Predicting Acute Kidney Injury Risk After Cisplatin Treatment: A Large-Scale Multicenter Study

Core Concepts

A novel risk prediction algorithm can accurately forecast the risk of moderate to severe acute kidney injury following the first dose of intravenous cisplatin treatment.

Abstract

This article discusses the development and validation of a risk prediction algorithm to forecast the risk of acute kidney injury (AKI) following the first dose of intravenous cisplatin treatment. The study was conducted across six US cancer centers, involving a large cohort of nearly 25,000 patients.
The key highlights are:

The researchers developed a risk score consisting of nine readily available clinical variables, including age, hypertension, diabetes, blood cell counts, and serum albumin and magnesium levels, to predict the risk of moderate to severe AKI (defined as a twofold or greater increase in serum creatinine or need for kidney replacement therapy within 14 days).

The risk score demonstrated strong predictive performance, with a C-statistic of 0.75 in the derivation cohort and 0.73 in the validation cohort, outperforming previous models.

Patients in the highest risk category had 24-fold higher odds of developing AKI compared to those in the lowest risk category.

Greater severity of AKI was associated with shorter 90-day survival, highlighting the clinical importance of accurately predicting this complication.

The risk prediction tool is available online for patients and providers to determine an individual's risk of cisplatin-associated kidney injury, enabling proactive management strategies such as closer monitoring or consideration of alternative treatments.

The study's strengths include the large, diverse patient population, external validation, and focus on moderate to severe AKI, which is the most clinically relevant form of kidney injury.

Overall, this study provides a robust and practical risk prediction algorithm to help clinicians and patients navigate the risk of cisplatin-induced kidney injury, a critical consideration in the management of cancer patients.

Stats

The incidence of cisplatin-induced acute kidney injury was 5.2% in the derivation cohort and 3.3% in the validation cohort.
Patients in the highest risk category had 24-fold higher odds of developing acute kidney injury in the derivation cohort and close to 18-fold higher odds in the validation cohort compared to those in the lowest risk category.
Greater severity of acute kidney injury was associated with shorter 90-day survival, with an adjusted hazard ratio of 4.63 (95% CI, 3.56-6.02) for stage III acute kidney injury versus no acute kidney injury.

Quotes

"This was truly a Herculean effort that involved physicians, programmers, research coordinators, and patients."
"While this is not the first attempt to devise a risk score, it is by far the biggest."
"The authors did not restrict patients with chronic kidney disease or other significant comorbidities and used the geographic diversity to produce a cohort that has an age, gender, racial, and ethnic distribution, which is more representative of the US than previous, single-center attempts to risk score patients."

Key Insights Distilled From

Can a Risk Score Predict Kidney Injury After Cisplatin?

by Marilynn Lar... at www.medscape.com 05-08-2024

https://www.medscape.com/viewarticle/can-risk-score-predict-kidney-injury-after-cisplatin-2024a10008vh

Deeper Inquiries

How can the risk prediction algorithm be further improved to better account for the dose-dependent nature of cisplatin-induced kidney injury?

To better account for the dose-dependent nature of cisplatin-induced kidney injury, the risk prediction algorithm can be further improved by incorporating the cumulative dose of cisplatin received by the patient. Currently, the algorithm focuses on predicting acute kidney injury (AKI) after the first dose of cisplatin. Including information on the total cumulative dose of cisplatin over the course of treatment can provide a more comprehensive understanding of the patient's risk for kidney injury. Additionally, considering the frequency and interval of cisplatin administration in relation to the total dose can also enhance the algorithm's accuracy in predicting dose-dependent kidney injury. By integrating these factors into the risk prediction model, clinicians can have a more nuanced assessment of the patient's risk profile and tailor monitoring and preventive strategies accordingly.

What are the potential barriers to the widespread adoption and implementation of this risk prediction tool in clinical practice?

Despite the potential benefits of the risk prediction tool for cisplatin-induced kidney injury, several barriers may hinder its widespread adoption and implementation in clinical practice. One significant barrier is the resistance to change or reluctance among healthcare providers to incorporate new tools or algorithms into their existing workflow. Clinicians may be accustomed to their current practices and may be hesitant to adopt a new risk prediction tool, especially if it requires additional time or resources.
Another barrier could be the complexity of the risk prediction algorithm itself. If the tool is not user-friendly or requires extensive training to interpret the results, clinicians may be less inclined to use it routinely. Moreover, issues related to data availability and interoperability within electronic health record systems could pose challenges to the seamless integration of the risk prediction tool into clinical workflows.
Furthermore, the need for validation and external verification of the risk prediction tool across diverse patient populations and healthcare settings is crucial for its acceptance and adoption. Without robust validation studies and real-world evidence supporting its efficacy, clinicians may be skeptical about relying on the tool for making clinical decisions.

How can the insights from this study be leveraged to develop similar risk prediction models for other nephrotoxic chemotherapies or medications?

The insights gained from this study on cisplatin-induced kidney injury can serve as a valuable foundation for developing similar risk prediction models for other nephrotoxic chemotherapies or medications. One approach to leveraging these insights is to identify common risk factors and predictors of nephrotoxicity across different drugs and treatment regimens. By analyzing the data on patient characteristics, comorbidities, laboratory values, and treatment parameters that contribute to kidney injury, researchers can establish a framework for building risk prediction models specific to other nephrotoxic agents.
Additionally, conducting multicenter studies with large patient cohorts, similar to the one conducted for cisplatin, can help validate the predictive accuracy and generalizability of the risk prediction models for other nephrotoxic chemotherapies. Collaborating with oncologists, nephrologists, and pharmacologists to gather comprehensive data on patients receiving different nephrotoxic medications can enhance the development and refinement of these predictive models.
Furthermore, incorporating feedback from patients and healthcare providers, as done in the cisplatin study, can ensure that the risk prediction models are clinically relevant, user-friendly, and aligned with the needs of the end-users. By applying the methodologies and learnings from the cisplatin risk prediction algorithm to other nephrotoxic agents, researchers can advance personalized medicine approaches and improve patient outcomes in oncology and nephrology settings.