OR WAIT null SECS
The Human Microbiome Project has increased our understanding of the relationship between humans and microorganisms. The authors offer a new perspective on how this knowledge should be considered in setting standards for pharmaceutical quality control in microbiology.
Submitted: Aug. 4, 2014. Accepted: Aug. 29, 2014.
One of the more exciting contemporary research activities in human biology has been the Human Microbiome Project (HMP) (1). The insights derived from the HMP have revealed details about the relationship between humans and the microorganisms we live with, and could not live without. In 2013, the first publications appeared that attempted to put the HMP reports into a pharmaceutical microbiology context (2). The purpose of this communication is to consider the HMP in the context of other recent insights into microbiological control and to explore how this knowledge could (or should) change the way standards are set for healthcare products. Microbiological analysis will also be explored with due consideration of what has been learned since the standards have evolved.
Fundamentally, the HMP provides far more expansive and accurate data regarding a topic that has long been of great interest, which is the nature, size (i.e., population), and scope of what’s been typically referred to as normal human flora (3). Normal human flora is a generic and loosely defined phrase used to describe the myriad organisms with which all healthy (and not so healthy) humans are colonized. In many industrial microbiological investigations, one may read that an organism recovered is a constituent of normal human flora (4). When an organism is categorized as a constituent of the normal human flora, that is generally understood as an indication that it is relatively harmless. So, normal human flora can often take on a distinct meaning for current good manufacturing practice (cGMP).
However, data in the published HMP reports have confirmed something many academic microbiologists have long suspected, that many humans also have in their normal flora known “opportunistic pathogens” as well. These are organisms that inhabit a quality control microbiology gray area in that, as their name implies, they are able in very rare circumstances to cause human disease. Before considering the gray area and its implication in detail, an examination of some additional history of the HMP is required.
The HMP was initiated by the United States National Institute of Health in 2007 and grew into a research consortium comprising some 80 global research sites (5). The initial study findings were released in June of 2012 in a set of articles published in Nature (2–4). These studies were conducted on 242 healthy volunteers (129 male and 113 female). A total of 15 sample sites were selected for the male subjects and 18 for females. Continuing research is underway on study cohorts with different health conditions, including children with frequent fevers, individuals with upper respiratory tract infections, and pregnant women. Additional HMP work will, therefore, be published in a steady stream during the next few years and probably beyond.
The coining of the term microbiome is generally attributed to Nobel Prize-winning microbiologist/molecular geneticist Joshua Lederberg. Lederberg hypothesized that microorganisms living in the various environmental niches available on humans and animals played a more significant role in health and disease than had been generally recognized. The human microbiome can be defined as “the ecological community of commensal, symbiotic, and pathogenic organisms that share our (the human) body space” (6).
It is useful to examine the contents of this definition a bit further, because it contains words and concepts that may not be familiar. The definition first mentions a community of microorganisms living within the human “ecology.” This implies interaction among the organisms that may be present. In other words, the microbiome idea as Lederberg saw it was one in which organisms cooperated and interacted with both each other and human cells. Next, the definition mentions organisms that could be commensal, symbiotic, or pathogenic. A commensal relationship is where one organism benefits from suitable living conditions on another living species without adversely affecting that organism. A symbiotic relationship is one in which both organisms, for example a human and a bacterial species, benefit from the interaction. A pathogenic relationship is one where the circumstance by which an organism lives on another causes a disease in the other organism. The term pathogen is most critically and significantly applied to a microorganism known to have caused a human infection. A true or frank pathogen is an organism which, when recovered from man, is almost always associated with an infection. True pathogens may be viruses, bacteria, parasites, or fungal species.
Pharmaceutical quality control microbiology, the goal of which is product safety, is quite properly fixated on the prevention of infection. Naturally, this leads to the focus on the role of microorganisms as pathogens. The principal reason for concern about microorganisms as contaminants is that failure to limit their transfer from a medication to patients could result in human disease. The fear of microorganisms causing disease as a result of using a medication takes us from the realm of industrial microbiology into another specialty of the multi-faceted discipline of microbiology, namely the study of infectious disease. The study of infectious disease is a different activity from industrial microbiology, although the two overlap when the discussion turns to infection risk.
