Frostbite

Overview

Frostbite is a two-stage tissue injury in which ice crystals form inside and between cells when exposed to temperatures below the freezing point of tissue, and during rewarming, microvascular thrombosis and inflammation are added. The damaged area is divided into three zones: irreversible coagulation zone in the center, surrounding stasis zone which can be salvaged with treatment, and the outermost hyperemia zone.

The entire purpose of treatment is to salvage the stasis zone; if this zone is not treated, it converts to the coagulation zone within hours. Ice crystals disturb the cell membrane and electrolyte balance; tromboxane A2 and prostaglandin F2-alpha released with endothelial damage deepen vasoconstriction and thrombosis.

All agents like ibuprofen, aloe vera, and iloprost target this arachidonic acid pathway. Frostbite most often involves the acral regions, namely the fingers and toes, ears, nose, and cheeks.

The patient group consists mainly of socially vulnerable people; Alcohol and substance use disorder, homelessness and psychiatric illness predominate in this population. In a study examining 148 frostbite patients admitted to a burn center, 18% of injuries resulted in amputation, and alcohol and substance use were found to be independent predictors of amputation (Endorf et al., Medicine, 2022).

Diabetes, peripheral artery disease, smoking, malnutrition, advanced age and previous history of frostbite are other known risk factors. The classical grade classification (first-fourth degree) similar to burns was on the verge of being abandoned because it did not give a prognosis in the first days.

The classification developed by Cauchy et al. from 70 cases of severe frostbite (Wilderness & Environmental Medicine, 2001) is based on the ANATOMICAL SPREAD of cyanosis after rewarming and early bone scintigraphy; predict the outcome without the 15-day wait required by previous classifications. In this study, the probability of amputation was found to be directly related to the proximal border of the lesion: approximately 1% if the cyanosis is confined to the distal phalanx, 31% if it extends to the middle phalanx, 67% if it extends to the proximal phalanx, 98% if it reaches the level of the metacarpal or metatarsal, and close to 100% if it reaches the carpal-tarsal level.

Amputation is not expected in a grade 1 lesion, grade 2 indicates soft tissue amputation, grade 3 indicates bone amputation, and grade 4 indicates extensive amputation accompanied by systemic effects. Since it is known that grade 2 lesions can progress to grade 3 in the first 24 hours, early staging should not be considered definite and should be repeated.

The diagnosis is clinical; the initial appearance does not reflect the true depth. The natural course follows a predictable timeline: edema begins 3-5 hours after rewarming and can last up to a week, blisters appear in 4-24 hours, eschar becomes evident in 10-15 days, and the demarcation line and mummification become clear in 3-8 weeks.

The role of imaging is to early estimate the depth and level of amputation. Technetium-99m bone scintigraphy is the best validated test for this purpose: in a study examining 92 severe frostbite injuries, a direct correlation was found between the uptake area in the phalanges and the final amputation level, with a positive predictive value of 0.84 and a negative predictive value of 0.99 reported; scintigraphy accurately indicated the amputation level in more than 84% of cases.

In the first days after injury, the initial scintigraphy taken early (e.g., on the 3 day) has high specificity; a second repeated scan around day 7 is more sensitive and informative. The addition of single-photon emission computed tomography (SPECT/CT) increases anatomical resolution and particularly facilitates the determination of the surgical margin in small fingers.

Angiography in the first 24 hours shows vascular obstruction and identifies the target for intra-arterial thrombolysis; magnetic resonance angiography, on the other hand, is an alternative to scintigraphy for assessing tissue viability.

