大数据能够战胜癌症吗

    “怎么可能与时俱进？”

    首轮资本公司（First Round Capital）合伙人乔什•科普曼说：“我们以前就和奈特及扎克合作过，也投资了他们上一个公司Invite Media。正是考虑到他们的一贯表现，我们决定投资他们现在的企业。”它也参与了Flatiron Health的两轮融资。

    科普曼补充说：“他们致力要解决的问题将对癌症治疗产生重大影响，这一点十分重要，也让我们产生了更强烈的投资兴趣。他们目前组建起来攻克这个难题的团队十分少有。”

    斯诺称，循证医学目前正发生着翻天覆地的变化。他说：“一个医生现在针对某个健康问题所能采取的治疗方案太多了，数量之大变得难以管理，尤其是癌症，因为这个领域十分复杂。”

    斯诺还称，使治疗更趋复杂的问题是，目前渠道在售的很多癌症治疗药物都做过生物标记。这就意味着，它们都是度身定制，只对特定人群、而不是所有人有效。

    斯诺说：“如果你只是个内布拉斯加州社区医院的肿瘤医生，怎么可能跟得上这些发展呢？”

    而另一方面，斯诺表示，Flatiron公司所谓的“96%”数据也“很唬人”。因为绝大多数临床试验都不会接受早期癌症病人，就算接受，一般也是找那些不太常见的癌症。

    斯诺说：“不是每位癌症患者都适合参加临床试验。所谓的4%参与率不是个很有意义的统计数据。”

    使问题更趋复杂的是：癌症患者一般并不知道自己适合参加哪些临床试验，斯诺称。“就算是斯隆•凯特灵纪念中心（Memorial Sloan Kettering，纽约著名癌症治疗中心——译注）的医生可能也不知道斯坦福大学（Stanford）的研究进展。”

    而特纳称，Flatiron的技术可以在病人适合参加特定试验时通知主治医师。

    “癌症领域的超级数据库”

    像斯隆•凯特灵和西奈山伊坎研究院（Mt. Sinai’s Icahn Institute）这样的大型癌症研究机构也在很多方面想利用自己的数据积累开展和Flatiron一样的业务。人们认为这类机构有更多懂得数据处理的专家，而Flatiron的主要问题是它只能用一个机构的数据。不管这个数据量有多大，都不足以解决问题。

    IDC公司Health Insights集团的研究总监阿兰•路易称，要深入研究所有电子病历也存在这个问题。

    路易称：“把所有数据集中到一起是一回事，能否有效利用它们却是另外一回事。必须能把生物信息学和对癌症及症状的深入了解结合起来才行。”

    大型癌症研究机构可能数据量相对有限，因为它们没法像Flatiron那样从多种渠道收集数据，他说：“但我认为这些机构的研究人员可能更懂得深奥的基因组学，也可能更有效地分析数据。”

    特纳称，在电子病历系统上Flatiron公司也是有对手的，如Epic。但就这些病历的分析来说，他表示：“我还真没看到多少对手，也许微软公司（Microsoft）的Excel算一个。”

    IBM公司同样也是对手之一，但“它们主要致力于为1%的客户服务，并专攻决策支持，”特纳称。而美国临床肿瘤协会（American Society of Clinical Oncology）的CancerLinQ算另一个。

    不过斯诺的看法有所不同。他说：“Flatiron更像是个信息服务提供商，而不像斯隆•凯特灵和伊坎这类专业研究机构。而斯隆•凯特灵和伊坎会说‘我们是专业卓越的中心，行业标准的起草者，我们为您提供授权，使您也能用上我们的卓越模式。’”

    而另一方面，斯诺表示，Flatiron的目标是“在癌症研究领域成为像Lexis-Nexis（美国顶级商业与法律数据库——译注）一样的顶级数据库。它主要业务就是收集数据、存储数据，再将数据和相关工具授权给相关研究机构，使其能咨询相关问题。”

    这意味着这家公司更核实的竞争对手是彭博社（Bloomberg）、Lexis-Nexis【隶属于里德爱思唯尔集团（Reed Elsevier）】，IMS Health和汤普森路透（Thompson-Reuters），而不是斯隆•凯特灵这类专攻癌症研究的机构。

    斯诺称：“我想该公司肯定能胜过那些信息销售服务商。这将是他们能纵横驰骋的领域，而目前这个领域的巨头进展缓慢，尤其是在医疗健康领域。”

    他说：“Flatiron的优势在于他们规模不大，聪明灵活。有了这一个亿，他们可能还能招到更出色的人才。”而这正是特纳的计划。（财富中文网）

    译者：清远

    ‘How are you possibly going to keep up?’

