The Prognostic Role of STAT5B Across Cancer Types and Comparative Analysis with STAT5A: A Systematic Review

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

In the systematic review entitled “The Prognostic Role of STAT5B Across Cancer Types and Comparative Analysis with STAT5A: A Systematic Review”, the authors conducted a systematic meta-analysis of STAT5B to evaluate its association with overall survival across cancers and to compare its prognostic role with STAT5A, that was studied previously. They verified that high STAT5B expression was significantly associated with favorable overall survival and that STAT5B was particularly protective in lung cancers and hematologic malignancies (in contrast, STAT5A demonstrated divergent effects, conferring favorable survival in breast cancer but poorer outcomes in hematologic cancers). They concluded elevated STAT5B expression is associated with improved survival in multiple cancers, supporting a tumor-suppressive role for it, evidencing the importance of isoform-specific STAT5 evaluation in cancer prognosis and highlighting STAT5B as a candidate biomarker and therapeutic target.

 

This manuscript is interesting in order to shed light upon STAT family role in cancers, and STAT5B in particular and the systematic review is clear, simple to read and well organized in paragraphs.

However, there are some issues to be addressed before to consider it for publication.

 

The authors should add an ABBREVIATIONS Section to the manuscript in order to clarify the meaning of such acronyms reported within the text (not only “cancer type abbreviations”.

 

Data presented are representative of those reported in online datasets (ie. GEPIA2). Are there any evidences about different STAT5B expression reached by the authors in experimental analysis? Have they analyzed medical as well as cell lines samples in order to support bioinformatics data?

Have the authors analyzed STAT5B protein levels and made functional assays in order to correlate STAT5B expression to cellular behaviour and fate.

 

In the DISCUSSION Section, the authors reported that “We propose that STAT5B’s divergent prognostic effects may reflect opposing activities of unphosphorylated STAT5B (uSTAT5B), which maintains chromatin stability, versus phosphorylated STAT5B (pSTAT5B), which drives proliferative signaling”. Are there any evidences for it or is this a speculation?

FIGURE 3 and FIGURE 4 are not in the correct order. Why FIGURE 4 is presented and commented before FIGURE 3. Please reorder them.

Could the authors add some schematic representations (a graphical abstract) of the pathways and genes involved in the molecular mechanisms they dissected in order to graphically show the possible STAT5B role in functional networks.

 

This manuscript in interesting in order to shed light upon mechanisms that may be harnessed to fight cancer but I think some analysis are too preliminary and need to be supported by some functional analysis to directly link STAT5B to tumor onset/suppression.

I think the manuscript cannot be considered for publicaton in its present form.

Author Response

“In the systematic review entitled “The Prognostic Role of STAT5B Across Cancer Types and Comparative Analysis with STAT5A: A Systematic Review”, the authors conducted a systematic meta-analysis of STAT5B to evaluate its association with

overall survival across cancers and to compare its prognostic role with STAT5A, that was studied previously. They verified that high STAT5B expression was significantly associated with favorable overall survival and that STAT5B was particularly protective in lung cancers and hematologic malignancies (in contrast, STAT5A demonstrated divergent effects, conferring favorable survival in breast cancer but poorer outcomes in hematologic cancers). They concluded elevated STAT5B expression is associated with improved survival in multiple cancers, supporting a tumor-suppressive role for it, evidencing the importance of isoform-specific STAT5

evaluation in cancer prognosis and highlighting STAT5B as a candidate biomarker and therapeutic target. This manuscript is interesting in order to shed light upon STAT family role in cancers, and STAT5B in particular and the systematic review is clear, simple to read and well organized in paragraphs.

However, there are some issues to be addressed before to consider it for publication.

The authors should add an ABBREVIATIONS Section to the manuscript in order to clarify the meaning of such acronyms reported within the text (not only “cancer type abbreviations”.

 

We included an ABBREVIATIONS section as suggested.

 

“Data presented are representative of those reported in online datasets (ie. GEPIA2). Are there any evidences about different STAT5B expression reached by the authors in experimental analysis? Have they analyzed medical as well as cell lines samples in order to support bioinformatics data?

Have the authors analyzed STAT5B protein levels and made functional assays in order to correlate STAT5B expression to cellular behaviour and fate.”

