虽然大多数分析师可能会同意以数十亿美元的销售预测来评估某些产品的价值，但这样的共识会隐藏基于商业风险判断而产生的不同观点。在目前已上市的产品中，风险最大的一些产品是面临生物类似物挑战药物，如强生用于类风湿性关节炎的英夫利西单抗(infliximab、Remicade)、罗氏用于肿瘤的赫赛汀（曲妥珠单抗、Herceptin）以及已经正在处于快速发展肿瘤领域竞争前沿的药物，如安斯泰来用于前列腺癌的恩杂鲁胺(enzalutamide、Xtandi)和强生用于套细胞淋巴瘤的依鲁替尼(ibrutinib、Imbruvica)。
通过销售预测都可以凸显要么被夸大要么太保守的资产，这取决于投资者们的风险癖好—尽管上市产品的风险的本质是在纯商业化的风险，而不是注册或临床。
生物类似物上位
这份报告中的特色是即将出现的生物类似物的风险的不确定性，主要是取决于生物类似物被市场接受的速度。去年，EMA批准了多个全球第一的“正确”的生物类似物，如该机构对Hospira和韩国Celltrion推出的强生公司的英夫利西单抗点头的时候。
因此，对获得英夫利西单抗美国市场共同销售权的默克的2020年的销售预测造就了分歧最大的一个产品。与evaluatePharma达成共识预测的17亿美元相比，投行高盛却给出了低的可怜的3.54亿美元。而与此同时，对强生该产品的2020年的销售预测却为53亿美元至79亿美元。
同样受到生物类似物威胁的是另一个抗肿瘤坏死因子单克隆抗体-α，艾伯维用于类风湿性关节炎、银屑病性关节炎和强直性脊柱炎等的阿达木单抗(adalimumab、Humira)，该药物已经成功登上2013年全球销售额最高药物的宝座，但是在该药物达到2017年顶峰之后，销售额预测的尖锐分歧也随之出现。生物类似物同样还将重创罗氏的赫赛汀（曲妥珠单抗、Herceptin），根据各大投行的数据库该药物2020年的销售预测将为33亿美元至72亿美元。
上表中的10只药物都是通过计算2020年销售预测最大值和最小值之间的差异确定，并且按百分比进行排序，然后再对这两个指标进行综合后进行排序。
本次的分析同样排除了一些项目，其中包含了被某些分析师们归入某一业务，而被其他分析师认为单一产品的复方制剂。诸如吉利德的索非布韦(Sofosbuvir、Sovaldi)，该药物由于价格策略的可持久性正在导致巨大的争议，此外还有该公司的艾滋病复方制剂Stribild(埃替拉韦、恩曲他滨、富马酸替诺福韦酯)以及Vertex用于罕见囊性纤维化的ivacaftor（Kalydeco）也没有出现在此次报告中。
肿瘤
在肿瘤领域，据投行美国美林证券的预测，罗氏的赫赛汀（曲妥珠单抗、Herceptin）2020年最大销售预测为72亿美元，该公司另外一方面却在唱空罗氏用于乳腺癌的帕妥珠单抗(pertuzumab、Perjeta)，给出了最小26亿美元的销售预测。
而投行摩根斯坦利的分析师们对这一观点并不认同，仍然坚信帕妥珠单抗(pertuzumab、Perjeta)将成为某些乳腺癌患者用药的新标准，同时认为该药物和Kadcyla(ado-曲妥珠单抗emtansine)将保持一定的持续性。该投行认为帕妥珠单抗(pertuzumab、Perjeta)和Kadcyla(ado-曲妥珠单抗emtansine)2020年的销售预测将分别达到63亿美元和75亿美元。
前列腺癌和白血病领域快速发展的竞争环境导致安斯泰来用于前列腺癌的恩杂鲁胺(enzalutamide、Xtandi)和强生用于套细胞淋巴瘤的依鲁替尼(ibrutinib、Imbruvica)也出现在此次榜单之中。后者的的销售预测共识正在被原研公司Pharmacyclics和共同的销售伙伴强生所分裂。目前看起来该药物能应对得了吉利德用于治疗套细胞淋巴瘤和白血病的Zydelig(idelalisib)，但是艾伯维的潜在药物venetoclax是一个巨大的未知数。
与此同时，赛尔基因的首个PDE4抑制剂银屑病关节炎治疗药物阿普斯特(apremilast、Otezla)2020年的销售额预测差异为6.25亿美元和29亿美元。投行Cowen的分析师们给出了最小值，他们认为在塞尔基因的领导下这是一只到2017年销售额刚达到4亿美元的小药物，且不会达到15-20亿美元的销售。
像依鲁替尼(ibrutinib、Imbruvica)一样，阿普斯特 (apremilast、Otezla)的高价格已经让人不满，但是和Pharmacyclics药物不同的是，近两年来该药物的长期销售预测已经暴涨了10倍。在目前的市场上，投资者们将会密切关注什么样的额销售预测将会发生变化以及各大投行之间存在分歧的观点。
Wide forecast spreads reveal industry’s big, risky bets
high degree of consensus among analysts is great if this accurately reflects a drug’s potential, but in making investment decisions it can be just as informative to find assets for which sellside expectations diverge.
