The diagnosis and management of a patient with acute pyelonephritis

Definition, aetiology and epidemiology

Acute pyelonephritis, known in clinical practice as an upper urinary tract infection (UTI), is defined as a tubulointerstitial disorder characterised by inflammation of the renal parenchyma caused by bacteriuria ascending from the bladder up the ureters to the kidneys (Choong et al, 2015). Bethel (2012) estimated that, annually, 1 in every 830 people in England develops pyelonephritis. However, the exact epidemiology and economic cost remain unknown because of the overlap of treatment in primary and secondary care (Bethel, 2012). In comparison, a lower UTI has an annual incidence of 3 in every 100 people, which accounts for over 224 000 hospital admissions each year (National Institute for Health and Care Excellence (NICE), 2014) with an associated £316 million healthcare cost (McDonald et al, 2014). It is important to emphasise that nearly all of these admissions have the pathophysiological possibility to develop into pyelonephritis if treatment is delayed or inadequate.

Presentation

Jane White presented with numerous pyelonephritis risk factors (see Table 1) which have been comprehensively explored in research. Pyelonephritis resulting in admission to secondary care is nearly five times more likely in females than males, with 11.7 versus 2.4 admissions per 10 000 cases respectively (Ramakrishnan and Scheid, 2005). Various reasons to justify this gender disparity have been explored, with Bethel (2012) confirming the primary reason being the difference in anatomy. The turbulence pressure generated in the female urethra allows backflow of bacteria into the bladder during micturition (Franz and Horl, 1999). Moreover, in females, a shorter urethra combined with the closer proximity of the urethral orifice to the anal region facilitates bacteria reaching the kidneys and increases susceptibility to pyelonephritis (Bethel, 2012). This is verified by the aetiology of pyelonephritis because in over 80% of female cases the causative agent is bacteria derived from the host's own bowel flora, namely Escherichia coli (Springall et al, 2002), as in this case study. Other organisms include Enterococcus faecalis, Klebsiella pneumoniae and Proteus mirabilis (Neumann and Moor, 2014).

Matuszkiewicz-Rowińska et al (2015) suggested that hormonal changes during the menstrual cycle increase the risk of pyelonephritis by reducing ureteral peristalsis, aiding the rapid colonisation of bacteria. A frequently cited journal article by Franz and Horl (1999) explained how oestrogen and the use of many oral contraceptives increase the risk. Oestrogen changes the quantity and quality of the mucopolysaccharide layer lining the bladder and urethra, which reduces the visceral smooth muscle tone and contractility, increasing the probability of bacterial attachment to the tubular epithelium (Franz and Horl, 1999). Regular sexual activity, another risk factor highlighted in this case, alters the normal lactobacillus-dominant vaginal flora and promotes invasion of E coli bacteria into the vagina, which acts as a bacterial reservoir (Scholes et al, 2005). Thus female anatomy, hormonal changes and regular sexual activity account for the increased incidence of pyelonephritis in women aged 18-49 years (Frassetto, 2018).

Patient history

The patient reported a history of lower UTIs, two in the past 12 months, which would predispose her to pyelonephritis (Scholes et al, 2005) (see Table 1). Evidence implicates genetic pathophysiological variations in patients with recurrent UTIs (Ambite et al, 2016). Ambite et al (2016) described the differences that negatively affect the host's immune response, which include altered recognition of receptor molecules, reduced cytokine release and decreased neutrophil activation, all of which lead to increased amounts of renal scarring. Renal scarring, a complication of pyelonephritis, causes dilation of the kidney tubules and changes the glomerular-capillary pressure, reducing the efficiency of the normal filtration role of the kidneys and predisposes to recurrence of the condition (Choong et al, 2015). Renal scarring can further contribute to the development of acute renal failure and renal hypertension (Bethel, 2012).

Diagnosis and complications

The mortality rate for acute pyelonephritis has significantly reduced over the past three decades and is attributable to revised clinical guidelines. It has decreased from an estimated near 10% in 1990 (Roberts et al, 1991) to 0.7%-1.65% in 2017 (Frassetto, 2018), but the complications of pyelonephritis, as discussed, remain unchanged. However, these complications can be significantly reduced with early diagnosis through thorough history taking at presentation (Frassetto, 2018). History taking, although an acquired clinical skill, is dependent on the patient's cognitive and sensory ability to explain their symptoms and the severity of the condition (Bethel, 2012). Norouzinia et al (2015) concluded that pain, discomfort and anxiety, as were noticeably present in the case of Ms White, can negatively affect communication and a patient's ability to convey information accurately. Therefore, prior to history taking, these elements were addressed by ensuring adequate analgesia and a discussion with the trainee advanced clinical practitioner (the first author) to prevent a delayed or incorrect diagnosis.