 

We appreciate the reviewer’s interest in experimental validation. This manuscript is designed as a systematic review and meta-analysis of publicly available gene-expression datasets; we did not perform new laboratory experiments or collect patient samples. Our aim was to synthesize existing transcriptomic data across many independent cohorts to evaluate the prognostic value of STAT5B in a harmonized manner. Because of this scope, we did not include original analyses of STAT5B protein levels, cell-line assays, or functional experiments. We have now made this limitation explicit in the revised Methods and Discussion, noting that protein-level validation and mechanistic studies are essential next steps to confirm the patterns observed here. We believe this clarification helps readers correctly interpret the scope and strength of the present study while pointing to important avenues for future research.

 

“In the DISCUSSION Section, the authors reported that “We propose that STAT5B’s divergent prognostic effects may reflect opposing activities of unphosphorylated STAT5B (uSTAT5B), which maintains chromatin stability, versus phosphorylated STAT5B (pSTAT5B), which drives proliferative signaling”. Are there any evidences for it or is this a speculation?”

 

We changed to “We speculate that STAT5B’s divergent prognostic effects may reflect opposing activities of unphosphorylated STAT5B (uSTAT5B), which maintains chromatin stability, versus phosphorylated STAT5B (pSTAT5B), which drives proliferative signaling.”

 

“FIGURE 3 and FIGURE 4 are not in the correct order. Why FIGURE 4 is presented and commented before FIGURE 3. Please reorder them.”

 

We placed Figure 4 after Figure 3 in the text.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

This manuscript presents a systematic review and meta-analysis investigating the prognostic role of STAT5B across various cancer types and comparing its impact with STAT5A. The study is timely and relevant, as STAT family members are increasingly recognized as critical regulators of tumor biology and potential therapeutic targets. The manuscript is well-organized, the methodology largely follows PRISMA guidelines, and the results provide meaningful insights into the isoform-specific functions of STAT5 proteins.

However, several issues must be addressed to improve the rigor and translational relevance of this study.

General Comments:

1. The observed heterogeneity in the meta-analysis (I² = 64%) is substantial, yet the exploration of its potential sources is insufficient.
2. While the results suggest clinical utility, the discussion lacks a detailed evaluation of how STAT5B detection could be integrated into current clinical practice.
3. The comparison between STAT5A and STAT5B, though mentioned, is not explored deeply enough to fully establish their divergent prognostic roles.

Specific Comments:

Abstract & Introduction:

1. The abstract clearly presents the findings but tends to overstate the conclusions (e.g., “STAT5B as a candidate biomarker and therapeutic target”). I recommend softening the tone to avoid over-interpretation and instead use wording such as “STAT5B may serve as a potential biomarker.”

Results:

1. The finding that STAT5B is downregulated in 42% of cancer types is compelling, but the lack of any tumor type showing upregulation should be highlighted as particularly striking.
2. The meta-analysis demonstrates that high STAT5B expression correlates with improved survival. However, given the modest effect size, the clinical significance should be more carefully discussed.
3. In hematologic malignancies, the protective role of STAT5B seems contradictory to prior reports of its oncogenic mutations (e.g., STAT5B^N642H). The results should be interpreted with greater caution, and this discrepancy should be explained in detail.

Discussion Section:

1. The discussion appropriately addresses the tumor-suppressive role of STAT5B but does not sufficiently explore its dual oncogenic roles in hematologic cancers. This may mislead readers into overgeneralizing STAT5B as uniformly protective.
2. The comparison with STAT5A is underdeveloped. The manuscript would benefit from a more thorough analysis of how the two isoforms diverge structurally, functionally, and prognostically.

Figures and Tables：

1. Figures (e.g., forest plots, KM curves) should include sample sizes and dataset identifiers for clarity.

Comments on the Quality of English Language

Overall, the language is clear but occasionally too definitive. Phrases such as “STAT5B is a tumor suppressor” should be revised to “STAT5B may function as a tumor suppressor in specific contexts.”

Author Response

“This manuscript presents a systematic review and meta-analysis investigating the

prognostic role of STAT5B across various cancer types and comparing its impact with

STAT5A. The study is timely and relevant, as STAT family members are increasingly

recognized as critical regulators of tumor biology and potential therapeutic targets. The

manuscript is well-organized, the methodology largely follows PRISMA guidelines, and the

results provide meaningful insights into the isoform-specific functions of STAT5 proteins.

