Frequency of Non-Type I/II CALR Mutations in Patients Undergoing Molecular Diagnostics for Myeloproliferative Neoplasms

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In this article “Frequency of atypical CALR mutations in patients undergoing molecular diagnostics for myeloproliferative neoplasms” by Tatiana Makarik et al., the authors have performed a retrospective analysis of atypical mutations in the calreticulin (CALR) gene on a large number of patient samples and attempted to find any link to the pathogenesis of myeloproliferative neoplasms (MPNs). They have found that most of the CALR mutations belong to the typical type I or type II class and only a few patients show atypical CALR mutations. Most of these atypical mutations are inframe and germline mutations. The authors state that the germline atypical mutations might not be a diagnostic marker for MPN patients and additional driver mutations could be involved in MPN pathogenesis. This attempt by the authors to look for additional mutations in the CALR gene in the context of MPNs is a welcome one. The manuscript is written well. However, there remain certain shortcomings in the author’s description of the results, rational of this study and in putting the findings into context with regard to MPN pathogenesis. The authors need to address the comments below to make the study relevant and robust.

Comments:

1. The significance of these atypical CALR mutations with respect to MPN pathogenesis is unclear from this analysis. The authors need to add a perspective to this central aspect.
2. Line 57, indicate what is “Phi-negative” for the benefit of the readers.
3. According to the demographic parameters in Table 1, vast majority of patients analysed (7785 out of total 8417) show no mutation in the CALR gene. Even among the 632 CALR+ patients, only a very minor fraction shows atypical mutation. In this context it is not clear what is the main focus of this study. The authors need to clarify their focus and accordingly modify the text.
4. The atypical mutations are highest in the area of allelic load of 45-50% as evident from Fig. 1. So also, the type I and type II mutations. Is there any significance to it? The authors could possibly discuss about it.
5. Regarding lines 116-118, does it correlate with the disease pathogenesis in any way? The authors need to discuss about it.
6. In Fig. 1, X- and Y coordinates need to be labelled clearly for a better understanding of the readers.
7. Lines 132-133, is there any correlation to disease pathogenesis or these additional mutations are just coincidence? Discuss.
8. Table 2 title should be "Clinical parameters......"
9. All abbreviations such as COSMIC, MDS RAIB-2 ETC., in the text should be expanded in full at the point of first use. Carefully look at these and amend them accordingly.

Author Response

Please see attached file

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript has some major problems in understanding the nature of pathogenicity of CALR mutations in MPN, as well as the principles in variant classification. 

The pathogenicity of CALR exon 9 frameshift mutations in MPN is to delete the negatively charged amino acids and/or to alter them into a stretch of positively charged amino acids in the C-terminal calcium binding site. Please see Klampfl et al NEJM 2013 Figure 2 B and C panels, as well as D Pietra et al. (PMID: 26449662) Figure 1. With this model, only variants detected in patients 2, 4, and 7 has this property. According to D Pietra et al, Patient 2 p.K368Rfs*51 is Type 1-like with stretches II and III deleted, Patient 4 p.E383Cfs*46 is Type 2-like with stretches I, II, and III preserved, and Patient 7 p.K374Rfs*52 may be defined as "Other type" intermediate between types 1 and 2 with only stretch III deleted. All the in-frame deletions of single amino acid in the other patients would not cause the alteration of the amino acid charges after stretch III, and they are unlikely pathogenic (likely benign) and should not be included in this cohort. 

Variant classification: Sequence deviations from the reference genome (hg19 or hg38) do not automatically mean that they are pathogenic or "mutations". They can also be benign variants that are rare but still circulating in the population. In fact, most of the in-frame deletion variants in this cohort can be found in gnomAD. Although hematologically pathogenic variants such as JAK2 V617F can also be seen in gnomAD, allele balance for heterozygotes and age distribution are very good indicators to differentiate germline circulating variants (likely benign) and recurrent somatic mutations (likely hematologically clonal/oncogenic). The presence in COSMIC does not mean they are recurrent somatic mutation but were most likely included from germline. As the authors rightfully suggested, there are no reliable reports of germline pathogenic variants that might cause MPN. 

Minor suggestion(s): 

Figure 2 is unnecessary and can be reported with one sentence clearly. 

For Patient 2, the BCR::ABL1 p210 VAF is very low. What is the fraction by FISH? Was it because the CML is low level incidental finding or the rearrangement was not covered by the PCR probe?

For Patient 4 and 7, the VAFs for JAK2 V617F are also very low, they might be a CALR-driven MPN with incidental JAK2 clonal hematopoiesis. 

