A Multivariable Prediction Model for Invasive Pulmonary Aspergillosis in Immunocompromised Patients with Acute Respiratory Failure (IPA-GRRR-OH Score)

Infectious Diseases Life Sciences Medicine Oncology Immunology
Invasive Pulmonary Aspergillosis Immunocompromised Acute Respiratory Failure Prediction Model Intensive Care

A Prediction Model for Invasive Pulmonary Aspergillosis in Immunocompromised Patients with Acute Respiratory Failure

Background Introduction

Invasive Pulmonary Aspergillosis (IPA) is a severe opportunistic infection commonly seen in immunocompromised patients, particularly those with immune dysfunction caused by hematological malignancies, stem cell transplantation, or long-term use of immunosuppressive agents. The diagnosis of IPA is often difficult, and once patients require mechanical ventilation, the mortality rate can reach up to 90%. Therefore, early diagnosis and timely treatment are crucial for improving patient outcomes. However, current diagnostic methods (such as lung biopsy and bronchoalveolar lavage) have many limitations in clinical practice, especially in critically ill patients, where these invasive procedures may exacerbate the condition. Additionally, the results of biological sample testing often take a long time and cannot provide diagnostic evidence quickly upon patient admission.

To address this issue, researchers have focused on developing a prediction model based on clinical data that can rapidly assess the risk of IPA when patients are admitted to the Intensive Care Unit (ICU), thereby guiding early antifungal therapy and reducing adverse consequences due to delayed diagnosis.

Paper Source

This paper was collaboratively completed by multiple researchers including Alice Friol, Guillaume Dumas, Frédéric Pène, and others from several renowned medical institutions such as Saint-Louis Hospital, Cochin Hospital, and Pitié-Salpêtrière Hospital in Paris, France. The paper was published in the journal Intensive Care Medicine on January 24, 2025, titled “A multivariable prediction model for invasive pulmonary aspergillosis in immunocompromised patients with acute respiratory failure (IPA-GRRR-OH score).”

Research Process

1. Study Design and Dataset

The study was divided into two phases: model development and model validation. The model development phase utilized data from 3,262 immunocompromised patients in the GRRR-OH (Groupe de Recherche Respiratoire en Réanimation Onco-Hématologique) database who were admitted to 12 ICUs between 2001 and 2017 due to acute respiratory failure. The model validation phase used data from another randomized controlled trial (HIGH trial) involving 776 patients.

2. Variable Selection and Model Construction

In the model development phase, the researchers first screened factors related to IPA through univariate analysis, then further identified independent predictors using a multivariate logistic regression model. To reduce bias caused by the low prevalence of IPA, the researchers adopted Firth’s bias-reduced penalized likelihood regression method. The final model included the following eight variables: type of immunosuppression, high-dose or long-term corticosteroid use, neutropenia, structural lung disease, time from symptom onset to ICU admission exceeding seven days, hemoptysis, focal alveolar pattern on chest imaging, and viral co-infection.

3. Scoring System Construction

Based on regression coefficients, the researchers converted these variables into a scoring system called the IPA-GRRR-OH score. The score ranges from -2 to 10 points, with higher scores indicating greater risk of IPA for the patient.

4. Model Validation

In the validation phase, the researchers applied the IPA-GRRR-OH score to an independent validation cohort and evaluated its diagnostic performance. The results showed that the performance of the score in the validation cohort was consistent with that in the development cohort, demonstrating good discrimination and calibration.

Main Results

1. Patient Characteristics and IPA Prevalence

In the development cohort, the prevalence of IPA was 4.5% (146⁄3262), while in the validation cohort, it was 3.3% (26⁄776). Hematological malignancies were the primary cause of immunosuppression, accounting for 51% and 38% of the development and validation cohorts, respectively.

2. Independent Predictors of IPA

Multivariate analysis showed that the following factors were significantly associated with IPA: - Type of immunosuppression (especially acute leukemia and allogeneic stem cell transplantation) - High-dose or long-term corticosteroid use - Neutropenia - Structural lung disease - Time from symptom onset to ICU admission exceeding seven days - Hemoptysis - Focal alveolar pattern on chest imaging - Viral co-infection

3. Diagnostic Performance of the IPA-GRRR-OH Score

In the development cohort, the AUC (area under the curve) of the IPA-GRRR-OH score was 0.72, while in the validation cohort, it was 0.85, indicating that the score had good discriminatory power. The optimal diagnostic cutoff value for the score was 4 points, with a sensitivity of 23.1%, specificity of 90.5%, and negative predictive value of 91.4%.

Conclusions and Implications

1. Scientific Value

This study developed and validated a new clinical scoring system (IPA-GRRR-OH score) that can rapidly assess the risk of IPA in immunocompromised patients with acute respiratory failure. The clinical application of this score helps in early identification of high-risk patients, thereby initiating timely antifungal therapy and reducing adverse consequences due to delayed diagnosis.

2. Application Value

The advantage of the IPA-GRRR-OH score lies in its simplicity and ease of use, requiring only clinical, biological, and imaging data available at the time of ICU admission. Moreover, its high negative predictive value (93.1%) allows it to effectively exclude low-risk patients, avoiding unnecessary invasive examinations and antifungal therapy, thus reducing waste of medical resources and potential toxic reactions in patients.

3. Research Highlights

  • Innovation: This is the first prediction model specifically designed for assessing the risk of IPA in immunocompromised patients with acute respiratory failure, filling a gap in this field.
  • Practicality: The scoring system is based on routine clinical data, making it easy to apply in the ICU.
  • Efficiency: The rapid calculation capability of the score allows it to provide diagnostic guidance immediately upon patient admission, reducing treatment delays.

Other Valuable Information

Although the study performed well in critically ill patients, its application in other clinical settings (such as non-ICU patients) still needs further validation. Additionally, with the widespread use of new immunosuppressive agents (such as BTK inhibitors and CAR-T cell therapies), the epidemiological characteristics of IPA may change; therefore, further research on specific patient groups is also necessary.

The IPA-GRRR-OH score provides an efficient and practical tool for diagnosing IPA in immunocompromised patients with acute respiratory failure, offering significant clinical importance and promotional value.