The National Biomarkers Development Alliance (NBDA)*
“AN INAUGURAL PUBLIC SCIENTIFIC FORUM”**
(* Startup funding for the NBDA provided by the Arizona State University; ** Workshop funding provided by the Virginia G. Piper Charitable Trust)
March 25, 2013
The Biodesign Institute Auditorium
Arizona State University
“Never doubt that a small group of thoughtful, committed people can change the world. Indeed, it is the only thing that ever has.”
The NBDA Concept – Current Status – Future Plans
Anna D. Barker, Ph.D., Director, NBDA, Co-Director, Complex Adaptive Systems Initiative; Professor, School of Life Sciences, Arizona State University
Dr. Barker opened the NBDA Forum, the organization’s first public event, by introducing the initiative as one step along the path toward personalized medicine. The science of disease biomarkers—discovering them, proving their clinical usefulness to both physicians and regulators, and then using them to help guide and improve patient therapy—is a critical component of personalized medicine, yet one that, for a number of reasons this forum will explore, is not fulfilling its promise, and hence is not facilitating the transition to an era in which chronic disease is treated more effectively and efficiently.
The reason the NBDA’s efforts come at a critical time is that the nation is facing an unprecedented economic and social crisis that will result from the growing impact of chronic diseases associated with aging, particularly Alzheimer’s disease, cancer, and diabetes. By the middle of this century, some 16 million Americans will suffer from Alzheimer’s disease—an increase from 5.4 million today—with a projected annual cost of $130 billion. The incidence of cancer is set to rise 30 percent over the next 20 years and cost over $260 billion annually. Diabetes has gone from a relatively infrequent disease thirty years ago to the seventh leading cause of death, and the rapidly rising incidence of diabetes shows no signs of abating.
These are not just problems for the United States. For example, the incidence of cancer in the developing world has increased by 50 to 100 percent since 2002. If current trends continue, by 2020 some 10.3 million people worldwide will die from cancer each year and 16 million new cases will be diagnosed worldwide. The economic and social burden on the developing world is expected to be extreme, even destabilizing.
Gene sequencing provides a new tool that will undoubtedly increasingly provide more detailed information about diseases such as cancer. With the cost of sequencing falling rapidly, patient sequencing could soon become routine: the National Genome Research Institute estimates that the total number of human genomes that will be sequenced will rise from about 30,000 in 2011 to one million by 2014. The information about disease that could be mined from that 28 petabytes of data could lead to revolutions in medical therapy for chronic disease, but only if research can turn that raw data into knowledge that can be applied in the context of each patient. Biomarkers must play a key role in making that translation.
What is a biomarker? It is simply a measurable biological or physical change (signal) in the body directly related to a normal or disease process. To be useful, a biomarker must be reproducibly measurable and quantifiable and directly related to biological processes or changes in those processes, whether those processes are associated with a disease state, a drug response, or other health indicator. Biomarkers have the potential to help uncover the underlying biology of disease and lead to new target discovery; be used as markers of toxicity and metabolism that can guide drug development; serve as surrogate endpoints for clinical trials; enable early detection of disease; identify the molecular basis of specific disease subtypes; help physicians make rational therapeutic decisions based on subtype classification or patient stratification; be used to assess treatment effectiveness; serve as prognostic disease markers; and be used in disease prevention.
One example is the mutation that is tied to the development of the anticancer agent Gleevec. Although the discovery of this mutation and its development as a clinically useful biomarker led to the development of perhaps the most successful targeted drug therapies for cancer, this success turned out to be one that is not broadly applicable to the discovery and development of other biomarkers and associated targeted therapies for cancer. As it has turned out, Gleevec and its biomarker were not the first in a wave of expected new generation of therapeutic interventions specifically targeted to a patient’s molecular profile. It was a suggestion of what might be possible but to date has not been reproduced in a broad sense across other types of cancer.
Currently, it is incredibly expensive to develop new drugs and most potential drugs fail in clinical trials because biomarkers fail. Few biomarker candidates are translating into clinical utility, and the rate at which the field is achieving success has been declining. While researchers have submitted over 1260 putative protein biomarkers to the Food and Drug Administration (FDA) for approval, and over 150,000 papers in the scientific literature have reported "biomarker discovery," less than one protein biomarker per year has been approved by the FDA since the early 1990s. In addition, the vast majority of the claims in these publications suffer from bias, lack of standards, and reproducibility across laboratories that make these discoveries questionable. It is currently estimated that fewer than 100 biomarkers are routinely used in the clinic today.