Diagnosis for pyelonephritis can be challenging because presentation of fever and nausea, considered two of the triad of predominant markers for pyelonephritis, is not specific to the condition (Bethel, 2012). Furthermore, the onset of symptoms can vary from hours to several days, with their severity substantially differing from patient to patient (Frassetto, 2018). Colgan et al (2011) stated that flank pain, conspicuous in this case, is universal in patients with pyelonephritis, but the key abnormal examination finding is tenderness on palpation of the costovertebral angle. Due to the anatomical position of the kidneys, palpation of Ms White's right costovertebral angle elicited renal tenderness, which strongly suggested an inflammatory process in the patient's right kidney (Faust and Tsung, 2017). Importantly, nephrolithiasis and ureterolithiasis, which also present with flank pain, do not typically present with tenderness of the costovertebral angle (Colgan et al, 2011). This allowed these differentials to be excluded upon examination.

Pain

The patient presented with acute flank pain (Table 1) caused by the activation of nociceptors on a cellular level (Pham et al, 2017). Nociceptors are stimulated by inflammatory mediators, specifically prostaglandins and cytokines, which are released by the kidney in response to cellular damage (Pham et al, 2017). Nociceptors are also stimulated in pyelonephritis by the smooth muscle distention of the renal capsule, caused by vascular leakage as part of the inflammatory process (Choong et al, 2015). Current research has suggested a protective role of toll-like receptor 4 (TLR4), the key receptor to E coli bacteria, as a mediator of pyelonephritis-associated pain responses that are independent of inflammation (Rosen and Klumpp, 2014). However, the mechanism remains unclear (Rosen and Klumpp, 2014). Once activated, the visceral afferent fibres synapse in the dorsal horn of the spinal cord and ascend via the lateral spinothalamic tract to the somatosensory area in the brain for localisation and conscious recognition of pain by the patient (Martini and Nath, 2009). Adding to the complexity of diagnosis, although not seen in this case, patients with pyelonephritis can present with referred visceral pain to the groin due to the convergence of different afferents on the same dorsal horn neurones in the spinal cord (Martini and Nath, 2009). Nausea and vomiting are associated symptoms of pain, as seen in Ms White's case.

Pyrexia

The patient presented with pyrexia and associated rigors. The generation of a fever occurs when the lipopolysaccharides of the bacteria stimulate the production of prostaglandin (PGE2), which acts in the organum vasculosum of the lamina terminalis (OVLT) in the hypothalamus (Walter et al, 2016). The increased synthesis of PGE2 slows the firing rate of the warm-sensitive neurones, resulting in an elevated body temperature. Walter et al (2016) stated that the maintenance of a fever is caused by cytokines, produced during an immune response, which directly stimulate the OVLT. In addition to rigors, Ms White was hypotensive and tachycardic. This is common in the acute phase of pyrexia, owing to the combination of redistributed blood flow, nitric oxide-induced vasodilation and hyperdynamic circulation (Walter et al, 2016).

Walter et al (2016) stated that pyrexia has three key benefits. First, an elevated body temperature inhibits bacterial replication because this occurs optimally at temperatures below 37°C (Walter et al, 2016). Second, pyrexia increases the efficacy of antimicrobials and, third, it enhances the host's immune response (Walter et al, 2016). However, pyrexia has a negative effect on kidney function. A morphological study by Vlad et al (2010) confirmed that glomerular capillary dilation, interstitial haemorrhage, small vessel vascular stasis and reduced renal blood flow occur with pyrexia, increasing the plasma creatinine and urea concentrations and reducing the glomerular filtration rate.

Treating pyrexia is controversial in the literature due to the discussed benefits being inhibited by prescribing medication (Walter et al, 2016). However, this is not recognised in practice because pyrexia is always treated. In this case, paracetamol was prescribed as recommended in the NICE guidelines (2019) due to its effectiveness at treating pyrexia and rigors. Furthermore, non-steroidal anti-inflammatory medications are not recommended due to their nephrotoxicity which can cause further renal impairment (Joint Formulary, Committee, 2020). Ultimately, patient comfort is the foremost priority.

Importance of urinalysis

Clinical guidelines recommend urinary dipstick tests to differentially diagnose acute pyelonephritis from conditions with similar symptoms (NICE, 2019). Ms White had a positive urine dipstick analysis for nitrites (see Table 2), which is a rapid screening test and surrogate marker of bacteraemia (Douglas et al, 2013). Nitrites are not normally found in urine but various bacteria, including E coli, convert urinary nitrates to nitrites by the enzyme reductase in response to nitrosative stress, which facilitates the organism's further colonisation (Al Majid and Buba, 2010). Importantly, because the dipstick reagent is susceptible to aerial oxidation which can produce a false-positive result, the nitrate dipstick was taken from a closed container by the practitioner (Simerville et al, 2005). Therefore, the presence of nitrites confirmed bacteriuria in the patient.

