3 Hormesis Once Marginalized, Evidence Now

Hormesis: Once Marginalized, Evidence Now Supports Hormesis as the Most Fundamental Dose Response Edward J. Calabrese Abstract The biomedical community made a fundamental error on the nature of the dose-response relationship early in the 20th century and has perpetuated this error to the present. The error was the byproduct of the conflict between homeopathy and traditional medicine. To deny support to homeopathy, leaders of the biomedical community rejected the hormetic biphasic dose-response model, the proposed explanatory principle of homeopathy. The threshold dose-response model was adopted as an alternative model, quickly becoming central to toxi- cology/pharmacology and their numerous applications. Despite its near-universal acceptance, no attempt was made to validate the ability of the threshold model to accurately predict responses in the below-threshold zone at the time of acceptance and throughout the 20th century. In contrast, the hormetic biphasic dose-response model became marginalized and was excluded from the mainstream of pharma- cological/toxicological teaching and practice, textbook development, professional society journal publications, annual meeting presentations, grant funding, and use in government risk assessment. Over the last decade there has been a resurgence of interest in hormesis due to findings indicating that hormetic responses are common, reproducible, and generalizable, as well as independent of biological model, endpoint, and chemical class/physical stressor. Large-scale studies have indicated that the threshold model fails to accurately predict responses below the threshold, whereas the hormetic dose-response model performs very well. These findings indicate that the biomedical community made an error on the nature of the dose-response relationship, compromising the accuracy of toxicological and risk assessment practices, including environmental exposure standards, and impeding drug discovery/development and drug safety studies. Keywords Hormesis · Hormetic · Biphasic · U-shaped · J-shaped · Dose-response relationship · Adaptive response · Preconditioning · History of science E.J. Calabrese (B) Department of Environmental Health Sciences, School of Public Health and Health Sciences, University of Massachusetts, Amherst, MA 01003, USA e-mail: edwardc@schoolph.umass.edu 15M.P. Mattson, E.J. Calabrese, Hormesis, DOI 10.1007/978-1-60761-495-1_2, C© Springer Science+Business Media, LLC 2010 16 E.J. Calabrese Introduction The dose-response relationship is the central concept within the fields of pharmacol- ogy and toxicology. It guides how studies are designed and biostatistical modeling is performed, the general focus of mechanistic research, drug efficacy and safety eval- uation, and governmental environmental risk assessment practices for protecting humans and other life against threats to food, water, air, and soil. The dose-response relationship is also a fundamental concept of biology, in that it is central to evolu- tionary theory and its underlying processes of mutation, DNA repair, and a plethora of integrative adaptive responses. Central to the biological and health sciences, the dose-response relationship is a scientific concept that seems as obvious as it is pro- found, being nearly universally understood based on common experience. Therein lies the trap into which the general public and the scientific community have fallen. Over the last century the scientific community accepted the threshold dose- response model as a description of how chemical and physical stressor agents affect the vast range of biological processes across essentially all forms of life and biolog- ical organization. This concept has become integrated into all biological disciplines and regulatory practices, quietly evolving into a fundamental concept. Reinforcing this “scientific” decision on the primacy of the threshold dose-response model is the general recognition of thresholds in the physical sciences, such as melting, boiling, and freezing points, and common experiences with medications and other products. The convergence of agreement on the dose-response model by the scientific com- munity and the general public is also as important as it is in reinforcing belief in the validity of this concept, hence its acceptance and status as a central pillar in various disciplines. Despite the history of science with its self-correcting features and the wisdom of the general public’s experiences and its integration of perceptions concerning the dose-response relationship, both science and the lay public have the relation- ship wrong. This error has profoundly affected the understanding of evolutionary biology, the nature of the body’s adaptive response, and the testing and assessment of drugs and chemicals, adversely affecting the health of individuals and popu- lations and even national and world economies due to misplaced priorities and extremely wasteful