Review Reports - Myelodysplastic Neoplasms (MDS) with Ring Sideroblasts or <i>SF3B1</i> Mutations: The Improved Clinical Utility of World Health Organization and International Consensus Classification 2022 Definitions, a Single-Centre Retrospective Chart Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors conducted a thorough study of the utility of the WHO classification for MDS with RS or SF3B1 mutations within their center. They focused on patients with SF3B1 mutations who met all other criteria for this WHO subtype and detailed their characteristics, including RS count, VAF, co-mutations, and cytogenetics, albeit without outcome. The study was impeccably planned and executed, and the manuscript is well-crafted. However, the absence of novel findings or a fresh research inquiry is notable. Given the established use of the WHO classification for some time now and previous descriptions of similar patient cohorts (e.g., Malcovati et al., Bernard et al.), as well as investigations into treatment response (such as the Medalist trial), the novelty of the results is somewhat lacking, although the references are accurately cited.

Furthermore, the authors' proposal for mutation analysis over morphology fails to acknowledge the necessity of excluding certain cytogenetic aberrations for defining MDS with SF3B1 mutation or RS. This limitation of their suggestion should be addressed in the discussion.

Minor revisions:

Line 51: This statement pertains to reference 12, not 13.

Line 109 and Table 1: The age range should be presented as 57 to 92, rather than 92 to 57.

Line 112: It is improbable that all patients had an IPSS score of 0. If even one patient exhibited 2 cytopenias, the upper range would increase to 0.5. At least the patient with inv(13q) would have at least 0.5 IPSS points.

A key finding of this study is the sensitivity of morphology, which is only 74% (as noted in line 188), and the characterization of patients without RS (as stated in line 136). While these patients have been described previously by Malcovati et al., the current study underscores the fact that not all patients with SF3B1 mutations exhibit RS.

While the manuscript meets technical standards and requires only minor revisions, the lack of novelty may raise concerns. Nevertheless, it still holds relevance for publication in Current Oncology. It underscores the importance of conducting immediate mutation analysis in confirmed cases of MDS to ensure appropriate treatment for patients with SF3B1 mutations, as these individuals may be overlooked based solely on morphology.

Author Response

Dear Reviewer,

Thank you for the careful and detailed review of our study and the opportunity to respond and address the questions and concerns. Please find below a point by point response with corrections and changes attached with the resubmission.

Sincerely,

Shamim Mortuza and Cyrus C. Hsia on behalf of the investigators.

Comments and Suggestions for Authors

Response: Thank you for the review of our manuscript and the suggestions. We will add this acknowledgement of excluding cytogenetic aberrations with mutation analysis over morphology in the discussion. However, it should be clarified, that we are not suggesting that molecular diagnostics only in all bone marrows tested for MDS but only appropriated triaged bone marrows. In other words, if a bone marrow shows MDS morphologically, then standard cytogenetics with chromosomal karyotyping is appropriate, but in bone marrows that do not morphologically fit the picture of MDS, then molecular diagnostics should be performed first before standard cytogenetics. See Kawata et al (2021 Feb;192(4):729-736. doi: 10.1111/bjh.16891. PMID: 32588428.).

Minor revisions:

Line 51: This statement pertains to reference 12, not 13.

Response: Reference 12 is Arber DA et al which is the ICC 2022 but reference 13 is Khoury JD et al which is the WHO2022 classification. This sentence referencing the WHO2022 Table 3 addendum states that “Detection of ≥15% ring sideroblasts may substitute for SF3B1 mutation.” So we believe reference 13 is correct.

Line 109 and Table 1: The age range should be presented as 57 to 92, rather than 92 to 57.

Response: Thank you. This has been corrected.

Response: Yes, this has been corrected. There were 7 patients with IPSS 0.5 and the remaining 32 patients with IPSS 0. The manuscript text and table have been corrected.

Response: We agree

Reviewer 2 Report

Comments and Suggestions for Authors

The Authors provide an interesting paper about MDS with SF3B1 mutation.

How many, if any, patients experienced a disease progression to MDS high risk (eg MDS/AML according to ICC 2022) or AML? What are the survival data (eg OS, LFS...?)

Author Response

Dear Reviewer,

Sincerely,

Shamim Mortuza and Cyrus C. Hsia on behalf of the investigators.

Comments and Suggestions for Authors

The Authors provide an interesting paper about MDS with SF3B1 mutation.

