Review Reports - Efficacy of Technology-Based Cognitive Rehabilitation Tools for Cancer-Related Cognitive Impairment in Non-CNS Cancer Patients: A Systematic Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

What are the exact keywords used in the search strategy? Although the authors reference the PICOS-based terms, the specific keyword strings are not clearly presented in the main text, despite an indication around line 115. This requires clearer reporting.
The results of the included 11 studies are not analyzed beyond descriptive synthesis. Since this is a systematic review, it is expected that the authors provide additional meta-analytical insights such as the most frequently contributing publishers, most common journals, geographical distribution of studies, and temporal trends in publication. These analyses are currently missing.
The potential sources of bias within the included studies are not clearly discussed. In the MMAT checklist, several items are assigned a score of “7,” yet the manuscript does not explain whether this refers to Section 7 or how these ratings translate into risk-of-bias categories. A more explicit explanation is needed.
The review highlights a disproportionately high representation of breast cancer participants (over 80% female), yet the findings are generalized to all “non-CNS cancer patients.” A more cautious interpretation is warranted, as the results may not be generalizable to male patients or to other cancer types such as colorectal or lymphoma.
The search period concludes in June 2025, but the authors do not justify why earlier digital rehabilitation studies (pre-2011) appear to be absent. It would strengthen the methodological transparency to clarify whether this reflects an intentional cutoff, possibly due to technological advancement, or a limitation of the search strategy.

Author Response

We thank the reviewer for the insightful and constructive comments. We carefully addressed each point in the revised manuscript to enhance methodological transparency, clarify search strategies, and provide a more nuanced discussion of generalizability and potential biases.

What are the exact keywords used in the search strategy? Although the authors reference the PICOS-based terms, the specific keyword strings are not clearly presented in the main text, despite an indication around line 115. This requires clearer reporting.

We thank the reviewer for this comment. In the revised manuscript, we made sure to further specify the keywords used for the initial string elaboration. In particular, we have added:

Records were therefore searched for using MeSH terms closely related to the PICOS criteria, such as: "Cognitive Impairment”, “Chemotherapy-Related Cognitive Impairment”, “Cancer Survivors”, “Neoplasms”, “Rehabilitation”, “Cognitive Training”.

The results of the included 11 studies are not analyzed beyond descriptive synthesis. Since this is a systematic review, it is expected that the authors provide additional meta-analytical insights such as the most frequently contributing publishers, most common journals, geographical distribution of studies, and temporal trends in publication. These analyses are currently missing.

We thank the reviewer for this remark. In the revised manuscript, we made sure to provide a more critical insight regarding the generalizability of our results, under the Discussion section. More specifically, you will find

“ A final remark needs to be made regarding the methodological rigor included studies and their temporal, geographical and demographic characteristics, which could all have impacted the generalizability of our results. First, given the disproportionate high representation of women with breast cancer patients in the included studies, our results might not be generalizable to the entire oncological population as well as to male oncological patients. Furthermore, 6 out of the 11 included studies were published later than 2020, with 5 studies published in the last two years. This temporal trend indicates a growing focus on cognitive deficits in oncological patients, which is expected to increase in future publications. Our contribution rests therefore on relatively new findings across literature. Moreover, even though the geographical distribution of the included study spans across four continents (Europe, North America, Asia and Australia), previous literature on barriers to inclusivity in oncological research highlights the possibility of biases and systematic practices that may limit the inclusivity and generalizability of results to the entire oncological population (Dahò et al., 2025). Lastly, a certain degree of methodological transparency of the included studies (See Supplementary Materials 2) might also have contributed to the weakening of the reliability, robustness and generalizability of our results.”

The potential sources of bias within the included studies are not clearly discussed. In the MMAT checklist, several items are assigned a score of “7,” yet the manuscript does not explain whether this refers to Section 7 or how these ratings translate into risk-of-bias categories. A more explicit explanation is needed.

We thank the reviewer for this comment. In the revised manuscript we provided what we hope will be a clearer explanation of the score assignment procedure; we also updated the Supplementary Material accordingly, as to make results clearer. More specifically, under the Quality assessment of the Included studies section, you will find:

“Quality was categorized as low (MMAT score, 2–4), moderate (MMAT score, 5–6), or high (MMAT score, 7).”

