Review Nr. 1: Usage of anti-NPM1 antibodies for diagnostics and research of acute myeloid leukemia: a systematic review

Reviewer: Laurens Zaschke

Article: Usage of anti-NPM1 antibodies for diagnostics and research of acute myeloid leukemia: a systematic review

Author: Maximilian Zuleeg

Summary and Recommendation

In his article, Maximilian Zuleeg presents the current state of research on the question of whether anti-NPM1 antibodies improve diagnostics of acute myeloid leukemia (AML). He compares mono- and polyclonal antibodies from different research groups with the current gold standard PCR in AML diagnostics with regard to sensitivity and specificity. Zuleeg concludes that while some antibodies have shown good sensitivity in detecting NPM1 mutations, none has yet made it to a commercially available true alternative to PCR. Zuleeg, however, sees potential applications for anti-NPM1 antibodies in basic AML research in tracking individual cell lines.

The paper is clearly and comprehensibly structured and introduces all the technical background necessary to understand the argument. The only thing missing in the introduction is a better explanation of the relevance of anti-NPM1 antibodies as a substitute to PCR.

The results of the study can be well understood on the basis of the methods described, but are rather sobering with regard to the diagnosis of AML, also because the studies reviewed are still partly too incomplete to be adequately classified with regard to the research question of the paper. Due to this lack of data, the analytical gain in knowledge seems rather low compared to the high descriptive value of the work. In addition, the presentation of results sometimes lacks coherence, as a few rather specific aspects are mentioned in the analysis of the different working groups cited (e.g. differences in Fc protocols or statistical evaluations), but these are not taken up again in the discussion and their relevance therefore remains unclear.

The language of the text is factual and comprehensible, but lacks precision in some places (see below), e.g., when referring to cited publications. In addition, there is a lack of references in some passages to support stated facts and conclusions.

All in all, Zuleeg’s article is an interesting work and a good summary of the current state of research on anti-NPM1 antibodies. After addressing the Major Issues mentioned below, I would therefore recommend to accept the manuscript. I am happy to review the paper a second time after revision.

Detailed Review

For each section of the paper, I summarize below particularly positive points (P) as well as criticisms, which I divide in Major (MJ) and Minor (MN) Issues and list in the order they came to my attention throughout the manuscript.

Title and Abstract

• P: Precise and meaningful title that summarizes well the topic of the study.

• P: Abbreviations are introduced before they are used.

• MJ: It seems to me that the statement "If mutated, it [NPM1] is one of the most common molecular causes for acute myeloid leukemia (AML)." is imprecise, since it is noted in section 1.3 that it is not clear whether mutation of NPM1 actually leads to a higher growth rate and thus could be considered a causal cause of AML, so to speak, or whether AML is merely associated with mutation of NPM1. I would rather put it as in 1.3 that NPM1 mutations have been found in many AML patients and therefore represent a potential biomarker for AML disease.

• P: Versatile description of the potential scope of NPM1 antibodies.

• MN: Somewhat lost in the detailed description of the applications of anti-NPM1 antibodies in general is what exactly is the analytical question of the study. As I understood it, the research question regards the comparison of anti-NPM1 antibodies with other diagnostic method for NPM1 mutations. This could be stated explicitly in the abstract.

Introduction

P: Basic information on functions of NPM1 and its role in AML pathophysiology is reasonably structured and presented in appropriate detail, so that one can understand the contents of the paper even without much prior knowledge of the gene in question and AML.

MN: The phrase "Additionally different publications show that NPM1 interacts with different tumor suppressors and proto-oncogenes, such as Adenosine diphosphate-Ribosylation Factor (ARF)." Implies several publications supporting the statement, but only one is cited. This contradiction could be easily resolved by a specific reference to the cited study or wording in the passive.

MJ: References are missing for the first 5, fact-filled sentences of section 1.2.

MN: "In particular, for NPM1-mutated AML, there is a clear lack of development of more specific targeted therapies." I feel there is a lack of basis for this sentence from section 1.3. For example, in the paper cited shortly thereafter [16], it is noted that: "NPM1 mutation status did not influence treatment effectiveness". If NPM1 mutated AML patients need specific therapies, I miss the reference on this. Also, the sentence mentions target therapies, but the rest of the paper does not mention any paper that would have been about NPM1 mutant proteins as a therapy target. Rather, it is primarily about the role of anti-NPM1 antibodies in diagnostics, and thus the weakly documented reference to therapy seems inappropriate to me.