A look at evaluatePharma’s recently introduced minimum versus maximum analyst forecast feature reveals some of these assets. Investors for whom high-risk/high-reward is an appealing strategy should, it seems, look at projects like Lilly’s novel basal insulin, the resurgent PCSK9 class and PD-1 inhibitors (see table below).
Among projects that have yet to reach the market the one with the greatest divergence in 2020 revenue expectations is Lilly’s pegylated basal insulin LY2605541; the mean expectation is for 2020 sales of $566m, but this is derived from a spread from just $112m to Goldman Sachs’s astonishing $1.3bn.
The project recently generated its first phase III data, showing superiority to Lantus but also throwing up a worrying liver toxicity signal (Safety concerns mean Lilly's novel Lantus competitor has much to prove, May 13, 2014). In an already well-served market it is perhaps unsurprising that this is seen as a risky bet.
A similar case can be made for the anti-PCSK9 antibodies, two of which – evolocumab and bococizumab – feature next on the list. While this drug class’s long road to the market should soon end, its true potential lies in long-term cardiovascular outcome studies – a fact that has not prevented some analysts from putting blockbuster numbers into mid-term models.
| Diverging sellside opinions: 10 controversial pre-market assets | | | | | 2020 sales forecasts | | | Project | Company | Drug type | Min ($m) | Max ($m) | Status | | Basal Insulin Peglispro | Eli Lilly | Insulin - long-acting Humalog | 112 | 1,333 | US and EU filings planned for Q1'15 | | Evolocumab | Amgen | Anti-PCSK9 MAb - high cholesterol | 167 | 2,540 | US filing planned for Q3 | | Bococizumab | Pfizer | Anti-PCSK9 MAb - high cholesterol | 50 | 1,125 | Phase III | | Cerdelga | Sanofi | Oral enzyme replacement therapy - Gaucher's disease | 172 | 1,167 | FDA decision pending | | Epratuzumab | UCB | Anti-CD22 MAb - lupus | 125 | 1,182 | Phase III | | Pembrolizumab | Merck & Co | Anti-PD-1 MAb - melanoma and other cancers | 2,400 | 7,600 | FDA decision due by Oct 28, 2014 | | Lynparza | AstraZeneca | PARP inhibitor - ovarian and other cancers | 179 | 1,100 | FDA decision due Jan 2, 2015 | | Palbociclib | Pfizer | CDK 4 & 6 inhibitor - breast and other cancers | 720 | 4,004 | US filing pending | | Sovaprevir | Achillion | Hepatitis C NS3 protease inhibitor | 375 | 1,585 | Phase II | | MM-398 | Merrimack | Topoisomerase I inhibitor - pancreatic cancer | 391 | 1,379 | US filing pending |
|
The top 10 above were identified by calculating both the actual difference between the lowest and highest forecast 2020 sales number, and the percentage spread. The projects, many of which are late-stage assets subject to regulatory rather than clinical risk, were then ranked on the sum of these two measures.
We excluded certain projects, including those that some analysts believe will not make it to market at all, like Merck & Co’s CETP inhibitor anacetrapib, whose inclusion would yield a minimum sales figure of zero. Drugs likely to be sold in various combinations are also excluded, as some analysts model these as a franchise and others as single products.
For the same reason certain portfolio extensions were omitted, like Sanofi’s novel insulin Toujeo, whose sales forecasts are frequently incorporated into the company’s Lantus franchise. The analysis comprises products wher at least three analyst forecasts are available, meaning that many under development by smaller companies, which attract less coverage, are excluded.
Small-cap bet
One that does make it in is Achillion’s sovaprevir. While this now looks like an also-ran, set to enter a highly competitive hep C field, it did recently have its US clinical hold removed, providing an investment trigger.
Forecasts vary widely, helped by the fact that Cowen analysts do not appear to have changed their expectations after the hold was lifted. Broad forecast spreads for epratuzumab and MM-398 reflect the discounting of risk by some analysts and not others in the high-risk areas of lupus and pancreatic cancer respectively.
Less easy to explain is the presence of Merck’s pembrolizumab in sixth spot, and that of the competing late-stage immuno-oncology projects Opdivo from Bristol-Myers Squibb and MPDL3280A from Roche not far outside the top 10. Rather than these being high-risk assets, their forecasts’ variability reflects a wide spread between very large sales expectations and massive ones.
Meanwhile, it is also worth looking more closely at Lilly, which in addition to LY2605541 has three other entries in our top-20 list of assets with the greatest sellside forecast variability: the Jak1/2 inhibitor baricitinib, necitumumab – an Erbitux successor that disappointed at Asco – and the Alzheimer’s disease anti-beta amyloid project solanezumab.
The last of these could be selling as little as $175m in 2020 or as much as $950m, and is perhaps the embodiment of a project that could be huge, but whose chances of actually making it to market at all are vanishingly small. Lilly’s prominence in this analysis suggests that the high-risk/high-reward label could be applied to more than just a few of its R&D projects.
A more general question, given the difference of sellside opinion for some assets, and Brean Murray Carrett’s $1.4bn 2020 forecast for MM-398, for instance, is whether expectations are in some cases running away from reality. No doubt investors will take account of several opinions before deciding which assets fit their methodology.