Key considerations

• Frostnip is a superficial cold injury that heralds frostbite; it causes pallor and numbness in the skin but since ice crystals do not form, sensation fully returns with rewarming, blisters do not develop, and no tissue loss occurs.
• Trench foot and other non-freezing cold injuries occur with prolonged wet exposure at temperatures ABOVE freezing; it can develop even at relatively warm temperatures such as 15°C if the foot remains continuously wet. Since wetness is required, it is not expected to occur simultaneously with frostbite, and ice crystals are not present.
• Chilblains (pernio) are an inflammatory response to recurrent DRY cold; they cause itchy, painful, red-purple nodules on the back of the hands and the front of the legs, resolve spontaneously, and unlike frostbite, do not cause permanent tissue loss.
• Cold urticaria and cold panniculitis produce urticarial plaques or subcutaneous induration in minutes in the area exposed to cold; the condition is temporary, cyanosis and demarcation do not develop.
• Acute arterial ischemia and advanced peripheral artery disease cause paleness, coldness, and pulselessness in the extremity even without cold exposure; in frostbite, pulses are often preserved and the history includes a defining freezing exposure.
• Raynaud's phenomenon manifests with a three-phase color change (paleness, cyanosis, redness) triggered by cold or stress and fully resolves with warming; unlike frostbite, it is episodic and tissue necrosis occurs only if there is an underlying connective tissue disease.
• Cryoglobulinemia and cold agglutinin disease cause acral cyanosis and retiform purpura in the cold; unlike frostbite, systemic findings, hemolysis, or vasculitis findings accompany it, and the exposure is not at a freezing level.
• Necrotizing soft tissue infection can also develop on the background of frostbite and differs from the predictable course of frostbite with pain disproportionate to the physical exam, rapidly spreading erythema, crepitus, and systemic toxicity; this condition requires surgical debridement within hours.
• Purpura fulminans and meningococcemia cause acral symmetrical necrosis; fever, hypotension, widespread intravascular coagulation, and petechial rash differentiate the condition from frostbite, and there is no history of cold exposure.
• Vasopressor-induced digital ischemia causes symmetrical acral necrosis in an intensive care patient receiving high doses of norepinephrine or vasopressin; the drug history and absence of freezing exposure distinguish it.
• Contact with compressed gas or cryogen (liquid nitrogen, propane) creates frostbite in the contact area within seconds; the lesion is sharply demarcated and the exposure time is very short, but the pathophysiology and treatment are the same as classic frostbite.

Why it matters

The framework of this protocol is based on the Wilderness Medical Society (WMS) Frostbite Prevention and Treatment Clinical Practice Guideline 2024 Update; since this guideline is the only comprehensive international guideline in the field, there is no classical conflict between guidelines, but in the points specified below, the sources diverge. It is recommended that the temperature of the reheating water be 37-39°C and maintained for about 30 minutes until the tissue softens and becomes flexible.

There is a real source discrepancy here: while StatPearls and some emergency medicine sources continue to provide the traditional range of 40-42°C, WMS has adopted the lower range of 37-39°C; the rationale is that a higher temperature in already insensate tissue would add thermal burns and pain. Canadian protocols allow for a wider range, such as 36-42°C.

Practical consensus is that the temperature should be tolerable to touch and kept constant throughout the bath. Dry heat, radiators, fire, and rubbing are prohibited; since insensate tissue cannot feel excessive heat, these methods add burn and mechanical trauma to frost damage.

If there is a risk of refreezing in the field, heating is NOT INITIATED. The reason is that the freeze-thaw-refreeze cycle is far more destructive than remaining continuously frozen; the edema and reperfusion produced during thawing create larger ice crystals and deeper necrosis when refrozen.

Walking on a frozen extremity causes less harm than walking on a thawed extremity. Accompanying systemic hypothermia is always corrected first, and the core temperature is raised above 35°C.

The rationale is the afterdrop phenomenon: warming the periphery first returns cold peripheral blood to the core through vasodilation, further lowering core temperature and potentially triggering ventricular arrhythmia. Therefore, the trunk is warmed before the extremities, and in moderate-to-severe hypothermia, the patient is kept horizontal.

Ibuprofen is given at a dose of 12 mg/kg/day divided into two doses (maximum 2400 mg/day). This dose is aimed not at analgesia but at antiprostaglandin effect: cyclooxygenase inhibition reduces thromboxane A2 production, limiting vasoconstriction and thrombosis in the zone of stasis.

For the same reason, topical aloe vera is applied every 6 hours, as aloe also inhibits thromboxane. In patients planned for thrombolysis, ibuprofen is not given because it increases the risk of bleeding.

The basis for the thrombolysis recommendation is the 2007 study by Bruen and colleagues in the Archives of Surgery: when tissue plasminogen activator (tPA) was administered within the first 24 hours after injury, the rate of digital amputation decreased from 41% to 10%. It should be noted that this study was retrospective and the number of patients receiving tPA was only seven.

