Using evolutionary principles to make clinical decisions: a case series of urinary tract infections

Abstract

The principles of evolutionary medicine have significant potential to be useful in a wide variety of clinical situations. Despite this, few demonstrations of clinical applications exist. To address this paucity, a case series applying evolutionary medicine principles to urinary tract infections, a common medical condition is presented. This series demonstrates how applying evolutionary medicine principles can be used to augment clinical decision-making.

Introduction

Since Nesse and Williams published The Dawn of Darwinian Medicine in 1991 [1], the field of evolutionary medicine has seen significant advances in anthropology, oncology, microbiology and other disciplines making diverse contributions to the broader project of public health. Amid this history of progress, direct clinical applications have been slower to appear. In an attempt to rectify this gap, the following case series will demonstrate how the application of evolutionary medicine principles and theory can augment clinical decision-making.

This is a case series of urinary tract infections (UTIs), which are frequently encountered in both hospital and outpatient medicine [2]. Over 50% of women will have at least one UTI in their lifetime, and 25% of those will have another UTI within 6 months of their first [3]. UTIs are also commonly acquired in the healthcare setting. Patients with urinary retention, chronically indwelling catheters, neurogenic bladder and incontinence are particularly vulnerable. Although most UTIs are not life-threatening, there is significant morbidity associated with UTIs and even mortality that can result from secondary bacteremias and kidney infections [2].

UTIs are frequently caused by a member of the host’s gastrointestinal microbiome, typically a Gram-negative rod such as Escherichia coli or Klebsiella [2]. Treatment of UTIs and suspected UTIs account for a large percentage of antibiotic use in adult medicine in both inpatient and outpatient settings and acquired antibiotic resistance over time is commonly seen in the microbiomes of patients with recurrent UTIs [4].

Despite their frequent occurrence, confirming a UTI diagnosis is often clinically challenging. Cultures can be positive without an actual infection due to colonization by benign flora. Urinary analysis can similarly show signs of bladder inflammation without an infection [5]. As patients can have bacteria in their urine without having an infection, diagnosis is reliant on symptoms—like burning with urination (dysuria), abdominal pain and feeling the urge to urinate more frequently (frequency). However, many patients most at risk for UTIs are unable to report these symptoms [6, 7]. For example, as described in cases below, a paraplegic with no sensation below the navel or a 90 year old who no longer speaks due to her dementia requires thinking outside of the diagnostic box [8].

Evolutionary concepts can be used to help answer some of these clinical questions and help better understand the biological processes happening in these disease presentations. Some of the evolutionary principles that are particularly useful to the discussion of this case series include selection, fitness and fitness cost, competition and mismatch. They are briefly discussed in Box 1.

Box 1. Evolutionary principles significant to these case discussions.

Selection is the main driver of evolutionary change, During selection, phenotypes within a population are found to have greater or lesser fitness in a given environment. These fitness differences result in a higher or lower probability of reproductive success, resulting in changes in the phenotype ratios in a population over time [9]. For example, bacterial populations exposed to antibiotics demonstrate increasingly resistant phenotypes over time. This concept was became viral on YouTube with Harvard’s MEGA prate experiments showing time-lapse videos of evolving resistance in the form of successive and visually distinct clonal expansions into areas of higher antibiotic concentrations (Fig. 1) [10].

Fitness is the degree to which an individual or a population is more or less likely to reproduce in a given environment. Thus, a fitness cost describes a phenotype that decreases the chance of reproduction an environment [11]. For example, it has been shown that many bacterial populations will lose their antibiotic resistance quickly after no longer being exposed to the antibiotic driving the resistance [12,13]. This suggests that the resistance in question can come at cost in environments without the corresponding antibiotic, and thus is selected against.

Competition results when organisms require the same limited resources in the same environment. This can be seen both between populations and within populations. This competition likely resulted in the evolution of the bacterial and fungal antimicrobials that we have now co-opted to use in medicine (such as Penicillin) [14]. An example of this type of competition is presented in a paper by Mashburn et al. that discusses a model of Pseudomonas aeruginosa and Staphylococcus aureus co-culture. In short, growing P. aeruginosa with S. aureus changed the iron acquisition strategies of P. aeruginosa, resulting in direct competition between the two species of bacteria as well as within the aeruginosa population [15].

Mismatch results when a previously adaptive (or even neutral) phenotype encounters an environment in which its traits are no longer favored [16]. Our current environment. especially With regard to healthcare, is considerably different from the one in which most of our genes were selected. Advances in healthcare, Such as surgeries, pharmaceuticals, and medical devices are allowing more people to live longer with more chronic conditions. This is not to say that our ancestors never reached advanced age or were not cared for, both the historical and archeological records confirm this. However, the means and degree has changed rapidly in recent decades. As these changes are recent and these medical conditions Often occur well after reproduction, selection has had neither the time nor the effect on reproductive fitness necessary to decrease heritable risks for many of the medical issues physicians commonly treat [17]. For example, an indwelling Foley catheter, and the risk of urinary tract infections that it brings was not something that was present in our environment until its invention in 1929.

