Kidney Damage After Sepsis: A Long-term Follow-up. An Experimental Study.

Background: Acute kidney injury (AKI) in sepsis is a common event. This study aims to evaluate the long-term impact of sepsis on renal hemodynamics and morphology. Methods: Wistar rats underwent sepsis and survivors (n = 24) were followed for up to six months, monitoring macro, regional and micro hemodynamics of the kidney, serum creatinine, and renal histology. The naive animals were used as the control group (n= 6), and sepsis was induced by E. coli e.v. inoculation. Surviving animals were monitored for up to six months. Results: Overall, the findings show that sepsis survivors have long-term hemodynamic and morphological compromise, as well as a progressive worsening of renal functional unit components over time. Even after six months of recovery from sepsis, severe renal hypoxia, chronic inflammation, evidence of increased vascular resistance, and renal fibrosis were observed in surviving animals. These alterations were present in animals with a healthy appearance and normal MAP. Conclusion: Those findings may represent a state of severely impaired physiology and be a contributing factor to the higher susceptibility to renal failure in the face of a new infectious challenge or to other pathological stimuli in the post-sepsis periods.

The main pathological hemodynamic characteristics in patients with sepsis and septic shock include distributive shock, myocardial depression, altered microvascular flow, and diffuse dysfunction of the endothelial barrier, which results in microvascular leakage, edema of tissues and organs. [8][9][10][11]. Microcirculatory dysfunction can induce an imbalance between systemic oxygen supply and demand, leading to global tissue hypoxia and shock. [11][12].
Acute kidney injury (AKI) remains a frequent event in severe sepsis due to systemic hemodynamic disorders associated with an exacerbated systemic inflammatory state. [13][14]. Up to 60% of patients with sepsis have AKI and patients with AKI associated with sepsis have significantly increased mortality compared to those with AKI of another etiology. [15].
Microcirculation dysfunction in sepsis may cause the development of AKI, because of the decrease in the supply of oxygen, leading to areas of hypoxia/ischemia, and a decreased nutrient supply. [16]. Localized hypoxia can generate changes in metabolic, inflammation, and tissue damage. [17]. Tubular injuries, apoptosis and necrosis have been associated with AKI in sepsis. [18][19].
Considering the complexity of the pathophysiological mechanisms of kidney changes in sepsis, the understanding of the consequences of the disease in survivors may bring new knowledge for the intensive care of patients with sepsis. This study was designed to assess whether the kidney injury due to sepsis recovers or can persist, such as residual damage, representing a condition that can limit the ability to have an adequate renal physiological response in the case of reinfection in survivors.

MAP
The MAP remained within the normal range (106±9 mmHg) in all groups at all time periods of the study.

Renal blood flow
Despite maintaining normal MAP levels, renal artery and vein flow were significantly reduced in the very acute phase of sepsis (6h) compared to the Naive group (n=6) data. (Figure 1). Figure 1, Surprisingly, in the post-sepsis periods, the surviving animals (n=24) persisted with a significant hypo flow even up to six months, despite presenting a healthy clinical aspect with good mobility and with weight gain.
Although there was a trend in the recovery in the first month, a drop of around 50% was observed by the third month, and there was a partial recovery in six months (30-40% decrease). The percentage change shows how big the change is relative to the initial value (control group). The percentage was calculated using the following  -80 % restoration of kidney blood flow was not achieved, both in the artery and in the renal vein, demonstrating that a single episode of sepsis causes significant long-term blood flow alterations in sepsis survival.

Cortical tissue perfusion unit (Laser Doppler)
Similarly, renal cortex tissue perfusion showed a matched hypoperfusion pattern ( Figure 3), suggesting that kidney tissue hemodynamics are directly influenced by the regional flow, even in the absence of macrohemodynamic changes (MAP). In all post-sepsis phases, cortical tissue perfusion was significantly reduced. Within 30 days, there was a partial recovery, followed by a gradual decline for up to six months.
When compared to the naive group, hypoperfusion was considerable at the end of six months (baseline). This pattern demonstrates that the renal cortical blood flow is significantly reduced and does not return to its baseline.

