All initially responded to therapy and were sent home, but were readmitted around a month later when the infection rebounded.
Samples of the parasite that causes malaria were analysed at the Malaria Reference Laboratory at the London School of Hygiene and Tropical Medicine.
Dr Colin Sutherland told the BBC News website: "It's remarkable there's been four apparent failures of treatment, there's not been any other published account [in the UK]."
All of the patients were eventually treated using other therapies.
Dr Sutherland added: "It does feel like something is changing, but we're not yet in a crisis.
"It is an early sign and we need to take it quite seriously as it may be snowballing into something with greater impact."
Two of the cases were associated with travel to Uganda, one with Angola and one with Liberia - suggesting drug-resistant malaria could be emerging over wide regions of the continent.
Dr Sutherland added: "There has been anecdotal evidence in Africa of treatment failure on a scale that is clinically challenging.
"We need to go in and look carefully at drug efficacy."
The malaria parasites all seemed to be evolving different mechanisms rather than there being one new type of resistant malaria parasite spreading through the continent.
The type of resistance is also clearly distinct from the form developing in South East Asia that has been causing huge international concern.
Dr Sutherland says doctors in the UK need to be aware the drugs might not work and argued current treatment guidelines may need to be reviewed.
Professor David Lalloo, Dean of Clinical Sciences and International Public Health at Liverpool School of Tropical Medicine, said more studies are needed.
"This is an interesting and well conducted study and again emphasises the incredible ability of the malaria parasite to rapidly evolve to become resistant to antimalarial treatment," he said.
"It is too early to fully evaluate the significance of these findings but the paper highlights the need to be constantly vigilant when treating patients with malaria and larger studies are certainly needed to explore this issue further."
Follow James on Twitter.