Acute Kidney Injury in Hospitalized Cancer Patients: Single-Centre Real-Life Analysis of Incidence and Clinical Impact

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the possibility to review the manuscript titled: “Acute Kidney injury in hospitalized cancer patients: single-centre real-life analysis of incidence and clinical impact.” This is an interesting and large study of 56,390 hospitalized patients, 6,723 (11.9%) had cancer diagnosis that evaluated AKI in cancer patients versus non-cancer patients. There are several points to consider in order to improve the manuscript:

-The discussion section is rather short and does not fully analyse the results of the study and limitations.

1) Diabetes is a general term and description of the comorbidity. Patients with diabetes may have well compensated disease with low Hb1Ac and overall good health, while others may have poor serum glucose management and multiple comorbidities. I think this should be mentioned in the discussion section that it was not possible to included a broader array of laboratory values

2) metastatic solid cancers is also a general term as metastases can be in different organs and tissues. Metastasis in the liver or kidneys significantly differs from a metastasis to spleen or other organs

I think these points should be highlighted in the discussion section and in the limitations of the study.

Please take into account the recommendations in the spirit of improving the quality of the submission

Author Response

Dear Editor and Reviewers,

We are pleased to resubmit the revised version of our manuscript entitled “Acute Kidney injury in hospitalized cancer patients: single-centre real-life analysis of incidence and clinical impact”

We sincerely thank you for your insightful and constructive comments, which have substantially improved the quality and clarity of our work. We have carefully addressed all points raised by the Editor and Reviewers and provided detailed, point-by-point responses in the rebuttal letter. All changes in the revised manuscript are highlighted in red for ease of evaluation.

In addition, we revised the wording in several sections—particularly within the Methods—to improve clarity and reduce textual overlap with our previous publications, while maintaining full consistency with the study design and results.

The final version of the manuscript has been reviewed and approved by all authors, who have ensured its accuracy and integrity. We confirm that no conflicts of interest, financial or otherwise, exist.

Thank you for your consideration.

 

Giacomo Garibotto § Pasquale Esposito

 

REVIEWER 1

Thank you for the possibility to review the manuscript titled: “Acute Kidney injury in hospitalized cancer patients: single-centre real-life analysis of incidence and clinical impact.” This is an interesting and large study of 56,390 hospitalized patients, 6,723 (11.9%) had cancer diagnosis that evaluated AKI in cancer patients versus non-cancer patients. There are several points to consider in order to improve the manuscript:

-The discussion section is rather short and does not fully analyse the results of the study and limitations.

1) Diabetes is a general term and description of a comorbidity. Patients with diabetes may have well-compensated disease with low Hb1Ac and overall good health, while others may have poor serum glucose management and multiple comorbidities. I think this should be mentioned in the discussion section that it was not possible to include a broader array of laboratory values

• Response 1: We agree with the reviewer that diabetes mellitus is a broad clinical entity encompassing a wide spectrum of disease severity. In our study, diabetes was identified through administrative data without considering laboratory values or diabetes-related complications. Admittedly, this approach may affect the estimation of the actual impact of diabetes on AKI risk, such as on patients’outcomes. We have explicitly addressed this limitation in the Discussion, emphasizing that the lack of association between diabetes and AKI risk may reflect the heterogeneity of this definition and the absence of detailed laboratory and clinical data rather than a true lack of effect.

2) Metastatic solid cancers is also a general term as metastases can be in different organs and tissues. Metastasis in the liver or kidneys significantly differs from a metastasis to spleen or other organs

• Response 2: We acknowledge that the term “metastatic solid cancer” represents a highly heterogeneous group, as metastatic involvement of different organs may confer substantially different risks for AKI. Due to the lack of detailed information on metastatic localization in our dataset, we were unable to further stratify this group. This limitation has now been clearly stated in the Discussion (where we also added a specific citation), highlighting that the heterogeneity of metastatic disease may have influenced risk estimation and interpretation of results in this subgroup.

Reviewer 2 Report

Comments and Suggestions for Authors

General Comments: This retrospective observational study aims to evaluate the incidence, risk factors, and outcomes of Acute Kidney Injury (AKI) in a large cohort of hospitalized cancer patients compared to a non-oncologic population. The topic is clinically relevant, and the sample size is substantial. However, there are significant methodological concerns regarding the definition of baseline creatinine, the lack of data on pharmacological treatments, and the interpretation of statistical analyses that need to be addressed. Furthermore, I have concerns regarding the potential overlap of this dataset with the authors' recently referenced work (Ref. 37, SIN-AKI Study).

