July 2020

Synopsis 1: Accumulating evidence supports the use of radiocontrast in patients with chronic kidney disease, but less is known about the potential for acute kidney injury in critically patients. This study was a retrospective observational study of 7,333 adult ICU patients across a six-hospital health system in Florida. Using propensity score (PS) matching to adjust for patient characteristics and comorbidities, admission type and diagnosis, and severity of illness, 2,306 PS-matched pairs were obtained. 90% of patients who received contrast were matched and patients with AKI stage 3 at reference and CKD stage 3-5 at reference were excluded from analysis. The primary outcome was absolute and relative difference in AKI according to contrast exposure. The median age was 64 years, 55% were male and median APACHE IV score was 46. Overall, there was no difference in rates of AKI in patients who received contrast (19.3%) vs. no contrast (18.0%) [relative risk 1.07; 95% CI 0.95-1.21; p=0.27]. In two exploratory multivariable analyses of the PS-matched patients, contrast exposure was not a significant risk factor for AKI after adjustment for demographic, comorbid and diagnosis (OR 1.11; 95% CI 0.95-1.29; p=0.189) or adjustment for demographics and severity of illness (OR 1.10; 95% CI 0.94-1.28; p=0.246).

Synopsis 2:  The Kompas trial was a multicenter, noninferiority, randomized clinical trial conducted at six hospital in the Netherlands with 523 patients with CKD stage 3 (eGFR 30-59 ml/min/1.73m2). Patients were randomized in a 1:1 ratio to receive no prehydration or prehydration with 250ml of 1.4% sodium bicarbonate as a 1-hour infusion prior to undergoing elective contrast-enhanced CT scan. The primary endpoint was mean relative increase in serum creatinine level 2-5 days after contrast exposure compared to baseline (noninferiority margin <10% increase). The median age was 74 years, 64.2% were men, 39% had diabetes, 35% had coronary artery disease and 35% had peripheral artery disease. 523 patients were included in the intention-to-treat analysis. Overall, the mean relative increase in creatinine at 2-5 days was 3.0% (SD 10.5) in the no prehydration group vs. 3.5% (SD 10.3) in the prehydration group (mean difference, 0.5; 95% CI, -1.3-2.3). Postcontrast AKI occurred in 7 patients (2.7%) in the no prehydration group and 4 patients (1.5%) in the prehydration group (RR 1.7; 95% CI 0.5-5.9; p=0.36).

Clinical Takeaway: Administration of iodinated contrast medium (ICM) is not associated with a risk of acute kidney injury in critically patients. Furthermore, the rates of acute kidney injury after ICM administration, even in patients with chronic kidney disease, is low and not reduced by prehydration. In summary, these studies add to the substantial and growing body of evidence refuting the argument that ICM administration should be avoided in patients for fear of renal complications.

Synopsis: Precision medicine is increasingly heralded as the future of improving outcomes for patients across a wide variety of disciplines. Recent advances in prognostic transcriptomics have enabled investigators to identify changes in gene expression that can distinguish and even predict the development of cancer, thrombosis, and sepsis among other conditions. These approaches are anchored by microarray technologies, namely the ability to measure changes in gene expression across 1000’s of genes at once from a sample of tissue or blood. In 2000, several investigators sought to apply this technology to trauma in hopes of better understanding the transcriptional response to injury. This NIH sponsored “Inflammation and the Host Response to Injury” study, otherwise known as the Glue Grant (www.gluegrant.org), was a large scale, multi-institutional undertaking led by several leaders in the field known as the Investigators of the Inflammation and Host Response. The Glue Grant used microarray technologies to identify 2078 genes that were differentially expressed in trauma patients with complicated outcomes compared to those who had a time to recovery (TTR) of less than five days. Of these, 63 genes exhibited a two-fold difference in expression among the patients who had a complicated course. More specifically, up to 70% of the blood leukocyte genome was altered at 28 days after severe blunt trauma, a phenomenon now referred to as “genomic storm.”

