Abstract

Introduction

Reticulin is a normal and common component of the bone marrow stroma. In clinical trials of adults with chronic ITP treated with the thrombopoietin mimetic romiplostim, a small number of cases of reticulin and/or collagen in bone marrow were reported. However, bone marrow biopsies were not systematically performed in these trials. Here, we report a multicenter study of adult ITP patients that evaluated bone marrow biopsies for reticulin and collagen pre- and post-treatment with romiplostim.

Methods

Eligible ITP patients had platelet counts <50x109/L, received ≥1 prior ITP therapy, and no collagen in baseline bone marrow biopsies (ie, before romiplostim). Three cohorts of patients were scheduled for biopsies after 1 (Cohort 1), 2 (Cohort 2), or 3 years of romiplostim, dosed to maintain platelet counts at 50–200x109/L. Bone marrow biopsies were also performed if patients discontinued early or failed to achieve/maintain a response to romiplostim, defined as platelet counts ≤20x109/L for 4 consecutive weeks at the maximum romiplostim dose of 10 µg/kg. Reticulin and collagen formation were measured using the modified Bauermeister scale (0 no reticulin; 1 occasional fine fibers/foci fine fiber network; 2 fine fiber network; 3 diffuse fiber network, scattered coarse fibers; 4 diffuse coarse fiber network with areas of collagenization). Collagen was detected by trichrome staining and reticulin by silver staining. Cohort 1 biopsy samples were reread as the modified Bauermeister scale had not been applied as specified in the study protocol. All samples were read by 2 hematopathologists at a central bone marrow laboratory with grading discrepancies reviewed by an independent bone marrow panel. Data cutoff for this analysis was August 10, 2012.

Results

Of the 50 patients enrolled in Cohort 1 (54% female, mean age 55.5 years, range 20–86 years), 39 patients received romiplostim and had bone marrow biopsies (Table). Mean (SD) dose was 3.8 (2.9) µg/kg. No patients with evaluable results developed collagen. No patients had an increase ≥2 grades from baseline in reticulin. Of the 50 patients enrolled in Cohort 2 (76% female, mean age 48.6 years, range 19–83 years), 39 patients received romiplostim and had bone marrow biopsies (Table). Mean (SD) dose was 4.1 (3.2) µg/kg. Again, of the patients with evaluable results, none developed collagen. Two patients had a 2-grade increase in reticulin (1 patient: baseline grade 0 to 2 at year 2; 1 patient: baseline grade 1 to 3 at end of treatment). The safety profile was similar to previous trials. In Cohort 1, 3 patients died (not attributed to romiplostim); causes were fungal sepsis in a patient with longstanding corticosteroid use, cerebral hemorrhage, and pulmonary hemorrhage in a patient with pulmonary thrombosis treated with acenocumarol. In Cohort 2, 2 patients died (not attributed to romiplostim); causes were acute renal failure, in a patient with a medical history of chronic renal insufficiency, and suicide. There were no neutralizing antibodies to romiplostim or thrombopoietin in the 2 cohorts.

Table 2

Romiplostim Cohort Results

Patients Cohort 1 N=50 Cohort 2 N=50 
Bone marrow biopsies after receiving romiplostim1 39 39 
Biopsies evaluable for collagen (trichrome stain)2 34 36 
Positive for collagen 
Biopsies evaluable for reticulin (silver stain)2 33 36 
Increase in reticulin by ≥2 grades or to grade 4 23 
Patients Cohort 1 N=50 Cohort 2 N=50 
Bone marrow biopsies after receiving romiplostim1 39 39 
Biopsies evaluable for collagen (trichrome stain)2 34 36 
Positive for collagen 
Biopsies evaluable for reticulin (silver stain)2 33 36 
Increase in reticulin by ≥2 grades or to grade 4 23 
1

3 patients in each cohort had bone marrow biopsies at end of treatment due to early study discontinuation.

2

Trichrome and silver staining could not be performed on all biopsies due to inadequate samples.

3

patient: baseline grade 0 to 2 at end of year 2; 1 patient: baseline grade 1 to 3 at end of treatment.

Summary/Conclusion

There was no evidence of collagen formation after 2 years of romiplostim treatment. The incidence of increase in reticulin was low, consistent with results of previous romiplostim studies. This ongoing study will provide additional data on bone marrow changes with up to 3 years of romiplostim treatment.

At completion of year 1 treatment, discontinuations included withdrew consent (n=5), non-responders (n=4), died (n=3), adverse event (n=1, severe joint pain), administration decision (n=1).

At completion of year 2 treatment, discontinuations included withdrew consent (n=6), non-responders (n=2), declined biopsy (n=2), adverse events (n=2, dermatitis and suspected non-Hodgkins lymphoma), died (n=1), required alternate therapy (n=1).

Disclosures:

Janssens:Amgen, GSK, Mundipharma, Roche: Membership on an entity’s Board of Directors or advisory committees. Rodeghiero:Amgen, GSK: Honoraria; Amgen, Eisai, GSK, LFB, Suppremol: Membership on an entity’s Board of Directors or advisory committees. Chong:Amgen, GSK: Research Funding. Pabinger:CSL Behring: Research Funding; Amgen, Baxter, Bayer, Boehringer Ingelheim, CSL Behring, Pfizer: Honoraria; Amgen, Baxter, Bayer, Boehringer Ingelheim, CSL Behring, Pfizer: Membership on an entity’s Board of Directors or advisory committees. Cervinek:Alexion, Amgen, GSK: Consultancy. Wang:Amgen: Employment, Equity Ownership. Lopez:Amgen: Employment, Equity Ownership.

Author notes

*

Asterisk with author names denotes non-ASH members.