HSCT with Mismatched Unrelated Donors (MMUD): A Comparison of Different Platforms for GvHD Prophylaxis

Round 1

Reviewer 1 Report

Peer review on

Massimo Berger et al.:

HSCT with Mismatched Unrelated Donors (MMUD): a Comparison of Different Platforms for GvHD Prophylaxis

Submitted to TRANSPLANTOLOGY                                                                    

The topic of the paper is of high clinical relevance – there are, however, despite the relatively small numbers of patients in each cohort, several (major or minor) issues with the manuscript that need to be addressed prior to publication.

Abstract:

• Were the ATG-treated patients uniformly given MTX/CSA post transplant? This information is later given in the Introduction, please specify briefly already in the Abstract.
• Please indicate the post SCT time for which the respective event probabilities were calculated (e.g., OS probability at xxx years was 42% vs. 64% for ATG and PT-Cy (P<0.0005).
• Please indicate in the abstract that the study is of retrospective nature, because the “ClinicalTrials.gov” registration number may suggest to the reader it might be a prospective study.

Methods

• What does this sentence mean: “The donors were randomly assigned to their patients, based on the availability of the Centers to carry out the transplant.” Above, the selection algorithm was described in a different way. Should this sentence describe the allocation of an individual patient to one of the two GVHD prophylaxis platforms? Were the respective, participating centers each performing one specific type of immunosuppression (i.e. all ATG based transplants in one center, and all PTCy based HSCT in the other(s)? Please clarify – this issue is of great importance for interpreting data.
• Patients with graft failure should not be censored. There is no reason to do so. In case only first transplants were considered for this study (as indicated), a patient with a second transplant might be followed as if no second HSCT had occurred. However, the cumulative incidence of graft failure (and a description of what happened with GF patients – incidence of retransplantation?) should be indicated for each platform (was there a higher GF rate in PTCy?). For example, if a patient is suffering (primary) GF (without evidence of primary disease persistence or recurrence), and will decease during a consecutive neutropenic infection, this is clearly a case of non-relapse-mortality. This subsequent event (treatment failure) cannot be censored.
• HLA Typing: (GvH=134 and HvG=6). Should be formulated as “GVH (including all bidirectional MM)=134 and …”

• “Relapse” should not include molecular relapse but only hematological or clinical relapse. Thus relapse-free survival should be defined accordingly. If the authors wish to consider molecular relapse as well, there may be an issue with not all patients having a molecular marker while others do so. One might define survival until any kind of relapse as “event-free survival”, while “relapse” should be reserved for the clinical definition. Please recalculate relapse incidence (and accordingly, NRM too) with only clinical/hematological relapse as the event of interest.

• Line 200-214: this paragraph seems not to belong to the manuscript…

Results

• The median FU is generally stated for surviving patients only. It is not necessary to indicate the median FU of deceased patients.
• EBMT score of 1 seems to be quite frequent, particularly in cohorts containing exclusively alternative donor transplants (accounting already for one point). 42% with EBMT score of 1 would mean that these 42% were below age of 20y and time dx to tx less than one year, and had no female donor to male recipient and had early stage disease. Was this indeed the case? It also seems to contrast with the respective proportions of intermediate and high DRI, which would suggest a higher EBMT score in the majority of patients, too, as does the high proportion of patients with time from Dx to Tx > 1 year. Please check.
• Reference numbers seem to be displaced (ref for risk scores should be 39-41 and not 40-42).
• Failure to achieve neutrophil /platelet engraftment: these numbers are quite high: it should be explained why these patients did not achieve full engraftment: was it because of early relapse/disease persistence? Or because of early fatal infections? Or were these patients considered primary graft failure, and if so, were they retransplanted, or did they recover blood counts at a later time point?
• Table 2) Please indicate the post tx time for the respective event probabilities (e.g. at 3 years?)

