Lab Soft News

Biomarkers Used to Assess Treatment Efficacy

As I think more about the role of biomarkers in the total healthcare delivery process, it appears to me that they can be organized into four categories. A recent press release described a company that is focusing its research on biomarkers to assess the efficacy of cancer treatments (see: Singulex to Develop Assays to Assess Efficacy of Cancer Therapeutics). Below is an excerpt from the article (boldface emphasis mine):

Singulex today announced that it secured the second phase of its Fast Track Small Business Innovation Research ...contract from the National Cancer Institute (NCI). The company will use the funds to develop biomarker assays that will be useful in determining whether a potential cancer therapeutic is effective. Because the levels of specific proteins vary depending on disease state, biomarkers, or proteins associated with a disease, can be measured to determine whether a disease is progressing. In addition, by measuring changes to disease-associated biomarkers following treatment with a specific drug, pharmaceutical developers can assess whether a drug candidate demonstrates efficacy...."Singulex's powerful assay to detect VEGF, a protein implicated in many cancers, overcomes the limitations of traditional biomarker detection technologies and allows drug developers to see normal levels of VEGF in blood and assess the efficacy of anti-VEGF drugs such as Avastin," said [the CEO] of Singulex.

In a recent note, I commented on the potential value of biomarkers for assessing therapeutic efficacy, the same direction that Singulex is moving (see: Moving Resources from the Therapeutic to the Diagnostic Silo). The measurement of therapeutic efficacy has a "downstream role" in the concept of personalized medicine as I described it in a recent note (see: Continuing Discussion of the Role of Lab Medicine in Personalized Medicine). Based on all of the above, here is what might be described as a first cut at the major biomarker functionality categories:

• Diagnose disease, particularly in its early stages (see references to the early health model)
• Function in concert with biotech drugs as companion diagnostics to select candidates for treatment
• Exclude patients from treatment who may suffer a serious side effect (see: Genetic Testing to Cull Out the "Un-Right Patient" as a Candidate for a Particular Drug Therapy; Personalized Medicine and Help/Harm Companion Lab Testing).
• Assess the efficacy on a continuing basis of treatment with a particular drug

Obviously, there is nothing novel about assessing efficacy of treatment or stage of disease with an analyte. Measurement of blood urea nitrogen (BUN) has been used for many decades to assess the severity of renal disease. Presumably these emerging efficacy biomarkers will show far greater sensitivity and specificity in this regard. Also unique these days is that certain companion diagnostics are being developed in tandem with some biotech drugs (see: Some Interesting Insights into Companion Diagnostics).

Biosurgery and Degenerative Joint Disease

When browsing the Genzyme home page, I came across the fact that one of their EVPs, David Meeker, M.D., is in charge of a business unit called Therapeutics, Biosurgery & Transplant. This reference to biosurgery caught my attention because I had not encountered the term before. Genzyme defines biosurgery on their web site in the following way (see: About Genzyme Biosurgery):

Genzyme Biosurgery is a leading developer of novel biotherapeutic and biomaterial products.  Our products and research pipeline harness the power of biology to improve the health of patients in the orthopaedics and post-surgical adhesive disease areas....The result: a suite of products and product development candidates designed to provide new tools to help improve patient outcomes, reduce the time and complications associated with surgery, and may in some cases, postpone the need for surgery.

Their biosurgery product list for Genzyme is focused on the treatment of degenerative arthritis. Here is a list of company products in advanced testing stages:

• Synvisc (hylan G-F 20) for hip – U.S. Phase 3. Genzyme is conducting a clinical study to evaluate the safety and efficacy of Synvisc in patients with pain due to hip osteoarthritis.
• Synvisc for ankle, shoulder– Europe. Phase 3. Genzyme is conducting several clinical studies designed to evaluate the safety and efficacy of Synvisc in patients with pain due to ankle or shoulder osteoarthritis.
• Hylastan for knee. Phase 2. Genzyme is conducting a clinical study to evaluate the safety and efficacy of hylastan in patients with pain associated with osteoarthritis of the knee.

i discovered after further investigation that biosurgery has two meanings: the first is the use of living organisms such as maggots to clean and disinfect a wound. The second, and obviously the use intended by Genzyme, is defined in the following way (see: Biosurgery):

A second sense has grown up in parallel with it to mean surgical techniques that employ a range of natural or manmade materials. Biomaterials are biologically compatible glues to seal surgical incisions, lubricants to help joint movement, and support on which living tissue is grown or shaped, but the term also covers biocompatible materials such as hip replacements and artificial pacemakers. It also includes biotherapeutic techniques such as gene and cellular therapies.