How many, if any, patients experienced a disease progression to MDS high risk (eg MDS/AML according to ICC 2022) or AML? What are the survival data (eg OS, LFS...?)

Response: Thank you for the review of this manuscript. The questions surrounding outcomes such as disease progression and survival data are interesting and important for patients with MDS with SF3B1 mutation but are beyond the focus of this paper which is to ensure that molecular testing for SF3B1 is being done despite the absence of RS.

However, given that we have been allocated more space for wording, we will include these additional requests. These findings on outcomes will have all the usual limitations from this retrospective study with missing information but also may not be fully representative of all MDS with SF3B1 patients at our centre as we may have more patients in the 282 patients who are lacking data. A description of the limitations will be included.

The 39 patient charts were reviewed and updated until the end of Feb 2024. There were 14 deaths (35.9%). This may seem like a high death rate for such a lower risk and good prognosis population. However, please keep in mind that some of these patients may have been diagnosed with MDS years before we had NGS testing available. Some patients were diagnosed in 2009 for example. Despite the number of deaths recorded, none of the patients had disease progression or documented death due to their MDS. In 10 patients the deaths were documented as unrelated due to falls and trauma (3), underlying cardiac issues (3), renal failure (1), Crohn’s disease (1), other malignancies (2). Four patients, the cause of death was unknown. Median duration of follow-up from time of diagnosis to last follow-up assessment or death was 35 months (range 3-180) in the 39 patients with median overall survival (OS) 37.5 months (range 3-180) in the 14 patients who died. Further, this length of follow-up may be underestimated as some patients may have had MDS prior but due to the pandemic, investigations including NGS and bone marrow testing were deferred.

So, what about disease progression (to higher risk MDS), progression to acute leukemia, or leukemia-free survival? We included only MDS with SF3B1 patients meeting inclusion criteria with lower risk MDS. This would underestimate the true number of patients who would have been eligible at time of diagnosis. The reason for this is that testing for SF3B1 mutation may have been much later (since 2018) than the initial diagnosis (prior to 2018) in some patients and if their disease progressed by the time they had NGS testing, they would not be included in our study. We had to use this exclusion of higher risk MDS as we wanted to show that MDS with SF3B1 patients could be candidates for lower risk treatments such as luspatercept. Unfortunately, with this retrospective study, we lacked the data to distinguish those patients with higher risk disease at the time of NGS testing but had lower disease at the time of diagnosis of MDS. Unfortunately, this also meant that we would be missing those patients that have MDS with SF3B1 mutation with lower risk disease at diagnosis who could have transformed.

So, as of last retrospective review of these 39 patients end of February 2024, other than the 14 patient deaths, there were no disease progression or transformation to acute leukemia in the remaining 25 patients.

We hope that the additional data and discussion added to the revision will satisfy this question.

Reviewer 3 Report

Comments and Suggestions for Authors

In the manuscript entitled “Myelodysplastic Neoplasms (MDS) with Ring Sideroblasts or SF3B1 Mutations: The Improved Clinical Utility of World Health Organization and International Consensus Classification 2022 Definitions, A Single Centre Retrospective Chart Review”, the authors collected a total of 6817 patients with suspected hematologic malignancy underwent molecular testing using a next generation sequencing based genetic assay and 395 MDS patients at their medical center from January 1, 2018 to May 31, 2023. They identified 39 evaluable patients with lower risk MDS with SF3B1 mutations and studied those cases. The authors stated that their findings support the revised 2022 WHO and ICC definitions of the molecularly defined subtype of MDS. Since there were many large-sized studies published on MDS with SF3B1 mutations and ring sideroblasts, this manuscript provides little new information in my opinion. I have some major concerns about their results as shown below.

1. In the Abstract and the Tables, they stated 39 cases of MDS with SF3B1 mutations. However, Figure 1 shows “Cases with SF3B1 mutations (n = 395)”. I am totally confused about the total number of MDS cases and the number of the cases of MDS with SF3B1 mutations.

2. What were the diagnoses on those excluded cases with SF3B1 mutations? I want to know this answer because the rate of SF3B1 mutations in the manuscript is much lower than the rates in those reported large series of MDS cases.

3. What does the term of “suspected hematologic malignancy” mean? How many were benign vs malignant? Lymphoid vs myeloid?

4. In Table 2, almost all of those cases had single lineage dysplasia (SLD), even those cases with multiple mutations and/or cytogenetic abnormalities such as del(20q) or del(13q). In our experience, such cases typically have multilineage dysplasia.

