Your "dark spot" serum won't fix those red marks. Here's what will.
A cosmetic chemist explains why post-acne red/purple marks are a blood vessel problem, not a pigment problem, and what actually fades them.
If you're reading this, you've probably got red or pink marks left behind after acne that just… won't go. The spot is long gone. The inflammation has calmed down. But the mark is still sitting there, staring at you, weeks or months later.
And you’ve probably been reaching for a “dark spot corrector” to fix it. Maybe something with kojic acid, alpha arbutin, or hydroquinone.
Here’s the problem: those ingredients target melanin. Your red marks aren’t melanin. They’re damaged blood vessels.
Those flat red or pink marks left after acne have a name: post-inflammatory erythema, or PIE, and they’re a completely different beast from the brown spots (post-inflammatory hyperpigmentation, or PIH) that most “dark spot” products are designed for. The colour you’re seeing is haemoglobin in dilated, damaged capillaries sitting in the upper layers of your dermis, visible through your thinned, healing skin.
Which means: you need a completely different approach. And just like melasma, PIE needs a protocol, not a product.
Let me give you the protocol first, and explain the science after.
Let’s get into it.
The protocol: what actually fades PIE
PIE is driven by a self-sustaining loop of vascular inflammation. Your skin has multiple things going wrong at once:
Capillaries are dilated and damaged from the inflammatory response to acne
New, excess blood vessels have formed where they shouldn’t be (angiogenesis)
Inflammatory signals (VEGF, NF-κB, TNF-α) are keeping the whole thing alive
The structural support around those vessels (collagen, elastin) has been degraded
Your epidermis is thinner after the inflammation, making the redness more visible
And UV exposure is reigniting the vascular damage cycle daily.
No single ingredient addresses all of that. You need a team.
Here’s your team:
Step 1: Azelaic acid (15–20%) — to calm the inflammation and shut down angiogenesis
This is your heavy hitter for PIE, and it has the strongest direct evidence of any topical ingredient.
A 2024 randomised, double-blind, placebo-controlled trial (72 enrolled, 60 completed, 12 weeks) showed that 15% azelaic acid gel significantly reduced PIE and, critically, decreased haemoglobin content in the lesions. That’s not just cosmetic improvement. That’s measurable vascular change. This is the only topical trial that has directly demonstrated blood vessel-level improvement in PIE.
Why does it work so well here? Azelaic acid suppresses NF-κB (the master inflammatory switch that keeps PIE alive), has anti-angiogenic activity demonstrated in psoriasis and rosacea models (the direct evidence in PIE-specific skin tissue is still emerging, but the mechanism is consistent), and is a potent antioxidant.
It also has anti-tyrosinase activity, which means if you’ve got red marks AND brown marks coexisting (common across all skin tones, but particularly challenging to treat in Fitzpatrick III–VI), it hits both.
The formulation caveat I gave you in the melasma article applies here too: azelaic acid has terrible water solubility (~2.4 g/L). The vehicle, pH, and particle size determine how much AzA actually reaches your skin. A well-formulated 10% can potentially outperform a poorly-formulated 20%.
My recommended products are the same ones I listed in
the melasma article: Finacea 15% gel, Skinoren 20% cream, RegimenLab Azelaic Advanced (fully solubilised), Facetheory Blemicalm 15%, Skin1004 (cyclodextrin encapsulated), Dr Sam’s Flawless Brightly, or Sesderma Azelac RU.
Step 2: Ectoine (1–7%) — to shut down the inflammatory cascade before it starts
I need to be upfront here: ectoine isn’t on any official PIE treatment protocol. There’s no PIE-specific clinical trial for it. I stumbled across it while reading studies on atopic dermatitis and lung inflammation, and I kept noticing that the inflammatory pathways it was suppressing: TNF-α, IL-6, IL-8, NF-κB were exactly the same ones driving PIE.
