some investors have proposed that enrollment of patients with less severe heart failure contributed to a robust placebo response on 6MWT in ATTRibute-CM. This hypothesis is fueled by the fact that only 14% of placebo patients and 17% of acoramidis patients had Class III heart failure at baseline in ATTRibute-CM (Exhibit 2). This contrasts with ATTR-ACT in which 36% of the placebo group and 30% of tafamidis patients had Class II heart failure. Our Thoughts: (1) We think it is possible that enrollment of a more healthy baseline population (based on NYHA heart failure classification) contributed to the placebo response in ATTRibute-CM. Consequently, we view the distribution of NYHA heart failure severity amongst patients enrolled in APOLLO-B as an important variable. (2) Our view is supported by Yap et. al's literature review which suggests patients with less severe heart failure perform better on the 6MWT (Exhibit 3). Although Pfizer's ATTR-ACT trial has proven patients with Class I and II heart failure do in fact decline over time, we believe enrollment of a healthier patient population at baseline would minimize the difference between placebo and drug arms. (3) However, Pfizer's tafamidis performed better in patients with milder heart failure and greater baseline 6MWT distance (link), suggesting a healthier population is actually beneficial for the treatment effect of a potential ATTR-CM treatment. In addition, BridgeBio did not find meaningful change in 6MWT in the placebo group across NYHA classes. However, we think healthy patients' performance on 6MWT likely serves as the upper limit to how much a drug intervention can slow the rate of decline on 6MWT. Given this upper limit, we believe the benefit of enrolling more healthy patients would be more prominent in the placebo arm. TRIAL QUALITY ASSURANCE The Concern: Another hypothesis for the miss in ATTRibute-CM is poor quality control in execution of the 6MWT. Since the BridgeBio trial, Alnylam investors have been interested in how 6MWT was conducted in ATTRibute-CM and how this may differ in APOLLO-B. The American Thoracic Society guidelines (here) identify numerous sources of variability within the 6MWT procedure itself, including but not limited to: shape of walking corridor, design of turnaround points, clothing, time of day. Some have proposed geographic distribution as a contributor to poor quality control. Our Thoughts on 6MWT Conduct: We think it is difficult to evaluate quality assurance without direct visibility into trial conduct. Nevertheless, we use overlapping trial sites with ATTRIBUTE-CM as a proxy for quality assurance risk in APOLLO-B. We estimate ~30% (21/66) of the APOLLO-B sites currently listed on clinicaltrials.gov are overlapping with ATTRIBUTE-CM (Exhibit 5). Notably, Alnylam's enrollment completion announcement (here) indicated APOLLO-B enlisted 90 sites worldwide, but only 66 of those sites are listed on clinicaltrials.gov. Despite an anticipated overlap in trial sites, Alnylam has noted the vendor used to conduct APOLLO-B is different than the vendor used by BridgeBio in ATTRibute-CM. Our Thoughts on Geographic Distribution: We think a skew towards more ex-US trial sites may be an important factor contributing to the failure of ATTRibute-CM, and as such, could add risk to APOLLO-B. Studies has suggested geographic variation in 6MWT (Exhibit 6). However, we do not know how geographic variation can impact 6MWT measured over time. Secondly, ex-US trial sites could potentially use variations of 6MWT protocol, which could have an impact on the results. Literature has described various adaptations of the 6MWT in low resource settings (link). We do not know whether an adaptation to the 6MWT protocol is allowed. As an aside, we note that one distinction between US and ex-US sites is the technetium radiotracer used for diagnostic imaging, as Tc-PYP is the only radiotracer used in the US and Tc-DPD is commonly used in other countries. Given there is robust evidence suggesting Tc-PYP and Tc-DPD have similar diagnostic accuracy, we do not think the use of different radiotracers geographically contributed to less accurate diagnosis of patients in ATTRibute-CM and do not suspect it would in APOLLO-B. FALSE POSITIVE DIAGNOSIS USING TECHNETIUM IMAGING The Concern: Cardiac imaging with technetium radiotracers (Tc-DPD, Tc-PYP, and Tc-HMDP) has emerged as a promising non-invasive diagnostic in ATTR-CM, although the more invasive endomyocardial biopsy has been the gold standard for years. Given the fact that technetium imaging was not used during the enrollment of Pfizer's successful registrational study and more recent trials from BridgeBio (ATTRibute-CM) and Alnylam (APOLLO-B) diagnose patients using both technetium imaging and traditional endomyocardial biopsy, some