The review highlights a disproportionately high representation of breast cancer participants (over 80% female), yet the findings are generalized to all “non-CNS cancer patients.” A more cautious interpretation is warranted, as the results may not be generalizable to male patients or to other cancer types such as colorectal or lymphoma.

We thank the reviewer for this important remark. In the revised manuscript, we made sure to add a paragraph in our discussion, addressing risk of bias and systemic practices that may weaken the robustness of our results. More specifically, you will find, under the Discussion section:

“A final remark needs to be made regarding the methodological rigor included studies and their temporal, geographical and demographic characteristics, which could all have impacted the generalizability of our results. First, given the disproportionate high representation of women with breast cancer patients in the included studies, our results might not be generalizable to the entire oncological population as well as to male oncological patients. Furthermore, 6 out of the 11 included studies were published later than 2020, with 5 studies published in the last two years. This temporal trend indicates a growing focus on cognitive deficits in oncological patients, which is expected to increase in future publications. Our contribution rests therefore on relatively new findings across literature. Moreover, even though the geographical distribution of the included study spans across four continents (Europe, North America, Asia and Australia), previous literature on barriers to inclusivity in oncological research highlights the possibility of biases and systematic practices that may limit the inclusivity and generalizability of results to the entire oncological population (Dahò et al., 2025). Lastly, a certain degree of methodological transparency of the included studies (See Supplementary Materials 2) might also have contributed to the weakening of the reliability, robustness and generalizability of our results.”

The search period concludes in June 2025, but the authors do not justify why earlier digital rehabilitation studies (pre-2011) appear to be absent. It would strengthen the methodological transparency to clarify whether this reflects an intentional cutoff, possibly due to technological advancement, or a limitation of the search strategy.

We thank the reviewer for this important observation. In the revised manuscript we made sure to clarify, under the Information Source and Search Strategy section, how we did not limit our search to publication year. As we stated:

“Moreover, the search was conducted without any restrictions on publication year.”

Reviewer 2 Report

Comments and Suggestions for Authors

Study title:

Both computer and digital are under the umbrella of Technology-based interventions for cognitive rehabilitation. The authors could clearly define, since it comes across as conceptually redundant in the title. Aren't all the studies included in the review digital anyway?

Introduction:

The claim in the introduction that “…the aim of the study is to critically evaluate …” needs to be clarified, since the findings are presented more descriptively and the discussion as well. The critical evaluation is missing.

Design:

Followed PRISMA guidelines

Search databases were appropriate.

PICOS criteria were followed. Some of inclusion and exclusions were not clearly defined.

For example, one study included lung cancer metastasized to the brain- was CNS cancer not an exclusion criterion

Line 96 page 3 /24 ; should cognitive assessment (subjective and objective) not be included? This was not stated as an inclusion criterion.

Search Strategy:

Search terms and strategy are not provided under the Methods section (only included as Supplementary). The main search terms can be mentioned (and the strings provided as supplementary, as they did). But they just mentioned Booleans AND and OR was used with the assistance of the Librarian and where applicable MesH terms were used, depending on data base, but Not sure which search Terms were used.

Based on the above PRISMA guidelines

Section 13 was not fully achieved. And authors need to clearly state these sections

13 a (no )

13 b (no )

13 c ( no

13 d (no )

13 e (no )

13 f (Yes MMAT)

Data analysis:

The authors did not clearly state that the Data Analysis method was used for analysis. This left the reviewer to assume that a narrative (qualitative) descriptive synthesis was conducted. It would enhance the quality of the manuscript to clearly state this in the Abstract and under Methods (describe the process)

Descriptive summary tables were provided and described. No quantitative analysis was conducted. Not that meta-analysis is mandatory, it would be important to make this explicit.

Discussion:

Reasonable based on the synthesis, but could be more explicit since MMAT was used -claims of efficacy must be contextualized. This is an issue of rigor that must be clear.

Limitations were discussed, but the benefit meta-analysis was not conducted.

Implications are reasonable, and future directions are reasonably discussed.

Author Response

We thank the reviewer for the detailed and constructive comments regarding methodological rigor and data analysis. We have revised the Discussion and Methods sections to clarify the narrative synthesis approach, contextualize efficacy claims based on MMAT quality ratings, and explicitly note the absence of a quantitative meta-analysis due to study heterogeneity.