However, several issues must be addressed to improve the rigor and translational

relevance of this study.

General Comments:

1. The observed heterogeneity in the meta-analysis (IÇ = 64%) is substantial, yet the

exploration of its potential sources is insufficient.”

 

We agree and have expanded the Results and Discussion to address potential sources of heterogeneity. Specifically, we now report subgroup analyses by cancer type (hematologic vs. epithelial, hormone-responsive vs. non–hormone-responsive) and discuss how differences in platform, probe selection, mutational background, and treatment regimens may contribute to between-study variability. We also clarify in the Methods that we used a random-effects model and note that interaction analyses were not feasible with the current dataset but represent an important focus for future work. These additions appear in the revised manuscript (highlighted in yellow).

 

2. While the results suggest clinical utility, the discussion lacks a detailed evaluation of

how STAT5B detection could be integrated into current clinical practice.

 

We have expanded the Discussion to outline potential clinical applications of STAT5B detection. The revised text now explains how STAT5B expression could be incorporated into existing prognostic models (e.g., alongside IPI in DLBCL or current lung adenocarcinoma classifiers) and how isoform-specific analysis may inform the use of emerging STAT5 inhibitors. We also caution that any clinical integration would require prospective validation, protein-level assays, and standardized cut-offs. These additions are highlighted in the revised Discussion.

 

3. The comparison between STAT5A and STAT5B, though mentioned, is not explored

deeply enough to fully establish their divergent prognostic roles.”

 

We have substantially expanded the Introduction and Discussion to emphasize structural and functional differences between STAT5A and STAT5B and to integrate additional references comparing their prognostic roles across cancer types (refs. 3,5,6,8–10,14–19). The revised section now explains that STAT5B contains a unique C-terminal extension absent in STAT5A, is the principal mediator of growth-hormone–responsive transcription, and shows predominantly tumor-suppressive associations across epithelial cancers, whereas STAT5A displays more variable, context-dependent effects. This new text appears in the revised manuscript and is highlighted in yellow.

 

“Specific Comments:

Abstract & Introduction:

1. The abstract clearly presents the findings but tends to overstate the conclusions (e.g.,

“STAT5B as a candidate biomarker and therapeutic target”). I recommend softening

the tone to avoid over-interpretation and instead use wording such as “STAT5B may

serve as a potential biomarker.”

 

We changed to “These findings underscore the importance of isoform-specific STAT5 evaluation in cancer prognosis and suggest that STAT5B may serve as a potential biomarker and therapeutic target.”

 

“Results:

1. The finding that STAT5B is downregulated in 42% of cancer types is compelling, but

the lack of any tumor type showing upregulation should be highlighted as particularly

striking.”

 

We added “Strikingly, no cancer type exhibited significant upregulation of STAT5B compared to normal tissue. This uniform absence of overexpression, together with the widespread downregulation, underscores STAT5B’s potential role as a broadly conserved tumor-suppressive factor, in contrast to many oncogenes that show tissue-restricted overexpression patterns.”

 

“2. The meta-analysis demonstrates that high STAT5B expression correlates with

improved survival. However, given the modest effect size, the clinical significance

should be more carefully discussed.”

 

We changed the text to “While the effect size (lnHR = –0.4009) is modest, its consistency across multiple datasets suggests a reproducible biological signal. From a clinical standpoint, the modest magnitude indicates that STAT5B expression alone is unlikely to serve as a stand-alone prognostic determinant, but rather may add incremental predictive value when integrated with established biomarkers and clinical variables.”

 

“3. In hematologic malignancies, the protective role of STAT5B seems contradictory to

prior reports of its oncogenic mutations (e.g., STAT5B^N642H). The results should be

interpreted with greater caution, and this discrepancy should be explained in detail.”