For me, the use of the term biosurgery in relationship to the injection of, say, Synvisc into a knee joint for the treatment of osteoarthritis is confusing because the only surgery involved is the introduction of the needle into the knee joint. At least in this case, the terms biomaterials and biotherapy make more sense.

Introducing the Virtopsy, a Variant of the Catopsy Theme

In two recent notes (see: Reinventing the Autopsy: CT Imaging as a Routine Part of the Procedure; Additional Discussion About Reinventing the Autopsy), I proposed that the classic autopsy could and should frequently be initiated with a total-body CT scan followed by selective tissue biopsies to confirm the initial CT impressions as well as to harvest tissue for tissue banks. I labeled this new approach to the post-mortem examination as the catopsy. It seems that a parallel approach on this same general theme is also being discussed: the virtopsy or virtual autopsy. In this latter case, the post-mortem examination is "bloodless" and limited to a various medical imaging modalities without any tissue sampling. Below is an excerpt of a relevant article (see: Virtopsy: the virtual autopsy):

Multi-slice computed tomography (MSCT) and magnetic resonance imaging (MRI), when used with 3-D imaging technology, create vivid images of the interior of the human body. Dr. Richard Dirnhoter and Dr. Michael Thali and their team of specialists at the University of Bern's Institute of Forensic Medicine, Switzerland are using these new imaging technologies to develop "Virtopsy"—a bloodless and minimally invasive "virtual autopsy" procedure to examine bodies for causes of death. Virtopsy detects internal bleeding, bullet paths, and hidden fractures hard to find in a traditional autopsy. The MSCT and MRI aid in picturing fracture patterns, bone and missile fragmentation, brain contusion, 3-D bullet localization, gas embolism, and blood aspiration to the lung. Unlike traditional autopsy, Virtopsy does not destroy human tissue. It can be used when religious beliefs prohibit, or families object to, the cutting open of the body. The developers of Virtopsy do not envision the procedure as a replacement for traditional autopsy but as a tool to be used in cases where dissection of the body is not feasible or where forensic evidence is particularly hard to visualize.

I do need to take issue to the implication in the article that the standard autopsy "destroys tissue." If anything, the procedure preserves tissue. Despite this quibble, the virtual autopsy (virtopsy) also seems to be taking hold at U.S. institutions such as Johns Hopkins (see: A GROWING NUMBER OF AUTOPSIES ARE DONE USING CT).

Frankly, I can't envision a turf war erupting between pathology and radiology over an expensive, time-consuming, and uncompensated procedure that many pathologists would like to avoid and that seems to to be falling into disfavor (see: The declining autopsy rate and clinicians' attitudes; Declining clinical autopsy rates versus increasing medicolegal autopsy rates in Halifax, Nova Scotia).

Although a virtopsy is better than no autopsy at all, I favor the hybrid approach, the catopsy, because of the opportunity to obtain tissue to confirm imaging impressions with firm histopathologic diagnoses. The autopsy suite is also better equipped to manage deceased patients than the radiology department. This statement, of course, assumes that suitable imaging equipment can be installed in the autopsy suite. Although older CT scanners can probably be obtained at a reasonable cost, substantial costs would also be incurred in retrofitting the autopsy suite for this additional functionality.

Continuing Discussion of the Role of Lab Medicine in Personalized Medicine

In response to my note of last Friday (see: Genetic Testing to Cull Out the "Un-Right Patient" as a Candidate for a Particular Drug Therapy), Dennis Dionne submitted the following comment:

Chemotherapy drugs, i,e.; 5-fluorouracil (5-FU), are dosed by BSA [body surface area]. In the May 2008 issue of the Journal of Clinical Oncology, Dr E Gamelin reported a randomized multi-center study among metastatic CRC [colo-rectal cancer] patients treated with continuous infusion 5-FU whose drug blood levels were monitored and dosing adjusted to achieve target concentrations. Gamelin demonstrated that patients were over-dosed and only 25% were actually in the optimal range. By practicing therapeutic dose management (TDM) to maintain the plasma level of 5-FU within the therapeutic window, patients did much better. They experienced significantly better objective tumor response and doubling of survival at two years, while experiencing significantly less toxicity. Therefore, once the right patient is placed on the right therapy (genetic testing), therapeutic drug monitoring can make sure the right dose correlates with the right plasma level exposure to achieve optimal value.