MN: Section 1.4 could be introduced in more detail, e.g. by not only writing "different types of antibodies", but by specifically distinguishing between polyclonal and monoclonal antibodies and mentioning that the antibodies in the class are again distinguished by the respective epitopes and whether they attack wild type or mutant. In its current version, the section is somewhat content-poor.

Methods

P: Naming of the period of research and concrete terms to be searched for

P: Reasonable inclusion and exclusion criteria

Results

MN: The flowchart Figure 1 showing the process of the review seems unnecessary in view of the comparatively simple procedure, which can also be well understood by the short text in 3.1. Nevertheless, the graphic design and labeling of the figure is well done and clearly arranged.

MN: In section 3.2, some references are missing in the second part. For example, in "Additionally, NPM1 mutations with aberrant cytoplasmic localization coupled with fluorescent proteins were detected by immunohistochemistry during different studies." several studies are mentioned, where it is not clear whether these are the studies mentioned in the following part of the paper or if further studies are meant. I also miss a reference for the description of the fluorescence detection of the antibodies, although it seems to be accurate to me.

P: Clear classification of the different antibodies, which makes the following presentation of the results much easier to understand.

MN: "The scientific community however has been looking for additional methods for diagnostics and research of NPM1 mutations." This is quoted without reference, although it does not seem self-evident to me. Why do we actually need anti-NPM1 antibodies in AML diagnostics when we already have a well-functioning gold standard, PCR? As indicated in the discussion, it seems to me to be a matter of cost, speed, and better spatial resolution with which to better describe tumor heterogeneity. The disadvantages of PCR could be mentioned earlier in the introduction (with references) so that it is clearer why alternative methods are being sought in the first place.

MN: What is the "most common" (Section 3.3) mutation that Gruszka et al. targeted?

MN: Is there a reason why Gruszka et al. used an Fc block protocol (section 3.3) and Tan et al. did not? And if so, is this information relevant to the comparison of the antibodies? This is not revisited in the paper and therefore should not be mentioned here if it has no further relevance.

MN: I see a similar problem in section 3.4, as it is not clear what relevance the normalization of mean fluorescence intensity by Du Pisani et al. has compared to the non-normalized parameter.

MN: Since in section 3.4 a chronological classification of the work of the two working groups is made (by which the use of different antibody kits is justified), it seems to me to make sense for the sake of clarity to include the year numbers also in the text, so that one does not have to take a look in the bibliography.

MN: The abbreviation MFI for "mean fluorescence intensity" is used without being introduced before.

P: Good separation between results and discussion. There are no interpretations in the results section.

Discussion

MN: The sentence "The defining factor of AML is the presence of 20% or more blasts in the bone marrow or in the peripheral blood." (Section 4.1) in my opinion fits better in the introduction, where AML has already been introduced as a clinical picture, but should in any case be backed up with a source.

MN: "The other publications [...]" (section 4.1) is too vaguely worded. Since there are only 2 others, I would name them specifically.

MJ: The second to last paragraph in section 4.1 falls a bit out of the line of argument and belongs for me rather in the introduction. The fact that detection of NPM1 mutations can spare cytogenetic diagnostics is a good motivation for improving NPM1 mutation diagnostics, but in my eyes does not play a role for the central question of the paper, whether anti-NPM1 antibodies are suitable for mutation diagnostics.

P: Good differentiation of the application of the different antibodies with regard to initial dignostics vs. MRD diagnostics in AML and comprehensible selection of the most promising candidates based on the results of the literature search.

P: Section 4.2 gives a nice outlook on how anti-NPM1 antibodies can be used to represent genetic heterogeneity of tumors in research and what aspects future studies on anti-NPM1 antibodies need to consider to be useful for this application.

Conclusion

P: Good referencing back to the question of what role anti-NPM1 antibodies can play compared to the gold standard PCR at the moment and concluding outlook on a possible alternative application in basic research.

MN: The theory that the two most promising antibodies of the considered research groups are no longer commercially available because the results were not reproducible, I consider to be speculative and to be irrelevant in the conclusion with regard to the research question.

Second Round of Review

Maximilian Zuleeg has taken into account the most important points of criticism from my review. In particular, significantly more references have been placed and irrelevant or misleading passages have been removed. Therefore, I now recommend the editors to publish the paper.