Navigating Complex Relationships Between Perioperative Hypotension and Delayed Graft Function in Kidney Transplantation

See Clinical Research on Page 1819

Delayed graft function (DGF) is typically defined as receipt of dialysis within 1 week of transplantation.¹ It is associated with a higher incidence of acute rejection and reduced long-term graft survival. Although many donor and recipient factors have been associated with increased risk of DGF, evidence from experimental models suggests that ischemia and the subsequent reperfusion of kidneys subjected to hypothermic preservation initiate a complex series of events that play a central role in its pathogenesis. Although defined by a simple clinical criterion, DGF is a composite end point that combines intersecting donor and recipient factors and events (many of which are unknown or unrecorded) that occurred over a donor’s lifetime (e.g., history and severity of hypertension) and in the hours or days after transplantation.

One potentially modifiable contributor to DGF (by exacerbating graft ischemia after kidney reperfusion) is recipient hypotension. Hypotension is a promising potential target because it is easy to detect, routinely measured in clinical data, and can be treated by multiple potential interventions. Unfortunately, like DGF, hypotension may be easy to define but is deceptive in its apparent simplicity. Retrospective studies of blood pressure and its impacts on short-term outcomes are challenging to conduct. As illustrated in this study, even the frequency of blood pressure measurement and the ability to detect hypotension vary by care setting, which is determined by patient needs and hospital practices and resources (e.g., availability of beds in an intensive care unit or ward).

In a recent issue of Kidney International Reports, Morrison et al.,² report the results of a retrospective observational study of kidney transplant recipients managed in a regional Canadian program to identify associations between recipient hypotension during 2 time periods—intraoperative (after reperfusion) and postoperative—and DGF. Among the 1020 patients captured in the study’s clinical and outcomes dataset, intraoperative hypotension (defined as a systolic blood pressure < 90 mm Hg during the last hour of surgery, after graft reperfusion was assumed) occurred in 209 recipients (20.5%), whereas postoperative hypotension (defined by the same systolic blood pressure measured during the first 2 postoperative days) was observed in 112 (11.0%). The incidence of DGF among cohort patients, the primary outcome, was 15.3%.

The authors determined that intraoperative hypotension was not associated with either DGF (adjusted odds ratio: 1.02; 95% confidence interval: 0.61–1.72) or prolonged hospital length of stay (adjusted odds ratio: 1.19; 95% confidence interval: 0.81–1.76) but was associated with significantly lower risks of graft loss. Conversely, the study found that postoperative hypotension was significantly associated with an increased risk of DGF (adjusted odds ratio: 4.01; 95% confidence interval: 2.24–7.19) and secondary outcomes, including prolonged lengths of stay and graft loss.

The study joins a growing body of work focused on this problem and patient population, in which recipient and donor risk factors for DGF are well-characterized. Intraoperative hypotension, though variably defined by blood pressure threshold and duration, is an established risk factor for DGF.3, 4, 5 Furthermore, this study identified postoperative hypotension, rather than intraoperative hypotension, as a significant predictor of DGF and other adverse outcomes. Although strong correlations between hypotension in the 2 time periods may be assumed, different potential sources of hypotension (e.g., bleeding and anaphylaxis) may be satisfactorily treated in the operating room or develop on the ward (e.g., sepsis) without relation to one another.

In addition to differing definitions and the incidence of DGF among studies, discrepant findings may be attributable to heterogeneity in study populations, the blood pressure thresholds and durations defining hypotension, and detection method (e.g., arterial or noninvasive). Even in individual studies, it is challenging to isolate the potential impacts of hypotension from the complex and heterogeneous ecosystem of donors, recipients, and causes of DGF other than graft malperfusion (such as acute rejection or graft thrombosis). The authors propose a simple conceptual model of key study exposures and outcomes in the manuscript’s Supplementary Material. Herein, we propose an alternative scheme highlighting complex (and sometimes cyclical) relationships between factors (Figure 1). This model illustrates the likely presence of confounding variables (including duration of dialysis, recipient body mass index, and intraoperative management), which were not available in the study data. Our model suggests that available data cannot ascertain cause and effect, even among clearly defined and isolated factors. Two issues deserve further discussion: intraoperative clinical management and optimal graft perfusion pressure.

Figure 1.

Conceptual model of factors contributing to perioperative hypotension and study outcomes. Factors italicized were not included in analyses by Morrison et al.²

First, though hypotension commonly occurred in this study, we cannot determine how or to what degree perioperative hypotension was prevented or treated by clinicians. For example, in the operating room, anesthesiologists consider not just patients’ baseline blood pressure and comorbid disease (including coronary artery disease and diastolic dysfunction) when selecting anesthetic drugs (which vary in their vasodilatory and myocardial suppression properties) and doses, but also when defining acceptable intraoperative blood pressure ranges. Kidney transplant recipients are unique from other patient populations in both their burdens of comorbid disease and an important caveat: management of perioperative hypotension with vasopressor drugs may have different effects on the denervated transplanted graft than on native organs and tissues.⁶ Although specific evidence guiding anesthetic management during kidney transplantation is limited,⁷ the best studied factors are mannitol, the type and volume of fluid administered, and vasopressors and inotropes, which may mitigate the risk of DGF attributable to multiple upstream causes, including hypotension.

Another potentially overlooked factor is the definition of hypotension as an absolute threshold, rather than relative to patients’ (and donors’) individual physiology. Although analyses were adjusted for comorbid disease (including chronic hypertension among recipients, which may lower the risk of DGF⁸) and sensitivity analyses examining multiple specific blood pressure thresholds were conducted, neither approach fully accounts for expected heterogeneity in donors’ and recipients’ baseline blood pressures. In addition, though work in other surgical populations has demonstrated that hypotension defined by threshold values or relative to individual patients’ baselines are comparable predictors of postoperative myocardial and kidney injury,⁹ we cannot ascertain whether perioperative pressures were sufficient to perfuse transplanted grafts.

Given these limitations, what can we take from this study and the greater body of work? Fundamentally, these studies aim to ascertain whether treating a common clinical problem, which may be addressed using readily available tools (such as vasopressors and fluids), may prevent or mitigate an outcome important to patients, clinicians, and health care systems. Although causal relationships cannot be established from this work, it illustrates that despite readily available tools, perioperative hypotension is a common condition among recipients of kidneys from deceased donors after brain death. In addition, though evidence of a direct relationship with DGF is limited and sometimes conflicting, the study cannot and should not be interpreted as justification for avoiding detection, prevention, or treatment of recipient hypotension, which has consequences beyond DGFS1 and is supported by recent guidelines.S2

In the future, more work is needed to determine how to best monitor and care for kidney transplant recipients during index hospitalizations. Although randomizing patients to hypotension versus no hypotension may not be ethical or scientifically sound, formal synthesis of past work is an opportunity to evaluate the strengths and weaknesses of contemporary evidence systematically. Future studies in this population may consider including living donor kidney transplant recipients or recipients of kidneys donated after circulatory death, randomizing recipients to patient-specific blood pressure ranges, measuring physiologic end points more likely to be directly impacted by recipient blood pressure (such as graft blood flow), and (as in this study) examining patient outcomes beyond DGF. An unknown is how effective perioperative hypotension prevention and treatment is relative to emerging and novel therapies for DGF³; as practices change and technology evolves, the role of hypotension will need to be further revisited.

Disclosure

EAV reports contracted research (paid to her institution) by eGenesis (Cambridge, MA). SM declared no competing interests.

Supplementary Material

Supplementary File (PDF)

Supplementary References.