Study title:

We thank the reviewer for this comment. We agree that “computerized” and “digital” interventions overlap conceptually under the broader category of technology-based cognitive rehabilitation. We retained both terms in the original title to reflect the diversity of interventions included, as some studies specifically used computerized cognitive training programs delivered via software installed on computers, while others employed broader digital tools, such as web-based platforms or tablet-based applications.

However, to address the reviewer’s concern and improve clarity, we have revised the title as follows:

“Efficacy of Technology-Based Cognitive Rehabilitation Tools for Cancer-Related Cognitive Impairment in Non-CNS Cancer Patients: A Systematic Review.”

Introduction:

We thank the reviewer for this insightful observation. We acknowledge that the original wording in the Introduction suggested a formal “critical evaluation,” whereas the manuscript primarily presents a descriptive synthesis of the included studies. To clarify this point and ensure consistency, we have revised the aim statement in the Introduction as follows:

“Building on the demonstrated effectiveness of digital and computerized cognitive interventions in other clinical populations, the present review aims to describe and synthesize the current evidence on the efficacy of technology-based cognitive rehabilitation interventions for CRCI in non-CNS cancer patients.”

Additionally, we have strengthened the Discussion section by highlighting methodological limitations, potential biases, variability in interventions, and gaps in evidence, thereby providing a more critical perspective on the current literature.

Design:

Followed PRISMA guidelines

Search databases were appropriate.

PICOS criteria were followed. Some of the inclusion and exclusions were not clearly defined.

For example, one study included lung cancer metastasized to the brain- was CNS cancer not an exclusion criterion

Line 96 page 3 /24 ; should cognitive assessment (subjective and objective) not be included? This was not stated as an inclusion criterion.

We thank the reviewer for this comment. We have clarified the inclusion and exclusion criteria in the Methods section. Specifically, studies were included if they involved computerized cognitive rehabilitation in non-CNS cancer patients and evaluated intervention effectiveness. Studies including patients with CNS tumors were excluded, but patients with brain metastases were included if the primary tumor was non-CNS. We also clarified that all included studies measured cognitive outcomes, and only studies with at least one objective cognitive assessment were considered eligible.

Search Strategy:

Based on the above PRISMA guidelines

We thank the reviewer for this comment. To improve clarity and comply with PRISMA guidelines, we have now included the main search terms in the Methods section. Specifically, we wrote:

“Records were therefore searched for using MeSH terms closely related to the PICOS criteria, such as: ‘Cognitive Impairment’, ‘Chemotherapy-Related Cognitive Impairment’, ‘Cancer Survivors’, ‘Neoplasms’, ‘Rehabilitation’, ‘Cognitive Training’.”

Section 13 was not fully achieved. And authors need to clearly state these sections

13 a (no )

13 b (no )

13 c ( no

13 d (no )

13 e (no )

13 f (Yes MMAT)

We thank the reviewer for this comment. We have addressed this point by specifying Section 13 of the PRISMA checklist and by revising the location in which the relevant items are reported, in accordance with the structure of the revised manuscript. Furthermore, we have also revised the text to further specify where and how data synthesis and analysis took place. More specifically, under the Data Extraction section, you will find:

Extracted data were then organized in a Microsoft Excel spreadsheet (version 2016) to facilitate comparison, data synthesis and data analysis.

Data analysis:

Descriptive summary tables were provided and described. No quantitative analysis was conducted. Not that meta-analysis is mandatory, it would be important to make this explicit.

We thank the reviewer for this important comment. We agree that the data analysis approach was not sufficiently explicit in the original version of the manuscript. To improve clarity and transparency, we have now clearly stated the nature of the data analysis both in the Abstract and in the Methods section.

Specifically, we have added the following statements to the manuscript:

In the Abstract, we specified that the review was based on a qualitative synthesis, stating:

“Data was synthesized using a narrative descriptive approach, and results were summarized in a descriptive table.”
In the Methods section (data extraction), we clarified the data analysis process as follows:

“Given the heterogeneity of study designs, interventions, outcome measures, and follow-up durations, a quantitative meta-analysis was not feasible. Therefore, a narrative (qualitative) descriptive synthesis was conducted. Data were summarized using descriptive tables and systematically compared across studies in terms of intervention characteristics, cognitive domains assessed, and reported outcomes.”

