Pancreatic stone protein predicts postoperative infection in cardiac surgery patients irrespective of cardiopulmonary bypass or surgical technique

Abstract

Introduction: We investigated the role of pancreatic stone protein (PSP) in predicting the occurrence of infection in the postoperative course of cardiac surgery patients. Several biomarkers indicating the presence of inflammation and infection are available in the clinical routine; yet, their utility in the postoperative course of patients following cardiac surgery remains uncertain. Moreover, cardiopulmonary bypass, also referred to as "on-pump surgery", increases the susceptibility to an exaggerated inflammatory state. However, the impact of such extracorporeal circulation on circulating PSP levels remains poorly understood.

Methods: In a prospective cohort of unselected patients undergoing cardiac surgery, we set out to elucidate the diagnostic accuracy of serum PSP levels as opposed to canonical biomarkers (CRP, WBC) of inflammation to discriminate between the presence of infection and surgical trauma,. In addition, we investigated whether the biomarkers were influenced by the surgical technique employed, i.e. on-pump vs. off-pump and minimally invasive surgery vs. sternotomy. Levels of circulating PSP and routine inflammatory biomarkers (CRP, WBC) were measured in samples taken from 120 patients at baseline as well as at postoperative day 1-3.

Results: Univariate analysis showed that among the biomarkers investigated, only PSP levels had discriminatory power to differentiate infection from surgical trauma in the postoperative course of the entire cohort of patients following cardiac surgery. With regard to cardiac surgical interventions, there was no significant association between the absence or presence of extracorporeal circulation and PSP levels. However, there was a significant difference in the slope of the rise of postoperative PSP between minimally invasive surgery as opposed to patients subjected to sternotomy.

Conclusion: In an unselected population of cardiac surgery patients, post-operative serum PSP levels were significantly associated with the presence of infection in both the on-pump and off-pump setting. Of note, the surgical technique employed (sternotomy vs. minimally invasive approach) had a significant impact on postoperative PSP levels.

Conflict of interest statement

Competing Interests: Rolf Graf (RG) is inventor of an assay covered by patent No: EP 2185937 B2 "METHOD FOR ASSAYING SEPSIS IN HUMANS", which is owned by the University of Zurich (Zurich, Switzerland). This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.

Figures

Fig 1. Association of PSP, CRP and WBC levels with age (A), gender (B) and diabetes (C).

Shown are values at baseline (preoperative values). The groups were compared by one-way ANOVA. Asterisks indicate significant differences (*p

Fig 2. Time course of biomarkers (A-C) following cardiac surgery.

All inflammatory markers investigated showed a statistically significant rise in the course of postoperative day 1–3 (p

Fig 3. Discriminatory power of biomarkers for postoperative infection.

PSP levels showed a significantly stronger postoperative rise at day 1–3 in patients receiving a diagnosis of infection during their hospital stay as opposed to patients exhibiting an uneventful course. (p

Fig 4. Time course of biomarkers as related to the absence or presence of cardiopulmonary bypass.

All inflammatory markers investigated showed a steeper rise in patients subjected to cardiopulmonary bypass without reaching statistical significance. (p>0.05 for interaction for all biomarkers, determined by a linear mixed effects regression model).

Fig 5. Time course of biomarkers as related to the surgical technique employed.

PSP levels showed a significantly steeper rise in the postoperative course of patients subjected to sternotomy compared to those receiving minimally invasive surgery (p

Fig 6. Receiver operating characteristics curve to predict infection for PSP (A), CRP (B) and WBC (C) at postoperative 1–3.

A cutoff of 48.1 ng/ml for PSP levels at postoperative day 2 reveals a sensitivity of 64% and a specificity of 70%. PLR (positive likelihood ratio) = 2.1. NLR (negative likelihood ratio) = 0.51.