iPLAT1: the first-in-human clinical trial of iPSC-derived platelets as a phase 1 autologous transfusion study

TO THE EDITOR:

Alloimmune platelet transfusion refractoriness (alloPTR) is a serious clinical issue present in ∼5% to 15% of patients who undergo platelet transfusion owing to sensitization against alloantigens such as HLA class I and human platelet antigens (HPAs).¹ To provide transfusable platelets for a patient suffering from alloPTR but with no compatible registered donor, we developed an induced pluripotent stem cell–derived platelet product (iPSC-PLT). We have previously established immortalized megakaryocyte progenitor cell lines (imMKCLs) by introducing doxycycline-inducible c-MYC, BMI1, and BCL-XL overexpression during differentiation from iPSCs.² The imMKCLs expand by the expression of 3 transgenes, and thereafter, they mature and generate iPSC-PLTs. We further produced clinically required, 10¹¹-scale competent iPSC-PLTs ex vivo by developing a turbulent flow–mediated VerMES Bioreactor, based on the discovery that turbulence regulates thrombopoiesis, and using several novel drugs:³ a thrombopoietin mimetic, TA-316;⁴ a Rho-associated kinase inhibitor, Y-39983 and an aryl hydrocarbon receptor antagonist, GNF-316, together enables adhesion-independent manufacturing;³ and an ADAM17 inhibitor, KP-457, which blocks glycoprotein Ibα (CD42b) shedding.⁵ (Figure 1A; supplemental Tables 1-8, available on the Blood website).

The iPSC-PLTs were produced based on GMP using newly established, patient-specific imMKCLs as the MCB and gamma-ray irradiated (25 Gy) to eliminate the tumorigenicity potential of iPSC-PLTs, which still included imMKCLs. Compared with donor platelets available from the JRC-PLTs, iPSC-PLTs were larger (approximate diameter, 2.5 μm vs 3.5-4 μm; Figure 1B), but the ultrastructures were comparable (Figure 1C), and in vitro quality and function were sufficient within 5 days (Figure 1D). Consistent with results using JRC-PLTs and iPSC-PLTs of a laboratory clone in rabbit models,⁶ the patient’s iPSC-PLTs showed stable circulation in vivo from 2 to 7 hours (Figure 1E) and 100% success in hemostasis (Figure 1F). Later tests using in vivo imaging system confirmed systemic circulation and distribution in the lung, liver, and spleen (Figure 1G), which resembles patterns reported with general platelet transfusions.⁷ In contrast, imMKCLs were mostly restricted to the lungs. Further details regarding the production and nonclinical tests confirming the quality and safety of the iPSC-PLTs are published in Blood Advances.

The clinical trial of an autologous transfusion of patient-derived iPSC-PLTs, the iPLAT1 study, was approved by Kyoto University and the Ministry of Health, Labour and Welfare, and this trial was registered at the Japan Registry of Clinical Trials, https://jrct.niph.go.jp/latestdetail/jRCTa050190117, as #jRCTa050190117 (supplemental Figure 1 and protocol). The enrolled patient was a 55-year-old Japanese woman with a history of 2 pregnancies and was diagnosed with severe aplastic anemia at the age of 47. After the first platelet transfusion as a pretreatment to the antithymocyte globulin administration, she developed a fever of 39.8°C and a generalized swollen rash, and no platelet increase was observed, followed by the diagnosis of alloPTR owing to anti–HPA-1a antibodies. However, the frequency of the patient’s HPA-1b/1b phenotype in Japan is estimated to be <0.002%,⁸ and such blood donors were not available in the JRC repository. Fortunately, cyclosporine monotherapy was effective and eventually completed at the age of 54. She also underwent a mastectomy with lymph node dissection for breast cancer at the age of 51. Postoperative chemotherapy was reduced to monotherapy with dose reduction to avoid serious cytopenia (Figure 2A).