A very relevant comment and I will return to this later.

In my mind, the definition of personalized medicine, also known as targeted therapy, is becoming more complex. The source of this additional complexity lies in the following new developments: (1) the growing importance of companion diagnostics; (2) the growing importance of medical imaging in both the diagnosis of disease, particularly tumors, and monitoring the efficacy of treatment; (3) the use of genomic and proteomic testing to cull out candidates for drug therapy who may be harmed by it; and (3) the use of medical imaging post initiation of chemotherapy to monitor its efficacy. The most obvious example of this latter approach is to determine with imaging whether a primary tumor or its metastases are shrinking.

Lets start this discussion of personalized medicine with what I considered to be a reasonable current definition from the Wikipedia presented below. To the right I also offer a diagram of an expanded flow diagnostic work flow diagram that encompasses some of the newly emerging nuances of personalized medicine, as I understand them. In the diagram, I include a box at the bottom to represent Dennis' very relevant comment about therapeutic drug monitoring (see above).

Personalized medicine [involves the] use of information and data from a patient's genotype, or level of gene expression to stratify disease, select a medication, provide a therapy, or initiate a preventative measure that is particularly suited to that patient at the time of administration. In addition to genetic information, other factors, including imaging, laboratory, and clinical information about the disease process or the patient play an equally important role.

In this diagram, I distinguish between "mandated lab companion testing" and additional genomic/proteomic testing because the former type of testing is frequently developed collaboratively by pharmaceutical companies and in-vitro diagnostic companies and often required or recommended by the FDA. The latter category may include newly discovered test methods that may clarify the therapeutic picture.

On the basis of this expanded understanding of personalized medicine (targeted therapy), I offer the following expanded definition that I have intentionally kept relatively brief for clarity:

Personalized medicine involves the use of laboratory-based molecular diagnostics and medical imaging to select suitable candidates for treatment with a particular drug(s), to rule out those patients who would suffer unacceptable side effects from the proposed drug treatment, and to monitor the health status of the patient post initiation of therapy to assess therapeutic drug levels and the continuing efficacy of the agent in suppressing or curing the disease.

Genetic Testing to Cull Out the "Un-Right Patient" as a Candidate for a Particular Drug Therapy

The goal for personalized medicine has been drummed into my head using the following mantra: selecting the right drug for the right patient at the right time. However, I have tended to avoid the use of the term personalized medicine, favoring the term targeted therapy. However and not unexpectedly, we have come to another turn in the road regarding the meaning of personalized medicine/targeted therapy medicine due to new scientific discoveries.

My goal for this note is to discuss what I will refer to as the need to cull out the un-right patient as a candidate for a particular drug therapy. A recent article in the New York Times (see: F.D.A. Urges Genetic Test Before Giving AIDS Drug) suggested this idea to me. Below is an excerpt from the article with boldface emphasis mine:

Seeking to prevent life-threatening side effects, the Food and Drug Administration is urging doctors to use a genetic test to screen patients before prescribing a drug widely used for H.I.V. infection and AIDS. In an advisory ..., the agency says that patients with a particular variation in an immune system gene should not be given the drug abacavir because they are at a far higher risk of a severe allergic reaction to the drug....F.D.A. officials said they believed that the genetic test, which is already available from some laboratories, would be quickly adopted for screening patients before prescribing abacavir....A small percentage of abacavir users suffer so-called hypersensitivity reactions, either when they start the drug or when they resume using it after some interval....Based on that data it was estimated that 61 percent of people with the genetic variant — called HLA-B*5701 — would suffer a hypersensitivity reaction, in contrast to only 4 percent without the variant, according to new information on the drug’s label.

