Association of Preoperative Platelet-Activating Factor and Postoperative C-Reactive Protein with Inflammatory Burden and Early Outcomes After Major Cardiac Surgery

Comments 1: The findings that preoperative PAF and postoperative CRP levels were not predictive of clinical outcomes (length of ICU stay and mortality) after open heart surgeries would indicate that the clinical impact of the study is limited.

Response 1: We appreciate the reviewer’s perspective. We have revised the manuscript to more explicitly define this study as exploratory in nature. Our primary goal was to explore the role of Platelet-activating factor (PAF) within the inflammatory cascade and translate these pathophysiological findings into the clinical setting of major cardiac surgery. We acknowledge that the current sample size is insufficient to provide the statistical power necessary to identify independent correlations with clinical outcomes such as acute kidney injury (AKI), ICU length of stay, and mortality, as outcomes in major cardiac surgery are highly dependent on a multitude of factors, including specific cardiac pathologies, operative variables, and baseline patient reserves.

Updated text:

Introduction:

“Given the profound complexity of the systemic inflammatory response following CPB, this study was designed as an exploratory investigation to translate established pathophysiological reasoning regarding the role of platelet-activating factor (PAF) in the inflammatory cascade into the clinical context of major cardiac surgery. We hypothesized that elevated preoperative PAF identifies a biological phenotype predisposed to an exaggerated postoperative inflammatory response and subsequent hemodynamic in-stability, independent of operative complexity or baseline cardiac function.

Our primary aim is to explore whether identifying these upstream mediators, such as PAF, can help clinicians anticipate the physiological needs following major cardiac surgery. Consequently, the objectives of this study were to: (1) Examine the association between preoperative PAF and postoperative CRP as a measure of inflammatory burden; (2) Evaluate the relationship between PAF and postoperative hemodynamic support; (3) Assess the prognostic utility of PAF and CRP for adverse postoperative outcomes, including AKI and mortality, in patients undergoing major cardiac surgery with CPB”

Discussion:

4.6. Clinical and Pathophysiological Implications

“The present findings suggest that PAF and CRP should not be interpreted as stand-alone prognostic biomarkers but rather as indicators of an underlying biological vulnerability characterized by heightened inflammatory and vascular responsiveness to surgical stress. Our findings demonstrate that higher baseline PAF levels correlate with increased vasoactive–inotropic requirements during the critical intermediate postoperative period. This association follows a clear pathophysiological reasoning: PAF is known to promote endothelial dysfunction and vascular hyporesponsiveness—processes that are central to the development of CPB-associated vasoplegia.

Furthermore, we acknowledge that the field of major cardiac surgery is highly dependent on a multitude of key factors, including specific cardiac pathologies, baseline patient reserve, and operative variables such as bypass duration. In such a heterogeneous environment, a single biomarker is unlikely to act as a stand-alone predictor of all adverse outcomes. However, from a pathophysiological standpoint, these results support a growing body of evidence implicating platelet-mediated inflammatory pathways in cardiovascular vulnerability and perioperative organ dysfunction [49–53]. By confirming that PAF’s role in the inflammatory cascade translates into increased clinical support requirements, this study provides the groundwork for personalized risk stratification and may help identify patients who would benefit from hypothesis-driven interventional studies. Ultimately, understanding these mechanistic pathways beyond conventional demographic or procedural variables, is a necessary step toward future therapeutic interventions designed to modulate the inflammatory burden of CPB.”

4.7. Limitations

“Several limitations to this study must be acknowledged. First, the retrospective, single-center design limits causal inference and generalizability. As an exploratory retrospective analysis, the sample size of 87 patients was insufficient to provide the statistical power required to identify independent correlations with outcomes such as acute kidney injury, ICU length of stay, or mortality. These clinical endpoints are influenced by a complex interplay of surgical, anesthetic, and institutional factors that may obscure the independent contribution of any single inflammatory mediator. Furthermore, biomarkers were measured at single time points, precluding the assessment of dynamic perioperative changes, and residual confounding cannot be excluded despite multivariable adjustment.