Figure 1.

An experimental device for studying microbial evolution in a spatially structured environment. (A) Setup of the four-step gradient of Trimethoprim (TMP). Antibiotic is added in sections to make an exponential gradient rising inwards. (B) The four-step TMP MEGA-plate after 12 days. E. coli appear as white on the black background. The 182 sampled points of clones are indicated by circles, colored by their measured MIC. Lines indicate video-imputed ancestry. (C) Time-lapse images of a section of the MEGA-plate. Repeated mutation and selection can be seen at each step. Images have been aligned and linearly contrast-enhanced but are otherwise unedited [10].

Cases

Case 1—selection and fitness cost

A 36-year-old female presents to the clinic with a history of recurrent UTIs—approximately two a year. She has no history of a resistant organism and no other medical history. She has no known antibiotic allergies. She reports dysuria, abdominal pain and frequency. She denies any flank pain, back pain or fevers. Vital signs including temperature, blood pressure and heart rate taken at the clinic are within normal limits. A physical exam is performed that is notable for lower abdominal pain on palpation but no other significant findings. Prior screening complete blood count (CBC, includes total white blood cells [WBCs]) and complete metabolic panel (includes kidney and liver function) taken 3 months ago are also within normal limits. Urinalysis (UA) is positive for WBCs, leukocyte esterase and nitrates; findings consistent with inflammation in the bladder.

As she has no symptoms of a more systemic infection such as pyelonephritis (infection of the kidneys) as noted by the lack of flank and back pain or fevers, she can be considered to have a simple UTI, also referred to as simple cystitis or infection of the bladder.

Per guidelines for treating simple cystitis, four medications are recommended for first-line empiric therapy: fosfomycin, nitrofurantoin, Trimethoprim/Sulfamethoxazole (TMP-SMX) and pivmecillinam (see Fig. 2) [7]. This patient has no contraindications to any of these medications. What is the best therapeutic choice for her?

Figure 2.

Guideline-based therapy of simple cystitis. As first published in [7]

Applying the principle of selection can help clinicians to decide between these options. Fosfomycin and nitrofurantoin are absorbed in the proximal small intestine and concentrated in the urine, with little drug reaching the colon [18, 19]. Relatedly, nitrofurantoin seems to have little impact on the gastrointestinal microbiome [20]. Few such publications exist for fosfomycin, however as it is similarly absorbed and concentrated in the urine, it is likely to have similar properties. This attribute of drug absorption should decrease the selection pressure for resistance in the gut, from where these infections are typically seeded. There might be other reasons a clinician mindful of selective pressures should choose nitrofurantoin; at least one paper has posited a high fitness cost associated with nitrofurantoin resistance [21]. There is empirical support for this hypothesis: resistance to nitrofurantoin remains relatively rare in clinical isolates despite it being an older and frequently used antibiotic. For example, a paper examining resistance rates in E. coli isolated from urinary tracts found a nitrofurantoin resistance rate of 3.8%, compared to over 20% for trimethoprim/sulfamethoxazole (TMP/SMX) and fluoroquinolones [22]. In the absence of contraindications to nitrofurantoin, an evolutionarily minded physician, given the information above, would likely have an easier choice in deciding among the four recommended treatments for simple cystitis, while simultaneously also decreasing the likelihood of the development of antibiotic resistance.

Case 2—competition

A 51-year-old male presented to the hospital per the recommendation of his urologist who was concerned about the need for urgent attention due to a positive urine culture obtained in the clinic. He has a pertinent medical history of paraplegia from a motor vehicle accident 20 years ago, an indwelling urinary catheter, and recurrent UTIs. The culture from the urologist grew Acinetobacter baumannii resistant to ciprofloxacin, levofloxacin, cephalosporins, carbapenems, piperacillin/tazobactam, TMP/SMX and sensitive to tobramycin. Per the urologist’s note, he has no symptoms of a UTI at the time, though his paraplegia makes the evaluation of symptoms challenging.

When he presented to the emergency room another urinary culture was obtained. This culture grew Proteus mirabilis resistant only to nitrofurantoin. He reported no typical symptoms, however, he explained that he often has nonspecific symptoms as he lacked sensation below the spinal cord level of T10 (below the ribs) and is thus unable to feel typical symptoms such as abdominal pain or dysuria. In addition, his urinary catheter prevented him from experiencing urgency or frequency. Vital signs reveal he was afebrile but did exhibit tachycardia (increased heart rate). A CBC showed a normal white blood cell count. A UA had elevated WBCs and leukocyte esterase.

This case presented clinical uncertainty regarding if the cultured Proteus should be treated as a genuine infection or considered an asymptomatic bacteriuria (bacterial colonization of the bladder) and not treated with antibiotics.

The practice of medicine always involves weighing risks versus benefits. While it was not possible to unequivocally rule out a true UTI in this scenario, the risks of exposing patients to antibiotics reflexively for asymptomatic bacteriuria must be accounted for.