The study of infectious disease is not new. The formal study of the infectious disease began with reports of research done by Robert Koch, a German physician of the late 19th and early 20th century. Koch, who won a Nobel Prize for medicine was effectively the founder of medical microbiology. Koch formulated a set of simple, but rigorous postulates that medical microbiologists and physicians use to positively attribute disease causation to a single species of microorganism. Four criteria that were established by Koch to identify the causative agent of a particular disease are as follows (7):
Scientists have had more than a century to investigate the causes of human infectious disease, and enormous progress has been made in developing both preventive medicines such as immunizations and treatments such as antibiotics for various infectious diseases. It is also possible to prevent or control deadly disease without knowledge of the causative organism. Two notable examples of this are smallpox and yellow fever. Edward Jenner created a vaccination for smallpox without knowing that the agent was a microorganism, or more specifically, a virus. He was able to do this in 1798, well over 100 years before the word virus entered the scientific lexicon and prior to Paul Ehrlich’s work, which led to the first understanding of a field that came to be known as immunology (8). The smallpox virus will not be found as a constituent of the HMP. In fact, no true (“frank”) pathogens were found in/on healthy humans, and conventionally, are found only in/on people suffering from a disease caused by that particular pathogen.
Consider for a moment that the HMP has found that humans have associated with them upwards of 10,000 different species of bacteria. The exact number of different species of bacteria on earth is unknown, but estimates range from hundreds of thousands to a billion. The pharmaceutical world often discusses “bugs” as microorganisms; bacteria and mold are viewed in the industry as though any recovery of these bugs above some numerical trigger point portends grave risk to the product’s end user. The vast majority of environmental isolates or “bugs,” enumerated from product or excipient tests, are completely harmless commensals or organisms from the environment external to the manufacturing plant. The overwhelming majority of organisms on our planet, including those that live in or on humans, are completely harmless. It may then be surprising to learn that the World Health Organization reports that more than 90% of all human infectious disease is caused by only six infective sources. Of these six diseases, four are caused by a single microorganism (9). The six types of infections that kill the most individuals worldwide in descending order of prevalence are:
The industrial microbiologist will notice immediately that the “bugs” responsible for these six primary infectious diseases are extremely unlikely to be found in healthcare products. Additionally, none of the viruses, or the malarial parasite, could ever be recovered by the prevalent microbiological analysis methods in use. Some of the viruses listed above are actually used to make healthcare products-live attenuated measles and influenza vaccines to cite to notable examples. Safety in the manufacture of those products obviously hinges on the ability to achieve effective attenuation. However, the majority of healthcare products, including all small-molecule pharmaceuticals and devices, could not harbor these organisms and therefore need not be tested for them.
Although studies on the viral component of the microbiome are underway, the viruses responsible for acute diseases are also unlikely to be found among the organisms recovered from normal healthy subjects. Some of the viruses that cause human illness are likely to be found, including herpes simplex I and varicella zoster. Per cGMP rules worldwide, people that are actively ill and manifest symptoms of infectious disease, such as fever, coughing, and skin lesions, should not be working in healthcare product production environments. However, it would be impossible to remove staff who carry herpes simplex 1 or varicella zoster. Attempting to do so would diminish the pool of qualified workers by more than 99%, because nearly every healthy human carries these viruses. Fortunately, it is possible in the modern world to immunize healthy humans against many of the common viral or bacterial diseases that continue to cause massive human misery.
The HMP data confirm that true or frank pathogens are not commonly present in normal human microflora. These data should reinforce the reality that real risk comes not from organisms that are part of our natural flora residing in the various environmental niches our bodies provide, but rather those that are associated with humans in significant numbers only when they are ill.
An obvious conclusion is that given their comparative rarity, those few species of microorganisms responsible for nearly all human infectious disease must have special characteristics; or those who are affected by these organisms must be particularly susceptible to them. Actually, both the condition of virulence (the ability to cause disease) and susceptibility (on the part of the individual) must be present for the result to be a serious clinical infection. Most human infections are either minor or completely asymptomatic, which is to say we don’t know we have them.