Despite these limitations, the strength of the association between preoperative PAF and the subsequent inflammatory and hemodynamic response provides a clear signal that this pathway warrants further, larger-scale prospective investigation to fully define its prognostic and therapeutic potential.”

Comment 2: In each individual patient, how was the correlation between PAF and CRP? This paired analysis would provide a better correlation between the two biomarkers.

Response 2: We thank the reviewer for this important clarification. We have updated the manuscript to explicitly state that the correlation and analysis was performed using paired individual patient data. Each data point in the scatter plot (Figure 1) represents the preoperative PAF level and the corresponding postoperative peak CRP concentration for a specific patient, demonstrating that preoperative PAF is a strong predictor of the subsequent inflammatory response

Updated text:

Results section

3.2. Association between PAF and Postoperative inflammation

To evaluate the relationship between biomarkers at the individual level, a paired analysis was conducted using each patient’s preoperative PAF and their corresponding peak postoperative CRP concentrations. Preoperative PAF levels were strongly and positively associated with postoperative CRP concentrations (Figure 1). In multivariable linear regression analysis utilizing individual-patient data, with CRP as a continuous outcome, PAF remained independently associated with postoperative inflammation after adjustment for age, body mass index, and CPB duration. Specifically, each 1000 pg/ml increase in preoperative PAF was associated with an adjusted increase of approximately 36 mg/L in postoperative CRP (β = 0.036, 95% CI 0.028–0.045; p < 0.001).

Comment 3: Two-thirds (2/3) of the surgeries were surgeries involving cardiotomies (i.e., opening the myocardium---indicating direct injuries to the myocardium). Sub-analysis of these patients would be helpful as well.

Response 3: We sincerely thank the reviewer for this insightful observation regarding the physiological impact of cardiotomies. As this is an exploratory study, our primary goal was to capture the broad biological signal of the 'vulnerable phenotype' across the shared inflammatory stimulus of cardiopulmonary bypass (CPB). All procedures in this study were performed using CPB, and off-pump procedures were strictly excluded. While we acknowledge that cardiotomy involves direct myocardial injury, the systemic response in these patients is primarily driven by blood–surface contact within the extracorporeal circuit and ischemia–reperfusion injury common to all on-pump surgeries. In this context, we believe our paired individual-level analysis (Figure 1) effectively demonstrates that preoperative PAF identifies a consistent susceptibility to this shared surgical stress. We focused this exploratory work on the pathway of inflammation triggered by the CPB circuit. Given the modest sample size of our cohort (n=87), we were concerned that further partitioning the data into smaller subgroups for sub-analysis might significantly reduce statistical power and increase the risk of Type II errors. However, we truly value your suggestion, as it helps define new objectives of our current findings and paves the way for the design of future works for our collective.

Updated text:

Methods:
“Inclusion criteria were age ≥18 years, performance of open cardiac surgery requiring CPB, and availability of complete perioperative and laboratory data in the institutional registry. Exclusion criteria included emergency, minimally invasive and off-pump surgeries, transcatheter procedures, and incomplete clinical or laboratory records.”

Discussion:

While a significant portion of our cohort underwent procedures involving cardiotomies, the predictive value of preoperative PAF appears robust across these different surgical techniques, likely due to this dominant and shared inflammatory driver of CPB.

Comment 4: The sample sizes in some groups of patients were very small. For example, only 3 in CPR 0-10 mg/L group and 6 in CRP >200mg/L group. This would introduce bias into the study. I would recommend increasing the number of study patients.

Response 4: We thank the reviewer for this observation. We acknowledge that the groups at the extremes of the inflammatory spectrum (CRP 0–10 mg/L and CRP >200 mg/L) are small.

However, our core findings—including the correlation analyses (Figure 1) and the multivariable regression models—were performed using the entire cohort of 87 patients as continuous variables, minimizing the risk of bias and ensuring that our conclusions were supported by the total sample size. As this is a retrospective exploratory study based on a specific clinical timeframe (2021–2022), our sample size is limited by the available historical data at our center. While we agree that a larger, multicenter cohort would strengthen the generalizability of these findings, we believe our current continuous analysis provides a robust initial signal. We have updated the Limitations section to more clearly address the uneven distribution across CRP groups and the need for larger prospective studies.