 

Major Comments:

1. Lack of Pharmacological Data (Crucial Limitation): The study completely lacks data on specific nephrotoxic agents, including chemotherapy regimens (e.g., platinum derivatives), immune checkpoint inhibitors, and iodinated contrast media. As the authors acknowledge, drug-induced AKI is a major mechanism in oncology. Attempting to analyze risk factors for AKI in cancer patients without adjusting for these treatments introduces a massive confounding bias. The inability to identify specific treatments makes the risk factor analysis—particularly the distinction between metastatic and non-metastatic groups—difficult to interpret clinically.
2. Definition of Baseline Creatinine: The authors defined the baseline serum creatinine (sCr) as the "lowest sCr value recorded during hospitalization". This methodology is problematic and may lead to an overestimation of AKI incidence. Hospitalized cancer patients often experience fluid overload (dilution) or have reduced muscle mass (sarcopenia/cachexia), which artificially lowers sCr. Using this nadir as a baseline meets the KDIGO criteria for "sCr increase" too easily.

3. Implausible Risk Factor Analysis in Metastatic Patients: The multivariate analysis indicates that in patients with metastatic solid cancers, none of the traditional variables (including sepsis and ICU admission) significantly predicted AKI. This finding is biologically counter-intuitive and raises concerns about the statistical modeling.

   • Concern: It is highly unlikely that sepsis is not a risk factor for AKI in this subgroup. This suggests potential issues such as model overfitting, collinearity, or insufficient statistical power despite the sample size. The authors need to re-evaluate the model and provide a more convincing explanation than "overwhelming systemic burden".

4. Potential Data Overlap (Salami Slicing Concern): The manuscript cites the "SIN-AKI Study" (Ref. 37) by the same group. The title and methodology appear very similar. The authors must clarify in the Methods section whether the current dataset is a subset of the SIN-AKI study. If this is a sub-analysis of a previously published database, it must be explicitly stated, and the unique contribution of this specific analysis must be justified to avoid concerns of redundant publication.

Author Response

Dear Editor and Reviewers,

We are pleased to resubmit the revised version of our manuscript entitled “Acute Kidney injury in hospitalized cancer patients: single-centre real-life analysis of incidence and clinical impact”

We sincerely thank you for your insightful and constructive comments, which have substantially improved the quality and clarity of our work. We have carefully addressed all points raised by the Editor and Reviewers and provided detailed, point-by-point responses in the rebuttal letter. All changes in the revised manuscript are highlighted in red for ease of evaluation.

In addition, we revised the wording in several sections—particularly within the Methods—to improve clarity and reduce textual overlap with our previous publications, while maintaining full consistency with the study design and results.

The final version of the manuscript has been reviewed and approved by all authors, who have ensured its accuracy and integrity. We confirm that no conflicts of interest, financial or otherwise, exist.

Thank you for your consideration.

 

Giacomo Garibotto § Pasquale Esposito

 

REVIEWER 2

General Comments: This retrospective observational study aims to evaluate the incidence, risk factors, and outcomes of Acute Kidney Injury (AKI) in a large cohort of hospitalized cancer patients compared to a non-oncologic population. The topic is clinically relevant, and the sample size is substantial. However, there are significant methodological concerns regarding the definition of baseline creatinine, the lack of data on pharmacological treatments, and the interpretation of statistical analyses that need to be addressed. Furthermore, I have concerns regarding the potential overlap of this dataset with the authors' recently referenced work (Ref. 37, SIN-AKI Study).

Major Comments:

1. Lack of Pharmacological Data (Crucial Limitation): The study completely lacks data on specific nephrotoxic agents, including chemotherapy regimens (e.g., platinum derivatives), immune checkpoint inhibitors, and iodinated contrast media. As the authors acknowledge, drug-induced AKI is a major mechanism in oncology. Attempting to analyze risk factors for AKI in cancer patients without adjusting for these treatments introduces a massive confounding bias. The inability to identify specific treatments makes the risk factor analysis—particularly the distinction between metastatic and non-metastatic groups difficult to interpret clinically.

• Response 1: We fully agree that the lack of detailed pharmacological data represents a major limitation of our study. Unfortunately, the administrative nature of our dataset did not include information on specific oncologic treatments or iodinated contrast media. Admittedly, as drug-induced AKI is a key mechanism in cancer patients, the absence of these variables may introduce residual confounding and limit the clinical interpretability of risk factor analyses. We have expanded the Discussion to clearly acknowledge this limitation and to emphasize that future prospective studies integrating treatment-related nephrotoxicity are needed. In addition, we revised the Conclusions to emphasize that analyses restricted to patient-related factors (e.g., age, sex, comorbidities, and acute complications) may not fully capture AKI risk in cancer patients, and that future prospective, longitudinal studies should incorporate detailed pharmacological exposures and other iatrogenic risk factors to better define AKI determinants in this population.

            2. Definition of Baseline Creatinine: The authors defined the baseline serum creatinine (sCr) as the "lowest sCr value recorded during hospitalization". This methodology is problematic and may lead to an overestimation of AKI incidence. Hospitalized cancer patients often experience fluid overload (dilution) or have reduced muscle mass (sarcopenia/cachexia), which artificially lowers sCr. Using this nadir as a baseline meets the KDIGO criteria for "sCr increase" too easily.