5. I highly recommend comparing the hematologic indices, transfusions, clinical courses and therapeutic responses between the SF3B1-only case and the cases with SF3B1 plus other genetic mutations.

Author Response

Dear Reviewer,

Sincerely,

Shamim Mortuza and Cyrus C. Hsia on behalf of the investigators.

Comments and Suggestions for Authors

In the manuscript entitled “Myelodysplastic Neoplasms (MDS) with Ring Sideroblasts or SF3B1 Mutations: The Improved Clinical Utility of World Health Organization and International Consensus Classification 2022 Definitions, A Single Centre Retrospective Chart Review”, the authors collected a total of 6817 patients with suspected hematologic malignancy underwent molecular testing using a next generation sequencing based genetic assay and 395 MDS patients at their medical center from January 1, 2018 to May 31, 2023. They identified 39 evaluable patients with lower risk MDS with SF3B1 mutations and studied those cases. The authors stated that their findings support the revised 2022 WHO and ICC definitions of the molecularly defined subtype of MDS. Since there were many large-sized studies published on MDS with SF3B1 mutations and ring sideroblasts, this manuscript provides little new information in my opinion. I have some major concerns about their results as shown below.

In the Abstract and the Tables, they stated 39 cases of MDS with SF3B1 mutations. However, Figure 1 shows “Cases with SF3B1 mutations (n = 395)”. I am totally confused about the total number of MDS cases and the number of the cases of MDS with SF3B1 mutations.

Response: Apologies for the confusion and thank you for catching this. The flow diagram is missing a block that will help explain this and is redrawn for clarity. Of 6817 patients, there are 395 with SF3B1 mutations. Of these 395 with SF3B1 mutations, 113 were excluded as they were not patients followed at our centre. Of the remaining 282 (395-113, this box has been added), there were 39 that had fully evaluable data. Unfortunately, for the 243 (282-39), these patients were excluded due to alternate diagnosis such as myeloproliferative neoplasms, had missing data (such as lack of a bone marrow aspirate/biopsy or it was done elsewhere that we do not have access to), or did not meet inclusion/exclusion (such as having higher risk MDS or another concurrent hematologic malignancy such as myeloma or lymphoma).

We hope that the revised figure is more understandable.

Please note that there are likely more MDS patients that do not have SF3B1 mutations that were not included in our study or analysis as they would have been excluded in the 6422 patients without SF3B1. Also, there are likely MDS with SF3B1 mutation patients excluded who were not assessed at our centre (in the 113 patients) and without complete evaluable data or not meeting inclusion criteria (in the 243 patients). Finally, our 39 MDS with SF3B1 patients meeting inclusion criteria is still an underestimate of the true number of patients who would have been eligible at time of diagnosis. The reason for this is that testing for SF3B1 mutation may have been much later (since 2018) than the initial diagnosis (prior to 2018) in some patients and if their disease progressed by the time they had NGS testing, they would not be included in our study. We had to use this exclusion of higher risk MDS as we wanted to show that MDS with SF3B1 patients could be candidates for lower risk treatments such as luspatercept. Unfortunately, with this retrospective study, we lacked the data to distinguish those patients with higher risk disease at the time of NGS testing but had lower disease at the time of diagnosis of MDS

What were the diagnoses on those excluded cases with SF3B1 mutations? I want to know this answer because the rate of SF3B1 mutations in the manuscript is much lower than the rates in those reported large series of MDS cases.
What does the term of “suspected hematologic malignancy” mean? How many were benign vs malignant? Lymphoid vs myeloid?

Response: We will address both of these questions/comments together. Of 6817 patients tested, there are 395 with SF3B1 mutations, and 6422 were excluded who did not have SF3B1 mutations. We have to back up and explain the dynamics of myeloid NGS panel testing at our centre. Since Jan 1, 2018, our centre provided access to the myeloid NGS panel for any patient “suspected of having a hematologic malignancy”. There were no restrictions on the ordering physician for which patients to test and they could order the test on both peripheral blood and bone marrow samples. Thus, any patient presenting with any cytopenias (anemia, thrombocytopenia, neutropenia, bicytopenia, or pancytopenia), cytosis (polycythemia/erythrocytosis, thrombocytosis, neutrophilia, or panmyelosis), work-up for symptoms or signs (such as fever, night sweats, weight loss, or splenomegaly), or other indications for testing could have had an NGS test done.