And unlike tranexamic acid (which has beautiful mechanisms but can’t get through the skin ( more on that later ), ectoine’s target is much more accessible. It’s a tiny molecule (142 Da) that, while hydrophilic like TXA, doesn’t need to reach the deep dermis. Its mechanism operates at the cell membrane surface of keratinocytes in the outer viable epidermis, a much shallower target. And crucially, we have clinical data from topical application on intact skin showing it actually reduces redness, which is something TXA simply cannot demonstrate for PIE.
So I tried it. And personally, I saw a noticeable reduction in redness. That’s anecdotal, not a clinical trial, and I want to be honest about that distinction. But the mechanistic case is strong enough that I think it deserves a place in this protocol.
What is ectoine? It’s a cyclic amino acid made by extremophilic bacteria, organisms that survive in volcanic springs and salt flats. They produce ectoine as a molecular shield to protect their cells from extreme heat, UV, and osmotic stress. When you apply it to your skin, it does something similar: it wraps a structured water shell around your cell membranes, stabilising them against stress-induced damage.
Why does that matter for PIE? Think of it this way.
The inflammatory cascade that creates PIE starts when acne bacteria (C. acnes) hit an alarm button on your skin cells called a Toll-like receptor (TLR-2). That alarm triggers a chain reaction: inflammatory signals flood out, blood vessels dilate, new vessels form, and the redness cycle begins. Most anti-inflammatory ingredients, including niacinamide, try to mute the alarm after it’s already gone off. They’re running into the room and turning down the volume on a siren that’s already blaring.
Ectoine does something different. By stabilising the cell membrane, it physically changes the environment where that alarm button sits.
Imagine putting a protective case over the fire alarm so it can’t be pulled as easily. The TLR receptors need to assemble into specific clusters in the membrane to fire, and ectoine’s membrane-stabilising effect disrupts that assembly. The inflammatory signal never gets properly transduced in the first place.
That’s upstream prevention, not downstream damage control. And for a condition like PIE, where the whole problem is a self-sustaining inflammatory loop, intervening at the earliest possible point makes a real difference.
The erythema data is direct and impressive. In a double-blind, placebo-controlled study conducted by bitop AG (the ectoine manufacturer, I should be transparent that this is proprietary technical data, not a peer-reviewed publication), 1% ectoine reduced skin redness by 38% after just 7 days in an SDS-irritation model, with efficacy equivalent to or exceeding 0.25% hydrocortisone cream. That’s a steroid-level result from a non-drug ingredient at just 1%.
The peer-reviewed evidence comes from atopic dermatitis studies: Trusova et al. (2019) showed 67% hyperemia reduction in mild and 44% in moderate atopic dermatitis, not PIE specifically, but the inflammatory erythema mechanism overlaps. In retinoid dermatitis, which is very relevant to post-acne skin, 7% ectoine cream resolved erythema comparably to dexpanthenol (non-inferior, not superior).
Clinical data also shows ectoine suppresses the exact inflammatory messengers keeping PIE alive, IL-6 down by 50%, TNF-α down by 40%, IL-8 down by 45% (reported in bitop AG’s technical literature).
You know transparency matters to me so let me be clear about two things:
these specific percentages come from the manufacturer’s data, not independent peer-reviewed publications.
the studies were conducted on lung inflammation and chemotherapy-induced mucositis, not skin specifically.
But these are the same cytokines, the same pathways, the same biology. Inflammation is inflammation, whether it’s in lung tissue or in the papillary dermis around a damaged capillary.
The TNF-α part is especially relevant. TNF-α doesn’t just maintain inflammation, it also drives the formation of new blood vessels by upregulating VEGF.
Remember, excess angiogenesis is one of the core engines keeping PIE alive. So when ectoine suppresses TNF-α, it may also be taking the foot off the accelerator of new vessel formation. That’s a two-for-one you don’t get from niacinamide.
And it’s incredibly gentle. This matters more than people realise. PIE lives on post-acne skin that’s already sensitised, barrier-compromised, and reactive. The last thing you need is an active that stings. Niacinamide, while generally well-tolerated, can cause stinging on barrier-compromised or rosacea-prone skin, and this is more about the individual’s barrier status and the formulation’s pH than a simple concentration threshold, but it does happen and it’s relevant for the PIE population.