In previous notes, I have commented at length about the use of companion diagnostics (see: Consideration of a Broader Definition for "Companion Diagnostics"; A Closer Look at Companion Diagnostics Strategies; Some Interesting Insights into Companion Diagnostics). My understanding of companion diagnostics has been that various lab tests are used to select appropriate candidates to receive a particular drug. In other words, the goal has generally been to select the patients for whom a  drug will be effective. I extended this concept in another note about therapeutic efficacy (see: Moving Resources from the Therapeutic to the Diagnostic Silo) in which I suggested that molecular diagnostics and medical imaging could be used to identify those patients for whom treatment with drugs, particularly expensive chemotherapeutic agents, would be effective. It was overtly stated, however, that this approach would also enable us to cull out those patients for whom the use of a drug would not be effective. In so doing, resources could be saved.

I now understand that we need to expand our understanding and definition of personalized medicine and targeted therapy to include the selection of the right patient AND the exclusion of the un-right patient. By the use of this latter term, I mean both the patient who will not be helped by the therapy and also the patient who will be harmed by the therapy. This may seem like splitting hairs but I do like the notion of using lab testing to identify both the right and the un-right patients.

Strategic Challenges Facing Hospital-Based Pathology & Lab Medicine Groups

When browsing the July issue of Laboratory Economics, I came across the following table: What are the biggest challenges pathology groups will face over the next 5 years?  I thought that such a list would be a good starting point for generating an expanded list of the major strategic challenges facing hospital-based pathology and lab medicine groups in the near-term. To launch this discussion, I now provide this table for your review.

It's clear to me that nearly all of the items in this table represent major challenges with the exception perhaps of pod labs that I believe are declining in strategic importance, as the table reflects, and will continue to decline because of legal and regulatory challenges. In addition, I will add three additional items to the list to reach an even ten: acquisition of capital funds, regulatory pressure, and the challenge of increased laboratory automation.

This leads me to offer the following list to readers of what I believe are the top ten strategic challenges facing hospital-based pathology and clinical labs:

• Declining reimbursement
• Competition from commercial labs
• Staffing shortages, particularly medical technologists
• Specialty physician groups insourcing surgical pathology services
• Demand for more sophisticated molecular diagnostic products
• Selection and deployment of new information technology
• State and federal clinical laboratory regulation
• Increasing pressure for enhanced laboratory automation
• Deploying point-of-care lab systems
• Evolution of the digital pathology department

I debated whether I should assume that digital imaging is included in the information technology category but decided to keep is as its own separate category because it poses a set of challenges far beyond technical solutions. I would be interested in any comments from readers of this blog about how to improve this list. It is my intention in future months to attempt to address many of these issues in greater detail.

The "Early Health Model" as an Example of Disruptive Innovation

I am pleased to announce that Dr.Jason Hwang has agreed to deliver a lecture at the next Lab InfoTech Summit that will take place in Las Vegas on March 16-18, 2009. He is a co-author with Clayton Christensen of a book (The Innovator's Prescription: A Disruptive Solution for Health Care) that discuses disruptive innovation as it relates to the healthcare industry entitled. You are going to have to wait for it to be published -- it won't be available until October 23, 2008. However, we will be distributing complimentary copies of this book to the first 100 conference registrants. Details about this offer to follow.

To launch this discussion about disruptive innovation, here are two brief quotes from the Wikipedia defining the term with boldface emphasis mine:

A disruptive technology or disruptive innovation is a term describing a technological innovation, product, or service that uses a "disruptive" strategy, rather than a "revolutionary" or "sustaining" strategy, to overturn the existing dominant technologies or status quo products in a market....Christensen replaced disruptive technology with the term disruptive innovation because he recognized that few technologies are intrinsically disruptive or sustaining in character. It is the strategy or business model that the technology enables that creates the disruptive impact.

"New market disruption" occurs when a product fits a new or emerging market segment that is not being served by existing incumbents in the industry.

