Platelet-rich plasma is not a single uniform procedure in cardiac surgery. The same term is used for topical platelet gel on surgical wounds, autologous plateletpheresis in perioperative blood management and experimental myocardial applications. These three fields of research must be assessed separately.
For this evidence assessment, PubMed/PubMed Central, ClinicalTrials.gov, relevant guideline and legal sources, and studies cited in key publications were specifically reviewed through 20 August 2026. Clinical studies representing the three research fields separated in this article were considered. Retracted publications are clearly marked and are not used as evidence of efficacy. The selection is narrative and does not claim the methodology of a systematic review.
Wound healing examines local biological effects at the sternum or vein harvest site.
Plateletpheresis primarily targets hemostasis and blood management around cardiopulmonary bypass.
Myocardial applications examine regenerative or angiogenic approaches and should not be included in the same efficacy assessment.
The graphic is deliberately designed as a selectable flow. On mobile devices, it rearranges vertically without overlapping elements.
Several large observational studies report marked benefits. Controlled and randomized data, however, have not yet confirmed reliable protection against sternal wound infections.
Key point: biologically plausible, but no robust randomized evidence of efficacy for infection prevention.
The term platelet-rich plasma was already used in transfusion medicine in the 1950s. An early and frequently cited milestone in autologous PRP use in cardiac surgery is the 1987 study by Ferrari et al.: PRP was prepared during cardiac surgery as part of an autologous transfusion strategy and later reinfused.[1]
The widespread narrative “dentistry first, cardiac surgery around 50 years later” is therefore chronologically incorrect. The dental and maxillofacial applications that became established from the 1990s onward represent a different development pathway. More important for today's assessment is that, historically, cardiac surgery focused not only on wound healing but also on blood management.
Deep sternal wound infections are among the clinically relevant complications after median sternotomy. Researchers therefore examined whether activated autologous PRP or platelet gel could provide an additional local benefit during wound closure. PRP does not replace established infection prevention measures.
However: sequential, non-randomized cohorts from different time periods. The authors calculated an NNT of 71 (95% CI 41.8–244.5) to prevent one deep infection. This figure is not a general proof of efficacy.[8]
In retrospective cohorts, SWI and DSWI were less frequent with PRP. In the randomized-study subgroup, however, there was no significant difference.[13]
Assessment: Not “ineffective”, but the higher-quality evidence is insufficient for a reliable preventive claim.
In the Patel cohort, whole blood was collected before skin incision, separated in a point-of-care system, the resulting PRP was activated and applied locally to the sternum and soft tissues during closure. These details show why such studies cannot be equated with any outpatient tube system.[8]
Study description, not instructions for use.
The overview deliberately includes positive observational studies as well. This makes the difference between study designs visible instead of selecting only one direction of the literature.
382 treated patients were compared with 948 concurrent and 929 historical controls. Superficial infections were 0.3% versus 1.8% and 1.5%, respectively; deep sternal infections were 0% versus 1.5% and 1.7%. The design remains susceptible to bias.[2]
Prospective and double-blind. No significant differences in major or minor wound complications; other perioperative endpoints were also comparable.[3]
Wound infections in 9/70 patients with topical PRP versus 8/70 controls; no significant benefit.[4]
DSWI requiring revision occurred in 6/97 patients with PRP and 3/99 in the control group. The difference was not significant. The publication lists Biomet Deutschland GmbH as the study funder.[5]
A negative result in a manufacturer-funded study is a relevant part of the evidence, without deciding the overall question on its own.All 11 severe DSWI occurred in the control group; none occurred in the intervention group receiving PRP, calcium-thrombin and vancomycin. Adjusted OR 0.05 (95% CI 0.01–0.50). The PRP effect cannot be isolated because vancomycin was part of the combined intervention.[6]
DSWI 1/422 (0.20%) with PRP versus 10/671 (1.5%) without PRP; p = 0,043. Superficial infections were also less frequent. Positive signal, but not a randomized comparison.[7]
Large sequential cohort analysis with positive results, but with a chronologically earlier control group and no randomization. The groups also differed in VAD/transplantation and emergency procedures. NNT to prevent one DSWI: 71.[8]