Biomarker development is slow and difficult for a number of reasons:
- The knowledge of the biological space is insufficient to understand how biomarkers reflect function.
- Biomarkers are often developed without attention to biospecimen and technology standards.
- Proteomics technologies and others are not standardized, with the result that the findings from one laboratory are often not reproducible in another laboratory.
- Biomarkers are not developed specifically for a specific application, that is, they are not “fit for purpose.”
- Biomarkers are not qualified before entering clinical trials.
- Few academic researchers understand the rigor required to develop a biomarker as a clinically useful diagnostic test, rather than simply as an analytical measure for routine laboratory use.
- The regulatory pathways are not clear, again because of a lack of standards to guide regulatory decisions.
- There is no end-to-end, standards-based system for biomarker development.
The NBDA is tackling this last issue as a mechanism to address all of these other shortcomings and has proposed the system shown in Figure 1 as the means to do so. Toward this end, the NBDA held its first workshop in December 2012 to assemble the overall collaborative construct that will be needed to create this pipeline. The NBDA is getting underway and planning to hold additional workshops and think tanks to thoroughly understand the problems. The NBDA is a unique consortium that will serve as a national "catalyst" to identify and resolve the problems that plague biomarker development. The NBDA will not reinvent wheels and plans to publicly share information and knowledge.
“The world we have created today has problems which cannot be solved by thinking the way we thought when we created them.” Albert Einstein
ASU, Personalized MEDICINE, and the NBDA
Michael Crow, Ph.D., President, Arizona State University
In his brief welcoming remarks, Dr. Crow described the design of Arizona State University (ASU) as being different from other universities of the same scale. ASU decided not to become an “elite” institution in terms of who it admits but rather to focus on broad admission and an "elite education" for who qualify. One driver for this change in attitude is that nationwide the top 20 percent of students in terms of academic ability who come from the lowest 20 percent of family income have less than a 10 percent chance of graduating college. This is a situation that must change to meet Arizona’s needs (and that of the nation) for an educated workforce.
ASU has also chosen to take a different path in terms of how it approaches developing solutions for the nation’s health care system, with the NBDA being a prime example of that approach. Dr. Crow noted that he wrote a commentary for the publication Nature, in which he proposed revamping the National Institutes of Health (NIH) into three institutes: a fundamental biomedical systems research institute, a health outcomes research institute, and a health transformation institute. Such reorganization would limit and perhaps eliminate the Balkanization of biomedical research that comes from having separate NIH units dedicated to particular diseases and instead reflect the convergence of thought and collaboration across a discipline, which is becoming the norm. He closed by expressing ASU’s commitment to the NBDA and its integrated, collaborative approach to biomarker development.
Are Biomarkers the Path to Personalized Medicine? If So, What Will It Take from the FDA’s Perspective?
Janet Woodcock, M.D., Director, Center for Drug Evaluation and Research, Food and Drug Administration (Video Cast)
Dr. Woodcock began her remarks by stating that it is clear today that improving the outcome of medical treatment requires biomarkers. Diagnosis is the foundation of medicine and biomarkers have the ability to extend the diagnostic reach of the best physicians in an era of personalized medicine. She characterized the current state of diagnostics as appalling given the vast increase in biomedical knowledge that the research enterprise has produced over the past decade. She noted, too, that the pharmaceutical industry talks about the challenges they face identifying suitable validated biomarkers as being the most difficult step in developing targeted therapies. The industry also blames the biomarker deficit as being a key reason why it has such great difficulty predicting the efficacy of drug in specific patients.
Clearly, the lack of good biomarkers is inhibiting medical progress and drug development. The main problem confronting biomarker development, she said, is that academics focus on discovery and publishing, not the grunt work that it takes to develop a replicable, applicable biomarker. In fact, too many of the publications in this field cannot be reproduced, pointing to the need for the development of standards, particularly for tissue acquisition and assay development. The field also needs to move beyond studies involving small numbers of patients to much larger-scale trials that demonstrate clinical correlation. She noted that the FDA has a data submission safe harbor provision that biomarker developers should take advantage of as they move toward large-scale clinical correlation. The FDA, as part of its Critical Path Initiative, offers advice to biomarker developers, but its resources are being strained by requests for such advice.