Some readers will be familiar with the staggering bacterial numbers reported by the various HMP studies that have appeared. With the identification of more than 10,000 bacterial species in the human microbiome so far, it is unsurprising that most of the known bacterial genera have been observed in association with healthy humans. The sheer number of total bacteria found in or on humans is an amazing ~1014, which is approximately 10-fold more than the number of human cells each of us contain. Obviously, the 1–2 kg of bacteria that constitute our normal flora are responsible for neither harm nor undue health risk within their particular niches or none of us would make it to adulthood. This should inform the reader that the sheer number of microorganisms in a product is unlikely to meaningfully change the population of bacteria present at the site of administration.
It is long known that certain bacteria help protect us from disease, and assist us in digesting food, as well as other positive contributions they provide. A new peer-reviewed scientific journal has appeared entitled Beneficial Microbes (10). This should not surprise anyone because humans routinely consume milk products supplemented with bacteria, yoghurt containing active cultures, or probiotics containing millions of live bacteria. The concept of beneficial microorganisms is clearly at odds with “the only good bug is a dead bug” mentality that so often prevails in the pharmaceutical industry. This concept also confirms that the consumption of hundreds of thousands or even millions of colony forming units (CFU) of bacteria in food or within a probiotic capsule can be safe given that humans have done this for thousands of years.
The HMP has irrefutably confirmed how ubiquitous microorganisms are on and in humans. More importantly, it must be acknowledged that the overwhelming majority of microorganisms living in and on humans and present in the environment are either helpful or harmless. We need not worry about microbial risk arising from vast number of organisms associated with a healthy human. Nor should they cause any additional concern when they are released into work environments as they inevitably must be. The HMP should not give any thoughtful microbiologist or standard setter cause to embark on a campaign for tighter standards, more monitoring, or more intensive product testing. Nor should it cause an increased drive to expand the lists of so-called “objectionable” organisms. Obviously, there are no more microorganisms associated with workers now than before 2012 when the first HMP data appeared. There are no more microorganisms associated with people in 2014 than there were in 1714 or 1914. The HMP hasn’t uncovered any increased or previously unknown risk potential-the only thing that has changed is an expanded knowledge of human biology. It may, in fact, lead to an understanding that the risk is not as great as some recent initiatives on microbial control for pharmaceutical products would indicate.
Instead of an increased and completely unnecessary fear of previously unknown microbial risk factors that might be present, the HMP results should cause us to marvel at the wondrous complexity of cross species interaction and to appreciate how much more interesting human biology is than previously imagined. We should instead feel comforted by the fact that the only organisms conspicuously absent in healthy human subjects during the HMP studies were frank pathogens. Although that is not new information, it is important to recognize that the average healthy human is colonized at all times by organisms that under certain circumstances may cause disease in some people. For example, approximately 30% of the healthy subjects in the HMP had Staphylococcus aureus present in their anterior nares. This does not mean that S. aureus should be tolerated in medicines. It does, however, teach us that even those who don’t carry this organism are exposed to it on a daily basis, which should help keep the risk in perspective and consider it on the basis of product type and route of administration.
S. aureus is present on the skin in some humans and behaves as a commensal organism unless it reaches an environmental niche where it can grow unchecked. This means that if it is present, but goes undetected, dangerous conditions are unlikely, because in their daily lives, humans must encounter this organism on a frequent basis. It also means that because three out of 10 humans have this “bug” in their noses, it must routinely be released into the workspace. Certainly the ubiquity of S. aureus in man confirms that the use of reasonable infection control procedures, such as routine hand washing and the use of protective masks in production areas, are reasonable precautions.
Perhaps ironically, the greatest risk from pathogenic strains of S. aureus has arisen from efforts to treat infections. This risk did not arise as a result of S. aureus in product, but rather from a misuse and over-use of antibiotics. Humans have forced the evolution of dangerous and hard to treat strains such as methicillin-resistant S. aureus (MRSA) by prescribing antibiotics to patients suffering from common viral upper respiratory diseases against which antibacterial agents have no value (11). The emergence of MRSA and other manifestations of antibiotic resistance is an example of the law of unintended consequences at work in medical microbiological risk abatement gone wrong. As it turns out, the misuse of prescription drugs has created much of the modern day risk regarding healthcare system acquired infections. Contrastingly, the actual microbial contamination of drug products has proven to be an insignificant contribution to human morbidity or mortality.