spending. The error originated in the fields of pharmacology and toxicology and, like a highly contagious disease, quickly infiltrated all biological disciplines, as well as government regulatory agencies, including their codified deci- sions with their non–self-correcting features. This error in judgment on the nature of the dose-response relationship became accepted in the early decades of the 20th century and has been perpetuated to the present time (Calabrese, 2005b, Calabrese, 2005c; Calabrese 2007; Calabrese and Baldwin, 2003a), reinforced by a dominant governmental regulatory and funding culture that strongly influences what scientific ideas will be studied. This chapter assesses the history of the dose-response relationship and the basis of the error by the scientific community concerning it. The chapter proposes Evidence Now Supports Hormesis 17 the most basic and appropriate dose-response relationship for the biological sci- ences along with supportive documentation and a perspective on its broad societal implications. Historical Antipathies, Rather Than Science, Determined Which Dose–Response Model Would Dominate Biology The error that determined what has been long considered the fundamental nature of the dose-response relationship was rooted in a scientific version of the Hundred Years’ War, that is, the prolonged and bitter conflict between homeopathy and what eventually came to be called “traditional” medicine. To citizens of the late 20th and early 21st centuries, this “medical” conflict would seem to be a minor event, given the overwhelmingly powerful victory of traditional medicine, and therefore not likely to be more than a historical footnote. However, this will be shown not to be the case. As a result of this medical science–based conflict, the basic dose- response relationship—that is, the biphasic dose-response model—got caught in the cross-fire and was victimized because it was a central and highly visible feature of homeopathy. The linking of the biphasic dose-response relationship to homeopathy was facil- itated principally by Hugo Schulz (1853–1932), a professor of pharmacology at the University of Greifswald in northern Germany. Schulz believed that the biphasic dose responses (i.e., low-dose stimulation and high-dose inhibition) he observed in laboratory studies (Schulz, 1888) assessing the effects of chemical disinfectants on yeast metabolism could be broadly generalized and serve as the explanatory principle of homeopathy. Schulz [1923, with English translation by Crump (see Crump, 2003)] emphasized the reproducible nature of his findings in an auto- biographic account of the discovery, a perspective that was strongly supported by detailed studies (Branham, 1929) specifically designed to reaffirm and gen- eralize his findings to a wider range of potential antiseptic chemicals. Chester M. Southam and John Erhlich (Southam and Ehrlich, 1943), forestry researchers at the University of Idaho who observed that low doses of extracts from the Red Cedar tree affected the metabolism of multiple fungal strains in a similar biphasic man- ner, renamed this dose response concept “hormesis” after the Greek word meaning “to excite.” Prior to his intellectually transforming studies with yeasts, Schulz was educated and trained along a traditional biomedical path, with strengths in chemistry and pharmacology. He was also mentored by Eduard Pfluger, one of the founders of modern physiology. However, Schulz was quietly open to homeopathic principles and practices due in large part to an admired and respected family homeopathic physician friend with whom he had a long and intellectually engaged associa- tion (Bohme, 1986). At about the time (1882) that Schulz started his career at 18 E.J. Calabrese Greifswald, research emerged indicating that veratrine, a homeopathic medicine, was a successful treatment for gastroenteritis. Because the causative bacteria had recently been identified and cultured, Schulz (Schulz, 1885) seized the opportu- nity to assess whether this drug acted via the killing of the bacteria. Extensive tests using a broad range of concentrations revealed that the drug was unable to do so. Although this observation failed to shake Schulz’s belief in the efficacy of the drug, it did compel him to conclude that the drug must act via a mechanism other than cell killing. Several years later when Schulz (Schulz, 1888) observed the biphasic concen- tration effects of a broad range of chemical disinfectants on yeast metabolism, he came to believe that he had determined how the veratrine might have been effec- tive in the treatment of patients with gastroenteritis. That is, Schulz claimed that at low doses the drug could induce adaptive processes that permitted the person to resist the infection and facilitate recovery. He soon extended this hypothesis to the broader homeopathic field, believing that he had discovered the underlying explanatory principle of homeopathy. Schulz quickly became