Researchers also need to pay more attention to analytical validation of biomarker assays, which again relates to standardization, and they need a better understanding about what the evidentiary standards should be at each stage of biomarker development. Although there is a large body of scholarly work on how specific assays are applicable on a population basis, there has been little work done establishing that a biomarker is meaningful. She noted that the field needs to establish the level of standardization and replication that must be met before publishing a biomarker finding. Today, journals accept as sufficient findings from just a few individual test subjects, and this bar for publication needs to be raised substantially. The lack of reliability in the published literature is a serious issue.
Dr. Woodcock noted that the President’s Council of Advisors on Science and Technology (PCAST) issued a report in 2012 on how to speed innovations in drug development, and one of the recommendations was that government work with various groups, such as the NBDA, that have initiatives in this area. She added that the Howard Hughes Medical Institute and the Institute of Medicine are running a workshop on late-stage biomarker evidentiary standards, and she hoped that efforts such as these could address the urgent need to develop consensus standards for biomarker development. The numbers of parties that are becoming interested in the biomarker problem hold promise because addressing the issues that confront biomarker development requires a coalition given the scope and complexity of the problem.
In closing, Dr. Woodcock said that the FDA would be delighted to part of the NBDA’s effort. Biomarker development is essential to realize the promise of personalized medicine, and speeding the development of biomarkers will require establishing evidentiary standards, something that the NBDA can accomplish. She added that NBDA could play a key role in biomarker development by taking a role similar to the Cochrane Collaboration in terms of developing evidentiary standards and using them to assign a level of evidence to a preliminary finding.
Upon completing her presentation, Dr. Woodcock responded to a question about how the FDA will accommodate biosignatures as opposed to single-entity biomarkers. She said that the agency’s current position is that there is a difference between an empirical finding—a biosignature—and a mechanistically driven result such as an individual protein or gene linked to a specific disease state. Empirical work requires more substantiation, she said, while mechanistic work is likely to be more robust.
A Facilitated Panel Discussion
Advancing Biomarkers: The Best of Times and the Worst of Times
Anahita Bhathena, Ph.D., Associate Director of Cancer Biomarkers, AbbVie Pharmaceuticals
The task of the forum’s first panel was to assess the state of the biomarker field. Dr. Bhathena discussed the troubling decline in the productivity of drug discovery and development. She cited figures from the National Institutes of Health (NIH) showing that when adjusted for inflation, the cost of developing a new drug doubles every nine years, and she reviewed some of the opportunities for biomarkers to increase the success rate across the drug development pipeline. In the discovery phase of drug development, biomarkers can help identify and validate new targets and refine chemistries of drug candidates. Biomarkers can be used to better understand the molecular mechanism of disease and the drug mechanism of action and resistance, identify the target patient populations, and provide proof that drug are reaching their target and that they have suitable pharmacokinetic properties.
During preclinical and clinical development, biomarkers are essential for improving clinical trial strategies that target the right patient population and that take advantage of adaptive trial designs. In particular, biomarkers can increase our understanding of target patient populations. Biomarkers can enable the single biggest increases in drug development productivity by enabling drug developers to select the right population for a clinical trial and to make a go/no-go decision at the earliest possible time in the clinical trials process.
The development pathway for a paired drug and diagnostic is relatively straightforward when the biomarker is identified pre-clinically. The development pathway becomes more challenging when the biomarker is discovered during Phase I or Phase II clinical trials. And of course the decision to use a biomarker determines the assay criteria and ultimate regulatory requirements for the trial. For example, laboratory developed tests, or LDTs, are usually suitable for internal decision making, while a biomarker that will be used as a companion diagnostic will have to satisfy the more stringent requirements for a Premarketing Approval (PMA) application.
Laura Esserman, M.D., M.B.A., Director, Carol Franc Buck Breast Cancer Center; Professor of Surgery and Radiation, University of California, San Francisco
Dr. Esserman expanded on the theme of where biomarkers can make a difference by discussing three promising opportunities. First, biomarkers can help influence the fate of disease regarding its stability and/or progression. Second, biomarkers can provide the means to assess the rate of disease progression and identify which patients are likely to respond to therapy. Third, biomarkers may afford the chance to slow or reverse disease progression by identifying those at risk. In terms of cancer, biomarkers can provide clinically critical information on tumor heterogeneity that could impact diagnosis and treatment. Biomarkers may also offer opportunities to address the need to test literally hundreds of new agents in the development pipeline.