Updated text:

Limitations:

The small subgroup sizes limit the interpretability of categorical comparisons, which we addressed by utilizing multivariable regression analyses for the entire cohort.

Comments 1:

Major Required Revisions

Language and grammar require thorough revision. The manuscript contains numerous errors, awkward phrasings, inconsistent hyphenation, and spacing mistakes. Please have the text edited by a native English speaker or a professional service. Examples include:

Inconsistent use of spaces around hyphens and parentheses (e.g., “vasoactive- inotropic” → “vasoactive-inotropic”)

Inconsistent spelling of institution names (Hatieganu vs. Hațieganu)

Awkward phrasing such as “the analysis used anonymous clinical data” (redundant)

Response 1: We thank the reviewer for the insightful suggestions, which have helped improve the clarity and completeness of the manuscript. We performed a comprehensive language revision of the manuscript.

Comment 2. Methods section is incomplete. Please add the following:

Manufacturer, catalog number, and assay characteristics (sensitivity, CV) for the PAF ELISA kit.

Response 2: We have added the details in the Methods Section 2.3 

Updated text: 2.3. Biomarker Measurements

“Plasma was separated by centrifugation and stored according to manufacturer recommendations until analysis. PAF concentrations were measured using a commercially available competitive enzyme-linked immunosorbent assay (ELISA) kit (St John’s Laboratory Ltd, London, UK; Catalog #STJE0011249), with colorimetric detection; optical density was inversely proportional to PAF concentration. The assay demonstrated a sensitivity of 224 pg/ml and a detection range of 468-30,000 pg/ml, with intra- and inter- assay coefficients of variation (CV) <10%.”

Comment 3. Exact timing of postoperative CRP measurement (postoperative day 1, 2, or peak?).

Response 3: Postoperative CRP represents the peak value recorded during the early postoperative period (first 48–72 hours).

Updated text: 2.3. Biomarker Measurements

Postoperative C-reactive protein (CRP) concentrations were obtained from routine laboratory measurements, with the reported value representing the peak concentration during the early postoperative period (48-72 hours).

Comment 4: Definition and calculation of the vasoactive-inotropic score (VIS), including which drugs, doses, and the specific time point (“intermediate postoperative” is vague).

Response 4: We expanded Methods Section 2.4 to include full formula for VIS calculation, which was measured at 24 hours postoperatively.

Updated text: 2.4. Outcome Definitions

“Postoperative hemodynamic support was quantified using the maximum vasoactive–inotropic score (VIS), calculated using the highest doses of vasoactive and inotropic medications administered during the first 24 hours after ICU admission.

VIS = dopamine (μg/kg/min) + dobutamine (μg/kg/min) + 100 × epineph-rine (μg/kg/min) + 10 × milrinone (μg/kg/min) + 10,000 × vasopressin (U/kg/min) + 100 × norepinephrine (μg/kg/min) + 50 × levosimendan (μg/kg/min)”

Comment 5. Sample size justification or power calculation. With only 87 patients and low event rates, the multivariable models are likely underpowered. Please report the number of events (e.g., AKI cases, deaths) per variable.

Response 5: This was an exploratory, hypothesis-generating study. Therefore no formal a priori power calculation was performed.

Number of AKI cases was added in the revised manuscript and number of deaths was also added in the revised manuscript.

Updated text:

2.5. Statistical Analysis

“Statistical analysis was performed using SPSS software version 23.0 (IBM Corp., Chicago, IL, USA). Given the exploratory and hypothesis-generating nature of this study, a formal a priori power calculation was not performed.”

3.4. AKI and Mortality Analyses

“During the study period, 19 cases of AKI (16.53%) and 5 in-hospital deaths (incidence 4,35%) were recorded.”

4.7. Limitations

“Several limitations to this study must be acknowledged. First, the retrospective, single-center design limits causal inference and generalizability. As an exploratory retrospective analysis, the sample size of 87 patients was insufficient to provide the statistical power required to identify independent correlations with outcomes such as acute kidney injury, ICU length of stay, or mortality. The small subgroup sizes limit the interpretability of categorical comparisons, which we addressed by utilizing multivariable regression analyses for the entire cohort. These clinical endpoints are influenced by a complex interplay of surgical, anesthetic, and institutional factors that may obscure the independent contribution of any single inflammatory mediator.”