 

Author Response

Please see attached file

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This brief communication piece describes atypical CALR variants in a cohort of patients with suspected myeloproliferative neoplasms (MPNs) and a control group. The strength of this manuscript is in the number of MPN patients tested (8417) and the presence of a control group. CALR type I and type II mutated MPNs represent a unique molecular subtype of MPNs. They are biologically distinct from other MPNs. The presence of these mutations aids in proving clonality and thus establishing a diagnosis of MPNs. This is important clinically for patients to obtain a correct diagnosis and have access to therapies. In this paper the team performs molecular testing in 8417 cases suspected to have an MPN and in a smaller control group. The lack of information for the control group (age, sex) is an issue but interestingly no CALR variants were found in the control group. The paper focuses on variants in exon 9 but does not mention much about variants outside exon 9. Overall, this paper adds to the current literature regarding atypical CALR variants in a large cohort of MPNs but could benefit from the addition of more granular data.

 

Major comments:

• Atypical CALR variants outside exon 9 were the most common but they are only mentioned once in the manuscript. Please provide more details about these. Are they mostly benign germline SNPs?
• Please provide general population frequencies for the atypical CALR exon 9 variants. Although they were not found in the control group, the control group is pretty small and that does not exclude them from being found in the general population. If they are found commonly enough in the general population, they could be benign germline polymorphisms as the authors alluded to.
• Did the laboratory provide germline classification for the germline CALR variants? The ACMG classification is the most accepted. Are these benign/likely benign or are they VUSs? Did the authors search for the variants in Clinvar? The authors cited a paper that reported germline variants in CALR as mostly benign, but laboratories will also deposit germline variants in Clinvar without necessarily publishing them in manuscripts. Clinvar is an important tool to investigate germline variants.
• There is no mention of family history for any of these cases. If the authors considerd the possibility that these germline variants increased the risk of MPNs then family history of MPNs would support or refute this hypothesis. Please provide family history if available.

 

Minor comments:

• Page 1 line 36: “In 2013, 35 the discovery of mutations in the CALR gene encoding calreticulin in 70-84% of patients 36 with essential thrombocythemia (ET) and primary myelofibrosis (PMF) significantly 37 improved the molecular diagnosis for these diseases”.

CALR mutations are in 70-80% of JAK2/MPL negative ET and MF. Please rephrase to make this statement accurate.

• Page 2, line 45: the reference for the 2022 WHO classification is for the lymphoid not the myeloid update. The correct reference is Khoury et al.
• Page 3 line 57: typo, this is supposed to be ph-negative not phi-negative
• I suggest testing for JAK2 exon 12 mutations in the triple negative cases. These are rare but are documented to occur in some cases of MPN. The authors did not test for these.
• Figures: figures have titles, but they do not have a figure legend. Also please label X and Y axis on figure 1
• Tables 2 and 3: I would combine these into one big table to make it easier for the reader to reference the variant and clinical characteristics
• Page 5: line 146 define RAIB-2, Line 148: are these CD markers positive on the blasts? Line 157: do you mean grade or stage of the cytopenias? Grade is more commonly used

 

Author Response

please see attached file

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

References need to be updated with most recent literature (recent years)

I think the diagnosis regarding patient n.6 has to be more precise, clear and has to be completed with morphological data about bone marrow (is it AML according to WHO 2022? which is the precise diagnosis?) It seems it lacks on information. As regards the therapy, has it been planned according to some protocol? wich one? it has to be specified, it is not clearly described.

Case n. 2 and n. 9 are CML? this seems to be according to molecular data, and not PMF or MPN as indicated in table n. 2

You surely have to review both laboratory parameters and diagnosis regarding patients with atypical mutations.

Author Response

Please see attached file

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors are quite receptive to the comments made on the original submission and have responded to the comments satisfactorily. The revised version is better in its focus and explains the observations in proper context.

Author Response

Dear Reviewer,

We are really grateful for the feedback and remarks that assisted us in enhancing the manuscript, as well as the positive decision.

Best regards, Andrey Sudarikov

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the response and revision. I agree with the authors that the in-frame changes can cause mis-attribution of the MPN to benign CALR variants. Here, there are some suggestions in clarifying this point. 

First, although the authors have added the description of Type I and Type II frameshift mutation of CALR, the authors should also describe the charge changes of the C-terminus rather than just "loss of KDEL domain".Please see Klampfl et al Figure 2 B and C panels. 