A 2023 systematic review and meta-analysis supported this finding: thrombolysis increased tissue salvage rates by an average of 25% (95% confidence interval 8-42) and the risk ratio for preserving fingers at risk was found to be 5 (95% confidence interval 3-7). The reason for the narrow time window is a period of progressive ischemia that begins after rewarming; this period is called warm ischemia time, and each hour of delay has been reported to significantly reduce tissue salvage rates.

Iloprost is a prostacyclin analogue; it causes vasodilation, inhibits platelet aggregation, and unlike thrombolytics, does not carry a bleeding risk. The basis is the open-label randomized study by Cauchy and colleagues including 47 severe frostbite patients: according to bone scintigraphy on day 7, the risk of amputation was 0% in the iloprost group, 19% in the iloprost plus tPA group, and 60% in the buflomedil group.

This study formed the basis for the United States Food and Drug Administration (FDA) approval of iloprost (Aurlumyn) in February 2024 to reduce the risk of finger amputation in severe frostbite in adults; this is the first drug approved for frostbite. In the same study, it was NOT SHOWN that the iloprost plus tPA group was superior to the iloprost alone group: randomization was imbalanced and most of the severely affected fingers were in the combination group, so no conclusion can be drawn that adding tPA provides additional benefit.

The time window for iloprost is the second point that varies among sources. WMS 2024 recommends iloprost as soon as possible, preferably within 48 hours and at the latest up to 72 hours after thawing, for distal interphalangeal joint or more proximal deep frostbite (grade 2-4); however, within the panel, it is not clear whether iloprost is necessary for grade 2 frostbite.

Some center protocols similarly use the 72-hour window for grade 2-4 involvement. The main advantage of iloprost over thrombolysis is that it maintains its effectiveness even with delayed presentation and does not require angiography; in contrast, tPA is only significant within the first 24 hours.

In countries where iloprost is not available, tPA is the only option for grade 4 and possibly grade 3 frostbite; when thrombolysis is contraindicated, iloprost is preferred. An important point for Turkey is this: iloprost (Ilomedin) is licensed in our country, but its indications are limited to Buerger's disease, severe peripheral arterial obstructive disease, and resistant Raynaud's phenomenon; its use in frostbite is off-label and requires the relevant approval process.

Abandoned teachings should also be known. Rubbing with snow and applying dry heat have been abandoned because they cause mechanical and thermal additional damage.

Hyperbaric oxygen therapy is NOT RECOMMENDED in frostbite; hyperoxia can cause vasoconstriction, worsen the zone of stasis, and its benefit has not been proven. Chemical or surgical sympathectomy is not recommended since it has not been shown to reduce tissue loss in the acute period and has been replaced by peripheral nerve blocks when necessary.

Intra-arterial vasodilators such as low molecular weight dextran, reserpine, and tolazoline have no proven clinical benefit. Prophylactic antibiotics are not routinely given; they are considered only in cases of significant contamination, crush injury, or wet gangrene.

Draining clear blisters and leaving hemorrhagic blisters intact is a common practice and is justified by the fact that the fluid in clear blisters contains thromboxane and prostaglandins that maintain tissue damage; however, WMS explicitly states that the evidence is INSUFFICIENT to recommend this practice. Hemorrhagic blisters indicate deeper damage, and opening them exposes the dermal matrix to infection and therefore they should remain untouched.

Early surgical amputation is avoided. Since the demarcation line and tissue viability become clear in 3-8 weeks, urgent surgery sacrifices tissue that could later be saved; black eschar often falls off leaving the underlying viable tissue.

This principle is the oldest and most solid rule of frostbite care. Exceptions are situations requiring emergency surgery, such as wet gangrene, infection, or compartment syndrome.