Comment 6: Reporting of regression coefficients is confusing. In the abstract and results, you state that each 1000 pg/mL increase in PAF is associated with a 36 mg/L increase in CRP, with β = 0.036. This is mathematically correct (0.036 mg/L per 1 pg/mL × 1000 = 36 mg/L), but the presentation is misleading to readers. Please either:

Report the β per 100 pg/mL or per 1000 pg/mL directly, or

Clearly state in the text: “β = 0.036 mg/L increase in CRP per 1 pg/mL increase in PAF (equivalent to 36 mg/L per 1000 pg/mL).”

Response 6: We agree. β = 0.036 mg/L increase in CRP per 1 pg/mL increase in PAF (equivalent to 36 mg/L per 1000 pg/mL).

Updated text:

Abstract - Results:

“Preoperative PAF levels increased progressively across postoperative CRP strata (p < 0.001) and were strongly associated with postoperative CRP concentrations in both univariate and multivariable analyses. β=0.036 mg/L increase in CRP per 1 pg/mL in-crease in PAF (equivalent to approximately 36 mg/L per 1000 pg/mL). Each 1000 pg/mL increase in preoperative PAF was associated with an adjusted increase of approximately 36 mg/L in postoperative CRP. Elevated PAF was also associated with increased inter-mediate postoperative vasoactive–inotropic requirements and modest prolongation of hospital length of stay (r=0.25, p=0.023). However, neither PAF nor CRP independently predicted AKI or mortality after adjustment for clinical variables. Discriminatory per-formance for mortality was modest for both biomarkers.”

Results – 3.2. Association between PAF and postoperative inflammation

“Specifically, β=0.036 mg/L increase in CRP per 1 pg/mL increase in PAF (equivalent to an adjusted increase of approximately 36 mg/L per 1000 pg/mL; 95% CI 0.028–0.045; p<0.001).”

Comment 7: Table 2 has issues:

The odds ratio for CRP (0.99, 95% CI 0.98–1.01) suggests a (non‑significant) protective effect, which is biologically implausible. Please comment on this or re‑examine the model.

The formatting of the confidence interval for PAF is messy (“1.0005-0.999–1.001”). Correct this.

Response 7: We agree this may appear counterintuitive. The OR reflects adjusted multivariable modeling. The effect is not statistically significant, possibly due to limited sample size.

We corrected the confidence interval to standard format.

Updated text:

Discussion – 4.5. Acute Kidney Injury and Mortality

“Although both PAF and CRP demonstrated univariate associations with adverse postoperative outcomes, neither biomarker remained independently associated with AKI  or mortality after multivariable adjustment, as the OR for CRP (0.99, 95% CI 0.99-1.01) was not statistically significant (p=0.32)."

Comment 8: Discussion contains redundancy. Several sentences repeat results verbatim (e.g., the β coefficient, the r = 0.25 for VIS). Please condense and focus on interpretation and comparison with prior literature.

Response 8: We have revised the Discussion to remove repeated reporting of β coefficients and correlation values and focus more on interpretation and comparison with prior literature.

Updated text:

Discussion – 4.1. Principal findings

“In this retrospective observational study of patients undergoing major cardiac surgery with CPB, we demonstrate a strong and consistent association between preoperative platelet-activating factor (PAF) levels and the magnitude of the postoperative inflammatory response, as reflected by C-reactive protein (CRP). Our results indicate that preoperative PAF is independently associated with CRP concentrations even after adjustment for relevant clinical variables and that PAF was also associated with increased postoperative vasoactive–inotropic requirements (VIS) and a modest increase in total hospital length of stay. These findings suggest that preoperative platelet-mediated inflammatory activation identifies a biologically vulnerable phenotype characterized by exaggerated inflammatory and vascular responses to surgical stress, rather than functioning as a stand-alone prognostic marker for adverse postoperative outcomes.

In contrast, neither PAF nor CRP independently predicted AKI or mortality after multivariable adjustment, and their prognostic discrimination for mortality was modest.”