Before interpreting the urine dipstick results, the practitioner measured the specific gravity of the sample, because this is directly proportional to urine osmolarity and can support false-negative or false-positive results (Kumar and Clark, 2017). Moreover, Franz and Horl (1999) emphasised that not all uropathogens are nitrate-reducing organisms, which means false-negative results can occur. For this reason, nitrite dipstick results should never be used in isolation to exclude bacteriuria. Owing to the low sensitivity of the nitrite test across all organisms, a urinary culture analysis was also performed because this is recognised as the gold standard in the national guideline (NICE, 2018).

Additionally, the patient's urine dipstick was positive for leucocytes, correlating with the presence of pyuria (Frassetto, 2018). The dipstick test result confirmed the presence of leucocyte esterase, which is an endoleukocyte enzyme released upon lysis of white blood cells (Hooton, 2012). Leucocytes are the predominant inflammatory cell in the immune response to bacterial pyelonephritis (Lager and Abrahams, 2013).

The cell walls of bacteria, specifically E coli, contain lipopolysaccharides, which are detected by TLR4, TLR5 and TLR11 found on the apices of the distal tubules (Choong et al, 2015). TLR4 signalling induces the production of cytokines, namely IL-6, IL-1 and tumour necrosis factor, which co-ordinate the cascading immune and inflammatory response. Neutrophils, a specific type of leucocyte, move down this chemotaxis gradient to the infection site and clear the bacteria via phagocytosis (Springall et al, 2002). The patient's blood results indicated that this pathophysiological process was occurring as her C-reactive protein (CRP) was raised (see Table 2). CRP is an infection biomarker synthesised by the liver in response to IL-6 and, along with the positive urine culture, justified the use of antibiotics in this case (Simon et al, 2004). Simon et al (2004) emphasised the clinical limitation of interpreting a CRP result in isolation because it is unable to differentiate between a bacterial and noninfective cause of systemic inflammation.

This significant neutrophil recruitment causes the release of proteolytic enzymes, which alter the physiology of the infected kidney by increasing vascular permeability and breaking down the epithelium (Choong et al, 2015). As a result, the endoleukocyte enzymes leak into the urine (Choong et al, 2015). Lager and Abrahams (2013) suggested that the significant presence of neutrophils in pyelonephritis can be seen in histological diagnostic features. These include pathognomonic pus cell casts found within the tubule lumens in a neutrophil-predominant inflammatory infiltrate (Lager and Abrahams, 2013). This was illustrated by a raised white blood cell count, namely neutrophil count, in the Ms White's blood results (Table 2).

As the patient was female, staff discussed whether this was a true positive result because leucocytes can originate from sites of inflammation in the genital tract as well as the kidneys (Franz and Horl, 1999). Furthermore, research has shown that leukocyturia can continue even after bacteriuria has symptomatically resolved (Franz and Horl, 1999). This could have been explained in this case by the patient's recent lower UTI. However, the positive leucocyte test, combined with the patient's examination and blood results, strengthened the diagnosis of pyelonephritis.

The patient's urine dipstick also indicated proteinuria. The pathophysiological phenomenon of proteinuria in cases of pyelonephritis without pre-existing renal disease has been extensively researched but with inconclusive results. A quantitative study by Mohkam et al (2014) concluded that the prevalence of proteinuria in pyelonephritis is 90.9%-98.7%, with variances due to the proximal or distal tubular leakage of proteins. Proximal tubular involvement causes a higher leakage level of tubular proteins and amino acids because of the increased effects of the inflammatory process (Mohkam et al, 2014). Conversely, Carter et al (2006) established in a systematic review that positive protein tests occur because of the reaction of the test pad with leucocytes and bacterial proteins. It was also concluded to be a result of alkalinisation of the urine caused by the bacterial breakdown of urea, rather than due to the tubular leakage of proteins. Further research is required in this area, particularly utilising immunoassay methods because these methods are not susceptible to the highlighted problems of urine test strips (Carter et al, 2006). However, such methods are expensive and would probably not be implemented routinely in today's NHS. Despite the exact nature of proteinuria being undetermined, it is a recognised feature of pyelonephritis.

The cellular structure of the nephron increases its sensitivity to infection. Choong et al (2015) stated that the nephron contains different types of epithelia: thin squamous cell epithelia line the glomerular capsule in the renal corpuscle being specific for filtration, while cuboidal cell epithelia line the tubular system and are specific for reabsorption. This variation in the epithelial structure and function throughout the different segments, combined with the changing chemical composition of the urine, has increased the colonising ability of bacteria (Choong et al, 2015).

