Thrombotic Complications in Immune Thrombocytopenia Patients Treated with Avatrombopag

Round 1

Reviewer 1 Report

Interesting case reports on thrombosis in patients on TPO-RA for chronic ITP.

In case #1 would include how many weeks post splenectomy patient was when she started on TPO-RA, specifically avatrombopaq as there is also risk of VTE post splenectomy.  

Would include what the platelet count was in both cases at the time that they developed thrombosis.  In most of the studies, TPO-RA are either held or reduced when plts >200k, so would be interesting to know how high plts were at time of thrombosis. 

Would include what the rationale was for changing from one TPO-RA to another when pt developed thrombosis, as thrombosis has been reported on studies of all 3, although whether causative remains unclear.  If anything, avatrombopaq has been reported to raise plt count without increasing platelet activation.  Can you offer explanation why pt would have developed thrombosis on avatrombopaq but changed to another TPO-RA and had no further ATE or VTE?

Would include/reference recent meta-analysis from 2022: Tjepkema et.al. Risk of thrombosis with thrombopoietin receptor agonists for ITP patients: A systematic review and meta-analysis.  Critical Reviews in Oncology/Hematology. Mar 2022.

Author Response

Reviewer #1:

1. In case #1 would include how many weeks post splenectomy patient was when she started on TPO-RA, specifically avatrombopaq as there is also risk of VTE post splenectomy.  

The reviewer raises an important question regarding the timing of the patient #1’s splenectomy.  Text was added to line 49 of p.2 to indicate her splenectomy was 19 months prior to her starting therapy with avatrombopag.

2. Would include what the platelet count was in both cases at the time that they developed thrombosis.  In most of the studies, TPO-RA are either held or reduced when plts >200k, so would be interesting to know how high plts were at time of thrombosis. 

Text was added on p.2 of the manuscript to indicated that both patients had normal platelet counts (151 x 10⁹/L and 150 x 10⁹/L, respectively) on the days that they presented with their thrombotic events.

3. Would include what the rationale was for changing from one TPO-RA to another when pt developed thrombosis, as thrombosis has been reported on studies of all 3, although whether causative remains unclear.  If anything, avatrombopaq has been reported to raise plt count without increasing platelet activation.  Can you offer explanation why pt would have developed thrombosis on avatrombopaq but changed to another TPO-RA and had no further ATE or VTE?

The reviewer raises a very important question regarding the choice to switch the patients to another TPO agonist and the concern for another thrombotic event.  It is certainly not clear that the thrombotic risk is the same for all TPO agonists, and the two patients reported had both previously taken TPO agonists without any thrombotic complications.  The first patient had previously done well with romiplostim, and given her lack of response to other ITP therapies, she elected to go back on therapy with the same.  The patient in case #2 had previously not responded to therapy with eltrombopag, and was concerned about the side effect profile of fostamatinib, therefore she elected to start treatment with romiplostim.  It is not clear why they developed a thrombotic event after treatment with avatrombopag but not previous TPO agonist therapy.  This suggests that avatrombopag may have a greater thrombotic risk than the other TPO agonists, but that conclusion cannot be made just on the basis of these two case reports without systematic study.  Text has been added at the end of the discussion section on p. 4 to address this issue in the manuscript.

4. Would include/reference recent meta-analysis from 2022: Tjepkema et.al. Risk of thrombosis with thrombopoietin receptor agonists for ITP patients: A systematic review and meta-analysis.  Critical Reviews in Oncology/Hematology. Mar 2022.

We have agree with the reviewer and have added this reference as suggested. 

Reviewer 2 Report

This article represents a summary of two case reports of thrombotic events in patients on the TPO agonist avatrombopag. The paper is overall well-written with some mild grammatical errors. At one point the authors do refer to eltrombopag by its brand name (Promacta, line 40) which should be changed. The novelty of the findings of thromboembolic events in patients treated with avatrombopag is certainly interesting and important to report, however, the author's presentation of the clinical cases is somewhat lacking in detail making interpretation of the events very challenging. Specifically, the authors do not report the platelet response to avatrombopag. Additionally, comorbidities (COVID19 infection, anti-phospholipid syndrome, PNH, etc), which might suggest alternative mechanisms for thrombotic events are not disclosed. Additionally, the pathophysiologic explanation of the lower extremity findings in case #2 should be supported with more evidence (did the patient have an eosinophilia, were inflammatory markers elevated etc?). Cessation of the symptoms in the absence of drug supports an association but I don't see how small vessel occlusions would subside solely with cessation of avatrombopag. The addition of more details to both case presentations (and a more illustrative figure for retiform purpura of case #2) would greatly support your conclusions if they fail to suggest alternative mechanisms of thrombosis. 