These additions explicitly clarify that a narrative descriptive synthesis was conducted and that no quantitative meta-analysis was performed, in line with PRISMA recommendations.

Discussion:

Reasonable based on the synthesis, but could be more explicit since MMAT was used -claims of efficacy must be contextualized. This is an issue of rigor that must be clear.

Limitations were discussed, but the benefit meta-analysis was not conducted.

Implications are reasonable, and future directions are reasonably discussed.

We thank the reviewer for these comments. In response, we have carefully revised the Discussion section to better contextualize the efficacy of included studies. Specifically, we have emphasized that while the majority of studies were of moderate to high methodological quality, caution is needed in interpreting efficacy due to heterogeneity in study designs, interventions, outcomes, and follow-up durations. Specifically, we have added the following sentence in the discussion section:

“Lastly, a certain degree of methodological transparency of the included studies (See Supplementary Materials 2), as well as the heterogeneity of oncological populations and sometimes small sample sizes, might also have contributed to the weakening of the reliability, robustness and generalizability of our results, which therefore need to be interpreted with caution.”

Additionally, we have clarified in the limitations section that due to this substantial heterogeneity, a quantitative synthesis or meta-analysis was not feasible. Consequently, the findings of the review are based on a narrative descriptive synthesis, which limits the ability to estimate pooled effect sizes and to draw firm conclusions regarding the magnitude of intervention effects.

These revisions aim to enhance the methodological rigor and transparency of our discussion, addressing the reviewer’s concerns about contextualization and clarity regarding the type of analysis conducted.

Reviewer 3 Report

Comments and Suggestions for Authors

The presented systematic review addresses a relevant and practically significant issue of cognitive rehabilitation in cancer patients. The methodological foundation of the work generally meets the standards for systematic reviews: the authors registered the protocol in PROSPERO and conducted independent selection and analysis of publications with the participation of four researchers, which enhances the reliability of the data selection procedure.

However, there are omissions in the structure and interpretation of results that may reduce the evidence-based value of the conclusions and their practical applicability.

The most significant shortcoming is the absence of a discussion of conclusions in accordance with intervening factors. The authors rightly note that some of the included studies have methodological limitations, even indicating the quality of the studies in Scheme 2. However, this is not utilized in the generalization of results: the text lacks an explicit correlation of these limitations with the obtained results. For example, contradictory data on the effectiveness in different cognitive domains (such as episodic memory) should be directly discussed in the context of the risk of bias in the corresponding studies. The authors should verify that all sections are described in this manner. For instance: digital therapy was effective in such-and-such studies, but they had small sample sizes or did not account for certain factors.
When describing the included studies, the authors do not indicate whether they assessed common conditions in cancer patients, such as anxiety and depression. These factors can significantly influence both cognitive functions and the subjective assessment of their improvement. The authors should either indicate this in the table or describe the absence of such data as a limitation of their review.
To enhance clarity and ease of analysis, it is advisable to combine the data from Tables 1 and 2 into one summary table. This table should include a separate column for the duration of therapy (in weeks or number of sessions), a column reflecting accounted intervening factors (e.g., control for depression), or note their absence.
The review states the fact of statistically significant improvements in neuropsychological tests. However, it remains unclear to what extent these changes are clinically significant for patients, i.e., whether they lead to a tangible improvement in the quality of daily life. This should be added to the discussion section.
The authors do not discuss at all the possible differences in the profile of cognitive impairments depending on different types of cancer. The authors consider all patients with "non-CNS cancer" as a single homogeneous population. While this may be acceptable for this specific research objective, discussing the varying pathophysiology of CRCI is certainly warranted. The pathophysiology of cognitive impairments between a patient with hormone-dependent breast cancer and a patient with lymphoma receiving high-dose chemotherapy can differ significantly. The authors should have at least highlighted this as a key limitation and called for the conduct of disease-specific studies in the future.
The concluding formulations about the promising nature of digital tools are general in nature. To increase the scientific and practical value, the conclusions need to be detailed. For which specific timeframes and cognitive domains, for which types of cancer is the effectiveness most convincingly proven. Also, for which follow-up periods (immediately after the intervention, 3 months, 6 months) there is data on the preservation of the effect.
In the "Future Directions" section, the recommendations are general in nature. They should be clarified by proposing examples of specific future studies. For example, conducting studies with long follow-up periods (which specific ones, how many months) to assess the sustainability of effects. Including in the study design an assessment of clinically significant changes and functional outcomes related to daily activities, among others.