Renal cortex microcirculation hemodynamics (SDF)
Videomicroscopy of the renal cortex microcirculation revealed significantly altered cortical tissue structures as well as microcirculatory flow. In the very acute phase of sepsis (6 hours), there were observed generalized alterations in the architecture of the renal tubules, with a reduction in the luminal space and edema of the epithelial cells.
The peritubular microvessels, which had venous characteristics, were congested or compressed as the tubular wall grew larger. These patterns of structural disarrangement produce the perception of a heterogeneous look due to tubularvascular zones that connect the preserved and compromised portions to variable degrees. (Figure 4-SDF). The hemodynamics of microcirculation in survivors improved gradually but only partially over time. However, even after six months, there was no complete recovery. This could be one of the factors driving chronic renal hypoxia/ischemia due to decreased blood flow in the renal artery and vein.

Histological findings
Naive: Homogeneous kidney structure distribution, with a normal aspect of the epithelial cells of convoluted tubules, preserved luminal space, peritubular microvessels with normal distribution, and with preserved intravascular space. The glomeruli had preserved urinary space and capillaries. (Figure 4-HE).  Owing to these histological findings, we next evaluated the collagen type.

Collagen findings
The use of Picrosirius red staining on the histology slides revealed that changes occurred over a one-month period, with Type III (green) collagen predominating over Type I (red) collagen in the walls of glomeruli and tubules. There was a considerable prevalence of Type I at three months, which increased even more in the sixth months,  state. [20]. The long-term renal effects of sepsis on survivors, on the other hand, are still unknown. [13,15]. Furthermore, it is unknown how much renal impairment caused by sepsis plays a role in hospital readmission or possibly death.
Hydroelectric homeostasis, blood volume management, blood pressure, pH balance, and plasma osmolarity are some of the kidney's physiological processes that help to repair the hemodynamic imbalance caused by critical illnesses. Injuries observed in AKI during sepsis include ischemia-reperfusion injury in the glomerulus, inflammation of sections of the nephron, oxidative stress, tubular injury mediated by cytokines and chemokines, and tubular and mesenchymal apoptosis. [16][17].
According to our findings, sepsis caused significant changes in the distribution of microcirculatory blood flow and severe morphological changes in the renal tissues by promoting microcirculatory dysfunction. The edema of tubule epithelial cells, which caused compression of their lumen and peritubular microvessels, resulting in microvascular congestion and restricted blood flow, was one of the most noticeable changes. When this occurs in a widespread manner, it creates congestion not just in the renal parenchyma's small and medium peritubular microvessels, but also in the glomerular capillaries of high blood flow. These effects were apparent from the acute phase of sepsis until the completion of a six-month recovery period.
Renal fibrosis and maladaptive repair may be associated to a condition of renal hypoxia/ischemia, an increase in renal vascular resistance, and the emergence of chronic pathogenic effects.
Several pathophysiological events occur simultaneously and in sequence during AKI, including microcirculatory failure, endothelial cell dysfunction, inflammation, recruitment of various leukocytes and cytokines, tubular cell damage, and tubular and renal venous congestion. [21][22]. Fibrosis is a feature of maladaptive repair, according to a crucial relationship between injury, improper repair, and the development of fibrosis. [23].
Pericyte to myofibroblast transition, endothelial failure, chronic inflammatory infiltrates, renin-angiotensin system activation, mitochondrial dysfunction, and epigenetic modifications are all pathological changes that contribute to renal fibrosis as a result of the maladaptive process. [24]. Other findings have highlighted the role of the proximal tubule and proximal tubular epithelial cell, a specialized tubular segment just adjacent to the glomerulus, as not only the target of injury but also an important mediator of renal fibrosis progression. [25].
Our findings regarding collagen repair also show that the fibrosis process is more evident in glomeruli, capillaries, and Bowman's capsule, and in the adjacent tubules. The growing prevalence of Type-I collagen over Type-III collagen and the progressive thickness of Type-I fibers, notably in the three and six post-sepsis periods, indicated the advancement of renal fibrosis, and the increase in renal vascular resistance.
These findings support the hypothesis that the first six months after sepsis represent a critical period for a new infectious insult. During this period, damage to the functioning kidney units provides a favorable environment for the disease. A study of patients aged 50 years or older than sepsis survivors demonstrated that they were 2.5 times more likely to be readmitted to the hospital for AKI within 90 days than patients with compatible comorbidity without sepsis. [26].
Furthermore, a significant and long-lasting reduction in renal artery and vein blood flow, in conjunction with microvascular dysfunction and tissue hypoperfusion, shows that the kidneys are continuously exposed to the hypoxia/ischemia factor in post-sepsis periods.