Author Response

Reviewer #2:

1. The paper is overall well-written with some mild grammatical errors. At one point the authors do refer to eltrombopag by its brand name (Promacta, line 40) which should be changed.

We thank the reviewer for his comments and suggested edits.  The change as suggested was made with additional proofreading of the manuscript.

2. The novelty of the findings of thromboembolic events in patients treated with avatrombopag is certainly interesting and important to report, however, the author's presentation of the clinical cases is somewhat lacking in detail making interpretation of the events very challenging. Specifically, the authors do not report the platelet response to avatrombopag.

Text was added to the case descriptions of each patient on p. 2 to indicate both patients responded well to treatment with avatrombopag, without the need for rescue medications of doses to be held due to elevated platelet counts. 

3.Additionally, comorbidities (COVID19 infection, anti-phospholipid syndrome, PNH, etc), which might suggest alternative mechanisms for thrombotic events are not disclosed.

The reviewer’s suggestions are very good and we would agree.  Additional text was added to the case reports for Case #1 and #2 on p.2 to describe the testing that was done at the time of their presentation and evaluation. 

4. Additionally, the pathophysiologic explanation of the lower extremity findings in case #2 should be supported with more evidence (did the patient have an eosinophilia, were inflammatory markers elevated etc?). Cessation of the symptoms in the absence of drug supports an association but I don't see how small vessel occlusions would subside solely with cessation of avatrombopag. The addition of more details to both case presentations (and a more illustrative figure for retiform purpura of case #2) would greatly support your conclusions if they fail to suggest alternative mechanisms of thrombosis.

As requested by the reviewer, additional information was added to the description of Case #2.  We have listed all of the relevant laboratory data that were available at the time of her presentation and follow-up.  As noted in the manuscript, her improvement followed the cessation of avatrombopag as well as starting therapy with apixaban.  Unfortunately, no additional images are available that would be more illustrative of the retiform purpura.   

Reviewer 3 Report

This is a well-written case report on thrombotic complications observed in chronic Immune Thrombocytopenic Purpura (cITP) patients after the administration of avatrombopag.

REMARKS

1] Conclusions and Discussion: The authors conclude that questions remain, regarding the thrombotic risks associated with the use of avatrombopag in ITP patients and propose further research to determine if there might be subsets of ITP patients, not safe to treatment with avatropombopag. In addition, in Case #2, a cutaneous microvessel thrombosis after the administration of avatrombopag is reported.

Recently, extensive microthrombosis was reported in COVID-19 patients after hospitalization, despite thromboprophylaxis (Koutsiaris et al, Clin. Hemorheol. Microcirc., 2022) and the existence of an unknown coagulation factor XC was suggested, related to the microvessel damage caused by COVID-19. In addition, a whole-genome sequencing (WGS) study (Kousathanas et al, Nature, 2022) reported evidence in support of a causal role for coagulation factors and platelet activation in critical COVID-19 cases.

A possible connection between the above studies and the submitted work could be discussed. Perhaps the same unknown coagulation factor XC is activated by avatrombopag in subsets of ITP patients.

2] Conclusions, Line 143: add “of ” before “avatrombopag”.

Author Response

Reviewer #3:

1. Recently, extensive microthrombosis was reported in COVID-19 patients after hospitalization, despite thromboprophylaxis (Koutsiaris et al, Clin. Hemorheol. Microcirc., 2022) and the existence of an unknown coagulation factor XC was suggested, related to the microvessel damage caused by COVID-19. In addition, a whole-genome sequencing (WGS) study (Kousathanas et al, Nature, 2022) reported evidence in support of a causal role for coagulation factors and platelet activation in critical COVID-19 cases.

A possible connection between the above studies and the submitted work could be discussed. Perhaps the same unknown coagulation factor XC is activated by avatrombopag in subsets of ITP patients.

The reviewer points of two very interesting articles regarding a possible novel mechanism for the microvascular thrombosis including the potential for a new thrombotic factor. Given that COVID-19 was  not thought to play a role in our patients (one tested negative, the other patient asymptomatic and not tested), this may be too speculative for this case report.

2. Conclusions, Line 143:add “of ” before “avatrombopag”.

Edit made as suggested.