The current article from Annals of Surgery seeks to build upon this work by determining if the gene expression changes that were associated with poor outcomes in the Glue Grant can be used to prospectively predict outcomes in severe blunt trauma. To do this, Raymond et al. employed a rapid transcriptomic platform that has already been FDA-approved for predictive genomic analysis in breast cancer. The NanoString nCounter evaluates changes in gene expression based on an interchangeable profile of genes and generates a single value within 24 hours that reflects those changes as a whole. The authors adapted this tool to trauma by querying the 63 genes that most dramatically changed in trauma in order to create a single value that they called the S63 metric. They then measured the S63 metric in 127 patients who suffered severe blunt trauma by examining blood collected upon admission and ~12 and ~24 hours later. Patients were then stratified by their time to recovery in order to determine if the S63 could prospectively predict clinical course in trauma. 

Here the authors found mean S63 levels were increased at 12 or 24 hours in patients with TTR of >5 or >14 days when compared to patients with a short TTR. Further, this S63 metric was better able to prospectively predict which patients would require an ICU greater than 5 days when compared to IL-6 and certain clinical injury indices. They also validated their findings in an existing data set of gene expression for 26 critically ill trauma patients and 6 controls. including evaluating changes in individual genes grouped by role. Interestingly, the authors found that changes in inflammation and endothelial activation improved at 24 hours, but changes in adaptative immunity continued to trend toward an increase in expression in patients who ultimately required more than 14 days to recover, suggesting a possible target for future therapies. Notably there are significant limitations to this study. The authors did not report the baseline S63 metric for any group, limiting the ability to exclude if there were inherent differences among cohorts. In addition, the study included only Caucasian males who suffered blunt trauma, excluding penetrating trauma and even traumatic brain injury. More work is needed to determine if this metric can be applied to other demographics and injury patterns. 

Regardless, this work lays the foundation for using rapid transcriptomics to predict an individual patient’s clinical course in trauma. The authors suggest that early identification of such patients will enable providers to more efficiently allocate resources and support. Regardless, this study suggests that gene expression analyses can be used not only to predict outcomes in trauma but potentially unravel mechanisms driving differences in disease severity and identify modifiable factors to improve patient care. 

Synopsis: Current guidelines support monotherapy with either aspirin or clopidogrel for secondary prevention of cardiovascular adverse events in PAD patients. Recent data from the COMPASS trial (Cardiovascular Outcomes for People Using Anticoagulation Strategies) showed that the addition of low dose Rivaroxaban (2.5 BID) to ASA decreased the incidence of major adverse limb events (and systemic cardiovascular complications) in patients with stable PAD. However, the benefit of anticoagulation after lower extremity revascularizations for symptomatic PAD has not been validated, despite being occasionally used by vascular surgeons for “high-risk” PAD patients.

The VOYAGER PAD trial (The Vascular Outcomes studY of ASA alonG with rivaroxaban in Endovascular or surgical limb Revascularization for Peripheral Artery Disease) was constructed to address whether the benefits of rivaroxaban extended to PAD patients after revascularization. This was a double blinded, multicenter trial that randomized 6,564 symptomatic PAD patients after surgical, open, or hybrid lower extremity infrainguinal revascularization to rivaroxaban 2.5 BID plus ASA vs placebo plus ASA. Critical limb ischemia was the indication for revascularization in 23% and claudication in 77%. Surgical revascularization was the approach in 35% and endovascular or a hybrid approach in 77%. 

The primary efficacy endpoint which was a composite outcome of acute limb ischemia, major amputation, ischemic stroke, myocardial infarction, ischemic stroke, or cardiovascular death was statistically lower in the treatment group at 3 years (estimated incidence 17.3% in the rivaroxaban vs 19.9% in the placebo group, HR 0.85, 95% CI 0.76 to 0.96, P=0.009). The absolute risk reduction was 1.5% at six months, 2.0% at one year and 2.6% at three years. The primary safety endpoint of major bleeding as defined by the Thrombolysis in Myocardial Infarction (TIMI) classification (intracranial bleeding, any bleeding with Hb drop >5, or fatal bleeding) was not significantly different between the 2 groups (2.65% in the rivaroxaban vs 1.87% in the placebo group at 3 years, HR 1.43; 95% CI, 0.97 to 2.10). Using the ISTH (International Society for Thrombosis and Hemostasis) definition, which has less strict criteria for major bleeding, significantly higher major bleeding events occurred in the rivaroxaban group compared to the placebo group (5.94% with rivaroxaban vs 4.06% with placebo (HR, 1.42; 95% CI 1.10 to 1.84; P = 0.007).