In Ms White's case, Gram-negative rods of E coli were isolated during microbiological analysis of the patient's urine culture. E coli has numerous virulence factors, which enabled its rapid colonisation in the kidneys in this case (Raeispour and Ranjibar, 2018). Examples of these factors are haemolysin A and P-fimbriae as previously discussed, but others include the syntheses of aerobactin and enterobactin, which are iron-binding proteins required for bacterial replication (Franz and Horl, 1999). P-fimbriae is crucial for mediating the binding to the epithelium in the early stages of colonisation, whereas type 1 fimbriae is essential for colonisation across the luminal centre of the tubule where there is no epithelium and significant hydrodynamic forces are identified from the filtrate (Choong et al, 2015). Type 1 fimbriae has adapted to be capable of inter-bacterial binding that occurs via a catch-bond mechanism, allowing the bacteria to bind and remain in the microenvironment rather than being washed out of the system.

From a clinical perspective, the infection-associated inflammatory response has an impact on normal renal function. Choong et al (2015) stated that bacterial colonisation causes sloughing of the tissue, clot formation, localised ischaemia and cellular blebbing (bulging) because the immune-inflammatory defence mechanisms are triggered to protect the patient from bacterial dissemination and systemic spread. These cause a degree of tubular obstruction and lead to changes in the hydrostatic pressure, negatively impacting the glomerular filtration rate and seen in this patient's blood results.

Kumar and Clark (2017) argued that hydronephrosis from obstruction, either by renal calculi, cysts, malignant tumours or infected collections, gives a causative reason for a patient's pyelonephritis. Consequently, in all patients, these differential diagnoses need to be excluded by carrying out an ultrasound of the kidneys and bladder, as was completed in Ms White's case. Ultrasound is the imaging method of choice for renal diagnosis because it has the two main advantages of avoiding both ionising radiation and the nephrotoxic intravascular contrast medium used in computerised tomography scans (Kumar and Clark, 2017). The patient's ultrasound confirmed no obstruction, strengthening the theory of bacteraemia being the solitary cause of the pyelonephritis.

Patient education

Patient education and the promotion of self-management have been identified as crucial treatment elements to reduce the risk of recurrence and promote health behaviours (Hooton, 2012). Health promotion information, endorsed by Kumar and Clark (2017), was provided by the practitioner and included recommendations such as discouraging the use of soaps, voiding after sexual intercourse, avoiding tight-fitting underwear and not delaying urination. However, a case-control study by Hooton (2012) revealed that these have no statistical significance in reducing the recurrence of pyelonephritis. Furthermore, increased fluid intake and drinking cranberry juice were recommended by the practitioner, both of which are supported by a Cochrane review that described these as biological mediators, inhibiting uropathogens' adhesions to uroepithelial cells (Jepson and Craig, 2008). However, a more recent randomised placebo-controlled trial concluded that there is no benefit in drinking cranberry juice or increasing fluid intake on the recurrence of pyelonephritis (Barbosa-Cesnik et al, 2011). Despite inconclusive clinical research, making recommendations about health behaviours and biological mediators remain an affirmed part of clinical practice in pyelonephritis management because they pose a low risk to patients.

Communication

Cole and Bird (2014) emphasised the importance of patient engagement through both clear communication during history taking and by promoting collaborative management in excluding differential diagnoses effectively. On analysis and clinical reflection by the practitioner, effective communication was facilitated during the patient's stay by clearly explaining the diagnosis and treatment, and by addressing any concerns arising due to the psychological effects of the diagnosis (see Table 3).

Verbal and written communication with the multidisciplinary team is crucial in pyelonephritis management to optimise treatment (Frassetto, 2018). In this case, discussions with the microbiologists and pharmacists were necessary to ensure effective antimicrobial prescription in line with antimicrobial stewardship (NICE, 2015).

Ms White was treated with intravenous antibiotics for 2 days until she was able to tolerate oral fluids and a light diet. Antibiotics were then switched to oral amoxicillin/clavulanic acid for a further 5 days (Table 2). In addition, Ms White was prescribed regular paracetamol for the first 3 days, then changed to ‘as required’ (PRN) thereafter. Non-steroidals are not advocated in acute pyelonephritis due to concerns regarding renal safety (NICE, 2015). Once oral antibiotics are commenced, patients are monitored for a period of 24–48 hours to ensure there are no further spikes in temperature. Because Ms White was tolerating diet and fluids she was discharged on day 5 with safety-netting advice to see her GP immediately or call 111 if symptoms recurred.

The patient discharge letter was comprehensive and included safety-netting recommendations for the GP in case of recurrence. An outpatient appointment with a renal consultant was also arranged to ensure that there were no residual effects on the patient's renal function.

Conclusion

It is evident that to diagnose pyelonephritis correctly, a comprehensive understanding of its pathophysiology is required, including the critical analysis of research. Clinically, effective history taking and the identification of risk factors can support the complexity of pathophysiology, aiding the main objective of successful patient care.

KEY POINTS

CPD reflective questions