A breakdown of the percentages of various conditions from benign vs malignant and lymphoid vs myeloid have previously been described in our initial few years cohort in (Bhai et al. 2022 doi: 10.1007/s40291-022-00581-7. PMID: 35381971).

In Table 2, almost all of those cases had single lineage dysplasia (SLD), even those cases with multiple mutations and/or cytogenetic abnormalities such as del(20q) or del(13q). In our experience, such cases typically have multilineage dysplasia.

Response: Thank you for pointing this out. We have gone back and corrected the IPSS scores based on cytopenias by the initial definition of hemoglobin < 100, platelets < 100, and neutrophils <100. There were 7 patients with IPSS 0.5 and the remaining 32 patients with IPSS 0. Further, we have re-evaluated the designation of SLD & MLD with WHO2016 criteria rather than the initial bone marrow aspirate report "favouring SLD" often. There are 23 RS-SLD, 7 RS-MLD, 7 SLD, and 2 MLD.

The manuscript text and table have been corrected.

I highly recommend comparing the hematologic indices, transfusions, clinical courses and therapeutic responses between the SF3B1-only case and the cases with SF3B1 plus other genetic mutations.

Response: The questions surrounding outcomes and comparing hematologic indices, transfusions, clinical courses and therapeutic responses are interesting and important for patients with MDS with SF3B1 mutation but are beyond the focus of this paper which is to ensure that molecular testing for SF3B1 is being done despite the absence of RS.

Of the 39 patients, there were 14 deaths (35.9%). There were no patients with dis-ease progression or documented death due to their MDS. In 10 patients the deaths were documented as unrelated due to falls and trauma (3), underlying cardiac issues (3), renal failure (1), Crohn’s disease (1), other malignancies (2). For four patients, the cause of death was unknown. Median duration of follow-up from time of diagnosis to last fol-low-up assessment or death was 35 months (range 3-180) in the 39 patients overall. The median overall survival (OS) was 37.5 months (range 3-180) in the 14 patients who died.

We identified 19 (48.7%) patients with isolated SF3B1 mutations; the median age was 76 years (57-90) with 10 females and 9 males. The median hematologic indices were WBC 7 x 109/L (4-12), hemoglobin 97 g/L (65-151), MCV 105 fL (90-124), platelets 245 x 109/L (24-651), and neutrophils 4 x 109/L (2-10). Prognostic scores showed IPSS score 0 and 0.5 in 18 and one patient respectively, R-IPSS very low and low in 5 and 14 patients re-spectively, and M-IPSS very low and low in 12 and 7 patients respectively. Nine patients experienced a transfusion while eight were deemed transfusion dependent and ten patients were exposed to an ESA and one patient received an EMA. Median duration of follow-up from time of diagnosis to last appointment or death was 37 months (11-137). There were seven deaths with median overall survival 44 months (19-137) from time of MDS diagnosis.

There were 20 (51.3%) patients with SF3B1 plus other molecular mutations; the median age was 79 years (64-92) with 9 females and 11 males. The median hematologic indices were WBC 6 x 109/L (2-12), hemoglobin 94 g/L (71-111), MCV 105 fL (89-123), platelets 398 x 109/L (46-398), and neutrophils 3 x 109/L (1-9). Prognostic scores showed IPSS score 0 and 0.5 in 14 and 6 patients respectively, R-IPSS very low and low in 5 and 15 patients respectively, and M-IPSS very low and low in 3 and 17 patients respectively. Nine patients experienced a transfusion while seven were deemed transfusion dependent, seven patients were exposed to an ESA and four patients received an EMA. Median duration of follow-up from time of diagnosis to last appointment or death was 19 months (3-180). There were six deaths with median overall survival 14 months (3-180) from time of MDS diagnosis.

We compared the group of 19 (48.7%) patients with only SF3B1 mutations to the group of 20 (51.3%) patients with SF3B1 plus other molecular mutations. Both were sim-ilar in age, sex distribution, hematologic indices, IPSS and R-IPSS scores. They differed in their M-IPSS scores as expected with the additional mutations in the latter group. Although both groups had similar percentage of exposures to transfusions and transfu-sion dependency, the latter group seemed to have a shorter overall survival. Median overall survival was 44 months in the seven patients with only SF3B1 mutations com-pared with 14 months in the six patients with SF3B1 plus other mutations who died. However, this was not statistically significant (p = 0.37).

We hope that the additional data with inclusion in the results and discussion added to the revision will satisfy this question.