4,692 patients; pooled results showed significantly fewer sternal wound infections and cases of mediastinitis. However, the authors emphasize the poor quality of the underlying data. No treatment effect was found in the RCT subgroup.[9]
DSWP 6/144 (4.2%) with platelet-leukocyte gel versus 13/118 (11%) in the historical reference group; p = 0.03. The authors cite, among other things, changes in antibiotic regimens and closure technique as limitations.[10]
Positive effects in retrospective cohorts; in the randomized subgroup, no significant difference in SWI or DSWI. The authors considered the evidence too weak for a firm conclusion.[13]
Less postoperative blood loss and lower requirements for allogeneic blood products; at the same time, substantial heterogeneity between studies.[11]
Retrospective cohort: after adjustment, aPRP was independently associated with more postoperative AKI (OR 1,729; 95% CI 1,225–2,440), while intraoperative transfusions were lower. This does not establish causality, but it is a relevant safety signal.[12]
Single-center RCT with fewer red blood cell transfusions. In the procedural description, the authors refer to the collected product as autologous platelet concentrate (APC); it should therefore not be equated with topical PRP gel. Higher calcium requirements and longer preparation time were also reported.[14]
2,150 patients, including 906 with aPRP. Reported outcomes included fewer reoperations for bleeding, lower platelet/cryoprecipitate use and more favorable ventilation endpoints; mortality, length of hospital stay and neurological and renal complications did not differ significantly. A large proportion of the evidence was observational.[15]
Published in Cureus. Four small groups: CABG 16, CABG+TMR 17, CABG+PRP 10, CABG+TMR+PRP 9. A GLS signal concerned the combination approach; postoperative atrial fibrillation was more frequent in the TMR+PRP group (67% vs. 25%). No isolated regenerative PRP effect can be inferred.[17]
Retracted in January 2025 after the peer-review process was judged to have been compromised. Not suitable as evidence of efficacy.[18]
The original article has PMID 38420690 and is explicitly listed by PubMed as a Retracted Publication . The retraction notice (PMID 39317945) was published in September 2024 because of a manipulated peer-review and publication process. Do not use as evidence of efficacy.[19]
During operations using cardiopulmonary bypass, platelets can be affected by hemodilution, contact with artificial surfaces and mechanical stress. Autologous plateletpheresis separates a platelet-rich autologous fraction before cardiopulmonary bypass, which is reinfused later. The primary objective is hemostasis and blood management – not regenerative wound healing.[11]
The meta-analysis of 1,002 patients found signals for less postoperative blood loss and lower requirements for several allogeneic blood products. Substantial heterogeneity limits generalizability.[11]
Among 660 patients with type A aortic dissection, adjusted analyses associated aPRP with more postoperative acute kidney injury (OR 1,729). At the same time, fewer intraoperative transfusions were reported. The design does not allow causal conclusions, but the signal belongs in the benefit-risk assessment.[12]
The single-center RCT by Gao et al., published in 2025, randomized 134 patients. In the procedural description, the authors refer to the collected product as autologous platelet concentrate (APC) within the plateletpheresis procedure. The overall RBC transfusion rate was 10/67 (14.9%) versus 34/67 (50.7%); higher calcium requirements and longer preparation times were also reported.[14]
The 2026 meta-analysis of acute type A aortic dissection included six studies with 2,150 patients. It reported, among other outcomes, fewer reoperations for bleeding, lower platelet/cryoprecipitate use and more favorable ventilation endpoints. Mortality, hospital stay and neurological and renal complications did not differ significantly. Most of the data were observational.[15]
ClinicalTrials.gov lists a randomized multicenter study with a planned enrollment of 250 participants. Results are not yet available.
The registry entry is listed as completed but currently contains no published results. It therefore cannot support an efficacy claim at present.
Blood volume, platelet and leukocyte content, anticoagulant, activation and separation vary by system. A prospective randomized, single-blind study published in June 2026 compared two point-of-care systems in 60 cardiac surgery cases. Mean platelet enrichment was 6.58-fold over baseline with the Magellan system and 6.31-fold with the Angel system.[16]
The study measures whether the preparation target was achieved, not clinical efficacy. That is exactly why it is useful: it shows why a result from one specific system cannot automatically be transferred to another.