Ectoine has shown zero irritation across all concentrations tested, including on children and on retinoid-treated skin. For PIE specifically, where your skin is already angry, that tolerability profile is a genuine advantage.
It also supports barrier repair, ectoine-containing formulations have shown improved barrier function (reduced TEWL, increased hydration) in clinical studies, likely through a combination of ectoine’s membrane-stabilising effects and co-formulated barrier lipids. A leaky, compromised barrier lets environmental triggers (UV, pollution, bacteria) re-activate the inflammatory cycle. Sealing it back up helps break the loop.
So why isn’t ectoine on every PIE recommendation list? Because the evidence, while mechanistically compelling, hasn’t been tested in a formal PIE-specific RCT yet. The erythema outcomes come from atopic dermatitis, retinoid dermatitis, and allergic inflammation models. The extrapolation to PIE is biologically sound: same cytokines, same vascular inflammation, same barrier compromise, but I want to be clear that we’re connecting dots between adjacent fields, not citing a direct PIE trial.
I’d love to see someone run that study. In the meantime, the mechanism makes sense, the erythema reduction data is real, the tolerability is outstanding, and my own experience was positive.
Step 3: A retinoid — to thicken your skin, rebuild the scaffolding, and block VEGF
Retinoids earn their place in a PIE protocol through three separate mechanisms:
1. They thicken your epidermis. Part of why PIE looks so dramatic is that post-inflammatory skin is thinner than normal; the underlying vascular damage becomes disproportionately visible through the surface. Retinoids stimulate epidermal turnover, rebuilding that thickness so the redness shows less.
2. They restore dermal collagen. The collagen and elastin around your capillaries got degraded by matrix metalloproteinases (MMPs) during the inflammatory response. Without that structural scaffolding, the vessels stay dilated. Retinoids stimulate collagen I and III synthesis, giving vessels the support they need to remodel back to normal.
3. They directly block VEGF production. Kim et al. 2006, Journal of Investigative Dermatology, demonstrated that all-trans retinoic acid blocks UV-induced VEGF production and angiogenesis via ERK pathway inhibition in human skin keratinocytes. An earlier study by Vega Diaz et al. (2000) independently established that retinoids inhibit VEGF expression in keratinocytes via anti-AP1 activity. That’s the exact growth factor keeping PIE alive. This makes retinoids the only topical that simultaneously addresses the structural, epidermal, and vascular components of PIE.
Start gently, retinoid dermatitis can temporarily worsen redness. Tretinoin 0.025–0.05%, retinaldehyde, or adapalene are all reasonable options. Build up slowly.
And this is where ectoine really shines as a companion, it was shown to resolve retinoid dermatitis comparably to dexpanthenol and even prevent it prophylactically. If you’re using a retinoid for PIE and your skin is struggling with the irritation, ectoine is the logical buffer.
Stability matters: retinol and tretinoin degrade rapidly on exposure to air and light. Opaque, airless packaging is the minimum.
Step 4: Centella asiatica — for wound healing support
Centella asiatica earns its place based on wound-healing and anti-inflammatory data, not PIE-specific trials.
Damkerngsuntorn et al. (2020) showed that a standardised centella extract (ECa 233, in which madecassoside constitutes approximately 51% of the extract’s composition) applied after fractional laser resulted in significantly less erythema; erythema subsided by day 7 versus day 28 on placebo. The mechanism involves madecassoside inhibiting IL-1β, TLR2, and NF-κB, plus stimulating collagen I and III synthesis.
This is post-procedural erythema data, not PIE data, so I’m being transparent about that extrapolation. But the pathways overlap substantially, and centella-based products are well-tolerated and support the healing environment PIE needs.
Look for products listing madecassoside or a standardised centella extract (TECA or ECa 233) rather than just generic “centella asiatica extract,” which can vary wildly in active triterpene content.