The I-SPY 2 trial, an ongoing study of women at risk for early systemic recurrence of breast cancer, uses an FDA-approved, 70-gene panel as a biomarker, along with several other biomarkers, to test a significant number of new agents using biomarkers to accelerate the rate of "knowledge turns" (to essentially create a learning system). This trial, and the innovative use of biomarkers to inform an adaptive trial design, is a major attempt to rewire the drug development process for the purpose of taking time off the clock and reducing costs.
George Poste, D.V.M., Ph.D., Regents’ Professor and Del E. Webb Chair in Health Innovation; Co-Director, Complex Adaptive Systems, Arizona State University
Dr. Poste discussed what he calls the “S's” for proficient biomarker discovery, validation, and adoption, stressing that systems, not silos, lead to solutions. Solutions must define the intended use of a biomarker, with a clear focus on utility versus the application of new technologies. Standards are critical to enable every step of the development pipeline, from discovery through regulatory and reimbursement approvals. It is critical that institutional review boards (IRBs), journals, and funding agencies assume responsibility for enforcing these standards, especially for the quality of the biospecimens. In that regard, the field must address issues involving biospecimens, the raw materials for all biomarker research, particularly with regard to quality and understanding the effects of preanalytical variables. In addition, the biospecimens must be linked to data on clinical phenotype, patient health measures, and clinical outcomes.
Biomarkers are key tools for segmentation or stratification of patients into relevant cohorts for clinical trials. Statistical methods are becoming increasingly important in biomarker science because of the high dimensionality problem associated with multiplexed assays. Scale is important, both in terms of the number of samples required for biomarker studies and the data that these studies produce. Biomarker studies need to become more sophisticated to accommodate the full complexity of biomarker validation and regulatory policies, and they need to include pharmacoeconomic measures to identify the savings that will sway payers to better value biomarkers. Finally, all biomarker assays need to be developed with an eye on seamless adoption into existing clinical workflows with demonstrable benefits.
David Parkinson, M.D., Venture Partner, New Enterprise Associates, Inc.
Dr. Parkinson acknowledged the promise of biomarkers over the long run but characterized the short-term situation as dire. He posed the question, “Why is it so hard to develop biomarkers that are useful?” We are still diagnosing cancer using pathology and characterizing it by organ, while most research today focuses on molecular targets that may have little correlation with pathological characteristics. These disconnects between diagnosis and biomarker research lead to a situation where patients are not being characterized effectively.
The FDA understands the value of molecular diagnostics and biomarkers, but the evidentiary standards for biomarker approval are high. However, the economics of biomarker development currently are not producing a return on investment, which makes it very difficult to pay for the studies needed to satisfy current standards. For example, a new drug was recently approved based on the development of a well-validated assay; however, the cost of drug therapy exceeds $100,000 per year, while the diagnostic assay costs $200. The situation is made worse by the fact that many laboratories can develop and run variations of the assay developed in-house, which lowers the return to the company that validated and submitted the assay for approval. The only remedy for this situation is to change policy to recognize the value of biomarkers and reward that value economically. One approach being taken in Europe is to only reward drug makers for clinical results, which incentivizes companies to develop drug-prediction biomarkers.
Summary (Key issue/desired change from the panelists)
- Ask better clinical questions that biomarkers can address and keep the clinical endpoint in mind when developing the biomarker.
- Ask regulators and payers to work together to realign incentives to reimbursement of molecular diagnostics.
- Increase the numbers of large-scale studies that serve as learning systems to link a clinical question to biomarker development.
- Create new models that integrate all of the resources and institutions needed to effect the needed changes in biomarker discovery, development, and validation.