There are inherent risks from any antibiotic use, but the application of evolutionary medicine illuminates additional consequences. The need to avoid evolving further antibiotic resistance in this patient with frequent UTIs cannot be overstated [4]. Furthermore, applying the principle of competition can help more accurately assess the risks and benefits of giving this patient antibiotics. The Infectious Disease Society of America, in its guidelines for asymptomatic bacteriuria, explains that treating this benign condition as a true infection may actually increase the risk of a symptomatic UTI [7]. However, this guideline declines to explain why or how this protective effect might occur.

Insights from microbial ecology can apply to this clinical scenario. Nonpathogenic bacteria in the bladder likely compete with, and thereby prevent, more pathogenic organisms from causing infections. The adage ‘nature abhors a vacuum’ is useful here—the less pathogenic organism takes space and resources while often employing its own antibacterial strategies [23]. Figure 3 shows examples of how these mechanisms may work [23]. In fact, the introduction of a nonpathogenic strain of E. coli or Lactobacillus has been shown to decrease the occurrence of UTIs in small clinical trials [23].

Figure 3.

Potential mechanisms of bacterial interference: The Bladder Inn. Abbreviations: G, good; B, bad [23]

In this patient, the risks of treating the Proteus in his urine and allowing another organism, such as the highly resistant Acinetobacter baumannii he is likely colonized with, to repopulate his bladder likely far outweigh the risks of not treating it.

Case 3—mismatch

An 88-year-old female with a pertinent medical history of advanced dementia presents to the hospital for altered mental status and is found to have a positive urine culture. She is minimally verbal, bedbound for the last year, and a long-term nursing home resident. This is her third similar admission in the last 8 months. An antibiotic is again prescribed based on culture results. Although it is questionable if these cultures represent actual infections, what processes underlie these recurrent positive urine cultures?

Applying the evolutionary/ecological principle of mismatch may help provide an answer. Which elements of her current environment are most pertinently different compared to the ones in which humans evolved?

Common risk factors for UTIs include constipation, urinary retention and/or incontinence (inability to voluntarily control urination), menopause, being bedbound, and having a chronic indwelling urethral or suprapubic catheter [8, 24]. Many of these risk factors represent conditions that were likely significantly less common in our ancestral environment.

While these risk factors for UTIs are widely known in medicine, the question of why they are risk factors and how to intervene is less clear. Guidelines from professional organizations have reached reasonable consensus on the treatment of UTIs but remain highly variable with regard to recommendations of prevention measures, and some even decline to discuss prevention [25]. Expanded examination of these risk factors may help to illustrate how evolutionary medicine can lead to enhanced understanding and prevention strategies. For Case 3, her risk factors include being bedbound and post-menopausal. A key feature of the human species is our vertical nature, including the adaptations concordant with this posture. Our voiding systems are evolved to employ gravity both for excretion and the urge to urinate [26].

Additionally, at 88 years old this patient is post-menopausal, a state that decreases vaginal estrogen resulting in vaginal dryness and atrophy as well as a decrease in UTI-protective Lactobacillus bacteria [23, 27]. These changes make post-menopausal women particularly susceptible to UTIs. Although our ancestors certainly could live well into their menopausal years, the frequency and duration of this experience has increased with the advent of modern medicine [28].

Despite the lack of consensus in guidelines and a lack of definitive trials, evolutionary medicine may provide and further reinforce potential management choices to prevent these infections. In this patient, correcting the decrease of vaginal estrogen after menopause with topical cream may restore her Lactobacillus microbiome and decrease the incidence of UTIs [29, 30]. This measure partially corrects a tradeoff of menopause—decreasing levels of estrogen prevents further investment in reproduction while increasing vulnerability to diseases such as UTIs. In addition, reinforcing the need for bedbound patients to be moved out of bed is critical to decreasing urinary retention (as well as other negative health effects like constipation) and the corresponding risk of UTIs [31].

Remaining cognizant of these mismatches will reinforce appropriate environmental modifications to decrease the chance of UTIs in these high-risk patients and thereby decrease the frequency of antibiotic treatments and their associated harms.

Discussion

This limited case series demonstrates the potential for evolutionary applications to assist in clinical decision-making and should be used as an example to encourage similar exercises. These exercises can be used as educational tools to reinforce evolutionary concepts as well as clinical skills. They also serve to highlight gaps between evolutionary theory and clinical application that suggest future areas of research and clinical trials. In addition, these approaches themselves should be studied to monitor their effect on clinical practice and medical education.

These concepts are meant to supplement guidelines and clinical trials rather than supplant them. It is common in clinical practice to encounter situations where treatment decisions are needed based on incomplete information, several appropriate treatment options exist, or the risk/benefit analysis regarding a certain treatment is particularly unclear.

In addition, professional guidelines can often be contradictory or lacking guidance in specific scenarios or questions, and evidence from clinical trials can be sparse or nonexistent. Especially in these circumstances, including evolutionary and ecological principles, like those outlined in this paper, can help resolve these uncertainties and assist clinicians in making the best clinical decisions for their patients.