A 2024 study published in Cureus randomized 52 patients to CABG alone (16), CABG+TMR (17), CABG+PRP (10) or CABG+TMR+PRP (9). The reported GLS signal after one year mainly concerned the TMR+PRP combination; LVEF did not differ significantly.[17]
A preliminary safety signal also belongs in a complete assessment: postoperative atrial fibrillation was more frequent in the TMR+PRP group than in the CABG control group (67% vs. 25%, p = 0.04). With only nine patients in this group, this is not robust evidence of harm, but it is another reason not to use the study as isolated evidence for “PRP regeneration of the heart”.[17]
These figures come from a single non-randomized sequential cohort analysis. They are not a generally applicable effect size.[8]
Incorrectly phrased. Plateletpheresis is being investigated to determine whether a patient's own platelets can be preserved, thereby reducing blood loss or the requirement for allogeneic blood products.
The Patel cohort included such patients, but it does not establish a specific PRP effect in VAD implantation or heart transplantation.
No. Kirmani 2017 found a positive pooled effect based predominantly on non-randomized data; Yao 2021 found the effect mainly in retrospective studies. Two later meta-analyses were retracted.
PMID 38420690 belongs to the article by Yifan Li and Zhong Wu, DOI 10.1111/iwj.14761. PubMed identifies this exact article as a Retracted Publication. The retraction notice has PMID 39317945 and was published in September 2024.[19]
The Zhu meta-analysis from 2023 was likewise retracted in January 2025.[18] Both papers remain visible in the explorer, but they do not contribute to any efficacy conclusion.
Cardiac-surgery studies use different separators, apheresis devices and preparation systems. A study using one specific system does not mean that another PRP tube, centrifuge or separator is intended for the same intraoperative use.
Under the MDR, the intended purpose of a medical device is determined by the manufacturer's information. CE marking is therefore not blanket proof for every PRP application described in the scientific literature.[22]
For the medicinal-product side, section 13 of the German Medicines Act (AMG) is relevant. Subsection 2b contains an exemption from the manufacturing authorization requirement for patient-specific preparation under direct professional responsibility for personal use; the general notification requirement under section 67 AMG and official supervision under section 64 AMG may also apply. Blood collection is subject to section 7 of the German Transfusion Act (TFG).[23][24][25][26]
Section 3 of the German Heilmittelwerbegesetz (HWG) is also relevant to external communication: misleading advertising, particularly unsupported claims of therapeutic efficacy or success, is prohibited. This article therefore makes no claim of product efficacy and no claim that any specific system is suitable for cardiac surgery.[27]
Observational data provide positive signals. Controlled studies have not yet confirmed robust protection against sternal infections. PRP does not replace established infection prevention.
Randomized data and aortic analyses show interesting transfusion and bleeding signals. A retrospective AKI safety signal points in the other direction. No routine recommendation can be made for every cardiac surgical procedure.
The small 2024 study examined combinations of CABG, TMR and PRP. An isolated PRP effect has not been established; atrial fibrillation was more frequent in the smallest combination group.
Not “Does PRP work?”, but: Which precisely defined PRP procedure shows reproducible clinical benefit for which endpoint in which patient group?
Not generally. Topical PRP is not listed as a standard component in established recommendations for preventing sternal wound infections. Plateletpheresis is an investigated blood-management approach; this article does not establish a general guideline recommendation for routine use.
Large positive studies are often historical or retrospective comparisons. Infection prevention, patient selection and surgical technique may have changed at the same time. Randomized studies reduce such biases but have so far been relatively small in this field.
No. Cardiac-surgery plateletpheresis uses different devices, blood volumes and process steps. In the Gao RCT, the product obtained during the procedure is described as autologous platelet concentrate (APC). Sharing an umbrella term does not mean the preparation or intended purpose is the same.
No. The specific intended purpose of the medical device is decisive. CE marking does not mean that every PRP application described in scientific literature is covered by the product.
No. Direct intra- or epicardial applications remain experimental research. The small 2024 study also examined combinations with TMR and does not permit conclusions about an isolated PRP effect.
The six links were checked against the live site. They lead to background information, wound healing and technical context – not to a cardiac-surgery product application.