Several important points were made during the discussion period:
Although many biomarkers are reported as "discovered" with little intent of clinical use, it was pointed out that the data should at least be reproducible. It was noted that the irreproducibility problem is prevalent across all of biomedicine, driven in large measure by only publishing positive results. Journals do not want to publish negative results, and there is little reward for team science/collaborative research in academic institutions. In this regard, the field needs to be critical of papers that proclaim associations between biomarkers and disease. It was pointed out that the journals could act as gatekeepers for standards, but in fact the reviewers for large-scale studies just don't exist—and the situation will get worse as we enter the new era of multi-dimensional data sets. One advance that would help to address all of these issues is to make data publicly available as soon as possible. Physics had to solve this problem—as have other areas of science—and biomedicine will have to solve it to reap the rewards of this "omics" revolution.
Payers are beginning to stop reimbursement for diagnostics for which there is no evidence to support their use. The field is in transition and requires well designed econometric analysis to set value-based pricing. The Center for Medicare and Medicaid Services (CMS) could help to solve the problem by employing a model that provides incentives for the development of a biomarker-driven evidence base for drug reimbursement.
Changing the value proposition for biomarkers (molecular diagnostics) will be required for the field to advance. Payers need to develop mechanisms to create a systematic approach to biomarker and drug development that involves large collaborations of investigators rather than individuals conducting small-scale studies. Payers could also begin to pay based on level of evidence: the greater the evidence the large percentage of reimbursement. The NBDA is advised to build the trans-sector networks needed to bring about the creation and implementation of the standards/processes needed to support end-to-end, evidence-based biomarker development.
A Facilitated Panel Discussion
Addressing the Biomarker Conundrum: Creating an End-to-End, Standards-
Based Critical Path for Biomarkers
Carolyn Compton, M.D., Ph.D., Professor, School of Life Sciences, Arizona State University
Dr. Compton focused on biospecimens. Biospecimens are the ultimate source of biomarkers, and therefore critical to the success of biomarker discovery, development, and clinical application. However, investigators often collect and use biospecimens in the absence of standards/best practices, which lead to artifacts that are wrongly interpreted as indicators of disease. The NBDA must set a high standard for biospecimen collection and processing and serve as an example for how others in the field should conduct their studies.
Setting standards for biospecimens is also critical because poor biospecimens can be responsible for incorrect diagnosis, which can prove catastrophic. It is the NBDA’s intention to look not only at the quality of the specimens used in its efforts but also at the quality of the biorepositories that will provide fit-for-purpose samples to researchers. The NBDA will collect samples based on biospecimen science-based protocols under the oversight of a quality assurance and quality control (QA/QC) process. The NBDA will also ensure that extensive information exists for each specimen, appropriately maintained in an appropriate information management system.
Michael Berens, Ph.D., Deputy Director, Research Resources, Translational Genomic Research Institute (TGEN)
Dr. Berens stated that the NBDA will create a new end-to-end biomarker discovery pipeline with NBDA-established protocols and standards. These will be made broadly available to all investigators with the intent that partners such as the FDA include such standards in their guidance. It was suggested that the NBDA’s initial efforts focus on the development of standardized questions for clinical trials. For example, the NBDA could form trans-sector teams to undertake projects such as testing survival biomarkers for use in stratifying patients to balance trial arms in equivalency trials; testing biomarkers to identify patients that have an intended target and are therefore suitable for inclusion in targeted agent clinical trials; and testing so-called futility biomarkers to provide an early indication of tumor response or non-response to therapy.
The NBDA could also adopt standardized practices for advancing pilot studies using clinically annotated biospecimens from no more than three sites to ensure minimal deviation from standard operating procedures. Further, the NBDA should use best-in-class analytical partners matched to the analyte and measure pharmacokinetic and pharmacodynamic correlative endpoints. In addition, the NBDA should use a parallel, blinded, limited competition model for maximal learning that will produce early wins. Finally, all assays should be tested across three laboratories for portability and reproducibility. Other recommendations included exploring what would be needed to improve and ensure reproducibility of laboratory-developed tests (LDTs) (this will be especially critical if and when the FDA elects to regulate the LDTs); harmonizing test validation between the FDA and CMS; and doing what is needed to support the FDA's requirements to conduct clinical trials on companion diagnostics. Everyone wins if the NBDA is successful!
Robert J. Penny, M.D., Ph.D., CEO, International Genomics Consortium, CEO, Paradigm Diagnostics
Dr. Penny stated that mapping the human genome and cancer genomes has created a path, albeit a complicated one, for developing new diagnostics that advance personalized medicine. However, poorly reproducible science combined with aggressive business development activity is creating a precarious situation for this field at a critical time. The NBDA’s model for biomarker development represents a tremendous opportunity to address this situation and create a new path based on sound science.