Discussion – 4.6. Clinical and Pathophysiological Implications

“The present findings suggest that PAF and CRP should not be interpreted as stand-alone prognostic biomarkers but rather as indicators of an underlying biological vulnerability characterized by heightened inflammatory and vascular responsiveness to surgical stress. Our findings demonstrate that higher baseline PAF levels correlate with increased vasoactive–inotropic requirements during the critical intermediate postopera-tive period. This association follows a clear pathophysiological reasoning: PAF is known to promote endothelial dysfunction and vascular hyporesponsiveness—processes that are central to the development of CPB-associated vasoplegia.

Furthermore, we acknowledge that the field of major cardiac surgery is highly dependent on a multitude of key factors, including specific cardiac pathologies, baseline patient reserve, and operative variables such as bypass duration. In such a heterogeneous environment, a single biomarker is unlikely to act as a stand-alone predictor of all adverse outcomes. However, from a pathophysiological standpoint, these results support a growing body of evidence implicating platelet-mediated inflammatory pathways in cardiovascular vulnerability and perioperative organ dysfunction [49–53]. By confirming that PAF’s role in the inflammatory cascade translates into increased clinical support requirements, this study provides the groundwork for personalized risk stratification and may help identify patients who would benefit from hypothesis-driven interventional studies. Ultimately, understanding these mechanistic pathways beyond conventional demographic or procedural variables, is a necessary step toward future therapeutic in-terventions designed to modulate the inflammatory burden of CPB.”

Comment 9: Reference list needs cleanup:

Reference 47 is broken (two references merged: Ribeiro et al. and Costa et al.). Separate them.

Standardize journal name abbreviations and add DOIs where possible.

Response 9: We have corrected references.

Comment 10: Minor Suggestions

In the abstract, the phrase “postoperative vasoactive- inotropic requirements” should be “postoperative vasoactive-inotropic requirements” (remove the space).

On page 5, “with a trend toward significance (p = 0.060).” – the period after the parentheses should be removed or placed correctly.

Consider shortening the title slightly; it is currently very long.

The phrase “biologically vulnerable phenotype” is used repeatedly. While acceptable, overuse weakens its impact.

Response 10: All minor issues have been corrected

Comment 11: Specific Questions for the Authors

What was the rationale for choosing the PAF cutoff of >3033 pg/mL? Was this derived from the data (e.g., median, ROC‑optimized) or from prior literature? Please state this clearly.

Response 11: The cutoff was data-driven, derived from receiver operating characteristic (ROC) analysis

Comment 12: How many patients had acute kidney injury and how many died? These numbers are not provided in the results section. Please add them.

Response 12: We have added AKI cases =19 and deaths =5 in the Results section

Updated text:

3.4. AKI and Mortality Analyses

“During the study period, 19 cases of AKI (16.53%) and 5 in-hospital deaths (incidence 4.35%) were recorded. In multivariable logistic regression analysis for AKI, age and body mass index were independently associated with AKI risk, whereas neither PAF nor CRP remained significant predictors after adjustment.”

Comment 13: Were there any patients with missing data for key variables? If so, how was this handled?

Response 13: We have clarified that there were no missing data for primary variables (PAF, CRP, key outcomes)

Comment 14: Conclusion

The core scientific question is relevant, and the primary association (PAF → CRP) appears robust. However, the manuscript cannot be accepted in its current form due to the missing figure, incomplete methods, language problems, and reporting inconsistencies. I recommend major revision. After the authors address the above points, the paper may become suitable for publication.

Response 14: We have thoroughly revised the manuscript to address all methodological, statistical, and language concerns. We believe these changes have significantly improved clarity and transparency.

Comment 15: Comments on the Quality of English Language

The English language quality of the manuscript requires substantial improvement. While the scientific content is understandable, numerous grammatical errors, awkward phrasings, inconsistent punctuation, and stylistic issues detract from clarity and professionalism. Below are specific examples and general recommendations.