The HMP’s findings regarding the content and staggering diversity of the human microflora increased the knowledge, but did not identify any new microbial infection risks. Personnel donning gowns to enter clean rooms or working in non-sterile product manufacturing don’t have more microorganisms on or in them than they did previously. In fact, it was already known that the numbers of bacteria carried by humans were enormous and that they sloughed large numbers of recoverable viable bacteria into the environment even when aseptically gowned (12). The contamination control challenge hasn’t changed, patient risk hasn’t changed, and a push for more regulation accompanied by new (and presumably more restrictive) microbial control requirements should not begin as a result of the HMP. What has changed is the understanding of human biology and this should manifest itself as a keener appreciation of the absolute futility of “the only good bug is a dead bug” mentality. The HMP should not be a cause of new cGMP requirements through the regulatory inspection process or as a result of compliance fervor.
What the HMP and other recent studies are confirming is that humans all have on, around, and in them, far more microorganisms than ever thought possible a decade ago. More importantly, it also demonstrates that despite the incredibly high numbers of species and overall population, nearly all of these organisms do no harm and may even play significant roles in keeping us safe from other bacteria that are not part of the natural ecology. Perhaps the fixation with the numbers of organisms in environments inhabited or presented to humans needs to be reconsidered. Worries about absolute numbers of human commensals and symbionts in the environment are not rational where they do not correlate with an increased concentration of true pathogens in products or a threat to the patient’s health.
Bacterial enumeration and patient safety
The healthcare industry has developed a number of standards that purport to set microbial “limits” for non-sterile products. The industry has also evolved target values for various production environments. The HMP findings suggest a reevaluation of how these limits were set in the first place. The more complete picture given by the HMP was made possible by the application of microbial analytical technology based on molecular biology. Previous attempts to evaluate and quantify “normal human flora” depended on the growth of bacteria on media and the enumeration of these organisms on solid media plates as introduced in Koch’s laboratories more than a century ago. In other words, all previous studies on human microflora came down to counting colonies (CFU). The reliance on the CFU as the standard of cell-count estimations must change as we move to the modern analytical tools that made the HMP possible. The modern methods used in the HMP are called somewhat imprecisely “rapid” microbiological methods (RMM).
Some readers may be puzzled as to why the switch to molecular biological methods would result in different numbers than those that were obtained using growth-based methods and reported in CFU. There is a prevalent belief structure in the compliance world that growth-based methods have good, even a nearly perfect, limit of detection. The existence of an analysis, boldly named the sterility test, implies that growth-based methods can and should be able to detect down to one cell. Recently, in personal communications, the authors have heard both regulators and industry representatives state that any new sterility test should have a limit of detection of one CFU. This is a clear example of a widely held compliance belief failing to come anywhere close to the scientific reality.
The problem with growth-based methods is quite simple; they can only “recover” organisms that will grow on the media selected under the incubation conditions offered. There are many presentations given in the industry in which there is an underlying expectation that Trypticase Soya Broth or Agar (TSB or TSA) will grow essentially all microorganisms if the right incubation temperatures and duration were selected. This statement has actually been known to be untrue for decades.
The “great plate count anomaly” was first reported in environmental bacteriology, but is now known to be generally applicable (13). Microbiologists noticed that when they viewed a sample preparation under a microscope and counted the bacteria present using a cytometer, there were often 100–1000 fold more cells present than grew on the agar plates. This observation led to a great deal of research regarding the formulation of media that would better recover a larger number of the cells present. Some of these efforts bore fruit and better recoveries were noted; a simple and limited example is the use of R2A agar rather than plate count agar for water analysis.
Changing to a different media formulation, however, will not eliminate the plate count paradox, it will only change the type of organisms recovered. One might recover organisms with a media change missed by TSA, but at the same time no longer recover some species that grew on TSA. One might increase recovery by 10-fold and still only recover a low percentage of the cells actually present. A laboratory could employ 10-15 different media each with different nutritional profiles and incubate them at different oxygen tensions or temperature ranges and still only manage to recover a limited amount of what actually may be present in a sample. This of course would be highly work intensive, and prohibitively costly in process, validation, or final product analysis. Such an effort could be logically deemed impractical given the limited expected improvement in results.
