Leukocyte-Rich vs Leukocyte-Poor PRP: Matching Cell Content to the Indication

Leukocyte-rich vs leukocyte-poor PRP is the single most consequential specification in platelet-rich plasma practice, and it is the one most likely to be missing from the chart note, the consent form and the device brochure. Two clinicians can run the same indication, the same number of sessions, the same interval, and the same reported platelet concentration, and inject biologically different products — because one is delivering concentrated neutrophils and monocytes into the tissue and the other is deliberately leaving them behind.

Tatiana Sarmiento states her position on this without hedging: "When I want to treat skin and hair, I would like to have no leukocytes, because I don't want to create inflammation. If you are treating patients for pain management, you may want to have some leukocyte percent."

That instinct is correct and it is supported by measured cytokine data. But the full picture is more interesting than a two-way split, because the musculoskeletal evidence itself divides — and knowing where it divides is what separates a clinician who has read the literature from one who has read a catalogue.

What leukocytes actually do to the preparation

The clearest quantification comes from a controlled laboratory study that made PRP from eleven human volunteers using two commercial systems, one that concentrates platelets while minimising leukocytes and one that concentrates both (Sundman, Cole & Fortier, Am J Sports Med, 2011;39(10):2135–2140).

The leukocyte-poor preparation contained platelets at 1.99× baseline and leukocytes at 0.13× — an actively depleted product. The leukocyte-rich preparation contained platelets at 4.69× and leukocytes at 4.26×.

What came out of those two tubes:

Both the anabolic growth factors and the catabolic cytokines rose together. That is not a coincidence of the devices — the study's correlation analysis shows why. TGF-β1 and PDGF-AB tracked with platelet count (r² = 0.75 and 0.60). MMP-9 and IL-1β tracked with neutrophil and monocyte counts (IL-1β versus neutrophils r² = 0.73; versus monocytes r² = 0.75).

So the two cell populations are contributing different halves of the biology. Platelets contribute the anabolic, pro-repair, pro-angiogenic signal. Leukocytes contribute a matrix-degrading, pro-inflammatory signal — matrix metalloproteinase-9 and interleukin-1β are the mediators of tissue breakdown and inflammatory amplification, not of quiet remodelling.

You cannot separate them by intent. You separate them at the bench, or you inject both.

The inflammation is not always the problem

It is worth being precise about the word "undesirable," because a catabolic phase is a legitimate and sometimes necessary part of repair. Inflammatory signalling recruits cells, clears debris and initiates matrix turnover. In a chronic, hypocellular, degenerative tendon that has failed to mount a healing response, deliberately inducing that phase is a defensible therapeutic strategy.

The question is never "is inflammation bad." It is: does this tissue, in this state, need a catabolic stimulus, and can it tolerate one?

That question has different answers in a degenerative tendon, in an arthritic joint, in facial dermis and in a miniaturising hair follicle. The rest of this piece works through those answers.

The classification frameworks, and which one to use

Because "PRP" names a category rather than a formulation, several groups have proposed classification schemes. They are not interchangeable and they capture different things.

Dohan Ehrenfest: the four-category architecture

The foundational scheme divides platelet concentrates on two axes — leukocyte content and fibrin architecture — producing four families: pure platelet-rich plasma (P-PRP), leukocyte- and platelet-rich plasma (L-PRP), pure platelet-rich fibrin (P-PRF) and leukocyte- and platelet-rich fibrin (L-PRF) (Dohan Ehrenfest, Rasmusson & Albrektsson, Trends Biotechnol, 2009;27(3):158–167). A later consensus paper extended and clarified it for orthopaedic and sports medicine use (Dohan Ehrenfest et al., Muscles Ligaments Tendons J, 2014;4(1):3–9).

This is the scheme to use when you need to say what family of product you are working with, and it is the one that makes the PRP/PRF relationship explicit — the fibrin axis is what distinguishes the liquid preparations from the clot-based ones covered in this cluster's PRF material.

PAW: platelets, activation, white cells

The PAW classification characterises a preparation by absolute Platelet count, whether and how it was Activated, and the presence or absence of White cells, with a sub-classification for neutrophils specifically (DeLong, Russell & Mazzocca, Arthroscopy, 2012;28(7):998–1009).

PAW's contribution is separating neutrophils from total leukocyte count — which matters, given that the Sundman correlations run most strongly through neutrophils and monocytes rather than through leukocytes as an undifferentiated mass.