The iPSC-PLTs, which were compatible for specifications (supplemental Figure 2; supplemental Tables 9 and 10) and later confirmed for ultrastuctures (supplemental Figure 3), residual additive concentration, and sterility (supplemental Table 11), were administered sequentially for 3 escalating doses of 1 × 10¹⁰, 3 × 10¹⁰, and 1 × 10¹¹ (1/20, 3/20, and 1/2 the dose of standard transfusions in Japan, respectively; Figure 2A-B). As for the primary endpoint of frequency and extent of adverse events, the patient did not present any significant clinical symptoms or signs throughout the study period (Figure 2B; supplemental Table 12). Laboratory data showed no remarkable change, but the level of a coagulation marker, D-dimer, and white blood cell count were slightly increased 24 hours after the transfusion in cohorts 2 and 3 (Figure 2C; supplemental Table 12). Nevertheless, lower extremity ultrasonography identified no deep vein thrombosis, and the D-dimer concentration spontaneously decreased. As for the secondary endpoint of CCI, the CCI at 1 hour and 24 hours were determined to be 0 even in cohort 3 (Figure 2B-C). However, a flow cytometry analysis of peripheral blood showed the existence of larger platelets after the transfusion, which gradually decreased (Figure 2D). The external Efficacy and Safety Assessment Committee confirmed the safety of the transfusion, approved transition to the next cohorts after the 28-day follow-up for cohorts 1 and 2, and concluded that the autologous transfusion of iPSC-PLTs is safe for this patient after the observation period of 1 year following cohort 3 (supplemental Figure 1).

The study outcome is of great significance given that anti–HPA-1a antibody could cause posttransfusion purpura, which even decreases platelet count after transfusion. The lack of an increase in CCI could be attributed to the difficulty in detecting the relatively small increase compared with the high pretransfusion platelet count and the common blood count device used, which could ignore large-sized iPSC-PLTs (Figure 1B). However, we observed populations of larger platelets in posttransfusion samples by flow cytometry, suggesting their circulation. Moreover, the CCI could have peaked at 2 to 6 hours later according to the kinetics in the rabbit study (Figure 1E), thus missing the optimal time point in this trial. The delayed peaking can be because of the fragmentation of large-sized iPSC-PLTs, as suggested in a previous study using mouse models.³ 

Meanwhile, the slight increase in D-dimer concentration after transfusion may indicate coagulation by iPSC-PLTs soon after the transfusion. The concurrent slight white blood cell increase may further indicate enhanced inflammation. Although the number of imMKCLs infused was small (<10⁷), such reactions may have been caused by the immune properties of megakaryocytes. However, only megakaryocytes in the lung parenchyma are reported to have the immune phenotype,^9,10 whereas the infused imMKCLs in this study were probably trapped in pulmonary circulation (Figure 1G). In addition, the compromised circulation could be owing to mild activation caused by irradiation (Figure 1D) or the aberrant status of glycosylation and sialylation,^11,12 which reflects the possible embryonic/fetal nature of iPSC-PLTs, as seen with erythrocytes.^13,14 We plan to assess these possibilities thoroughly in future studies.

In summary, we succeeded in the clinical scale manufacturing of GMP-based, 10¹¹-scale autologous iPSC-PLTs for a patient who did not have a compatible platelet donor and completed the first-in-human clinical trial of iPSC-PLTs, which assured the safety of the product through the dose-escalation study. At the same time, the study revealed a discrepancy in the circulation between animal models and the human participant.

Significant improvements in the efficacy and cost of producing iPSC-PLTs could realize the practical personalized medicine of autologous platelet products, which are essentially nonrejectable. Furthermore, the production system and the observed findings in this study should contribute to the clinical application of allogeneic iPSC-PLTs for off-the-shelf use. Given that the main causative antigen of alloPTR is HLA class I, iPSC-PLTs genetically depleted of HLA class I could serve as an optimal product for mass production,^15-17 with the additional genetic conversion of HPA being an option.¹⁸ These advancements will allow many patients with thrombocytopenia, especially those with alloPTR, to benefit from this combined modality of transfusion medicine and regenerative medicine.