Stem cells are immature cells which can become different types of adult tissue.
It is thought that in this instance the stem cells may have generated new immune cells which do not target the pancreas, helping to safeguard what remaining insulin-producing cells the patient has left.
However, it is also possible that the treatment may have led to the growth of new insulin-producing beta cells in the pancreas.
A third possibility is that the treatment stimulated an as yet unknown mechanism which stopped existing beta cells being destroyed.
The results do throw up the possibility that stem cell therapy could have a dramatic impact.
But the study was small, did not monitor the patients for very long and did not compare them with similar patients who were given alternative therapy or remained untreated.
Dr Iain Frame, of the charity Diabetes UK, warned against raising people's hopes on the basis of a "very preliminary" study.
"It is well known that there is often a honeymoon period of relative remission after the onset of Type 1 diabetes that complicates the interpretation of results such as the ones shown in this study.
"All these issues need to be addressed through more research before there are any conclusive findings in this area."
Dr Richard Burt, of Northwestern University, Chicago, who worked on the study, said: "I do not use the word cure, or the word breakthrough, but this is a step forward.
"It is the first time in which a stem cell therapy has been used effectively in this disease."
Previous research has suggested that stem cell therapy can benefit patients with other auto-immune diseases, such as rheumatoid arthritis and Crohn's.
Of the 2.2 million people with diabetes in the UK, only around 300,000 have the type 1 disease.
However, for unknown reasons, the number of British children under the age of five developing type 1 diabetes has risen five-fold in the last 20 years.