-by: Peter J Smith

In the late 1940s, childhood leukaemia was still an incurable and fatal disease. Duration of survival had been improved marginally by the use of blood transfusion and the early antibiotics, with up to 10% of children entering partial remission, ^{but survival was rarely for more than a few weeks or months. At that time most doctors felt that attempts to prolong the lives of such children simply added to their suffering and that the compassionate course was to relieve pain and offer simple palliative treatment.}

In 1947, Sidney Farber, pathologist and Chairman of the Division of Laboratories and Research at the Children's Medical Center, Boston, noted that children with acute leukaemia who were given very large doses of folic acid in an attempt to correct their anaemia developed an "acceleration phenomenon" in the progression of their disease. From this he deduced that folic acid antagonists might have an antileukaemic effect. He acquired a series of folic acid antagonists that were made available to him by Dr Y Subbarow and colleagues (Lederle Laboratories and Calco Chemical Division, American Cyanamid Co) and administered these to children with leukaemia. This work culminated, in 1948, with the publication of his observations on five of 16 children with leukaemia who were treated with aminopterin.^{2These children entered temporary remission, one lasting 43 days without further chemotherapy. However, Farber cautioned that the drug was very toxic and that, while the results were promising, they did not justify talk of a "cure" for childhood leukaemia.}

This is a good example of an important clinical outcome emanating from a strong clinical/basic research interface, but the rapidity of transfer of the chemical compounds from the laboratory to the clinic (without, for example, intervening animal studies) would of course be impossible today given the requirements of government regulatory agencies and institutional ethics committees.

John Colebatch, a Melbourne paediatrician, was excited by Farber's results, although his excitement was tempered by the knowledge that the Farber study reported selected patients with no controls; that some improvements in survival were being seen with supportive therapy alone; and that there was very strong feeling in the medical community against attempting to treat children with leukaemia.

It was also about this time that the Medical Research Council Group published the results of its controlled trial on the use of streptomycin in the treatment of pulmonary tuberculosis, ^{thus firmly establishing the concept of using controlled clinical trials to investigate therapeutic questions.}

In order to test Farber's observation on folic acid antagonists, Colebatch proposed to use the relatively new methodology of the controlled clinical trial, with children allocated either to receive or not receive aminopterin, based on the wishes of their parents. All patients would receive standard supportive treatment. Colebatch told the parents he did not know if aminopterin treatment would be effective and did not know how toxic the drug was.

Over a two-year period he enrolled 26 patients in the study -- the parents of 18 children gave permission for treatment, the parents of the other eight declining the treatment. The results of this clinical trial were published in the Journal on December 16, 1950. ^{The trial demonstrated that, despite its toxicity, high doses of aminopterin were effective in inducing remission in children with leukaemia. Thirteen out of 18 treated patients, but only two out of eight control patients, survived for longer than three months. Colebatch confirmed Dameshek's observation  that aminopterin was more effective against lymphoblastic leukaemia than non-lymphoblastic leukaemia. He also observed that the control patients had done much better than he would have expected and commented, "It is . . . obvious that no reliable assessment of a new therapeutic agent for leukaemia can be attempted unless appropriate arrangements are made for controlling the investigations", thus further articulating the case for concurrent controls in a clinical trial.}

The outcomes for patients in Colebatch's trial are, of course, quite modest by modern standards, and unfortunately all of the children treated with aminopterin ultimately died -- some from toxicity associated with the drug and others following subsequent relapses. While cautioning that antileukaemic treatment of this type should only be undertaken in institutions with appropriately trained staff and facilities, he also observed that ". . . there are glimmers of other and perhaps better therapeutic agents on the horizon". Although the controversy about whether to treat children with leukaemia continued to rage for many years, it was this view to the future that guided Colebatch and the many others who have worked in this field. Colebatch went on to organise the first multicentric randomised controlled clinical trial of cancer chemotherapy in Australia -- again studying childhood leukaemia.^{The randomised controlled trial format became the standard method of assessing new treatments and measuring improvements in cancer treatment.}

By the end of the 1960s, median survival for children with acute lymphoblastic leukaemia was about two years, and it was possible to begin to conceive of treatment regimens that might offer cure to children with this disease. Today, this dream has largely been realised, with expectations of a 75% event-free survival for children presenting with acute lymphoblastic leukaemia. ^{The national or international collaborative randomised clinical trial has become the "standard of care" in both paediatric and adult medical oncology.}

Peter J Smith
Stevenson Professor, and Head, Department of Paediatrics
University of Melbourne, The Royal Children's Hospital
Parkville, VIC
smithpATcryptic.rch.unimelb.edu.au