    “We have worked with Nat and Zach before, having funded their last company Invite Media, and given their track record, we wanted to be involved in their next venture,” said Josh Kopelman, a partner withFirst Round Capital, which also participated in both of Flatiron Health’s funding rounds.

    “The fact that they’re going after solving a problem that will have major impact in cancer care is hugely important and made it even more compelling for us to be involved,” Kopelman added. “It’s pretty rare to have a team like the one they’re building go after this kind of idea.”

    There is a “seismic shift” happening toward evidence-based medicine, Snow said. “The number of possibilities for what a doctor should do about a health problem is getting unmanageable, he said, “especially in cancer, which is very complex.”

    Further complicating the process is the fact that so many cancer drugs in the pipeline are biomarked, Snow added, meaning that they are tailored to physiologically impact only certain groups of people, rather than the entire population.

    “If you’re an oncologist practicing in a community hospital in Nebraska, how are you possibly going to keep up with all that?” Snow said.

    On the other hand, Flatiron’s “96 percent” statistic is “very deceiving,” Snow said. Most clinical trials won’t accept patients in the early stages of cancer, and even if they do, they typically are looking for a relatively uncommon form of the disease, he said.

    “Not every cancer victim is a candidate for clinical trials,” Snow said. “The fact that four percent participate is not a very meaningful statistic.”

    Compounding the problem: Cancer victims typically don’t know which trials they’re eligible to be in, Snow said. “Even a doctor at Memorial Sloan Kettering may not know what’s going on at Stanford,” he said.

    Flatiron’s technology alerts physicians when their patients are eligible for a particular trial, Turner said.

    ‘The Lexis-Nexis of cancer’

    Large cancer institutes such as Sloan Kettering and Mt. Sinai’s Icahn Institute are in many ways trying to do the same thing as Flatiron with their own sets of data. Those institutions are considered to have more expertise to deal with the data; Flatiron’s fundamental position is that the use of only one institution’s data, however large, is insufficient to fully address the problem.

    It’s the same issue at hand with exploiting electronic medical records overall, said Alan Louie, research director with IDC’s Health Insights group.

    “The fact that you have the data all in one place is one thing, but whether you can make it usable effectively is another,” Louie said. “You have to really combine the bioinformatics with a knowledge of cancer and the manifestation of the disease.”

    The large cancer institutes may have more limited data sets, given that they don’t aggregate data from a variety of sources the way Flatiron does, “but I’d argue the researchers there may be more aware of the subtleties of genomics and probably can analyze the data more effectively,” he said.

    There are other EMR systems that Flatiron competes with, Turner says, such as Epic. But for analytics, “I honestly haven’t seen too many,” he said. “Maybe Microsoft Excel.”

    IBM is also a player, but “they’re very geared toward the top one percent and focused on decision support,” Turner said. The American Society of Clinical Oncology’s CancerLinQ is another.

    Forrester’s Snow saw it differently. “Flatiron is more like an information service provider and less like a Sloan Kettering or Icahn,” he said. “The Sloan Ketterings and Icahns are saying, ‘We are the centers of excellence, the creators of the protocols and the content, and we will sell you the license to have access to our model of excellence.’ “

    Flatiron, on the other hand, “wants to be the Lexis-Nexis of cancer,” Snow said. “Aggregate the data, store it, and license it to the community with tools so you can ask the questions you need to ask.”

    Which means the company is better suited to compete with Bloomberg, Lexis-Nexis (owned by Reed Elsevier), IMS Health, and Thompson-Reuters rather than organizations trusted on the subject on cancer like Sloan Kettering.

    “I think they absolutely can succeed against those who sell information as a service,” Snow said. “That’s the landscape they can prevail in, and that’s the landscape where the incumbents are moving slowly, especially in health.”

    “Flatiron’s trick is they’re little, nimble, and smart,” he said, “and with $100 million, they can probably recruit good talent.” Which is exactly what Turner said he plans to do.