Nail fungus (onychomycosis): oral and topical treatment, and do home remedies work?

A thickened, yellowish, brittle or lifting nail is not proof of a fungus. Nail psoriasis, injury, repeated micro-trauma in tight shoes, lichen planus or age-related changes can look similar. That is why clinical trials of drugs include in their analyses only people whose diagnosis is confirmed by microscopy, culture or histology. Treatment without confirmation exposes a person to side effects and costs with no guarantee that there is any target at all.

The result has an important consequence: a negative culture does not rule out a fungal infection, because this method picks up roughly half of cases. A positive PAS stain, in turn, confirms the presence of fungal elements in the nail but does not settle the species. Another study using latent class analysis (106 patients) showed that individual methods differ in sensitivity and specificity and that each centre should assess its own strategy. In practice a doctor may therefore combine two tests, and the patient should know that one negative result does not always close the matter.

One more reason to confirm: the authors of the 2023 safety review of oral drugs recommend that after a first treatment failure the diagnosis be confirmed again and antifungal susceptibility testing considered. So if someone still has a changed nail after a full course, it does not always mean the fungus is resistant: it may be an entirely different condition.

For someone who simply wants to know whether oral drugs work, the answer is: yes, and clearly so. The difference from placebo is large, and the quality of evidence in this comparison was rated high. It is worth looking at it honestly, though: a risk ratio tells you how many times more often cure occurs, not what proportion of people will be cured in practice. The Cochrane review does not give a single simple figure such as 'X in 100 people', and cure rates in individual trials depended on the definition (clinical, mycological or complete cure) and on severity.

Terbinafine turned out to be probably more effective than azoles, with a similar risk of adverse events. Griseofulvin, an older drug, had similar efficacy to azoles but caused adverse events more often and is no longer preferred in this disease. This agrees with the 2023 review, which notes that ketoconazole and griseofulvin are no longer used in onychomycosis because of safety concerns and lower efficacy.

The same first author, in a separate network meta-analysis of monotherapy from 2020 (26 trials, 8,136 patients), showed that the odds of mycological cure with continuous terbinafine 250 mg or continuous itraconazole 200 mg were significantly higher than with topical treatment, whereas fluconazole, pulse regimens and topical drugs did not differ significantly from one another. The two analyses therefore do not give an identical picture, and it is worth remembering that: they come from different sets of trials and comparisons, and neither replaces a doctor's decision based on the individual patient, their medicines and coexisting diseases.

The doctor sets the duration and dose according to the product information; here we give only the conclusions from the studies. The practical sense of these data is this: there is no clear advantage of a pulse regimen over a continuous one, so the choice may depend on convenience, drug interactions and tolerance, and terbinafine performs at least as well as the alternatives in the analyses. Sometimes the choice is decided by another medicine the patient takes, because terbinafine and itraconazole interact with other drugs.

Concern about the liver with antifungal drugs is justified and at the same time often exaggerated. The review of the safety of oral drugs in onychomycosis (Gupta et al., 2023) notes that terbinafine, itraconazole and fluconazole have numerous cytochrome P450 interactions, which can promote adverse effects in the liver, kidneys, skin, circulatory, nervous and blood-forming systems. The authors recommend liver function tests before starting any oral drug, particularly when clinically indicated.

A few rules follow for the reader. Before treatment, tell the doctor about all the medicines and supplements you take, and about liver disease and alcohol. Liver test results (ALT, AST, GGT) taken before treatment and, at the doctor's discretion, during it, allow a problem to be caught early; we discuss their interpretation in our entry on liver function tests. Symptoms such as marked fatigue, dark urine, jaundice, persistent nausea or pain under the right ribs need urgent contact with a doctor, who will decide whether the drug should be stopped.

What follows from this? Topical preparations make sense mainly for mild forms, when part of the plate is affected and the matrix is not involved, and for people who cannot take oral drugs. They demand persistence: most trials lasted 48-52 weeks, and applying a lacquer or solution daily for a year can be harder for patients than a course of tablets for a few months. The results of different preparations are not directly comparable, because they rest on single trials, and some comparisons had low-quality evidence.

It is worth knowing that topical treatment is sometimes combined with filing or removing the affected part of the plate (debridement); the authors of the tea tree oil trial discussed below considered such a combination a reasonable initial strategy. It should also be remembered that efinaconazole and tavaborole more often caused local reactions (dermatitis, vesicles, application-site reactions), although the severity was usually mild.

This is one of the few randomised trials of tea tree oil in nail fungus; the mycological cure rate (18%) is low and based on a comparison with an equally weak preparation. Another small study (60 people) described 80% cures after a cream combining butenafine (an antifungal drug) with the oil, but that result cannot be attributed to the oil, because butenafine is itself an antifungal drug, and a separate comment on the paper was published in the same journal. If someone wants to use the oil as a supplement, they should know that it can irritate the skin around the nail and does not replace treatment whose effectiveness has been shown.

Laser is a newer and more expensive method. The 2020 Cochrane review relied on three trials (112 participants) for the 1064 nm Nd:YAG laser and found no confirmed mycological effect, with very low certainty regarding adverse events. It does not follow that no laser will ever work, only that current evidence from well-controlled trials does not allow cure to be promised. Before paying for a series of sessions, it is worth asking whether the clinic relies on randomised trials with a control group.

Nail fungus is not just a cosmetic problem for everyone. People with diabetes, poor peripheral circulation, neuropathy or weakened immunity should take changes in the toenails more seriously, because a damaged, thickened plate can encourage abrasions and inflammation of the skin of the foot. The 2026 meta-analysis we cite describes a higher frequency of onychomycosis in people with metabolic disorders (53.96% among tested samples), but these are low-certainty data from centre-based studies, not proof that every second person with diabetes has a fungal infection.

For people with diabetes the practical advice is not about choosing a drug, but about whole-foot care: regular inspection, proper footwear, glycaemic control (we write about diabetes markers in our entries on diabetes and glycated haemoglobin) and seeing a doctor before starting self-treatment. Choosing an oral drug for people taking many long-term medicines requires checking interactions and liver function.

Similar caution applies to pregnancy and breastfeeding, liver disease, heart failure and the use of drugs that interact with cytochrome P450. We do not give detailed dosing advice here, because it would require reading the product information of a specific preparation; we only signal that in these groups the decision about oral treatment is made by the doctor, and topical treatment and observation can be a safer starting option.

Data on recurrence are weaker than on cure itself. In the 2017 Cochrane review, low-quality evidence suggested that terbinafine and azoles may reduce the proportion of recurrences compared with placebo (for example RR 0.05; 95% CI 0.01-0.38 for terbinafine in one study of 35 people), and no difference was shown between them. The numbers come from single, small trials, so they should not be used to calculate your own chances.

In nail fungus the order matters: diagnosis first, then treatment, and experiments only at the end. The evidence best supports oral terbinafine and itraconazole, with topical preparations as an option for mild changes or when tablets are not suitable. Laser and home remedies are best treated as a possible extra, not the basis of therapy, and not paid for as if cure were guaranteed. The greatest enemy of success is stopping treatment too early and ignoring the source of re-infection, namely the skin of the feet and the footwear.