Critical complications

• Compartment syndrome appears with intense edema developing after rewarming in deep frostbite; pain disproportionate to examination, pain increasing with passive stretch, tense swollen compartment, and progressive sensory loss are warning signs. Since classic findings are unreliable in an extremity that has already lost sensation, compartment pressure is measured if suspected; if the diagnosis is confirmed, emergency fasciotomy is performed and all compartments are opened.
• Rhabdomyolysis and resulting hyperkalemia with acute kidney injury are seen in deep frostbite involving a large muscle mass and on the basis of compartment syndrome. Creatine kinase, potassium, creatinine, and urinary myoglobin are monitored; treatment consists of aggressive isotonic fluid replacement, hourly urine output monitoring, and standard management of hyperkalemia.
• Infection ranges from cellulitis to wet gangrene and sepsis; fever, spreading erythema, purulent discharge, foul odor, and gray-edematous tissue appearance confirm the diagnosis. Conversion to wet gangrene, which is expected to be dry gangrene, requires emergency surgical debridement and systemic antibiotics. Prophylactic antibiotics are not routinely given; therefore, infection must be detected through clinical follow-up.
• Thrombolysis bleeding is the most important iatrogenic complication; in approximately 8.4% of patients receiving tPA, bleeding requiring a change in treatment has been reported. Patients are monitored for catheter entry site, gastrointestinal system, and central nervous system bleedings; if bleeding occurs, the infusion is stopped and standard thrombolysis bleeding management is applied. This is also the reason for stopping ibuprofen in patients planned for thrombolysis.
• Side effects associated with Iloprost infusion are expected and dose-dependent; according to the brief product information, headache occurs in approximately 68.8% of patients, facial flushing in 58%, and gastrointestinal complaints in 29.7%. Hypotension, palpitations, tachycardia, nausea, and vomiting are also common. Management is reducing the infusion rate; if the condition does not improve, the infusion is stopped and continued at the highest tolerated rate.
• Refreezing (freeze-thaw-refreeze cycle) is entirely an iatrogenic and preventable complication; it occurs in the field with an early decision to rewarm and produces much deeper necrosis than remaining continuously frozen. It is recognized through history; prevention is achieved by not starting rewarming before evacuation is secured.
• Afterdrop and ventricular arrhythmia during warming are iatrogenic conditions that occur when the extremities are warmed before the trunk in a hypothermic patient. Core temperature and rhythm are monitored; prevention involves keeping the patient horizontal, warming the trunk first, and not proceeding to limb baths until the core temperature exceeds 35°C.
• Long-term sequelae are a common outcome of frostbite; in a comprehensive review where 493 patients were individually monitored, permanent sequelae were detected in 341 patients (69%), with neurologically-originating complaints (chronic dysesthesia, chronic pain, sensory loss, cold hypersensitivity, and hyperhidrosis) forming the predominant group. Since there is no effective treatment, management is limited to neuropathic pain medications and protection from cold; conditions mimicking complex regional pain syndrome can develop.
• Frostbite arthritis appears as early-onset osteoarthritis in the affected joints and is clinically indistinguishable from ordinary osteoarthritis; years later it manifests as joint stiffness, pain, and decreased manual dexterity. Treatment is standard osteoarthritis management.
• Autoamputation and mummification are part of the natural course observed while awaiting demarcation and are not considered complications on their own; however, early surgical amputation is an iatrogenic error because live tissue is often present under the black eschar, and premature surgery sacrifices salvageable tissue.