 

We added the following to Results “The protective association of high STAT5B expression was particularly strong in hematologic malignancies, especially in diffuse large B-cell lymphoma. This appears paradoxical given prior reports of oncogenic STAT5B mutations, such as STAT5B^N642H, which drive proliferation in T-cell leukemias and certain myeloid neoplasms. One possible explanation is that bulk expression of wild-type STAT5B may reflect preserved differentiation programs or intact chromatin regulatory functions that are favorable for prognosis, whereas oncogenic mutations represent rare, gain-of-function events confined to specific subtypes. Thus, our findings highlight a dichotomy between wild-type STAT5B expression as a potentially protective factor in some hematologic contexts versus mutant or hyperactivated STAT5B as an oncogenic driver. This context-specific duality warrants careful interpretation and emphasizes the need for integrated analyses combining expression with mutational and activation status.”

 

Discussion Section:

1. The discussion appropriately addresses the tumor-suppressive role of STAT5B but

does not sufficiently explore its dual oncogenic roles in hematologic cancers. This may

mislead readers into overgeneralizing STAT5B as uniformly protective.

 

We added to Discussion: “Although our meta-analysis highlights the predominantly tumor-suppressive associations of STAT5B across epithelial cancers and lymphomas, it is equally important to recognize its established oncogenic functions in specific hematologic malignancies. Activating STAT5B mutations, particularly STAT5B^N642H, have been identified in T-cell large granular lymphocytic leukemia and other hematologic cancers, where they drive constitutive JAK–STAT signaling and malignant proliferation [14,15,18,19]. This apparent contradiction underscores the context-dependent nature of STAT5B biology: while bulk expression of wild-type STAT5B may reflect preserved differentiation or tumor-suppressive chromatin functions, hyperactivated or mutant STAT5B can exert potent oncogenic effects. Overgeneralization of STAT5B as uniformly protective should therefore be avoided, and future studies must integrate expression levels with mutational and phosphorylation status to clarify its role in distinct hematologic subtypes.”

 

2. The comparison with STAT5A is underdeveloped. The manuscript would benefit from a

more thorough analysis of how the two isoforms diverge structurally, functionally, and

prognostically.

 

We added to Discussion: “Direct comparison with STAT5A further illustrates the isoform-specific biology of STAT5 proteins. Despite their ~90% sequence homology, STAT5A and STAT5B diverge in several critical respects. Structurally, STAT5B contains a unique C-terminal transactivation domain that modulates DNA binding and transcriptional activation [3,6], whereas STAT5A is more closely linked to prolactin-responsive signaling and mammary gland development [10]. Functionally, STAT5A has been implicated in promoting oncogenesis in breast and hematologic cancers [5,19], while STAT5B demonstrates more consistent tumor-suppressive associations across epithelial malignancies and protective effects in B-cell lymphomas. Prognostically, STAT5A’s effects are highly context-dependent, showing both favorable and unfavorable associations depending on tumor type, in contrast to STAT5B’s predominantly protective profile. These distinctions suggest that isoform-specific functions extend beyond subtle differences in transcriptional targets, and highlight the need for comparative studies that examine STAT5A and STAT5B side-by-side in the same tumor contexts.”

 

Figures and Tables：

1. Figures (e.g., forest plots, KM curves) should include sample sizes and dataset

identifiers for clarity.

 

We made new KM curves and included sample sizes and dataset identifiers in these figures.

 

Overall, the language is clear but occasionally too definitive. Phrases such as “STAT5B is

a tumor suppressor” should be revised to “STAT5B may function as a tumor suppressor in

specific contexts.”

 

We made changes accordingly.

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

1. The author should provide the structural difference and the mode of action difference, or others, between STAT5A and STATB in the introduction section.
2.   The author has mentioned using the microarray databases for miRNA expression for these two markers. Why did the author use only microarray data instead of RNA-Seq (bulk or scRNA)? Please describe. 
3. Table -1 is not necessary until there is some special highlight. If so, please mention them. 
4. Please provide more detailed results in each figure legend.  
5. Please correlate the findings with more previous studies. Given that this paper identified heterogeneous roles of markers across different cancer types, please compare both effects. 
6. The discussion should be without a heading, and it should be in a paragraph. 
7. Please define the inclusion and exclusion criteria in more detail for the study selection. 

Author Response

1. The author should provide the structural difference and the mode of action difference, or others, between STAT5A and STATB in the introduction section.”

