Cost considerations in determining the affordability of adjuvant trastuzumab in breast cancer.
A topical issue is the affordability and value of adjuvant trastuzumab for 1 year after surgery for localised human epidermal growth factor receptor 2 (HER2)-positive breast cancer, particularly in low- and middle-income countries. Affordability and value differ in patient groups with different baseline prognoses. This is illustrated below using the hazard ratio (HR) of survival rates obtained from a Cochrane review  and personal communication with Roche.
The HR is the ratio of the relative survival of two patient groups, with and without the test therapy. This ratio will not vary much over time. Patients can be divided into prognostic groups with different outcomes, depending on tumour stage and nodal status. For an assessment of affordability, a specific time point needs to be determined for the calculation of baseline survival. Expected baseline survival rates (BLSRs) for periods up to 15 years can be determined with breast cancer outcome calculators, which are available on the internet (e.g. https://www.adjuvantonline.com/ and http://www. lifemath.net/cancer/breastcancer/therapy).
Number needed to treat (NNT) to benefit one patient at any selected endpoint
For clinical purposes, this may be approximated from the BLSR and the HR. The HR for overall survival in the calculation below was 0.69 (95% confidence interval (CI) 0.58-0.81).
* The number of patients who will benefit when 100 patients are treated = (100 - BLSR) x (1 - HR).
* Therefore, for a BLSR survival rate of 90%, the number who will benefit = 10 x 0.31 = 3. The survival rate therefore increases from 90% to 93%.
* The NNT to benefit one patient = 100/3 = 33.
More accurate numbers can be determined using more sophisticated formulae. 
The relative risk (RR) for increased toxicity for congestive heart failure = 5.11 (90% CI 3.00 - 8.72), and for decrease in left ventricular ejection fraction = 1.83 (90% CI 1.36 - 2.47).
Toxicity risk occurs in all treated patients, which means that the higher the BLSR, the higher is the proportion of patients who are free of disease who are exposed to potential toxicity. The toxicity also has cost implications, as all patients need to have their left ventricular function evaluated objectively with an echocardiogram or multigated acquisition (MUGA) scan every 3 months while on treatment.
The drug costs for 1 year on adjuvant trastuzumab
This analysis will restrict itself to a review of immediate drug costs alone, which are the largest cost factor. It excludes costs related to the facility fee, professional fees or associated investigations. A full economic analysis will include these costs as well as downstream costs for possible toxicity and savings from benefit.
The drug cost of 17 cycles of trastuzumab, allowing for vial sharing, for 1 year is ZAR413 000.
There are proponents for the use of adjuvant trastuzumab in all scenarios. The NNT to benefit one patient at a selected endpoint and the related drug cost to benefit one patient are shown in Table 1. The table indicates that the drug cost to benefit one patient ranges from ZAR13 752 900 (BLSR 90%) to ZAR4 006 100 (BLSR 60%).
It is a principle of clinical ethics that the clinician should provide the best treatment possible with available resources, provided there is evidence of benefit and the clinician is prepared to undertake the treatment.
In the Cochrane analysis,  the comment is made that the data indicate that the same efficacy is achieved with <6 months of treatment compared with 12 months of treatment and that there is less cardiac toxicity, but that this is not statistically significant with the numbers in the relevant studies. These studies are underpowered and there is relatively little incentive to undertake them. Although this is a lower level of evidence compared with that found with higher numbers, it is not an absence of evidence. Where funders are not able to provide 12 months of trastuzumab, as happens at present in state institutions and the lower-level medical schemes, <6 months of trastuzumab should be considered as a treatment option to benefit patients.
Clinicians rightly wish to extend the benefits from advances in cancer treatment to all suitable patients. This requires an appreciation of the costs involved and an ongoing engagement of state and healthcare institutions and the pharmaceutical industry.
[1.] Abratt RP. Modelling cost-effective therapies. S Afr Med J 2015;105(11):884. DOI:10.7196/SAMJ.2015.v105i11.9880
[2.] Moja L, Tagliabue L, Balduzzi S, et al Trastuzumab containing regimens for early breast cancer. Cochrane Database Syst Rev 2012, Issue 4. Art. No.: CD006243. DOI:10.1002/14651858.CD006243
[3.] Altman DG, Andersen PK. Calculating the number needed to treat for trials where the outcome is time to an event. BMJ 1999;319(7223):1492-1495. DOI:http://dx.doi.org/10.1136/bmj.319.7223.1492
Accepted 14 June 2016.
R P Abratt, FC Rad Onc (SA), MMed (RadT)
Head of Clinical Governance, Independent Clinical Oncology Network, South Africa; and Emeritus Professor of Radiation Oncology, Faculty of Health Sciences, University of Cape Town, South Africa
Corresponding author: R P Abratt (firstname.lastname@example.org)
Table 1. NNT to benefit one patient and the related drug costs with adjuvant trastuzumab in patients with breast cancer  Survival NNT to Drug cost (ZAR) Patient baseline rate with benefit one to benefit one survival rate, % trastuzumab, % patient patient 90 93 33.4 13 794 200 80 86 17.5 7 227 500 70 78 12.2 5 038 600 60 70 9.7 4 006 100
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||CLINICAL ALERT|
|Publication:||South African Medical Journal|
|Date:||Oct 1, 2016|
|Previous Article:||Button batteries in the oesophagus: a surgical emergency.|
|Next Article:||Sudden and unexpected childhood deaths investigated at the Pretoria Medico-Legal Laboratory, South Africa, 2007-2011.|