Discussion

• It should be discussed that the indicated CSA target trough level of 100-250 ng/ml may be somewhat low, particularly during the first weeks after transplant, in comparison with a published consensus recommendation (Ruutu et al, Bone Marrow Transplantation (2014) 49, 168–173). Liberal CSA management may have contributed to the higher incidence of GVHD and NRM in the ATG cohort. Please discuss.
• Another previous publication (Clausen et al, Biomedicines 2017, 5(2), 13)

has addressed the issue of the significant risk for mortality and NRM associated with an HLA MM even in ATG-based transplants. This reference should be cited and discussed, since the present study is supporting this previous finding. PTCy may circumvent the issue of NRM associated with an HLA MM, although at the price of a higher relapse rate. However, in the haploidentical setting (with generally more than 1-2 HLA-MM, a higher relapse rate may no longer be an issue with the PTCy concept. This might also be discussed.  

• The finding of a protective effect of an HLA DQ MM is quite interesting. Please check whether this effect can be separately revealed in the PTCy cohort, and in the ATG cohort (analyzed separately). It has been previously shown for PTCy based haploidentical transplants that class II mismatches may be protective towards relapse or NRM, respectively (Solomon et al. Biol Blood Marrow Transplant 24 (2018) 789–798.) Please include this reference to the discussion.

.

Author Response

In attach the reply to the reviewer 1

Author Response File: Author Response.docx

Reviewer 2 Report

General comments

The overall design of the study seems appropriate. The role of PT-Cy GVHD prophylaxis is an area of great interest in the context of MRD/MUD and MMUD HSCT transplant. 

The introduction and methods are well described (see specific comments below). The main comment is related to the results that should be better explained, tables and figures need more precise legends and data to better understand the outcomes.

A main concern is the missing CI of limited and extensive cGHVD (cGVHD evaluated only as a whole). I reccomend the evalutaion of the severity of cGVDH (possibly according to NIH) as this could represent a major difference between the 2 GHVD prophylaxis platforms.

I would not overspeculate on the role of different HLA mismatches (ie HLA-B, HLA-DQB1) as much higher numbers are needed to evaluate these effects. The results obtained on 39+40 patients could be not reliable.

I would alltogether shorten the article (mainly intro/methods).

I would review english language and style.

Specific comments

82-88 The information about the children's hospital is not relevant considering that only adults results are presented, I would eliminate this info as it is confusing.

200-214 Eliminate this part (guidelines to authors).

239-240 Clarify. You just say in the two groups (specify if ATG or PT-Cy).

242-242 Is this a repetition?

253 I would specify in the text the GVHD incidence in the 2 groups.

259 According to the table 3 older age seems to be protective for GVHD II-IV. Is this correct?

Table 2. A clear legend is needed. A clear heading is needed (are those CI of the different outcomes?). In the text also HLA mismatch are cited as factors influencing outcomes (ie line 289) but they are not in the table. A better graphic could make the table easier to read.

Table 3. A clear legend is needed. A clear heading is needed (are those OR? please specify). Are the missing data (-) not presented because not significant? Please specify how to read the table.

286. I reccomend to evaluate cGHVD severity evaluation (NIH?).

296. I would specify the OS in the 2 groups.

308. Why do you refer to figure 3?

309. I would specify the TRM in the 2 groups.

318. I would specify the RI in the 2 groups.

377. Where the number 31 come from? 

386. Specify the meaning of 'role'?

Author Response

In attach the reply to the reviewer 2

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Review of Manuscript revision I

All suggestions have been implemented in the revised version of the manuscript.

• As for the discussion of Reference 58 (Clausen et al), I would suggest discussing it in the context of the

”increased risk for NRM associated with an HLA-MM even in the setting of ATG-based transplants, as previously shown in retrospective study of 775 patients.... (Ref 58) .  PTCy may circumvent the issue of NRM associated with an HLA MM, although at the price of a higher relapse rate, as suggested by the present findings.”  

The part of the sentence “given a sibling/unrelated or mismatched unrelated” should be deleted.

• Grammar and spelling should be checked.