We agree and have revised the Introduction to briefly summarize the key structural and functional differences between STAT5A and STAT5B. Specifically, we now state that although the two isoforms share ~90 % sequence identity, STAT5B contains a unique 20–22-amino-acid C-terminal extension absent in STAT5A, which influences DNA-binding affinity, transcriptional potency, and growth-hormone–responsive genes, whereas STAT5A is more tightly linked to prolactin signaling and mammary gland biology (refs. 3,6,8–10). This sets up the rationale for our isoform-specific meta-analysis.

1.   The author has mentioned using the microarray databases for miRNA expression for these two markers. Why did the author use only microarray data instead of RNA-Seq (bulk or scRNA)? Please describe. 

We have added a statement to the Methods and the end of the Introduction clarifying that PrognoScan and GEPIA2 provide uniformly processed microarray datasets with survival annotations across many cancer types, which allows meta-analysis with consistent probes and clinical covariates. Comparable large-scale, harmonized RNA-seq datasets with survival data for STAT5B were not available when the analysis was initiated; single-cell and bulk RNA-seq repositories remain more fragmented and often lack mature survival follow-up. We note this as a limitation and highlight RNA-seq–based validation as an important future direction.

Specifically, we added the following to Introduction. “We therefore performed a systematic meta-analysis of STAT5B expression and overall survival across diverse cancers, leveraging large, uniformly processed microarray datasets with survival annotations from PrognoScan. Microarray data were used because, at the time of analysis, comparably harmonized bulk or single-cell RNA-seq datasets with mature clinical follow-up and consistent probe annotations were not available, whereas the microarray platforms allowed standardized cross-study comparisons.”

1. Table -1 is not necessary until there is some special highlight. If so, please mention them. 

We have changed Table 1 to Table S1 and moved it to Supplementary Information.

1. Please provide more detailed results in each figure legend.  

All figure legends have been expanded to include dataset identifiers, sample sizes, group allocations, hazard ratios with 95 % CI (where applicable), and the statistical test used. This ensures that readers can interpret each plot independently of the main text.

1. Please correlate the findings with more previous studies. Given that this paper identified heterogeneous roles of markers across different cancer types, please compare both effects. 

We have expanded the Discussion to integrate additional references comparing STAT5A and STAT5B across tumor types (refs. 5,14–19), emphasizing that STAT5A shows context-dependent oncogenic or tumor-suppressive effects, whereas STAT5B appears largely tumor-suppressive in epithelial cancers but can be oncogenic in selected hematologic malignancies. This section, Contextualizing STAT5B Findings with STAT5A and Previous Reports, now explicitly discusses how our results align with or diverge from prior reports.

1. The discussion should be without a heading, and it should be in a paragraph. 

We appreciate the reviewer’s suggestion to present the Discussion as a single paragraph. Because our Discussion is relatively long and covers several distinct issues — (i) summarizing key findings, (ii) contrasting STAT5B with STAT5A, (iii) reconciling apparent contradictions in hematologic cancers, (iv) clinical implications, and (v) limitations and future directions — we found that using brief subheadings markedly improves readability and helps readers navigate the main points. We have therefore retained concise subheadings (“Isoform-specific roles,” “Clinical implications,” etc.) but ensured that each section now reads as a continuous narrative without redundancy or abrupt breaks. We believe this format still meets the journal’s style guidelines while making the manuscript clearer for a broad readership. However, if the Editor prefers the Discussion as a single uninterrupted section, we will be happy to reformat accordingly in the final stage.

1. Please define the inclusion and exclusion criteria in more detail for the study selection. 

The revised Methods section (Study selection criteria) now explicitly defines inclusion criteria (public datasets from PrognoScan reporting STAT5B expression with overall survival, minimum sample size ≥30, annotated clinical covariates) and exclusion criteria (datasets lacking survival follow-up, incomplete clinical annotation for multivariate analysis, or non-human studies). We also clarify probe-selection (smallest log-rank p-value) and handling of missing data. These details are reflected in the revised PRISMA flow diagram.

Round 2

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

In the 2^nd version of the manuscript entitled “The Prognostic Role of STAT5B Across Cancer Types and Comparative Analysis with STAT5A: A Systematic Review”, the authors replied to all my questions.