Despite the favorable composite outcome with rivaroxaban, digging deeper into the tables reveals that all-cause mortality and cardiovascular mortality was higher in the treatment group (11.1% vs 10.9% and 7.1% vs 6.4% respectively), though not statistically significant, which the authors do not comment about. Furthermore, whether the portrayed benefit described in this study is offset by the bleeding risk is dependent on the definition used for bleeding. More data is needed on the safety and efficacy of rivaroxaban before vascular surgeons commit their patients to lifelong anticoagulation after lower extremity revascularization. Also important to note is that the overwhelming majority of patients underwent intervention for claudication and not for critical limb ischemia, a trend that should invoke further study and scrutiny of interventions offered to this subset of patients.

Introduction: Surgical informed consent is an ethical imperative, legal duty, and professional obligation. The process of informed consent involves an on-going discussion between the patient and surgeon about the nature of the procedure, risks, benefits, alternatives (including no treatment), and recuperative issues. States may mandate additional information be provided. The patient must have an opportunity to answer questions, and a procedure-specific consent form must be signed. Although such informed consent is an aspirational goal, the process in practice often falls short. In the last 25 years, close to 70 randomized controlled trials (RCT) have attempted to improve the nature of informed consent with varying results.

The Recent Paper: Schwarze et al adds to the informed consent literature with a large multicenter randomized controlled trial in which patients and family were sent—in advance of the surgical consultation—a question prompt list (QPL); 11 questions that could be asked of surgeons by patients/family about the surgery (N=223). The control arm was ‘usual care’ (N=223). About half as many family participants were enrolled. Demographics were used only to document the two groups were the same; not as independent variables. The overall conclusions were that the study was null related to the main outcomes of patient engagement and well-being.

However, some additional observations should be made. Even when the analysis was restricted to patients with ‘clear evidence’ of QPL review (not systematic) the same number of questions were asked. Patients who had ‘usual care’ expressed significantly higher support for autonomy in the Health Care Climate Questionnaire ((HCCQ), waived off by the authors as a ‘ceiling effect’). The Measure Yourself Concerns and Wellbeing (MYCaW) scores were higher with the QPL for family members—but not patients—at six weeks. Secondary outcomes included the Patient-reported Outcomes Measurement Information Systems (PROMIS) and the Psychological Illness Impact (PII); scores were better for patients in the ‘usual care ‘group. The Perceived Efficacy in Patient-Physician Interactions (PEPPI) scores were no different. Only the Observer-measured shared decision-making (OPTION5) scoring was higher in the QPL group; but observers were not blinded so this measure is subject to bias. Adding the QPL to the consultation didn’t take more time (as might be expected since the QPL group didn’t ask more questions).

Limitations: Beyond those noted in the paper, limitations included whether surgeons had a set response or even general guidelines to answer the QPL. There was no mention of standardization of the ‘usual care’ consent discussion. There was no documentation of routine assessment if patients/families looked at the QPL. The paper did not discuss the possibility that QPL questions might have been answered ahead of time by patients/families searching the internet (internet savvy was not a demographic). Although regret was not different in the two groups, other work suggests that regrets tend to occur only in the patients with complications and analysis was not restricted to this group. Interestingly, a very similar set of questions are available on the American College of Surgeons website https://www.facs.org/education/patient-education/patient-resources/prepare/10-questions, and it was not made clear why a separate set of questions might be better. 

Summary: A large multicenter randomized clinical trial prompting patients to use a series of questions in an attempt to improve patient engagement and well-being showed null results. However, much could be adapted for future trials. The statistical analysis was well done, and much of the data collected came from validated previously published measures. Clearly, provision of questions relevant to the patients’ surgical procedures did not enhance engagement or well-being. Emphasis on the surgeons with regard to communication may be necessary.