DEPA: dose, efficiency, purity, activation

DEPA was designed to fix what the earlier schemes left out: final volume, red cell contamination, and production efficiency. It scores Dose of injected platelets, Efficiency of production, Purity of the preparation and degree of Activation (Magalon et al., BMJ Open Sport Exerc Med, 2016;2(1):e000060).

DEPA is the most useful of the three for clinical practice, because it reports the thing that is actually delivered — platelet dose — and because its purity axis captures erythrocyte contamination, which the leukocyte-focused schemes ignore. Red cells in a preparation are not neutral: haemolysis products are cytotoxic and pro-inflammatory in their own right.

MARSPILL, and why more axes keep appearing

Later schemes such as MARSPILL add further descriptors — including whether the preparation was handled under image guidance, whether red blood cells were present, and light activation — reflecting continued dissatisfaction that reported protocols remain irreproducible. The proliferation of frameworks is itself the finding: after two decades, the field still cannot reliably compare one study to another.

Practical rule. Whatever scheme you cite, report at minimum: platelet concentration and absolute dose, leukocyte concentration (with neutrophil fraction if you have it), red cell presence, final volume, activation status and anticoagulant. That set makes your protocol reconstructible by someone else, which is the entire point.

What the musculoskeletal evidence actually says — and where it splits

The common shorthand is "leukocyte-rich for musculoskeletal, leukocyte-poor for skin." That is too coarse, and following it will lead you wrong inside musculoskeletal practice itself.

Intra-articular work favours leukocyte-poor. A Bayesian network meta-analysis of nine studies and 1,055 patients with knee osteoarthritis found that leukocyte-poor PRP produced significantly better WOMAC scores than hyaluronic acid (mean difference −21.14; 95% CI −39.63 to −2.65) and than placebo (−17.84; 95% CI −34.95 to −0.73). Leukocyte-rich PRP achieved no such separation (−14.28; 95% CI −44.80 to 16.25). Leukocyte-poor ranked highest on both efficacy measures. Local adverse reactions were more common with PRP than hyaluronic acid overall, but did not differ between the two PRP types (odds ratio 0.78; 95% CI 0.05–11.93) (Riboh et al., Am J Sports Med, 2016;44(3):792–800).

Intratendinous work has positive leukocyte-rich data. A double-blind randomised trial of leukocyte-rich PRP for chronic gluteus medius and minimus tendinopathy showed improvement over corticosteroid injection at 12 weeks, sustained at two-year follow-up (Fitzpatrick et al., Am J Sports Med, 2018;46(4):933–939; two-year data, 2019;47(5):1130–1137).

So the split inside musculoskeletal medicine is not tissue-agnostic. It looks like a compartment distinction: a synovial joint is a closed space lined with cells that respond badly to IL-1β and MMP-9, whereas a degenerative tendon is a poorly vascularised structure that may need a catabolic stimulus to restart a stalled repair sequence.

Stated as a working principle: leukocyte content is a decision about whether the target tissue needs a controlled inflammatory stimulus, and whether the compartment can clear it.

Platelet concentration is not monotonic either

A related finding deserves a place here, because it complicates "more is better" in the other direction.

In leukocyte-reduced PRP applied to tendon explants, increasing platelet concentration did not produce increasing matrix synthesis. Beyond a point, higher platelet concentrations decreased collagen gene synthesis (Boswell et al., Am J Sports Med, 2014;42(1):42–49).

There is, in other words, a dose-response curve with a peak, not a ramp. This is the strongest available argument against the marketing claim that a higher concentration factor is inherently a better product, and it should temper any assumption that the goal of a preparation protocol is maximisation.

Skin and scalp: why Tatiana's position holds

Now apply the framework to aesthetic and hair indications.

The target tissues are thin, superficial and visible. Facial dermis and scalp dermis are millimetres deep, richly vascular, and every inflammatory response is on display. An IL-1β-driven erythematous or oedematous response in a knee is invisible; on a face it is the complication.

The intended endpoint is quiet remodelling, not repair of an injury. As Tatiana frames the facial indication: PRP is "skin quality support, not structural replacement" — signalling that "will encourage collagen remodelling," experienced as texture, tone and luminosity over weeks. That is a slow anabolic process. MMP-9 is a matrix-degrading enzyme. Adding it to a protocol whose stated goal is matrix accumulation is working against yourself.