 

Due to the character of the systematic review I understand what authors mean about the lacking of an experimental validation and I accept their reply.

 

I think the manuscript may be considered for publication in its present form.

Author Response

"In the 2^nd version of the manuscript entitled “The Prognostic Role of STAT5B Across Cancer Types and Comparative Analysis with STAT5A: A Systematic Review”, the authors replied to all my questions.

Due to the character of the systematic review I understand what authors mean about the lacking of an experimental validation and I accept their reply.

I think the manuscript may be considered for publication in its present form."

Response:  We thank the reviewer for their time and effort. 

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

After carefully reviewing this paper, I find the authors' response to be rigorous, logically coherent, and robustly documented, significantly enhancing the study's credibility and methodological transparency. I offer one additional, small but constructive suggestion for further improvement:

The second half of the Discussion discusses non-canonical STAT family signals (such as uSTAT5B regulating heterochromatin) at length, and it is recommended to be slightly refined.

 

Author Response

"The second half of the Discussion discusses non-canonical STAT family signals (such as uSTAT5B regulating heterochromatin) at length, and it is recommended to be slightly refined."

We appreciate the reviewer's comments, and have further refined the Discussion to make it more coherent with improved clarity and flow. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study explores the roles of STAT5A and STAT5B, two closely related transcription factors, in cancer progression and prognosis. Although these isoforms share a high degree of sequence similarity, they perform distinct, and sometimes opposing, functions across different cancer types. The findings emphasize the complex and context-specific nature of STAT5 activity in cancer, and highlight the need for further research to clarify their individual biological and clinical significance. While the meta-analysis presented offers valuable insights into the potential role of STAT5B in cancer, it remains incomplete. The proposed biological mechanisms are largely speculative, and some findings appear inconsistent with existing literature. Additional experimental studies are necessary to verify the suggested tumor-suppressive role of STAT5B and to better understand its function in oncogenesis.

One of the key challenges in interpreting the results of this study is the inconsistency surrounding the role of STAT5B in cancer. While the authors emphasize STAT5B’s tumor-suppressive role highlighted by its downregulation in 42% of analyzed cancers and its association with improved survival outcomes, this finding contrasts with previous literature suggesting that STAT5B can also act as an oncogene, particularly in hematologic malignancies like leukemia. For instance, some cited studies (references 14 and 15) describe STAT5B as promoting proliferation in certain blood cancers, raising questions about its dual functionality.

This contradiction becomes particularly evident in B-cell lymphomas, where the current study finds STAT5B to be beneficial, even though related research suggests it can play a tumor-promoting role in similar contexts. A possible explanation may lie in the biological complexity of STAT5B's function; perhaps its role shifts depending on mutation status, interacting partners, or activation through pathways like JAK-STAT. Unfortunately, the study does not explore the distinction between wild-type and mutant or activated forms of STAT5B, which could help clarify these opposing observations.

Similarly, while the authors discuss the divergent prognostic roles of STAT5A and STAT5B in cancer, they fall short of explaining why these highly homologous transcription factors behave so differently. STAT5A, for example, is described as protective in breast cancer but detrimental in blood cancers. The study acknowledges this contrast, yet does not explore whether STAT5B might show similar duality. Although STAT5B’s interaction with oncogenic partners such as STAT3 is briefly mentioned, no in-depth analysis is offered. This raises an important mechanistic question: could the clinical impact of STAT5B be influenced by heterodimerization with STAT5A or STAT3?

From a methodological standpoint, the study relies heavily on publicly available datasets such as TCGA and GTEx for gene expression analysis, and on Prognoscan for correlating STAT5B expression with survival. While this approach is practical, it comes with inherent limitations. Public databases can suffer from batch effects, inconsistent normalization, or lack of detailed clinical annotations. More importantly, the study does not include any form of experimental validation (such as qPCR or immunohistochemistry) to confirm the computational findings, which weakens the biological interpretation.

In the meta-analysis component, the authors pool data from 42 datasets, but the reported heterogeneity (I² = 64%) indicates significant variability in patient populations, treatment protocols, and detection methods. While subgroup analyses by cancer type help mitigate this, certain subgroups, such as ovarian cancer, are underpowered due to a small number of datasets. Furthermore, the study does not account for potential treatment effects; patients with higher STAT5B expression might have received more effective therapies, which could bias survival outcomes.