Specific Issues Observed

Inconsistent hyphenation and spacing

“vasoactive- inotropic” (abstract, page 1) → should be “vasoactive-inotropic” (no space after the hyphen)

“blood- surface contact” (page 2) → should be “blood–surface contact” or “blood-surface contact”

“pre- existing” (page 7) → either “preexisting” or “pre‑existing” (consistent form)

Spaces before units: “1000pg / ml” (page 1) → “1000 pg/mL” (standard format)

Awkward or unclear phrasing

“The analysis used anonymous clinical data” (page 9) – redundant; better: “de‑identified clinical data” or simply remove “anonymous”

“the analysis used anonymous clinical data” – the subject “analysis” does not “use” data; rewrite as “the study used anonymized clinical data”

“the upstream determinants of interindividual variability in inflammatory burden remain incompletely understood” (abstract) – wordy; consider “...are not fully understood”

“a biologically vulnerable phenotype characterized by exaggerated inflammatory and vascular responses” (abstract and discussion) – repeated verbatim multiple times; this level of repetition is unnecessary

Typographical and punctuation errors

“(r=0.25(r=0.25 p=0.023)p=0.023)” (abstract) – spacing inside parentheses is incorrect; should be “(r = 0.25, p = 0.023)”

“(p=0.060)(p=0.060) .” (page 5) – period placed after a space; should be “(p = 0.060).”

Missing commas in compound sentences (e.g., page 2: “PAF is a potent phospholipid mediator with well- established roles...” – the hyphen after “well” is incorrect; should be “well‑established” without a space)

Verb tense consistency

In the abstract: “We conducted a retrospective observational study” (past tense) followed by “Preoperative plasma PAF levels... were measured” (past) – fine. But later: “Each increase... was associated” – also fine. However, in the discussion, the authors switch inconsistently between present and past when describing their own findings (e.g., “we demonstrate” vs. “we demonstrated”). Please review and standardize.

Word choice / vocabulary

“modest prolongation of hospital length of stay” (abstract) – “prolongation” is acceptable but “increase” is more direct.

“discriminatory performance for mortality was modest” – “discriminatory” is fine, but “discriminative” is more common in biostatistics.

“standalone prognostic biomarkers” (page 9) – “stand‑alone” is preferred in academic writing.

Redundancy

The phrase “identifies a biologically vulnerable phenotype” appears in the abstract, discussion, and conclusions almost word‑for‑word. Please rephrase or reduce repetition. The results regarding the β coefficient (36 mg/L per 1000 pg/mL) are repeated in the abstract, results, and discussion. Consolidate.

Response 15: The manuscript has been professionally proofread by a native English speaker, and the necessary corrections have been made.

Comment 16: Overall Recommendations

Copyediting by a native English speaker or professional service is strongly recommended. The manuscript is readable but contains enough errors to distract reviewers and readers.

Pay special attention to punctuation around parentheses and spacing before/after mathematical symbols.

Use a consistent style for units (space before unit, e.g., “10 mg/L”, not “10mg/L”).

Avoid repeating the same phrases across multiple sections.

With thorough language revision, the clarity of this otherwise interesting study will be greatly enhanced.

Response 16: We appreciate the reviewer's wise recommendations, which have enhanced the manuscript's comprehensiveness and clarity. We thoroughly revised the manuscript's language.

Comments 1:

There are some questions that arise when reading this work.

In Table 1, where the characteristics of the groups are given, it is advisable to specify in which group which surgical interventions were performed. The materials and methods indicate only the total number of patients

Response 1: We thank the reviewer for the positive assessment of our work and for the insightful suggestions, which have helped improve the clarity and completeness of the manuscript. We address each comment below.

We have expanded Table 1 to include the distribution of surgical procedures (CABG, valve surgery, aortic surgery, combined procedures) within each CRP group. This allows readers to better assess whether differences in inflammatory response could be influenced by procedural heterogeneity.

Comment 2. What was the postoperative LVEF? Were there any differences in this indicator? It would be interesting to evaluate the difference between the initial and postoperative L It is possible that there may also be interesting correlations of PAF and CRP with this parameter.

Response 2: Postoperative LVEF measurements were not systematically available for all patients, as echocardiographic reassessment was performed selectively based on clinical indication rather than as part of a standardized protocol. We have acknowledged this limitation in the revised manuscript

Updated text: 4.7. Limitations

“Furthermore, biomarkers and echocardiography parameters were measured at single time points, precluding the assessment of dynamic perioperative changes, and residual confounding cannot be excluded despite multivariable adjustment.”