investors have questioned whether diagnosis via technetium imagining contributes to placebo effect by enrolling patients that are misdiagnosed with ATTR-CM. Our Thoughts on Technetium Imaging: We think it is unlikely that technetium imaging contributed to false positive diagnosis of patients in ATTRibute-CM, and if Alnylam follows guideline recommendations, we would not see use of technetium imaging diagnosis as a risk in APOLLO-B. For context, technetium imaging is a two-step process that involves intravenous administration of technetium radiotracer followed by imaging to assess cardiac retention. Given the risk of confusing myocardial retention with either light chain amyloid (AL) cardiomyopathy or blood pooling, guidelines (here) recommend serum/urine testing to rule out AL cardiomyopathy and use of both planar and SPECT imaging to rule out blood pooling. Importantly, ATTRibute-CM followed technetium imaging guidelines for ruling out blood pooling or light chain amyloidosis. In our view, this minimizes the risk of false positive diagnoses. Alnylam has not disclosed whether both planar and SPECT imaging were used to rule out blood pooling in APOLLO-B, but mgt. has indicated urine/serum testing was used to exclude light chain amyloidosis. Our Thoughts on Perugini Scoring: We think it is possible that enrollment of patients with grade 1 cardiac uptake of technetium augmented placebo effect in ATTRibute-CM. However, we believe it is unlikely that a "low single digit" percent of grade 1 patients skewed the entire placebo cohort towards 6MWT performance in line with healthy patients. For Alnylam, we do not think there is any risk related to Perugini scoring, as APOLLO-B is only enrolling grade 2-3 patients. There is limited data available on the relationship between cardiac uptake scoring and 6MWT, although studies indicate there is no difference in survival between patients with grade 1-3 abnormal cardiac uptake of technetium. DISPROPORTIONATE NUMBER OF WILD TYPE PATIENTS The Concern: Guided by a generally more positive prognosis in wild type patients versus hereditary (median survival: 3.5 years vs. 2.5), some investors hypothesize that the placebo response in ATTRIBUTE-CM is related to 90% of the patients having non-hereditary ATTR-CM. This contrasts with Pfizer's successful ATTR-ACT study that enrolled only 76% wild type patients. We note Alnylam has an enrollment goal of 80% wild type and 20% hereditary patients, but has not disclosed the actual distribution following enrollment completion. Our Thoughts: We think wild type patient enrollment above the 80% goal would not have a meaningful impact on placebo effect in APOLLO-B. Our view is informed by analysis (Exhibit 3) of long-term extension data from Pfizer's ATTR-ACT study (here) which suggests there is no difference between wild type and hereditary patients on 6MWT at 12 months. We also note Lane et. al's natural history study (here) shows no difference between wild type and hereditary patients on 6MWT at 12 months. Overall, we do not think 12 months is enough time to see a notable difference between placebo response in wild type and hereditary patients, which lessens concern regarding "over-enrollment" of wild type patients in our view. TRAINING EFFECT The Concern: Training effect, or improved performance on the 6MWT due to repeated testing before evaluation, has also been proposed as a contributing factor to the robust 6MWT placebo response in ATTRIBUTE-CM. The American Thoracic Society notes training effect may be due to "improved coordination, finding optimal stride length, and overcoming anxiety." Prior to randomization in ATTRibute-CM, patients were required to complete two tests with results that were ≥150 meters and within 15% of each other. This is in contrast to APOLLO-B in which patients only took one test at baseline screening. Our Thoughts: We think it is unlikely training effect from multiple baseline tests in BridgeBio's ATTRibute-CM trial was durable enough to impact results at 12 months. However, the screening could have impacted other baseline characteristics which could have enrolled a population that increased the placebo response. For APOLLO-B, we think the protocol of one test at baseline minimizes the likelihood of training effect. Although studies have confirmed a learning effect on 6MWT when multiple tests are performed on the same day (Exhibit 11) and one day apart (here), training effect from tests repeated after longer periods of time has not been reported. We suspect that training effect would wear off within weeks to months. |
return to message board, top of board |
Msg # | Subject | Author | Recs | Date Posted |
35077 | Re: MS downgrade | Steve_382 | 0 | 5/2/2022 11:33:10 AM |
35078 | Re: MS downgrade | sherk | 1 | 5/2/2022 11:55:50 AM |