a leader within the homeopathic community, devot- ing the remainder of his professional life to its further study and intellectual expansion. Because Schulz was well known in the pharmacological and medical communities, with numerous publications, as well as active participation on edito- rial boards of leading professional journals (e.g., Naunyn-Schmiedeberg’s Archives of Pharmacology) (Starke, 1998), the homeopathic community looked to him to challenge traditional medicine in hopes of legitimizing their medical practices. This also meant that Schulz, his findings, and his interpretations became central in the conflict and the object of considerable criticism by those opposing homeo- pathic perspectives. The intellectual opposition, that is, traditional medicine, in the form of pharmacology and eventually its scientific offspring toxicology, could not accept Schulz’s scientific findings because this would appear as an endorsement of homeopathy. A careful analysis of Schulz’s experimentation (Schulz, 1888) would have revealed that it was not directly relevant to homeopathic medical treatment the- ory and practice. The vast majority of medical treatments are performed to reduce existing symptoms of illness and prevent their recurrence. This occurs when the individual becomes ill and seeks medical assistance. The homeopathic treatment would normally be expected to be administered after the onset of the illness. In Schulz’s work and the overwhelming number of examples of hormesis in the pub- lished literature, the investigations did not involve exposures after the onset of disease or chemically induced injury. Even though Schulz believed that his findings were at the core of homeopathic understanding, the scientific community made a critical error in not challenging his interpretation. However, instead it challenged the reliability of Schulz’s findings and his dose-response generalization, a decision that would prove to have far-reaching implications for pharmacology and toxicology. Given this strategic, although incorrect decision on how to challenge Schulz, two courses of action emerged: (1) the Schulz biphasic dose-response model (called the Evidence Now Supports Hormesis 19 Arndt–Schulz law at the time) had to be marginalized, and (2) a credible alterna- tive had to be formulated, and this becoming the threshold dose-response model, the model on which 20th century clinical pharmacology, toxicology, and risk assessment would be based. The most notable critic of Schulz was Alfred J. Clark (1885–1941), a highly accomplished pharmacology researcher and scholar, who had considerable influence among academics and government regulators (Verney and Barcroft, 1941; Gaddum, 1962). Nearly 70 years after his death, Clark remains a highly respected figure in pharmacology, with graduate fellowships and a distinguished chair in pharmacol- ogy at Edinburgh named in his honor. Clark (Clark, 1933, 1937) was the author of several highly influential, multiedition textbooks that criticized Schulz and his dose-response theories in highly dismissive ways (Calabrese, 2005a) while also linking him with the “extremist” elements within homeopathy (Clark, 1927). In fact, Clark’s Handbook of Pharmacology was highly regarded, being published as late as 1970, nearly three decades after his death, and influenced several generations of pharmacologists and toxicologists. Clark’s professional successes were due in considerable measure to his careful and objective evaluation of data and his capacity to obtain and integrate massive amounts of complex and technical information in scientifically valid and insight- ful ways. In the case of his analysis of Schulz, such thoroughness and objectivity were surprisingly below his normally high standards, with a retrospective eval- uation (Calabrese, 2005a) revealing that Clark was very selective in his use of the published literature to support his position while failing to report substantial independent findings that supported Schulz’s work with yeast and disinfectants (Branham, 1929), as well as his general biphasic dose-response concept (Calabrese and Baldwin, 2000a, Calabrese and Baldwin, 2000b, Calabrese and Baldwin, 2000c, Calabrese and Baldwin, 2000d, Calabrese and Baldwin, 2000e). Of particular note is that Schulz was not in a position to defend himself, given that Clark’s criti- cisms intensified after Schulz entered retirement in the early 1920s, and Schulz died (1932) before the first editions of Clark’s two critical books (Clark, 1933, 1937). Furthermore, when the prominent surgical and biomedical researcher August Bier came to his defense, political forces were quickly mobilized to strongly criticize the once-esteemed Bier (Goerig et al., 2000), who had been nominated for the Nobel Prize in Biology and Medicine on multiple occasions, sending a not-so-subtle mes- sage to other scientists, even those of considerable achievement and reputation, who might similarly wander from the “party