Comment 3. There is a lack of an outcome table where 7 columns could be made: the type of outcome, the total number, regardless of the groups, columns for each group, and the p-value.

Response 3: We have added a new table 3 summarizing postoperative outcomes, including: Outcome type (e.g., AKI, mortality, stroke, arrhythmias, etc.); Total incidence in the cohort; Distribution across CRP strata; Corresponding p-values

Updated text:

3.5. Postoperative Outcomes and Complications

“Postoperative outcomes and complications across the C-reactive protein (CRP) strata are summarized in Table 3. While individual event rates for major complications such as mortality, stroke, and acute kidney injury (AKI) were relatively low in this cohort, the distribution of these outcomes demonstrated no significant differences between the CRP groups (all p>0.05). The incidence of hemodynamic instability requiring support was assessed via the VIS as detailed in previous sections.

Table 3. Postoperative outcomes and complications stratified by peak postoperative C-reactive protein (CRP) levels.”

Comments 1: Regression coefficient presentation (critical): The reporting of the association between PAF and CRP remains confusing and mathematically inconsistent. In the Results section, the authors state: "*β = 0.036 mg/L increase in CRP per 1 pg/mL increase in PAF (equivalent to an adjusted increase of approximately 36 mg/L per 1000 pg/mL)*" and then separately write "*1000 pg/mL increase in preoperative PAF was associated with an adjusted increase of approximately 36 mg/L in postoperative CRP (β = 0.036...)*". This is incorrect because a β of 0.036 cannot simultaneously represent a 1 pg/mL and a 1000 pg/mL change. Please rescale the predictor (e.g., express PAF per 100 pg/mL or per 1000 pg/mL and report the corresponding β) OR clearly state in the text that the β = 0.036 is per 1 pg/mL and provide the 1000 pg/mL effect as a separate calculation without labeling it as β.

Response 1: We apologize for the inconsistent reporting of the regression coefficients. We have corrected the PAF predictor to units of 1000 pg/mL throughout the manuscript.

Updated text:

Abstract:
“Specifically, each 1000 pg/mL increase in preoperative PAF was associated with an ad-justed increase of 36.0 mg/L in postoperative CRP (β=36.0; p<0.001).”

Results 3.2.:

“Specifically, for every 1000 pg/mL increase in PAF, there was an adjusted increase of 36.0 mg/L in CRP (β=36.0; 95% CI 28.0–45.0; p<0.001).”

Table 2 was also adjusted.

Comment 2. Table 3 typo: The number of patients in the "CRP 40-200 mg/L" column is listed as n=159. The total cohort has only 87 patients. This should almost certainly be n=59 (consistent with Table 1). Please correct.

Response 2: Thank you for this comment. We corrected it.

Updated text: -

Comment 3. Reference 47 is broken: The reference list incorrectly merges two separate citations. "Ribeiro A, et al. ... Prolonged Systemic Inflammatory Response Syndrome After Cardiac Surgery. J Cardiothorac Vasc Anesth. 2024..." and "Costa J, et al. Inflammation and outcomes after cardiac surgery. Heart. 2017;103(24):1942-1948" appear as one entry. Please separate them into two distinct numbered references.

Response 3: Thank you for your observation. We changed it.

Updated text:

Comment 4: Language and typographical errors: Please correct the following:

Page 9, line 277: "plate-lect-activating factor" → "platelet-activating factor"

Page 7, Figure 2 legend: "VIS1" is used but the Methods only define "VIS". Please harmonize.

Page 8, Table 3 title: "CRP mg/L (n=3)" is missing the "0-10" range. Should read "CRP 0-10 mg/L (n=3)".

Response 4: We corrected the recommended changes.

Comment 5. Multiple comparisons: The authors test several outcomes (CRP, VIS, ventilation, hospital stay, AKI, mortality) without adjustment for multiple comparisons. Please add a brief sentence to the Statistical Analysis or Limitations section acknowledging that secondary analyses are exploratory and that the risk of type I error should be considered.