Despite the formal adherence to PRISMA methodological standards, the presented systematic review does not meet the expectations placed upon it as a tool for forming clear, practically applicable conclusions.

The conclusions are too general and optimistic to serve as a reliable basis for clinical decisions or planning further research. To correct the situation, a revision of key sections is required — especially the discussion and conclusion — towards greater methodological rigor, specificity, and a skeptical assessment of the obtained results.

Author Response

However, there are omissions in the structure and interpretation of results that may reduce the evidence-based value of the conclusions and their practical applicability.

The most significant shortcoming is the absence of a discussion of conclusions in accordance with intervening factors. The authors rightly note that some of the included studies have methodological limitations, even indicating the quality of the studies in Scheme 2. However, this is not utilized in the generalization of results: the text lacks an explicit correlation of these limitations with the obtained results. For example, contradictory data on the effectiveness in different cognitive domains (such as episodic memory) should be directly discussed in the context of the risk of bias in the corresponding studies. The authors should verify that all sections are described in this manner. For instance: digital therapy was effective in such-and-such studies, but they had small sample sizes or did not account for certain factors.

We thank the reviewer for this important remark. In the revised manuscript, we added a new paragraph within the Discussion section to address possible risks of bias within our contribution. For instance, you will find:

“A final remark needs to be made regarding the methodological rigor included studies and their temporal, geographical and demographic characteristics, which could all have impacted the generalizability of our results. First, given the disproportionate high representation of women with breast cancer patients in the included studies, our results might not be generalizable to the entire oncological population as well as to male oncological patients. Furthermore, 6 out of the 11 included studies were published later than 2020, with 5 studies published in the last two years. This temporal trend indicates a growing focus on cognitive deficits in oncological patients, which is expected to increase in future publications. Our contribution rests therefore on relatively new findings across literature. Moreover, even though the geographical distribution of the included study spans across four continents (Europe, North America, Asia and Australia), previous literature on barriers to inclusivity in oncological research highlights the possibility of biases and systematic practices that may limit the inclusivity and generalizability of results to the entire oncological population (Dahò et al., 2025). Lastly, a certain degree of methodological transparency of the included studies (See Supplementary Materials 2), as well as the heterogeneity of oncological populations and sometimes small sample sizes, might also have contributed to the weakening of the reliability, robustness and generalizability of our results.”

When describing the included studies, the authors do not indicate whether they assessed common conditions in cancer patients, such as anxiety and depression. These factors can significantly influence both cognitive functions and the subjective assessment of their improvement. The authors should either indicate this in the table or describe the absence of such data as a limitation of their review.

We thank the reviewer for this comment. We carefully re-examined the included studies to verify whether anxiety or depression were assessed. This information has now been added to Table 1 when available. However, we found that several studies did not report a systematic assessment of these conditions. We have therefore explicitly highlighted the lack of consistent evaluation of anxiety and depression as a limitation of the included studies in the Discussion section (page 27). Specifically, we have added as follow:

“Finally, although anxiety and depression were assessed in most of the included studies, three out of eleven did not evaluate these conditions, which may have influenced both cognitive outcomes and the subjective perception of cognitive improvement.”

To enhance clarity and ease of analysis, it is advisable to combine the data from Tables 1 and 2 into one summary table. This table should include a separate column for the duration of therapy (in weeks or number of sessions), a column reflecting accounted intervening factors (e.g., control for depression), or note their absence.

We thank the reviewer for this helpful suggestion. We carefully considered the possibility of merging Tables 1 and 2 into a single summary table. However, we chose to keep the two tables separate in order to preserve clarity and readability, as they serve different purposes. Table 1 focuses on study and patient characteristics, whereas Table 2 provides a detailed description of intervention procedures and results.