Second, the authors should use the correct terminology to ensure clarity. "Mutation" should be avoided, although I am ok if it means somatic (different from the constitutional/germline sequence of the individual) or de novo (different from parents) variants. Mutations should not be used to describe the (benign) germline variants/polymorphisms such as those in-frame changes in CALR. Please refer to https://www.gimjournal.org/content/authorinfo or https://www.sciencedirect.com/science/article/pii/S109836002104082X?via%3Dihub

In addition, the term "atypical CALR mutations" is very confusing. 

Third, The methodology describing JAK2 V617F and MPL W515L/K is unacceptable. Please specify what assays are these two "kits". Are they PCR, allele-specific PCR, dd-PCR, Sanger sequencing, or other type(s) of assays? Most of readers are not familiar with Syntol's products and the product lines are always subjected to change for the same company. Also, please include detection limit of these two assays, since low level somatic variant of JAK2 V617F is common even in healthy individuals. 

 

Comments on the Quality of English Language

In multiple places, germline variants/polymorphisms was misspelled as "germinal", which means very different things. 

Author Response

Dear Reviewer,
We are deeply thankful to you for your careful attention to our modest work and thorough review, which helps us to correct all the mistakes and inaccuracies we overlooked. We made an effort to correct everything, and you can find our replies to your remarks below marked in red.
Comments and Suggestions for Authors
Thank you for the response and revision. I agree with the authors that the in-frame changes can cause mis-attribution of the MPN to benign CALR variants. Here, there are some suggestions in clarifying this point.
First, although the authors have added the description of Type I and Type II frameshift mutation of CALR, the authors should also describe the charge changes of the C-terminus rather than just "loss of KDEL domain". Please see Klampfl et al Figure 2 B and C panels.
We have added these clarifications and the corresponding reference (lines 46-49 of the manuscript)
Second, the authors should use the correct terminology to ensure clarity. "Mutation" should be avoided, although I am ok if it means somatic (different from the constitutional/germline sequence of the individual) or de novo (different from parents) variants. Mutations should not be used to describe the (benign) germline variants/polymorphisms such as those in-frame changes in CALR. Please refer to https://www.gimjournal.org/content/authorinfo or https://www.sciencedirect.com/science/article/pii/S109836002104082X?via%3Dihub
We absolutely agree with this comment and replaced the term “mutation” with “variant” in all instances pertaining to the germline condition. We also appreciate the reference to a highly valuable discussion of terminological mistakes in genetics, some of which we were unable to avoid.
In addition, the term "atypical CALR mutations" is very confusing.
The term "atypical" might indeed appear too vague in this context, so we substituted it with "non-type I/II".
Third, The methodology describing JAK2 V617F and MPL W515L/K is unacceptable. Please specify what assays are these two "kits". Are they PCR, allele-specific PCR, dd-PCR, Sanger sequencing, or other type(s) of assays? Most of readers are not familiar with Syntol's products and the product lines are always subjected to change for the same company. Also, please include detection limit of these two assays, since low level somatic variant of JAK2 V617F is common even in healthy individuals.
The assessment of JAK2 V617F and MPL W515L/K was conducted via quantitative (JAK2) and qualitative (MPL) allele-specific real-time PCR. For JAK2 V617F kit they declare detection limit of 0.2% of variant allele. We also have UKE1 cell line in the lab and use DNA from it as a positive (VAF 100%) control. The relevant text has been inserted in lines 115-117 of the manuscript.
For JAK2 V617F, we participated in several interlaboratory comparative studies, including international (https://link.springer.com/article/10.1007/s00277-018-3570-8). Thus, we are confident in the assay we use. As for MPL W515L/K kit they also declare 1% detection limit, however we cannot approve it. Undoubtedly, the amplification efficiency and specificity of W515L and W515K alleles is different, and K can be misinterpreted as L, especially at low VAFs values. Therefore, we are forced to confirm with NGS all requests for MPL analysis with questionable allele-specific PCR result.
Comments on the Quality of English Language
In multiple places, germline variants/polymorphisms was misspelled as "germinal", which means very different things.
Thank you for mentioning this. Fixed.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have sufficiently addressed my comments.

Author Response

Dear Reviewer,

Thank you for the criticism and comments that helped us improve the manuscript and your positive decision.

Best regards,  Andrey Sudarikov

 

Reviewer 4 Report

Comments and Suggestions for Authors

all changes are appropriate

at line 159 the term decorated has to be modified because it could be not understood

Author Response

Dear Reviewer,

Thank you for your positive assessment of our efforts.  We believe the manuscript greatly benefited from your feedback. We overlooked an odd statement regarding megakaryocytes being "decorated".  Thanks for noticing.  It is now changed to "An immunohistochemical study revealed that granulocytes express MPO and CD15, megakaryocytes express CD31, and myeloblasts express CD117."

Best regards, Andrey Sudarikov.