Step 5: SPF 50+ — every single day, non-negotiable
UV radiation is a potent inducer of VEGF, the exact growth factor maintaining the abnormal vasculature in PIE. This isn’t theoretical. Brauchle et al. (1996) first demonstrated it in keratinocytes. Yano et al. (2004) showed that UV creates an “angiogenic switch” by simultaneously upregulating VEGF and downregulating thrombospondin-1 (your skin’s natural anti-angiogenic factor). Hirakawa et al. (2005) showed that blocking VEGF reduced UV sensitivity entirely.
Even sub-erythemal UV doses, well below the sunburn threshold, upregulate VEGF. Every time you skip SPF, you’re feeding the exact vascular pathway keeping those red marks alive.
UV doesn’t “darken” PIE the way it darkens PIH, it’s not making more melanin.
But it increases vasodilation, triggers additional angiogenesis, sustains inflammation through reactive oxygen species, and can also trigger PIH development alongside PIE. So you end up with red marks that also start turning brown. Double trouble.
Daily broad-spectrum SPF 50+. If you have PIE coexisting with melasma or PIH, a tinted sunscreen with iron oxides is ideal (for visible light protection).
If it’s pure PIE, any well-formulated SPF 50+ will do. Mineral filters containing zinc oxide offer inherent anti-inflammatory properties, which is a nice bonus on compromised post-acne skin.
Your actual daily routine (putting it together)
Morning: Gentle cleanser → Ectoine serum or cream (1–7%) → Optional: Centella-based moisturiser or a serum → SPF 50+ (tinted if you also have PIH or melasma).
Evening: Cleanser → Azelaic acid (15–20%) → Retinoid (tretinoin, retinaldehyde, or adapalene — start low, build up) → Ectoine serum or moisturiser (to buffer retinoid irritation and continue suppressing the inflammatory cascade overnight) → Moisturiser if needed (look for ceramide complex or centella for additional barrier support).
Important notes: Ectoine is your daily constant: morning AND evening, indefinitely. In the morning, it’s calming, barrier-repairing, and targets the upstream inflammatory cascade driving your PIE. In the evening, it doubles as your retinoid buffer, the retinoid dermatitis study showed ectoine not only resolved retinoid-induced erythema comparably to dexpanthenol, but also had a prophylactic effect when used from the start.
If your skin struggles with retinoid tolerance, ectoine is the logical companion. Your azelaic acid is your primary active treatment, the only ingredient with direct PIE-specific vascular evidence.
Your retinoid is your ongoing structural rebuilder, don’t skip it.
Your SPF is your lifelong non-negotiable, every unprotected UV exposure is feeding the vascular damage.
Control active acne first, every new breakout restarts the PIE cycle.
How long will this actually take?
I’m going to be straight with you.
PIE from superficial inflammatory acne (small papules, pustules) typically resolves in 6–12 months without any treatment at all. Yes, it does fade on its own. Slowly.
With an active topical protocol, visible improvement begins at 8–12 weeks, with meaningful resolution at 3–6 months.
More persistent marks, especially from nodular or cystic acne, can stick around for 1–2 years or longer. The depth relationship is straightforward: deeper inflammation = more extensive dermal damage = more VEGF-driven angiogenesis = more capillary proliferation = longer vascular remodelling time.
Picking at your skin makes everything worse. Manual manipulation of acne lesions substantially intensifies local inflammation and tissue trauma, creating more severe and longer-lasting PIE. I know you know this. But I’m saying it anyway.
The most comprehensive systematic review found that pulsed dye laser can produce significant improvement after a single session at one-month follow-up. Topicals take longer but are accessible to everyone.
Give your protocol a minimum of 12 weeks before you decide whether it’s working.
The ingredients that won’t fix your red marks (despite the marketing)
This is the most actionable section in this article and the most commonly misunderstood.
Hydroquinone, kojic acid, alpha arbutin, tyrosinase inhibitors in general:
These work by slowing melanin production. Hydroquinone competes with tyrosine at the tyrosinase active site. Kojic acid chelates copper at tyrosinase’s catalytic centre. Alpha arbutin inhibits tyrosinase through structural mimicry. None of these mechanisms have anything to do with dilated blood vessels, damaged capillaries, VEGF-driven angiogenesis, or vascular remodelling. They are mechanistically irrelevant to PIE.