Sepsis-induced AKI, according to Gomez et al., is the first clinical and
biochemical expression of adaptive tubular cell response to a detrimental inflammatory danger signal [17]. They proposed that the interplay of inflammation and microvascular malfunction defines and amplifies this signal, and that tubular cell mitochondria orchestrate metabolic dysregulation in response.
In conclusion, a single sepsis episode was able to trigger a cascade of changes that can impact renal physiology over an extended period of time, raising the question of whether this can happen in other organs commonly affected by sepsis. Certainly, the combined dysfunction of the damaged organs can impair the defense mechanisms' response capacity to prevent harmful outcomes.
To better comprehend this idea, we are looking into other organs that are involved in sepsis. Although experimental animal studies may not fully replicate the etiopathogenesis in human sepsis survivors' illnesses, the ethical concerns of invasive procedures in humans can be solved in experimental investigations, allowing new theories to contribute to post-sepsis disease understanding.

Experimental groups
Female Wistar-EPM rats, weighing on average 250 g, were obtained from the University Animal Colony (CEDEME-UNIFESP). The animals were allocated to plastic cages, given free access to a standard diet and water ad libitum in environmentally All the methods were performed in accordance with the relevant guidelines and regulations. After 6 hours, 30, 90, and 180 days, the surviving animals were submitted to macro, regional and microcirculation monitoring in vivo and the kidneys were collected for histological assay. All surgical procedures were carried out using aseptic and antiseptic procedures. After experiments, the euthanasia of the animals was performed by the section of the aorta under general anesthesia.

Macrocirculation monitoring
Under general anesthesia, the animals were trichotomized in the abdominal region, immobilized on a heated surgical board (37 o C) in the supine position. For the evaluation of mean arterial pressure (MAP), a small skin incision was made in the inguinal region, followed by the dissection and isolation of the femoral artery. Following the opening of the femoral artery, a catheter was inserted and fixed, and connected to a pressure transducer. For post-sepsis monitoring, rats were anesthetized with isoflurane 3% via inhalation and maintained throughout the in vivo monitoring with a 1.5mL/minute flow of oxygen.

Renal artery and vein blood flow.
The monitoring of blood flow of the renal artery and vein was realized by midline abdominal incision, isolation of renal vessels and placing a perivascular 0.7mm probe of the Transonic System Inc. TS420 connected to the flowmeter, under general anesthesia.

Kidney cortical area tissue perfusion
The kidney cortical tissue perfusion was measured by the BLF 21 Laser Doppler flowmeter, under general anesthesia. Three measurements per organ/per animal were obtained, and the results were expressed by the average of the tissue perfusion unit (TPU) readings.

Kidney microcirculation monitoring
The microcirculation of the kidney cortical area of all groups was obtained at 6h, 30,90, and 180 days after sepsis induction, by Sidestream Dark Field Imaging (SDF) videomicroscopy (MicroScan, MicroVision Medical Inc., Amsterdam, The Netherlands) [28]. Five measurements were performed per organ per animal. The results were compared between groups. All procedures were carried out while under general anesthesia.

Histological analysis
At the end of all the monitoring procedures, the kidneys were collected for a histological Microscope. All images were acquired following identical camera settings, including white balance, gain, and exposure. Images were taken at 40x magnification and the intensity of red (Type III) and green (Type I) collagen fiber staining was analyzed with the "rectangle profile" tool of the Zen software, utilizing similar methods described previously by Souza et al. [30][31].

Creatinine values
During the laparotomy, a venous blood sample was collected from the vena cava for analysis in a rapid test using a creatinine kit by I-STAT System (ABBOTT).

Statistical analysis
The nonparametric statistical test Kruskal-Wallis with Dunn correction for multiple comparisons [24] was used to quantify collagen Types I and III, as well as blood flow between study periods. GraphPad Prism 5.0 was used to perform collagen statistical analyses. Differences were considered statistically significant at a p-value < 0.05.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Disclosures
No conflicting relationship exists for any author.