Another point of concern is the assumption that STAT5B downregulation at the mRNA level equates to loss of function. No protein-level data (e.g., via Western blot) are provided, and post-translational modifications (like phosphorylation) can significantly alter STAT5B activity regardless of transcript abundance. This is a crucial gap, as functional conclusions drawn from mRNA expression alone may be misleading.

On the biological front, the authors hypothesize that STAT5B may act through chromatin regulation mechanisms, similar to STAT5A and STAT3. While this is an intriguing idea, it remains speculative in the absence of direct experimental evidence. For example, no chromatin immunoprecipitation (ChIP-seq) or HP1 interaction assays are presented. Moreover, if STAT5B indeed plays a role in heterochromatin stability, it is unclear why such a protective effect would manifest only in certain cancer types.

Despite STAT5A and STAT5B sharing over 90% sequence homology, the study does not delve into potential structural or functional differences that might explain their distinct roles. STAT5B has a unique C-terminal domain, which could affect DNA binding affinity or interaction with co-factors, yet this aspect is not investigated. Likewise, no analysis is provided on whether the two isoforms regulate different gene sets or signaling pathways.

From a statistical perspective, some issues also arise. The authors conduct a publication bias check using Egger’s test, which yields a non-significant result (p = 0.116), suggesting no major bias. However, given that several cancer subtypes are represented by fewer than five datasets, funnel plots and tests like Egger’s may be underpowered and unreliable in this context. Additionally, negative or null results, where STAT5B is not prognostic, may be underreported in the literature and thus absent from the meta-analysis.

The multivariate analysis includes adjustments for age, stage, and tumor grade, which is a solid starting point, but it leaves out other potentially important confounders such as mutational status or treatment regimens. The analysis also does not include interaction terms; for example, whether STAT5B’s prognostic value depends on the presence of p53 mutations or other oncogenic drivers limiting its interpretive depth.

Lastly, while the authors claim to follow PRISMA guidelines for systematic reviews, some key methodological details remain vague. It is unclear how probes were selected across different microarray platforms, which could significantly affect gene expression estimates. Similarly, there is no discussion of how missing data were handled, a common issue in public datasets.

In conclusion, this study offers thought-provoking but incomplete insights into the role of STAT5B in cancer. It raises important questions: Is STAT5B a universal tumor suppressor, or does its role vary with cellular context and molecular background? Why does it appear protective in lymphomas yet possibly oncogenic in leukemias? Could treatment differences or molecular subtypes confound the observed survival benefits? And does STAT5B truly regulate chromatin structure, or is this merely inferred from related transcription factors? These questions remain open and warrant further investigation through targeted, experimental studies.

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript addresses an original and important topic by exploring the distinct roles of STAT5 isoforms in cancer prognosis. Its timely focus on differentiating between STAT5 isoforms contributes valuable insight to the existing literature and aligns with current research trends. The presentation is mostly clear and well-organized, effectively guiding the reader through complex datasets. Notably, the study demonstrates strong clinical relevance by linking STAT5B expression to patient outcomes across multiple cancer types. Overall, the manuscript makes a comprehensive and impactful contribution to the fields of cancer biology and translational medicine.

However, one key area for improvement is the clarification of the study’s aim and its alignment with the discussion and conclusions. While the study is described as a systematic review, it is conceptualized like a meta-analysis, with results interspersed with interpretive commentary that belongs in the discussion. Additionally, the discussion itself is brief and underdeveloped. Despite the manuscript’s title referencing both STAT5A and STAT5B, the analysis is limited to STAT5B, with no comparative evaluation of STAT5A. This comparison should be included to fulfill the stated scope of the study. The discussion would also benefit from a more thorough examination of the potential applications of these findings. The authors should elaborate on how the insights from this study, alongside their previous work (reference 5), could inform future diagnostic or therapeutic strategies.

 

Minor points:

Please review the formatting of Table 1 and ensure Figure 2 is of sufficient resolution for clarity.

In Figure 4, the meaning of study numbers (1, 5, and 6) should be clarified.

The "Summary of Key Findings" appears misplaced at the end of the Results section and would be better integrated into the Discussion.