The different environmental niches on and in the human species cater to organisms with widely different nutritional requirements, and these niches also include certain symbiotic relationships between organisms that prove difficult to reproduce in a laboratory. The identification of molecular survival factors for organisms living in a given niche is an area of active research. Molecular genetics and biochemical analysis together can provide the means to better describe what niches exist within the human organism that favor the colonization by some species, but not others, in ways that would be impossible were they to rely on growth methods. Niche suitability can depend on availability of suitable nutrition, presence or absence of oxygen, perhaps pH, or the presence of other organisms that may produce needed nutrients or modify the niche to make it suitable for a symbiotic species.
Microbiology is a far more complex science than is allowed by the generalizations made in pharmaceutical microbiological standard setting and quality control analysis. Such a wondrously complex field of work does not easily yield to convenient assumptions.
Implications of the HMP on the patient
The implications of the HMP on the patient can be summarized as follows:
Implications of the HMP for microbial control of oral and topical dosage forms
If the typical oral dosage of a tablet or capsule would be accompanied by perhaps 120 mL of water (four fluid ounces) using the current water quality limits for drinking water (in CFU), this would equate to no more than 300 CFU/mL or 36,000 CFU. Assuming that most water falls well below the limit and is generally not more than 150 CFU/mL, this means that along with a tablet or two, the patient would receive ~18,000 CFU of bacteria in the water they used to take the medicine. With that in mind, limits in the range of 103 CFU for oral solid dosage products are extremely conservative.
Recent studies on potable water using molecular biological methods for microbial cell enumeration, however, indicate that 120 mL of drinking water is more likely to contain 100–1000 times more cells than those recovered using traditional plate counts as reported in CFU. Logically, then, testing non-sterile pharmaceutical products with more modern microbiological methods results would likely result in higher observed cell counts; however, there would be no added patient risk in a product historically known to be safe.
Similarly, any suggestion that we need to be more concerned with sloughing of bacteria from personnel more now than before the HMP is clearly wrong, if anything, the obverse is true as the patient has a comparable number of microorganisms present on/in them as well. Newer molecular biological technology has improved the ability to identify and roughly enumerate a far broader range of cells.
The sheer numbers of bacteria counted on man or in the water we drink should not be sources of increased concern among regulators or microbiologists. They should not result in calls for new standards or new compliance requirements. They do not suggest wholesale changes in what we do, but they do suggest that we should be less fixated on the total numbers of bacteria around or on us. Therefore, we should not fear that which we have been ignorant because nothing has changed with respect to our exposure. As we learn more about microbes and their relationship with man, we must recalibrate our own thinking and allow scientists, including healthcare professionals, suitable discretion to make reasonable risk judgments. The following points should be considered:
Implications of the HMP for microbial control of sterile products
The implications of the HMP for microbial control of sterile products can be summarized as follows:
Further considerations on product safety and human biology
There is a valuable lesson to be learned from the events associated with the contamination of aseptically manufactured steroidal injections made by an unapproved pharmaceutical manufacturer that has so far resulted in more than 50 deaths and approximately 730 injuries. In this outbreak, steroidal injections were contaminated with an environmental mold that would under normal circumstances be non-pathogenic. This outbreak, however, was in some respect, a perfect storm, because the injections of these products were made into the spinal cord and cerebrospinal fluid that have no normal flora. In addition, the immune response within the central nervous system is limited in its capacity to deal with microorganisms that would not normally be found there.
Humans come in contact each day with millions of mold spores of the type that caused the infections within the spinal cord. Humans are evolutionarily well adapted to dealing with such spores on their skin or drawn into their upper respiratory tracts. However, the manner of use of these contaminating steroidal injections and the particular characteristics of the injection target resulted in substantial risk. The combination of patient type and injection location with product made under substandard conditions in a facility colonized by mold resulted in a tragedy of substantial proportion. Further contributing to this disaster is the fact that treating a mold infection of this kind in the spinal cord is extremely challenging.