I have posted a number of previous notes about the early health model (EHM), defined as pre-clinical, pre-symptomatic diagnosis of disease, which I believe is an example of disruptive innovation. I put the question to Jason about the extent to which the that the EHM is disruptive for the various major participants in our healthcare system. Here is his answer:

One of the key tests of whether something is disruptive is to ask if it serves the interests of the industry’s most favored customers and, conversely, if it serves a market that was previously disenfranchised or altogether ignored. Obviously, healthcare is bit more complex, because [the term] “customers” is not so clearly defined. However, here’s my take on the early health model:

• [The early health model] is almost certainly disruptive to MDs and hospitals, because they have little to gain (and almost certainly will lose income) when they diagnose diseases pre-symptomatically. More importantly, if the early health model implies that someone with less skill and training, especially the patient herself, can do some of the diagnostic workup, then it is clearly disruptive.
• The answer is not so simple with insurers. I think all of them would say they would prefer to pay for pre-symptomatic and preventive care, so it’s difficult to say they would be disrupted. I’m also not convinced that their business model necessarily prevents them from paying for such a model of care, other than the fact that they could potentially alienate their existing network of providers. My best prediction is that it will take a new insurer that primarily promotes early health through the use of health savings accounts, combined with a more traditional provider for acute and complex care, paid through the use of catastrophic insurance.
• Pharma and other suppliers would indeed object, because this naturally squeezes their opportunities for downstream care, much like it does for the providers. Profit margins would also be predictably lower at the preventive end of care, so it would be difficult for their business model to adapt. Needless to say, they could be the ones being disrupted.
• Finally, the new markets that would open up would primarily benefit patients and purchasers. However, the first customer segments would be those who have the right incentives to seek out early care, such as patients paying out-of-pocket, self-insured employers, or proactive/educated patients. From their perspective, I’d agree this is a new-market disruption, since they have no real alternative otherwise.

Joel Saltz Appointed as Emory Healthcare’s Chief Medical Information Officer

In a previous note, I reported Fred Sanfilippo's departure from Ohio State (see: Fred Sanfilippo, Noted Pathologist, Moves to Emory). Joel Saltz, a leader in pathology informatics, is now making a similar move to Atlanta (see: Bioinformatics Pioneer Will Lead New Initiatives at Emory University). He currently serves as professor and chair of the Department of Biomedical Informatics and professor in the Department of Computer Science and Engineering at The Ohio State University, Davis Endowed Chair of Cancer at OSU, and a senior fellow of the Ohio Supercomputer Center. Below is an excerpt from the article:

Joel H. Saltz, MD, PhD, a pioneer in the fields of high-performance computing and biomedical informatics, will join Emory University’s Woodruff Health Sciences Center in September as director of the Center for Comprehensive Informatics and as Emory Healthcare’s Chief Medical Information Officer. The announcement was made by Fred Sanfilippo, MD, PhD, Emory executive vice president for health affairs, CEO of the Woodruff Health Sciences Center and chairman of Emory Healthcare....As chief medical information officer for Emory Healthcare, Saltz will direct strategic planning and implementation of the comprehensive Emory Medical Information Enterprise. He will guide recruitment, research and resource allocation for informatics programs across academic departments. Additionally, he will lead the further development of Emory’s external partnerships in bioinformatics, including those with the Georgia Institute of Technology, Children’s Healthcare of Atlanta, Morehouse School of Medicine, the Atlanta Veterans Affairs Medical Center, the Georgia Research Alliance and the Georgia Cancer Coalition.

Epic Flexes Its Political Muscle in Wisconsin with Boycott

Some of my recent notes have focused on the corporate culture of Epic Systems (see: Epic Systems and Its Corporate Culture; More on the Epic Culture: Is This a Cult or a Company?). This activity has brought to my attention a political tempest that has been brewing around Epic in its own back yard in Wisconsin (see: Epic Systems Should be WMC's Biggest Fan). Epic recently announced that it will no longer do business with companies associated with Wisconsin Manufacturers & Commerce (WMC), a Wisconsin trade and lobbying organization.  The reason for this boycott is Epic's opposition to WMC's involvement in the election of Michael Gableman to the State Supreme Court. This same article makes the following comment about Judy Faulkner, the CEO and founder of Epic (boldface emphasis mine):

Epic has always marched to its own beat.  Founded and led by Judy Faulkner, the company has blossomed into one of the top five providers of automated medical record systems in the nation....She likes telling people that the company is growing 30% annually and has zero debt.... While on paper she has a number of advisory groups, no company decision of any consequence is made without her....[She refuses to] advertise or market [and shuns] traditional business planning....But she can do that because it is her company. She has no board of directors to answer to, no proxy votes to worry about and no annual shareholder meetings where she has to explain herself.  In fact, the reclusive Ms. Faulkner feels little need to explain herself about anything.