We have nevertheless implemented the key suggested improvements. Specifically, we have added a column reporting the assessment of intervening factors such as anxiety and depression (or their absence) to the relevant table. In addition, the duration of the intervention is already explicitly reported in Table 2 under the “Procedures” section.

We believe that this approach allows for a clear and structured presentation of the data while avoiding excessive table complexity.

The review states the fact of statistically significant improvements in neuropsychological tests. However, it remains unclear to what extent these changes are clinically significant for patients, i.e., whether they lead to a tangible improvement in the quality of daily life. This should be added to the discussion section.

We thank the reviewer for this insightful comment. We agree that statistical significance does not necessarily imply clinical relevance. We have therefore expanded the Discussion section to explicitly address the distinction between statistically significant improvements in neuropsychological test performance and their clinical relevance to patients' daily functioning and quality of life. This has been added to the Future research section, as follow:

“ Importantly, future research should place greater emphasis on the clinical relevance of cognitive changes by integrating objective neuropsychological assessments with patient-reported outcomes and functional measures. This approach would help determine whether statistically significant improvements in test performance translate into meaningful benefits in daily functioning and quality of life. Establishing clinically meaningful thresholds for commonly used neuropsychological instruments would further enhance the interpretation of cognitive outcomes from a patient-centred perspective.”

The authors do not discuss at all the possible differences in the profile of cognitive impairments depending on different types of cancer. The authors consider all patients with "non-CNS cancer" as a single homogeneous population. While this may be acceptable for this specific research objective, discussing the varying pathophysiology of CRCI is certainly warranted. The pathophysiology of cognitive impairments between a patient with hormone-dependent breast cancer and a patient with lymphoma receiving high-dose chemotherapy can differ significantly. The authors should have at least highlighted this as a key limitation and called for the conduct of disease-specific studies in the future.

We thank the reviewer for this important observation. We agree that patients with non-CNS cancers cannot be considered a fully homogeneous population with respect to the mechanisms and profiles of cancer-related cognitive impairment. While the present review focused on non-CNS cancers as a broader category in line with its research objective, we acknowledge that the underlying pathophysiology of CRCI may vary substantially across different cancer types and treatment regimens. We have now explicitly highlighted this issue as a key limitation of the review and emphasized the need for future disease-specific studies in the Future Directions section. We have added as follow:

Limitation: “ Another important limitation of this review is the inclusion of patients with non-CNS cancers as a relatively heterogeneous group. Indeed CRCI may arise from different pathophysiological mechanisms depending on cancer type, disease biology, and treatment exposure (Wefel et al., 2015). For example, cognitive changes observed in hormone-dependent breast cancer may differ substantially from those experienced by patients with hematological malignancies undergoing intensive chemotherapy (Yin et al., 2023; Fardell et al., 203). The lack of disease-specific analyses limits the ability to characterize distinct cognitive profiles and mechanisms across cancer types.”

Future directions: “Finally, future research should prioritize disease-specific and treatment-specific studies to better characterize CRCI and to develop more targeted and personalized cognitive rehabilitation approaches.”

The concluding formulations about the promising nature of digital tools are general in nature. To increase the scientific and practical value, the conclusions need to be detailed. For which specific timeframes and cognitive domains, for which types of cancer is the effectiveness most convincingly proven. Also, for which follow-up periods (immediately after the intervention, 3 months, 6 months) there is data on the preservation of the effect.

We thank the reviewer for this insightful comment. To address it, we have enhanced the Discussion section to provide greater specificity regarding the effectiveness of digital cognitive rehabilitation. In particular, we have inserted details on the follow-up periods, cognitive domains, and types of cancer studied.

Specifically, we have inserted in the text as follows:

“Specifically, improvements were most consistently observed immediately after the intervention and at follow-ups of 3 and 6 months, while effects at 12 months were more variable across studies.” , to clarify follow-up timing and sustainability of effects.
“Overall, the evidence is strongest for working memory and attention, while effects on episodic memory are less robust and require further investigation.” , to specify which cognitive domains show the most convincing improvement.
“Specifically, the strongest evidence for effectiveness is observed in working memory and attention domains, particularly in breast cancer patients, with improvements most robust immediately post-intervention and at 3- and 6-month follow-ups. Effects on episodic memory are more variable, and evidence for other cancer types remains limited. These results suggest that digital cognitive rehabilitation can provide meaningful cognitive benefits, but further studies are required to establish long-term efficacy, optimal duration of training, and applicability across diverse cancer populations.”, to provide a more detailed and clinically informative conclusion.