If your marks are red or pink and they blanch when you press a glass against them (the diascopy test, try it), these ingredients are doing nothing for you. Save your money.
The one nuance: if you’ve got red marks AND brown marks side by side, tyrosinase inhibitors address the brown component. But they won’t touch the red. This coexistence of PIE + PIH is particularly challenging to treat in Fitzpatrick III–VI skin, where PIE can present as purple or violet rather than pink/red, making it easier to mistake for PIH. PIE is also underdiagnosed in deeper skin tones because the erythema is less visually apparent, not because it’s absent. This is where azelaic acid shines, it’s one of the few ingredients that straddles both pathways: anti-angiogenic + anti-tyrosinase.
Topical tranexamic acid (on intact skin): the penetration problem.
I need to address this one directly, because TXA is everywhere right now, including in PIE recommendations.
If you read my melasma article, you already know I’m sceptical of topical TXA. For PIE, the problem is even worse. Here’s why.
TXA has a log P in the range of −1.3 to −1.7 (the exact value varies by database and prediction method, but all agree it’s strongly negative). The ideal range for passive skin penetration is 1 to 3. That means TXA partitions into the stratum corneum’s lipid matrix orders of magnitude less efficiently than an optimally lipophilic compound. It’s a permanent zwitterion across the entire physiological pH range, simultaneously carrying positive and negative charges, with essentially no neutral fraction available for passive diffusion. It’s freely soluble in water but basically insoluble in anything lipophilic. From a formulation science perspective, this is one of the worst possible penetration profiles imaginable.
Ng et al. (2020) performed the key permeation study using Franz diffusion cells with human skin and found that a standard branded 2% TXA cream delivered barely detectable concentrations to the epidermis. Even with a specialised co-enhancer system, the same cream still needed 8.5x more TXA just to reach estimated therapeutic concentrations at the dermo-epidermal junction. Winn et al. (2025) confirmed it in vivo using confocal Raman spectroscopy: standard TXA largely accumulated in the outer stratum corneum and needed ester prodrug modification to reach the viable epidermis at all.
For melasma, TXA at least has an epidermal target (keratinocyte uPA signalling at ~50–100 µm depth). For PIE, the target is even deeper: damaged capillaries in the papillary dermis at 100–200+ µm. If TXA can’t reliably reach the basal epidermis, it certainly can’t reach the papillary dermis.
And when you actually look at the studies cited for “topical TXA” in PIE, most of them aren’t truly topical on intact skin. Bazargan et al. (2023): intradermal injection. Sharara et al. (2024): microneedling + TXA (barrier disruption). Agamia et al. (2022): triple combo with two vasoconstrictors, so you can’t isolate TXA’s contribution.
The only true topical-on-intact-skin PIE study is Jakhar & Kaur (2020): a brief therapeutic pearl with no controls, no blinding, no statistical analysis.
The 2024 meta-analysis I cited in the melasma article tells the same story: topical TXA at 8 weeks showed P = 0.92, nowhere near statistical significance.
Bottom line: TXA has genuinely interesting anti-angiogenic and anti-plasmin mechanisms that are directly relevant to PIE. But those mechanisms need TXA to actually reach the target tissue. On intact skin, with standard formulations, it almost certainly doesn’t. If your dermatologist delivers TXA via microneedling or mesotherapy, that bypasses the barrier entirely and the mechanism becomes relevant. But dabbing a TXA serum on your face for red marks? The pharmacokinetics say no.
Vitamin C (alone, as a PIE treatment): Vitamin C has antioxidant and collagen-synthesis value, and I’m not telling you to throw it away. In one small study, a 5% vitamin C concentrate showed erythema reduction of 9% at 2 weeks, 16% at 4 weeks, and 21% at 6 weeks. That’s something. But it’s modest. If you’re only going to use one serum for your red marks and you’re choosing vitamin C over azelaic acid, you’re bringing a butter knife to a sword fight.
Vitamin C is better positioned as a supporting player: antioxidant protection, collagen support, anti-pollution shield (as I covered in the melasma article). Not your lead PIE ingredient.