Thus, we are reminded again that introducing organisms into an area of the body with no natural flora and with limited immune response has substantial inherent risk. A product that should have been made with great attention to quality standards given the inherent dangers was instead produced under woefully inadequate contamination control conditions. The result was a human tragedy of a magnitude rarely associated with the use of commercially manufactured medicines.
The HMP research that has been reported should not serve as an excuse for new regulation, new performance standards, and certainly not new compliance initiatives. If we accept what the HMP is informing us about the reality of human/microbial interaction, we may instead see a redirection of our microbial control efforts to those areas where patient safety can be improved, rather than on those where the potential for harm is slight, if present at all.
It is also a time to reflect on human biology and to appreciate how humans and thousands of microbial species have evolved together over millions of years. We must consider how our health is dependent upon the establishment and maintenance of a healthy human microbiome, and that this microbiome plays a vital role in keeping us healthy and disease free. If there was ever work that should reduce the fear the typical layman has of microorganisms, it should be the HMP.
Finally, we must be cognizant that what makes the results of the HMP strikingly different from previous assessments of human microflora is the analytical approach taken in the study. The HMP was the first comprehensive effort to apply modern molecular biological methods to the assessment of human/microbiological interaction. The result was the discovery that the human flora, in terms of sheer numbers of bacterial species and total organism number, was more diverse than previously imagined. Over the past few years, reports have emerged that the world’s potable water supplies contained hundreds of thousands more organisms than previously thought to be present, representing a signal of microbial viability than had not previously been recognized. The common denominator in these studies is the use of molecular biological methods rather than traditional growth-based methods reporting results in CFU. This is no different scientifically from astronomers and astrophysicists finding more celestial objects using the Hubble telescope than they were able to perceive using earth-based optics. Science waits for no man and abides by no human-invented regulation.
The finding that the human organism is colonized by a larger number of organisms does not equate to greater risk, it only equates to greater knowledge, which must be applied wisely. As molecular microbiological methods are more widely used, we are likely to have other surprises, but this does not require us to fear new methods or to shy away from their implementation, it merely requires us to understand that if our processes haven’t changed, then patient risk must not have changed either. We haven’t uncovered new objective dangers, because the nature of humans and their environment hasn’t changed, only our knowledge of it has changed.
1. Human Microbiome Project, The NIH Common Fund, March 8, 2012.
2. A. Barbara et al., Nature 486 (7402) 215-221 (2012).
3. H. Curtis et al. Nature 486 (7402): 207-214 (2012).
4. T. Sandle, K. Skinner, and E. Yeandle, Eur. J. Parent. & Pharmaceut. Sci. 18 (3) 84-90 (2013).
5. P. Turnbaugh et al., Nature 449 (7164) 804-810 (2007)
6. L. Hooper and J. Gordon, Science 292 (5519) 1115-1118 (2001).
7. R. Koch, Cohns Beitrage zur Biologie der Pflanzen (in German) 2 (2) 277-310 (1876).
8. S. Kaufmann, Nature Immunology 9, 705-712, 2008.
9. The Global Epidemiology of Infectious Diseases, Global Burden of Disease and Injury Series, Volume IV (World Health Organization, Geneva, Switzerland, 2004).
10. Beneficial Microbes, Wangeningen Academic Publishers, published quarterly ISSN 1876-2883 Paper Edition, ISSN 1876-2891 Online Edition.
11. P.C. Appelbaum, Clinical Microbiology and Infection 12 (s1) 16-23 (2006).
12. B. Ljungqvist and B. Reinmueller, “Dispersion of Airborne Contaminatns and Contamination risks in Clean rooms, Chapter 7,” in Guide to Microbiological Control in Pharmaceuticals and Medical Devices, S. P. Denyer and R. M. Baird, Eds, (CRC Press, 2006).
13. A. D’Onofrio et al., Chemistry & Biology 17 (3) 254-64 (2010).
14. USP <1115> Bioburden Control of Non-Sterile Drug Substances and Products (December 1, 2014).
Article DetailsPharmaceutical Technology
Vol. 39, No. 4
Citation: When referring to this article, please cite it as J. Akers and J. Agalloco, “The Human Microbiome Project and Pharmaceutical Quality Control Microbiology,” Pharmaceutical Technology39 (4) 2015.