In the "Future Directions" section, the recommendations are general in nature. They should be clarified by proposing examples of specific future studies. For example, conducting studies with long follow-up periods (which specific ones, how many months) to assess the sustainability of effects. Including in the study design an assessment of clinically significant changes and functional outcomes related to daily activities, among others.

We thank the reviewer for this valuable and constructive comment. To address it, we have enhanced the Future Directions section with more specific recommendations and examples, as follows:

Follow-up periods: We specified multiple follow-up time points (6, 12, and 24 months) to evaluate both short-term and long-term sustainability of cognitive improvements.
Clinical significance and functional outcomes: We emphasized that future studies should assess functional outcomes related to daily activities, work performance, and social participation. This ensures that statistically significant improvements correspond to meaningful benefits in patients’ daily lives.
Integration of objective and patient-reported measures: We clarified the importance of combining standardized neuropsychological assessments with patient-reported outcomes to better capture the real-world impact of cognitive rehabilitation.
Disease- and treatment-specific studies: We highlighted the need for future research to prioritize disease-specific and treatment-specific designs, providing examples such as hormone-dependent breast cancer versus lymphoma treated with high-dose chemotherapy, to account for differing pathophysiology of CRCI and to develop more targeted, personalized interventions.

In particular, in future research we have added as follow:

“Specifically, follow-up assessments of 6–12 months could provide valuable information on the sustainability of cognitive improvements.”

“Future studies should also systematically include measures of clinical significance, such as assessments of functional outcomes related to daily activities, work performance, and social participation, to ensure that improvements in test scores translate into meaningful benefits for patients’ daily quality of life.”

We thank the reviewer for this important and constructive comment. We agree that adherence to PRISMA standards alone does not guarantee clinical applicability, and that overly optimistic conclusions should be avoided when evidence is heterogeneous and methodologically limited.

To address this concern, we revised the Discussion section to adopt a more cautious and critical interpretation of the findings. Specifically, we:

- Explicitly distinguished between statistical significance and clinical relevance of cognitive improvements;

- Highlighted the limited durability of effects, particularly at long-term follow-up;

- Discussed the impact of methodological heterogeneity, attrition rates, and sample imbalance on the robustness and generalizability of results;

- Clarified that current evidence supports digital cognitive rehabilitation as a promising but still exploratory intervention, rather than a clinically established approach.

Accordingly, the Conclusions have been reworded to reflect a more skeptical and methodologically grounded assessment, emphasizing the need for disease-specific, well-powered trials with clinically meaningful and functional outcomes before firm clinical recommendations can be made.

Reviewer 4 Report

Comments and Suggestions for Authors

1. Keywords should be written alphabetically.
2. The years covered in the Methods section should be specified.
3. The number of articles included in the systematic review is very low. A comprehensive study should be conducted, including the Web of Science database. Only 11 articles were included. In its current form, this could be considered a mini-review.
4. The study does not specify cancer types and disease burden.
Disease burden negatively affects cognitive processes and emotions.
The presence of metastasis in cancer patients, metastasis to the central nervous system, or tumor-related secretions reaching the brain as a result of tumor cell secretion can impair cognitive processes.
5. The type of cancer is not specified in the tables prepared in the review. The oncological patient group is very diverse. There are too many question marks in this systematic review. It should be written in much greater detail.

Author Response

We thank the reviewer for the thoughtful and detailed comments. We carefully addressed each point in the revised manuscript, providing clarifications on keywords, study selection, cancer types, and population characteristics to improve transparency and methodological rigor.

Keywords should be written alphabetically.

We thank the reviewer for this comment. The keywords have now been reordered alphabetically as requested.

The years covered in the Methods section should be specified.

We thank the reviewer for this comment. We have added a clarification in the Methods section specifying the time frame covered by our search, we have added this as follows: “Moreover, the search was conducted without any restrictions on publication year.”

The number of articles included in the systematic review is very low. A comprehensive study should be conducted, including the Web of Science database. Only 11 articles were included. In its current form, this could be considered a mini-review.