“Dark spot correctors” built around tyrosinase inhibitor blends: Products marketed as dark spot treatments that combine kojic acid + alpha arbutin + vitamin C are targeting melanin from every angle. Great for PIH. Useless for PIE. If your marks are red, not brown, these are the wrong products entirely.
Quick note on the diascopy test — how to tell if you have PIE or PIH
This is the simplest diagnostic you can do at home. Press a clear glass (or the back of a clear phone case) firmly against the mark.
If it blanches (the colour temporarily disappears or fades): that’s PIE. The colour is coming from blood in dilated vessels, and pressing the glass displaces it. Your marks are vascular.
If it doesn’t blanch (the colour stays): that’s PIH. The colour is from melanin deposits. Your marks are pigment-driven.
If it partially blanches: you’ve got both. Address the vascular and pigment components separately.
Parraga & Mayo (2025) showed that dermoscopy can reveal positive blanching even in deeper skin tones where it’s hard to see macroscopically. If you’re unsure, a dermatologist can do this more precisely.
Professional treatments that complement topicals
If your PIE is extensive or persistent, these are the energy-based options to discuss with your dermatologist or aesthetician:
Pulsed dye laser (595 nm) is the first-line device for PIE. It targets oxyhaemoglobin via selective photothermolysis, it finds the red and zaps it. Yoon et al. (2008) showed 90% of patients achieved clinical improvement with 57.6% reduction in lesion counts after two sessions. This is the gold standard.
IPL with a vascular filter is effective, Wu et al. (2022) reported 81.7% complete or partial clearance, but carries higher risk of PIH in darker skin types than PDL.
Fractional microneedling radiofrequency is the only treatment with histological evidence of VEGF and NF-κB reduction in actual PIE tissue. That’s uniquely compelling from a mechanistic standpoint.
LED red light (633 nm) plays a supporting role: reduces inflammation and promotes wound healing, but doesn’t directly target vascular pathology. Think of it as creating a better healing environment, not resolving established PIE.
OK, so what IS PIE actually? (The science, explained simply)
You’ve made it this far, so let me explain what’s actually happening under your skin. I’ll keep it simple.
When you get an acne breakout, your immune system launches an inflammatory attack. The bacteria (Cutibacterium acnes) trigger an alarm system (TLR-2 receptors), and your body floods the area with inflammatory signals: TNF-α, IL-1β, IL-6, IL-8, recruiting immune cells. To get those immune cells to the area quickly, your blood vessels dilate. Blood flow increases. That’s the redness you see during active acne.
This is all supposed to be temporary. Once the acne resolves, the inflammation should calm down, the vessels should return to normal, and the redness should disappear.
In PIE, it doesn’t.
Here’s why — and this is the key part. The inflammatory response doesn’t just dilate existing vessels. It triggers the formation of entirely new ones (angiogenesis) through a growth factor called VEGF (vascular endothelial growth factor). At the same time, the inflammation releases enzymes (matrix metalloproteinases) that degrade the collagen and elastin supporting those vessels. So now you’ve got extra blood vessels that lack structural support, can’t remodel back to normal, and just... stay dilated.
And then it gets worse, because PIE is self-reinforcing. The damaged vasculature produces nitric oxide, which activates COX-2 to produce prostaglandin E2 (PGE2). PGE2 is itself a potent VEGF inducer. And VEGF drives more angiogenesis. And the new abnormal vessels produce more nitric oxide. It’s a loop. The damaged blood vessels generate the signals that maintain the damaged blood vessels.
This is why PIE can persist for months or even years after the acne is completely gone. The original trigger (acne) has left, but the vascular damage has become self-sustaining.
On top of that, your epidermis over those areas is still thin and healing. Even mild residual capillary dilation, which might be completely invisible through thick, healthy skin, creates striking red or pink marks through that thinned surface.
The only study with actual histological and immunohistochemical data from PIE tissue confirmed all of this directly: successful treatment significantly downregulated VEGF, NF-κB, and IL-8 in PIE lesions.