It is not surprising that the medical regulations of 20 to 30 years don’t work today to support a clear regulatory path for biomarker development. The NBDA can be a powerful force to change this situation by providing the communities with the needed standards and processes that can in turn support the FDA's efforts in guidance development. However, the NBDA must not create bureaucracy but rather strive to enable standards-based processes that speed approvals. Specific goals that the NBDA should undertake include developing standards for the level of evidence needed to identify clinical relevance, developing reagents standar, and enabling strategies to further improve assay specificity and sensitivity.
Raymond DuBois, M.D., Ph.D., Executive Director, Biodesign Institute, Arizona State University, Co-Leader, Prevention Program, Mayo Clinic
Dr. DuBois discussed colorectal cancer and the role of biomarkers. Approximately one million people develop colorectal cancer annually and 50 percent will die of systemic disease within five years of diagnosis. Although a number of genes have been discovered and implicated in colon cancer, only one, the KRAS gene, has translated successfully into clinical use. Currently, KRAS is used as a futility biomarker for therapy and to this point has saved hundreds of millions of dollars in unnecessary therapy.
It was noted that there is a long list of DNA-discovered, but not validated, biomarkers for colorectal cancer. There are a number of problems with using any colon cancer biomarkers since much of the DNA in stool samples come from microbial flora, not the patient. Interestingly, it took many years to gain sufficient data to determine that KRAS was a valid biomarker for predicting drug response. The NBDA could pursue a strategy to look for promising futility biomarkers such as KRAS in its trials.
The KRAS Gene. Although KRAS is a powerful biomarker, only about 22 percent of patients are tested, and patients are often still treated even in the presence of a positive test. The idea was posited that perhaps the NBDA could do the types of studies needed to set stringent cut points for making decisions as to when to use KRAS results to guide treatment.
Biospecimens and Biomarkers. As to why it's so difficult to get biomedical scientists to use standards, guidelines, and standard operating procedures (SOPs), especially for biospecimens, it was pointed out that it's really hard to acquire samples, period, and as a result any samples that can be accessed are often not good enough. One example offered was the National Cancer Institute’s pilot program to standardize prostate tissue collection across the 11 prostate cancer Specialized Programs of Research Excellence (SPOREs). Although the SPOREs were willing to fully comply, the institutions were not, as the standardization affected local work flow and the economics of their operations
The question was raised as to the possibility of creating a dual path for tissue that does not impact the routine clinical process, i.e., where one path follows the local work flow and the other is established for the research path. It was the opinion of the experts (i.e Dr. Compton) that it would be possible, but the research path must establish processional responsibility for the tissue from the moment it is removed from the patient. Moreover, the only real way to enable such a path would be to ensure some level of reimbursement (financial incentives for doing it). This is the norm for breast cancer, where the standards are high since the pathologist is in the operating room and takes charge of the sample immediately, ensuring high quality. It was also noted that the NBDA should also focus on procuring normal biospecimens, which are extremely difficult to obtain. Finally it was noted that establishing standards for biospecimen procurement is a policy issue that could be solved by the Federal government requiring adherence to standards as a condition of its grant awards. The NBDA's plans to create a "learning environment" that can inform evidence-based biospecimen science could help this situation immeasurably and provide sufficient evidence to enable the NIH to act. Clearly the development/assembly of the standards needed for the acquisition and stewardship of high quality fit-for-purpose biospecimens will be a major goal for the NBDA.
Anna D. Barker, Ph.D.
The NBDA was created to build a "coalition of the willing" to significantly improve the success of biomarker discovery development and validation for the benefit of patients. Although a monumental undertaking—a head-exploding problem—solving a problem of this magnitude begins with first steps. With the collaboration of the affected communities, the NBDA will take the steps needed to assemble and/or create the information required to set standards that can change the dismal success rate of biomarker research and development. Accomplishing these audacious goals will require harnessing the power of networks to assemble needed data and build knowledge. When it comes to biomarkers, biomedicine has not been held accountable, which has to change. The NBDA must work to shift the focus of the long path of biomarker development to one that is focused squarely on patient benefit. The NBDA's mission needs to become a national movement that holds the promise of ultimately delivering state-of-the-art healthcare for all patients!