Special considerations

• Iloprost is CONTRAINDICATED during pregnancy and lactation; the Turkish short product information lists pregnancy and lactation as absolute contraindications. In these patients, the option in severe frostbite shifts to thrombolysis, but tPA is also relatively contraindicated in pregnancy, and the decision must be made by a specialist committee weighing fetal and maternal benefit-risk balance. Ibuprofen is not used in the third trimester due to the risk of early closure of the ductus arteriosus; during this period, the antiprostaglandin target is abandoned, and paracetamol with opioids is preferred for analgesia.
• Frostbite in children can directly damage the growth cartilage (physeal); early closure of the epiphyseal cartilage can lead to shortening in the phalanges, angular deformity, osteoarthritis, and amputation. Therefore, long-term orthopedic follow-up in pediatric frostbite is mandatory, unlike in adults, and growth monitoring is required even if the initial appearance is mild. The anti-prostaglandin dose of ibuprofen is given as 12 mg/kg/day divided into two doses. The pediatric dose of iloprost is NOT standardized and its use is left to the specialist's decision.
• Elderly patients have a lower frostbite threshold; reduced subcutaneous fat tissue, impaired thermoregulation, low basal metabolic rate, and frequently accompanying peripheral artery disease result in both shorter exposure and deeper damage. Cognitive impairment and dementia delay the recognition of exposure, so presentation is often late and the thrombolysis window is missed; in this group, the 48-72 hour wide window of iloprost provides a practical advantage.
• In kidney or liver failure, the iloprost dose is reduced; the brief product information usually recommends starting at HALF the suggested dose because iloprost clearance is significantly slowed in these patients and hypotension and headache occur more frequently. Ibuprofen, on the other hand, should have its anti-prostaglandin indication re-evaluated in kidney failure, as it can aggravate acute kidney injury, and if administered, it should be used for the shortest possible period.
• In patients with diabetes, peripheral artery disease, and smoking, underlying microvascular insufficiency reduces the chance of saving the stasis zone, and the same exposure results in a higher amputation rate. Because diabetic neuropathy dulls cold sensation, exposure can continue unnoticed. Nicotine causes vasoconstriction, so during the acute phase cigarettes and nicotine products must be completely stopped; this is the only acute intervention in which the patient can contribute.
• Homeless patients, those with alcohol or substance use disorders, and patients with psychiatric diagnoses constitute the majority of the frostbite population, and their outcomes are significantly worse. In a cohort of 148 patients, those with unplanned readmissions had a lower rate of stable housing at 48.7% compared to 75.0%, higher rates of substance use disorder at 60.2% compared to 85.0%, and higher rates of psychiatric illness at 50.9% compared to 70.0%. In a series of 54 patients in a burn center, those with psychiatric diagnoses had an average additional number of surgeries of 1.8 compared to 0.6 and a length of stay of 16.0 days compared to 7.7 days. Therefore, the discharge plan should include not only wound care but also housing, substance use, and psychiatric follow-up; a patient without a warm place to stay is directly discharged at risk of re-freezing.
• Iloprost requires caution in patients with coronary artery disease; unstable angina, myocardial infarction within the last six months, and severe coronary disease are listed as contraindications in the product information. Since iloprost-induced vasodilation and reflex tachycardia may disrupt myocardial oxygen balance, options in this group should be re-evaluated.

Follow-up plan

Daily evaluation of an inpatient includes capillary refill, color, temperature, sensation, and movement, as well as signs of infection; the extremity is continuously elevated to reduce edema, weight is not applied to affected feet, and gauze is gently placed between the toes to prevent maceration. Aloe vera is reapplied every six hours, and ibuprofen is continued until the wound heals or a surgical decision is made.

Bone scintigraphy is generally performed on days 7-10 to predict amputation level; the initial assessment is not considered definitive and is repeated because early-stage progression from grade 2 to grade 3 can occur within the first 24 hours. Surgical decisions are postponed until the demarcation line and tissue viability become clear, typically 3-8 weeks; this waiting period is a fundamental rule of frostbite care and is not altered except in cases of wet gangrene, uncontrollable infection, or compartment syndrome.

Follow-up is maintained by plastic surgery or a burn center; in severe frostbite with grade 2-4 involvement, specialist consultation should be obtained within the first 72 hours. For patients who are compatible, without risk of superficial freezing or refreezing, outpatient follow-up is appropriate; discharge criteria include the return of sensation, absence of blisters, pain controlled with oral analgesics, having a warm place to stay, and being able to come for a check-up within 24-48 hours.

Patients without assured shelter do not meet this criterion. The return instructions for the patient are as follows: the patient should seek medical attention without delay if fever, spreading redness, purulent or foul-smelling discharge, progressively increasing pain, new bruising or blackening on the finger, worsening loss of sensation, or feeling of tightness in the extremity develop.

The patient is instructed not to expose the area to cold again, to quit smoking and nicotine products, and not to puncture blisters by themselves. The patient should be clearly informed that long-term, cold sensitivity, numbness, chronic pain, increased sweating, nail deformity, and color changes may persist; persistent neuropathic complaints occur in approximately two-thirds of patients and there is no completely effective treatment.

Early osteoarthritis may develop in the affected joints over the years. In children, orthopedic follow-up is maintained for deformity and growth retardation due to damage to the growth cartilage.

In hand involvement, early hand therapy and range-of-motion exercises reduce joint stiffness; in patients requiring amputation, a prosthesis and rehabilitation plan are implemented.