We sincerely thank the reviewer for this valuable observation. While we acknowledge that the number of articles included is relatively low, we believe this highlights the originality and emerging nature of the topic. Although the work may be considered a mini-review in terms of quantity, we conducted a rigorous systematic review following established methodology, thereby highlighting that the limited number of studies reflects the current underexplored state of this field and emphasizes the relevance and novelty of our contribution. Furthermore, by applying very specific inclusion criteria focused on the topic, several studies did not meet these criteria for various reasons, which we have explicitly detailed in the manuscript. More specifically, under the Study selection and characteristics section, we added:

“Of these: 19 studies were excluded due to wrong study design, as they were conference abstracts or study protocols; 1 study was excluded as it was not written in English language; 1 study was excluded because it employed eye-tracking as an assessment tool rather than standardized neuropsychological tests; 9 studies were excluded as they did not include a digital cognitive rehabilitation in their intervention; 17 studies were excluded as they investigated the acceptability or feasibility of the intervention rather than the efficacy of the intervention itself. Therefore, due to non-compliance with the predefined eligibility criteria, 46 studies out of 58 were excluded and ultimately a total of 11 studies were retained for inclusion in the review.”

The study does not specify cancer types and disease burden.

Disease burden negatively affects cognitive processes and emotions.

The presence of metastasis in cancer patients, metastasis to the central nervous system, or tumor-related secretions reaching the brain as a result of tumor cell secretion can impair cognitive processes.

We thank the reviewer for this important observation. We would like to clarify that cancer types have been specified both in the Table and in the manuscript, reported as follows: “The majority of included studies (N=6) only focused on breast cancer populations, whereas the other 7 studies did not limit the intervention to breast cancer patients cohort, but also included other oncological diagnosis, such as: ovarian cancer, Hodgkin's lymphoma, non-small cell lung cancer with brain metastases, supraglottic/laryngeal cancer, throat/tongue cancer, anaplastic oligodendroglioma, colon cancer, prostate cancer, lymphoma, multiple myeloma, endometrial cancer survivors, colorectal cancer, and lymphoma.”

Moreover, we fully agree that the aspect of metastases is an important factor that could have a certain impact on the study of cognitive impairment, and it represents a limitation of the current evidence. To address this, we have added a section in the manuscript discussing the characteristics of metastases in the included studies. We have added a short section, in the sociodemographic characteristics paragraph, as follows: “The majority of included studies (N=7) excluded patients with metastases, while some studies (N=3) did not provide any information on this aspect. Only two studies included patients with metastases: in one study (Dos Santos et al., 2020), 15 patients (9% of patients included) had non-brain metastases, and in another study (Peterson et al, 2018), only one patient with non-small cell lung cancer had brain metastases.”

The type of cancer is not specified in the tables prepared in the review. The oncological patient group is very diverse. There are too many question marks in this systematic review. It should be written in much greater detail.

We thank the reviewer for this comment. We would like to clarify that the type of cancer is specified both in the manuscript text and in the tables. Moreover, we agree that the patient populations are very diverse; accordingly, we have added a brief statement in the limitations section of our review regarding the heterogeneity of the included populations. We have added this part as follows: “Another important limitation of this review is the inclusion of patients with non-CNS cancers as a relatively heterogeneous group. Indeed, CRCI may arise from different pathophysiological mechanisms depending on cancer type, disease biology, and treatment exposure (Wefel et al., 2015). For example, cognitive changes observed in hormone-dependent breast cancer may differ substantially from those experienced by patients with hematological malignancies undergoing intensive chemotherapy (Yin et al., 2023; Fardell et al., 203). The lack of disease-specific analyses limits the ability to characterize distinct cognitive profiles and mechanisms across cancer types.”

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed the comments accordingly.

Author Response

We thank the Reviewer for their time and availability in assessing our work. We made sure to revise the manuscript as requested.

Reviewer 3 Report

Comments and Suggestions for Authors

Authors answer all my comments and significantly improve their manuscript

Author Response

We thank the Reviewer for their time and availability, we made sure to submit the revised version of our manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

This manuscript was improved and can be published.

Author Response

We thank the Reviewer for their time and availability, the revised manuscript is being submitted.