The follicle is an immune-privileged structure. The hair follicle maintains a relative immune privilege, and inflammatory infiltrate around the follicular unit is a pathological finding in multiple alopecias, not a therapeutic one. Deliberately concentrating neutrophils and monocytes into a peri-follicular plane is difficult to justify mechanistically when the goal is to support follicles that are already miniaturising.

The honest caveat. There is, at present, essentially no head-to-head randomised evidence comparing leukocyte-rich with leukocyte-poor PRP for androgenetic alopecia. The recommendation above rests on mechanism, on the intra-articular meta-analysis by analogy, and on clinical experience — not on a trial that answers the question directly. A 2025 scoping review of regenerative therapies for androgenetic alopecia reached the same broad conclusion about the field's evidence quality (Gupta et al., Med Sci, 2025;14(1):5). Say so to patients and to colleagues. Overstating this is unnecessary; the mechanistic case stands on its own.

One comparative signal does exist on preparation route: a retrospective analysis of 747 PRP preparations and 163 androgenetic alopecia patients compared buffy-coat-derived with apheresis PRP, finding higher platelet concentration from apheresis at the cost of substantially greater blood volume (Huang et al., Clin Cosmet Investig Dermatol, 2026;19). Note what "buffy coat derived" implies about leukocyte content — the buffy coat is the leukocyte layer.

How you actually control leukocyte content at the bench

This is where the theory becomes a hand movement.

Leukocytes concentrate at the buffy coat, immediately above the erythrocyte layer. Therefore:

• Collecting down into the buffy coat produces a leukocyte-rich preparation. Some devices are designed to do exactly this and market it as a feature.
• Collecting the plasma layer while stopping short of the interface produces a leukocyte-poor preparation, at the cost of leaving behind the densest platelet population sitting just above that interface. This is the real trade — leukocyte depletion costs platelet recovery.
• Spin count does not determine leukocyte content. Both single- and double-spin protocols can be run either way. What determines it is where the collection stops. The companion spoke on single spin versus double spin covers why this is often misattributed to the spin protocol.
• Closed, purpose-built devices make this reproducible by removing the operator's judgement from the collection step. Open, manual protocols make leukocyte content a function of who is holding the syringe and how rushed they are.

If you intend to run leukocyte-poor PRP for skin and scalp, the operational question is not which centrifuge you own. It is whether your collection step is specified, trained and repeatable — and whether you have ever verified it with a cell count.

What to do differently

1. Name the cell content in your protocol and your chart note. "PRP" is not a specification. "Leukocyte-poor PRP, single spin, 6 mL, unactivated, sodium citrate" is.
2. Set leukocyte target by indication, not by habit. Skin and scalp: leukocyte-poor. Intra-articular: the meta-analytic evidence favours leukocyte-poor. Intratendinous: leukocyte-rich has positive randomised data.
3. Verify once. Run a cell count on your own preparation at least once per device and per operator. Most clinicians using PRP have never confirmed that their product is what they believe it is.
4. Stop equating concentration with quality. The tendon explant data show a peak, not a ramp.
5. Report enough to be reproducible. Platelet dose, leukocyte concentration, red cell presence, volume, activation, anticoagulant.

These preferences reflect Tatiana Sarmiento's clinical practice as taught in Empire Medical Training's hands-on curriculum. Technique is learned under supervision; this article is educational and is not a substitute for training.

Cell content is decided at the bench, in a step that takes about four seconds and is almost never taught explicitly. Empire's Platelet Rich Plasma Training and Medical Hair Loss Treatment, PDO Threads and PRP Hair Restoration Training teach preparation and collection alongside delivery, which is where this parameter is actually controlled. For how PRP sits against the other regenerative options clinicians are being asked about, see exosomes vs PDRN vs PRP.

Related guides in this cluster

Part of Regenerative Injectables: PRP and PRF.

Clinical GuideSingle Spin vs Double Spin PRP: What Actually Changes in the TubeSingle spin vs double spin PRP compared for clinicians — what the second spin concentrates, why volume compensates, and how processing Clinical GuideDiagnose Before You Treat: The Hair Loss Workup That Must Precede PRPThe hair loss workup before PRP, step by step — onset history, pattern recognition, trichoscopy, category-level labs, and the diagnoses Clinical GuideSequencing PRP with Microneedling and Laser: Recovery and Remodelling SupportSequencing PRP with microneedling and laser — same-session versus staged delivery, what the recovery evidence supports, and the order o