FOR cancer treatment to avoid over-treating or undertreating a patient, it must be based on sound scientific evidence and clinical judgement by the treating team.

The issue of appropriate and evidence-based chemotherapy dosing remains a very important, controversial and often-discussed part of clinical oncology.

The popular belief that “more chemotherapy must be better” has driven cancer treatment and research for over 40 years. However, it remains unsupported by the evidence. Despite many tens of thousands of clinical trials, there has never been level 1 evidence for a dose–palliation or dose–cure relationship for chemotherapy in solid cancers, even in very chemotherapy-sensitive and very curable cancers, such as early breast cancer.

All the evidence points consistently to a dose threshold effect for survival and palliation, so that a dose of chemotherapy above a certain – sometimes quite low level – adds no benefit, but increases cost and toxicity.

The recent chemotherapy dosing incident at St Vincent’s Hospital in Sydney has again highlighted concerns around choosing between protocols that use the minimum effective dose of chemotherapy versus the maximum tolerated dose. This debate will influence future oncology practice in Australia and elsewhere, where no oncologist or oncology unit wants to use protocols that they fear may deliver suboptimal treatment.

The current investigation by the NSW Health Department may mean that most oncology units and their multidisciplinary meetings will invariably use protocols and doses that err on the side of too much, rather than too little, just to be sure. This could have far-reaching consequences for ill and desperate patients with advanced cancer.

It may make some oncologists more concerned about the possible medico-legal implications of ever reducing protocol treatment doses.

This may cause a greater patient treatment toxicity and more recovery time in hospital with its larger costs and less quality time at home with families, particularly, with an ageing population and the associated and rising incidence of cancer in this group.

The confusion often arises because all early phase 1 and 2 trials with cancer drugs are designed by the pharmaceutical company to establish the maximum tolerated doses and anticancer activity, or tumour response rates, in very highly selected patients who have to pass rigorous tests of overall health. This means that they are usually very different to the often sick patients who come to our clinics.

These response rates are assessed as the proportion of patients who achieve a 50% or greater reduction in the two-dimensional diameter of tumours, which is always a result gratefully accepted by patients and their care team.

However, while there is a dose-response relationship for many chemotherapy drugs in cancer, shrinking cancers a bit more with ever higher and more toxic drug doses rarely, if ever, has meaningful benefits for patients.

This is because even a 90% shrinkage in the size of a cancer is only a one log cell kill and will only reduce the cancer cell population from 10 000 000 000 to 1 000 000 000 cells, for example, when clearly a 10- log cell kill is required for cure.

Therefore, the belief by some that it would be expected that on a population basis, a failure to adhere to “the protocol chemotherapy dose” is likely to result in higher rates of local recurrence and higher overall mortality may not necessarily be true. Its veracity depends on the strength of the evidence underpinning the protocol.

An example of this is the much discussed treatment used at St Vincent’s Hospital, Sydney, of combined carboplatin-based chemotherapy given as a concurrent radio–sensitiser in locally advanced head and neck cancer, which has certainly been established as a reasonable standard of care (here, here, here, and here).

The controversial part was the non-protocol low flat dose of carboplatin employed, and whether patients were undertreated.

Surprisingly, despite many years of research, we still don’t know the best dose of carboplatin to use in any patient. One study evaluated various doses of carboplatin in head and neck cancer and found that the efficacy for the lowest dose compared favourably to the highest one. The minimum effective dose has still not been established for carboplatin and may be much lower than is conventionally used.

In addition, chemotherapy given with radiation is probably not acting primarily as a cytotoxic and is very effective at low dose because it “sensitises” the cells to radiation by a variety of mechanisms (here and here). The radio-sensitising chemotherapy is the support treatment and must not increase the treatment toxicity to the level that it delays the primary radiation treatment. Recent data reveal that it is the completion of the scheduled radiotherapy within the defined protocol time and not the completion of the protocol chemotherapy that determines the outcome.

In view of these considerations, it will be important to analyse if the long term outcomes of the patient group treated at St Vincent’s Hospital differ significantly from an equivalent risk group based on the best available literature of evidence-based treatment during the same period. It will be impossible to do this assessment for individual patients.

As a high-profile lead editorial in the Journal of Clinical Oncology stated in 2000: “A ‘new’ paradigm for dosing chemotherapy … [uses] low-dose continuous chemotherapy … Public or underwriter pressure, allure of high-dose therapy, and technical capabilities for the sake of technology (eg, supportive care) should not drive the treatment algorithms unless they are based on sound scientific data. More is not always better, and this is high time for low-dose.”

The St Vincent’s Hospital episode is a good opportunity for all of us involved in caring for patients with cancer to re-examine the evidence underpinning current practice.

Clinical Associate Professor Ian Haines is a medical oncologist with the Alfred Medical Research and Education Precinct’s Department of Medicine at Monash University and Cabrini Health, in Melbourne.


Latest news from doctorportal:



A dose of chemotherapy above a certain level adds no benefit
  • Strongly agree (56%, 19 Votes)
  • Agree (35%, 12 Votes)
  • Neutral (9%, 3 Votes)
  • Disagree (0%, 0 Votes)
  • Strongly disagree (0%, 0 Votes)

Total Voters: 34

Loading ... Loading ...

3 thoughts on “Chemotherapy: more is often not better

  1. David Henderson says:

    I have always thought that the complaint may have lacked substance. I think this thoughtful review is very important, but it may not counter the weight of the outrage. The use of trial protocols has been extremely useful in advancing the treatment of cancer, but in practice there are many situations where patients do not fit into evidence based or protocolised regimens. Although protocols are useful, their use still requires clinical judgement in diagnosis, assessment and treatment, just as a road map is only a guide to a journey and does not give you all the information you need to travel.

  2. Anonymous says:

    A refreshing return to the tradition of seeking to learn from every circumstance in clinical medicine.
    Much of this article could also be applied to the dosing of cytarabine for a small group of patients with acute myeloid leukaemia in South Australia. The clinicians involved were named in Parliament and referred to AHPRA by the Premier and Minister for Health, there have been several inquiries, and an ongoing media witch hunt.
    Yet no one seems to have looked at the matter from a clinical viewpoint – the majority of the patients who were “undertreated” are still alive, and hence presumably cured. Although patient numbers are small, this is remarkable survival for AML.

  3. Anonymous says:

    this article represents a triumph for common sense. the current NSW chemotherapy debate has all the hallmarks of a media-driven witch hunt not related to patient welfare considerations

Leave a Reply

Your email address will not be published. Required fields are marked *