line.” Clark’s criticism of homeopathy and Schulz occurred at a time when homeo- pathic medicine was severely criticized by the so-called Flexner report (Flexner, 1910), which, together with the ongoing efforts by its author, with the backing of the Rockefeller Foundation, over the next two decades effectively led to the closing of the vast majority of homeopathic medical schools in the United States (Berliner, 1985). The final intellectual component of the tipping point regarding the dose- response concept occurred when colleagues of Clark’s (Gaddum, 1933; Bliss, 1935) (note that Clark’s assistance was acknowledged in the Bliss paper) independently 20 E.J. Calabrese derived the probit dose-response model to account for responses above the so-called toxicology/pharmacology threshold. A critical statistical refinement offered by the esteemed biostatistician R. A. Fischer in an appendix of the Bliss (Bliss, 1935) paper utilized the maximum likelihood estimate to constrain responses to asymptotically approach control-group values in the low-dose zone. In effect, any response below the control group was to be judged as variation, thereby denying the possible bio- logical reality of the J–shaped or inverted–U-shaped dose-response curve. Multiple forces therefore converged and were to control how toxicology and pharmacology considered the dose-response relationship for the next 80 years. Acceptance of the threshold dose-response model would drive the development of these fields, includ- ing the selection of animal models, study designs, and risk assessment practices and government regulation. The exclusion of the hormetic-like dose-response relationship from the main- stream of pharmacology and toxicology during the 20th century was strikingly suc- cessful even though there were a substantial number of high-quality research papers supporting the hormetic perspective (Calabrese and Baldwin, 2000a, Calabrese and Baldwin, 2000b). Despite such supportive scientific studies, the hormetic dose- response relationship became marginalized as traditional medicine established its control and directions on the field, transforming the discipline of toxicology in the process. By essentially denying the existence of the biphasic dose-response rela- tionship (Calabrese and Baldwin, 2000c, Calabrese and Baldwin, 2000d, Calabrese and Baldwin, 2000e), 20th century scientific leaders molded toxicology into a high- dose, few-doses discipline. A consideration of the historical development of the reliance of the U.S. National Cancer Institute’s (NCI) cancer bioassays on only two doses—the maximum tolerated dose (MTD) and MTD/2—to define the toxic- ity spectrum illustrates the impact of Clark’s dictum on 20th century chronic toxicity and carcinogen evaluations. By truncating the focus of toxicology to above- threshold responses, what did the field miss? Before that question can be answered, it is necessary to establish that biphasic dose responses exist and are reproducible and to discuss their mechanistic foundations and frequency. The Hormetic Dose-Response Relationship An important factor in the evaluation of hormetic-like biphasic dose responses is that numerous investigators have reported observations supportive of this dose- response relationship in highly diverse biomedical fields with a notably increased frequency from the mid 1970s and early 1980s to the present. These observations have been associated with various types of technological improvements, includ- ing the markedly enhanced capacity to measure lower and lower concentrations of chemicals in various media, thereby permitting toxicological and pharmacological evaluations over a far greater dose range than previously envisioned. It also has been related to major developments in the area of cell culture, including the use of 96-well and higher plates, which permit the assessment of large numbers of chemi- cals over a broader range of concentrations in a highly cost-effective manner. These Evidence Now Supports Hormesis 21 advances have been particularly evident in the evaluation of chemically induced immune responses (Calabrese, 2005b), as well as in the assessment of the responses of human tumor cell lines to a wide range of endogenous and exogenous agents (Calabrese, 2005c). Even though hormetic-like biphasic dose responses have been widely and increasingly reported, it has been common for various biological subdisciplines to use unique descriptors/terms for biphasic dose-response relationships, often spe- cific to each discipline (Table 1). Table 2 provides a historical time line of the citations of some of the main terms used to describe hormetic-like biphasic dose responses based on the Web of Science database. This table documents that the biphasic dose-response concept has shown an increase in citation frequency over the last several decades, that is, long after the dose-response concept had been firmly established and administratively “fixed” within the fields of