Now do you see why a “dark spot corrector” containing kojic acid isn’t going to touch this? The colour isn’t melanin. It’s haemoglobin in damaged blood vessels sustained by a self-reinforcing inflammatory-vascular loop. You need anti-inflammatories (azelaic acid, ectoine), structural rebuilders (retinoids, centella), and UV protection (to stop feeding the VEGF cycle). And you need to understand which ingredients actually reach their target tissue, which is why I excluded topical TXA despite its interesting mechanism.
PIE vs PIH vs acne scars — why people confuse them
These three things are routinely mixed up, but they’re completely different problems requiring different solutions.
PIE is vascular. Red/pink marks. Caused by damaged blood vessels. Blanches with pressure. Treat with anti-inflammatories, anti-angiogenics, retinoids, SPF.
PIH is melanin-driven. Brown/tan marks. Caused by overactive melanocytes triggered by inflammation. Doesn’t blanch. Treat with tyrosinase inhibitors (hydroquinone, azelaic acid, alpha arbutin, kojic acid), retinoids, SPF.
Acne scars are structural. You can feel them, texture changes, indentations (ice pick, boxcar, rolling) or raised tissue (hypertrophic, keloid). Caused by collagen loss or excess. Require mechanical remodelling: microneedling, fractional lasers, subcision, fillers.
All three can coexist on the same face. And there’s molecular crosstalk: VEGF upregulation (the PIE driver) triggers prostaglandin increases that activate melanocytes. This means untreated PIE can lead to PIH over time. Addressing the redness early may help prevent the brown marks from developing.
The bottom line
PIE is a vascular condition, not a pigment problem. The red marks you’re seeing are damaged, dilated blood vessels sustained by a self-reinforcing loop of inflammation and angiogenesis.
The right protocol targets four things simultaneously:
Azelaic acid (15–20%) as your primary active, the strongest evidence for topical PIE treatment, with anti-inflammatory, anti-angiogenic, and antioxidant properties in one ingredient.
Ectoine (1–7%) to shut down the inflammatory cascade at its most upstream point, stabilising cell membranes, disrupting TLR signalling, suppressing the cytokines (TNF-α, IL-6, IL-8) that sustain vascular damage, with clinical erythema reduction data from adjacent inflammatory skin conditions and zero tolerability concerns on sensitive post-acne skin.
A retinoid to thicken your epidermis, rebuild dermal collagen scaffolding, and block VEGF production.
SPF 50+ daily, because UV directly feeds the VEGF cycle maintaining your red marks.
Centella asiatica as wound healing support, collagen synthesis, anti-inflammatory, well-tolerated.
Give the protocol 12 weeks minimum. Be patient. PIE will fade, it’s not permanent, but it takes time, and the deeper the original acne, the longer the marks persist.
Be sceptical of any product marketed as a “dark spot corrector” for your red marks. If it’s built around tyrosinase inhibitors and nothing else, it’s the wrong tool for the job. And be equally sceptical of topical TXA serums for PIE, the mechanism is beautiful on paper, but the molecule almost certainly can’t reach the damaged blood vessels through intact skin. Ask your derm about microneedling delivery or mesotherapy if you want TXA’s benefits; a serum won’t cut it.
And stop picking. I’m begging you.
You’ve got this. x
Marina is a cosmetic chemist, biochemist, and science-based skincare creator and educator. Follow her on TikTok, Instagram and YouTube @4complexion for evidence-based skincare content.
Disclaimer: This article is for educational purposes only and is not medical advice. I’m a cosmetic chemist, not a doctor. If you have persistent post-acne marks that aren’t resolving, please see a dermatologist — especially if you’re unsure whether you’re dealing with PIE, PIH, or scarring.
Thank you so much, so thorough and informative. Any thoughts on Paula’s Choice 10% azaleic acid booster?
Hi Marina, thank you for your incredibly helpful advice regarding my skin condition, but I'm still concerned about using azelaic acid and retinoid together in my routine. Will it overload my system or negate the effects of either ingredient? Also, can I replace azelaic acid with a PAD (for example, Geek & Gorgeous's aPAD)? Could you give me some advice? Thank you so much.