• Response 2: We thank the reviewer for this important methodological comment. We recognize that defining baseline serum creatinine as the lowest value recorded during hospitalization may lead to AKI overestimation, particularly in cancer patients with fluid overload, sarcopenia, or cachexia. We are aware of these limitations, which we acknowledged also in the original version of the paper. However, it should also be admitted that the application of the full KDIGO criteria is challenging in real-world retrospective studies, especially due to the lack of urine output data and incomplete temporal information on creatinine measurements. Indeed, for example, the lack of urine data is common among epidemiological studies (see doi:10.1016/S0140-6736(15)00344-X). In this context, our approach represents a pragmatic solution. Importantly, the AKI incidence in our population is consistent with prior reports from the same geographical area (DOI: 10.1038/s41581-018-0052-0). This supports the validity of our approach for defining AKI in the large-scale general population. We have clarified this methodological choice and its limitations in both the Methods and Discussion sections.

 

         3. Implausible Risk Factor Analysis in Metastatic Patients: The multivariate analysis indicates that in patients with metastatic solid cancers, none of the traditional variables (including sepsis and ICU admission) significantly predicted AKI. This finding is biologically counter-intuitive and raises concerns about the statistical modeling.

• • Concern: It is highly unlikely that sepsis is not a risk factor for AKI in this subgroup. This suggests potential issues such as model overfitting, collinearity, or insufficient statistical power despite the sample size. The authors need to re-evaluate the model and provide a more convincing explanation than "overwhelming systemic burden".
• Response 3: Thank you for this comment, which highlights an important point that may indeed be misinterpreted. We agree that, at first glance, the absence of statistically significant predictors of AKI in patients with metastatic solid tumors may appear counterintuitive. However, our findings do not suggest that sepsis or ICU admission are biologically irrelevant in this subgroup. Rather, the direction and magnitude of the associations remain consistent with established pathophysiology (e.g., sepsis OR 4.28), while the lack of statistical significance is explained by wide confidence intervals, indicating limited precision and statistical power, likely due to the small number of metastatic patients with sepsis and ICU admission. Moreover, patients with metastatic disease represent a high-risk population in whom severe disease burden, critical illness, and multiple risk factors frequently coexist. In this context, the incremental effect of additional acute insults may be attenuated, potentially resulting in a ceiling effect. To clarify these points and avoid over-interpretation of non-significant results, we have: (a) added a clarifying sentence in the Results section; (b) revised Figure 3 to include confidence intervals; (c) revised the Discussion to better articulate this interpretation; and (d) revised the Conclusions to emphasize the complexity of AKI risk assessment in advanced cancer and the need for prospective longitudinal studies specifically designed to capture dynamic clinical variables and treatment-related exposures.

     4. Potential Data Overlap (Salami Slicing Concern): The manuscript cites the "SIN-AKI Study" (Ref. 37) by the same group. The title and methodology appear very similar. The authors must clarify in the Methods section whether the current dataset is a subset of the SIN-AKI study. If this is a sub-analysis of a previously published database, it must be explicitly stated, and the unique contribution of this specific analysis must be justified toavoid concerns of redundant publication.

• Response 4: Thank you for this observation. We confirm that the database used in the present study represents an extract from the larger SIN-AKI project promoted by the Italian Society of Nephrology. However, while the SIN-AKI study was multicenter, the current analysis focuses exclusively on patients hospitalized at our single center in Genoa. This approach was chosen to minimize center-related heterogeneity (also in oncological diagnosis) and allow a more homogeneous analysis of oncologic subgroups. Importantly, the SIN-AKI publication did not focus on cancer: it did not perform dedicated analyses in oncologic patients, nor did it stratify subjects according to malignancy type (hematologic vs solid, metastatic vs non-metastatic). The unique contribution of the current manuscript is the cancer-specific characterization of AKI incidence, risk factors, and outcomes across distinct oncologic categories within a uniform single-center setting. We have clarified this point in the Methods section and explicitly stated the relationship between the two datasets to avoid any concern regarding redundant publication.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have made all of the necessary conclusions

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have satisfactorily addressed all my concerns raised in the previous review.

1. Regarding the lack of pharmacological data: I appreciate the authors adding a clear statement in the Discussion and Conclusions acknowledging this limitation. This allows readers to properly interpret the results within the constraints of the administrative dataset.

2. Regarding the baseline creatinine definition: The inclusion of the discussion on the potential overestimation of AKI due to factors like sarcopenia and the lack of urine output data is appropriate.

3. Regarding the risk factor analysis in metastatic patients: The revised explanation concerning the wide confidence intervals and reduced statistical power clarifies why "traditional" risk factors did not reach statistical significance in this subgroup. The modification to Figure 3 to display confidence intervals is also helpful.

4. Regarding the dataset overlap: The clarification added to the Methods section regarding the relationship with the SIN-AKI study sufficiently distinguishes the unique contribution of this manuscript.

I have no further comments. The manuscript has been significantly improved.