Forbes.com had the following to say in a recent article about this same issue (see: Madison area company targets lobbying group):

[WMC's] participation in the recent state Supreme Court race won by ... Judge Michael Gableman has drawn the ire of Epic Systems, a medical records company in Verona that doesn't belong to the group. Epic said in a statement it was upset that WMC spent an estimated $1.8 million on that race and may pull business from local vendors who support the group. Those vendors included...J.P. Cullen & Sons, the contractor for Epic's more than $200 million campus expansion. CEO David Cullen resigned from WMC's board of directors on June 9 and his company dropped its membership....Epic's threat not to work with another company based on an election campaign appears to be the first of its kind nationwide, said Howard Schweber, a professor of law and political science at the University of Wisconsin-Madison. Its action is perilously close to a type of illegal boycott that typically arises in labor disputes, Schweber said. "We should be uncomfortable when private businesses have enough power to coerce businesses or other organizations to change their political views or affiliations or keep them secret," Schweber said.

Here's more about Epic's actions from Professor Schweber (see: Howard Schweber: Epic's power should not be used to silence others). He himself is very critical of WMC but makes some very interesting points about primary versus secondary boycotts and this is the basis for his criticism of Epic. In the case of a secondary boycott, an attack is mounted on the supporters of the primary target.

What I find the most interesting about this controversy is that large health systems such as Kaiser are willing to spend hundreds of millions of dollars for mission-critical clinical software supplied by a privately-held company like Epic that is so tightly controlled by its executive office and founder. This same issue may occasionally arise during the purchase of an LIS but the stakes are never as high.

Proposed Naming Schema for All Clinical Lab Testing

A recent post (see: Comparison of "Traditional" Lab Testing with Direct Access Testing) marked an effort on my part to begin to establish a naming convention for all clinical lab testing. Here's a quote from the note:

The key difference between traditional lab testing (TLT) and direct-access-testing (DAT) is that physicians order the former and they are always performed by someone other than the patient/consumer. Patients/consumers order and frequently pay out-of-pocket for the latter and they may or may not perform the tests themselves.

In response to this note, I received the following comment from Mark Terry:

I'm the author of the article [quoted in the blog note]....Although I don't disagree with your definitions in general, they are very, very broad, which poses a lot of difficulties in terms of market research and analysis. DTC, Direct-To-Consumer, would include pregnancy tests, ovulation tests, even blood sugar urine strips and test kits, which is a very different market (and market size) than DAT, Direct Access Testing, which is typically referring to Minute Clinics and Doc-in-The-Box or online services that provide laboratory testing without a physician's order, but generally send the samples to an independent laboratory (most often LabCorp and/or Quest Diagnostics).

I think that I will hold firm to my original idea to label all lab tests not ordered by physicians as direct-access-testing (DAT) instead of direct-to-consumer testing. After all, the consumer is "directly accessing" the tests without a physician as an intermediary. Having made this first cut of physician-ordered versus consumer-ordered tests, it seemed necessary to continue this naming schema to include additional levels. Below is a diagram of this work-in-progress.

There will be little question or argument about the left side of the diagram: traditional lab testing (TLT). On the right side of the diagram is what I refer to as direct-access-testing (DAT), all of which is ordered by patients/ consumers. What is slightly confusing about this right side of the diagram is that some of this testing involves involves self-collection and submission to a professional lab (e.g., submission of cheek swabs to the DNA labs), some involves self-collection and home testing with a kit or home instrument like a glucometer, and some involves collection in a patient service center (PSC) and test performance in a professional lab. This latter category, as well as DNA testing, is the "classic" form of DAT advertised and brokered by a web site.

I anticipate that there will be more variations on this DAT theme in months to come despite the attacks on genomic testing by the California lab regulatory body (see: Direct Access Genetic Testing in California). As noted above, this naming and classification schema is a work-in-progress so I am interested in any comments from readers of this blog about whether it makes sense and also how it can be improved.