Response 5: Thank you for your comment. We have added a statement to the Statistical Analysis and Limitations sections to acknowledge this and to ensure the results are interpreted with the appropriate level of caution.

Updated text:

2.5. Statistical Analysis

“Given the exploratory and hypothesis-generating nature of these secondary analyses, formal adjustments for multiple comparisons were not performed; therefore, results should be interpreted with consideration of the potential risk for Type I error.”

4.7. Limitations

“Furthermore, because no formal adjustments were made for multiple comparisons across these secondary clinical endpoints, the findings should be viewed as hypothe-sis generating, and the risk of Type I error must be considered when interpreting the strength of these associations.”

Comment 6: Events per variable: For the multivariable logistic regression of AKI (19 events) and mortality (6 events), please report the number of events per variable in the model and explicitly acknowledge that the low event rates limit statistical power and may lead to unstable estimates.

Response 6: We thank the reviewer for the comment. We addressed this topic in the Results section and we acknowledged the mentioned risk in the limitations section.

Updated text:

3.4. AKI and Mortality Analyses

“For the AKI multivariable model, including five predictor variables (PAF, CRP, age, BMI, and CPB duration), the ratio of events per variable was 3.8.”

4.7. Limitations

“The low number of events relative to the number of variables limits the statistical power and may lead to unstable estimates.”

Comment 7: Discussion redundancy: The statement that "PAF and CRP should not be interpreted as stand-alone prognostic biomarkers" appears twice within two paragraphs (pages 9-10). Please remove the duplicate.

Response 7: We thank the reviewer for pointing out this mistake. The duplicate was removed.

Updated text: -

Comment 8: Over-interpretation: The repeated phrase "identifies a biologically vulnerable phenotype" is strong for a hypothesis-generating retrospective study. Please soften this to something like "may help identify a biologically vulnerable phenotype" or "suggests the presence of a vulnerable phenotype."

Response 8: We thank the reviewer for this important point regarding the interpretation of our data. We agree that as a retrospective, hypothesis-generating study, the language should reflect a more cautious interpretation of the biological patterns observed.

Updated text:

Abstract:

“Although PAF and CRP were not independent predictors of adverse outcomes, they may help identifiy a biologically vulnerable phenotype characterized by exaggerated in-flammatory and vascular responses to surgical stress.”

Section 4.1. Discussion – Principal Findings

“These findings suggest that preoperative platelet-mediated inflammatory activation may identifiy a biological vulnerability characterized by exaggerated inflammatory and vascular responses to surgical stress, rather than functioning as a stand-alone prognostic marker for adverse postoperative outcomes.”

Section 4.6. Discussion – Clinical and Pathophysiological Implications

“The present findings suggest that PAF and CRP should not be interpreted as stand-alone prognostic biomarkers but rather as indicators of a potential underlying biological vul-nerability characterized by heightened inflammatory and vascular responsiveness to surgical stress.”

Conclusions

“While PAF and CRP are not independent predictors of adverse outcomes, they may help identify a biologically vulnerable phenotype characterized by exaggerated inflammatory and vascular responses to surgical stress.

Comment 9: Consider adding a post-hoc power analysis or sample size justification to the Limitations section, acknowledging that the study was underpowered for rare outcomes such as mortality.

Response 9: We thank the reviewer for this suggestion. We acknowledge that the low number of rare events, limits the statistical power of the study to detect significant associations for these specific endpoints.

Updated text:

“As an exploratory retrospective analysis, the sample size of 87 patients was insufficient to provide the statistical power required to identify independent correlations with outcomes such as acute kidney injury, ICU length of stay, or mortality, meaning that the study was statistically underpowered to detect anything but very large effect sizes for these rare outcomes.”

Comment 10: The English language quality of the manuscript requires substantial improvement. While the scientific content is understandable, numerous grammatical errors, awkward phrasings, inconsistent punctuation, and stylistic issues detract from clarity and professionalism. Below are specific examples and general recommendations.

Response 10: Thank you for your comment. The text has been thoroughly edited to correct grammatical errors, refine awkward phrasing, and standardize punctuation and medical terminology.

Updated text: -