Let's talk about cytotoxic chemotherapy dosing: unravelling adjustments and off‐protocol prescribing

Individualised dosing is important in cancer treatment in real‐world settings and may require departure from trial‐based protocols

Adequate and appropriate dosing of cytotoxic chemotherapy is an important issue, highlighted recently by public concerns about underdosing.1,2,3 Most cytotoxic agents have a steep dose–response curve and narrow therapeutic index. Underdosing may have significant implications for patient outcomes, while overdosing can result in life‐threatening adverse effects.1 Cancer chemotherapy dosing differs from drug prescribing in other areas of medicine. Phase 1 trials in patients with cancer identify safety and dose‐limiting toxicities, and determine the maximum tolerated dose and recommended phase 2 dose. Phase 2 trials focus on efficacy assessment within specific tumour types.4,5 Phase 3 trials, considered the key criteria for approval by drug regulatory agencies, validate the efficacy of the drug compared with a standard of care. A challenge has emerged with new molecularly targeted agents and immuno‐oncology drugs frequently having different toxicity profiles to chemotherapy, which may not be dose or time dependent. Accordingly, the traditional phase 1 and 2 model may not be as adaptable, resulting in evolution of trial design, such as phase 1 trials combining safety and efficacy expansion cohorts.6

As individuals have variable capacity to metabolise and eliminate drugs, owing to factors such as genetic polymorphisms, there is a substantial degree of inter‐patient variation in drug exposure.7 Chemotherapy dosing is mainly calculated using body surface area (BSA), while area under the curve is used for carboplatin. Flat dosing is used for chemotherapy agents such as bleomycin, targeted agents including erlotinib, imatinib and sunitinib, and immunotherapy checkpoint inhibitors (eg, pembrolizumab). BSA dosing minimises inter‐individual variation in drug exposure, calculated using a formula relating to weight and height. The first BSA formula was based on a nine‐patient cohort.8 Despite lack of rigorous validation, BSA‐based chemotherapy dosing remains the de facto standard for most cytotoxic agents and some therapeutic monoclonal antibodies. However, BSA‐based dosing may be associated with inter‐individual variation exceeding 30%, particularly in obese and morbidly obese patients.7

Chemotherapy dosing in obese patients can be heterogeneous. The narrow therapeutic index of most cytotoxic agents raises concern that obese patients will experience excessive toxicity when doses are based on actual body weight. In a questionnaire sent to 315 Australian medical oncologists, more than two‐thirds of respondents capped doses at a BSA of 2.0 m² or used ideal body weight. Only 6% used actual body weight for obese patients.9 There is little evidence to support the claim that dosing obese patients based on actual body weight increases toxicity.10,11 Rather, data from retrospective studies show that chemotherapy underdosing in obese patients is associated with inferior outcomes.11 The American Society of Clinical Oncology published clinical practice guidelines in 2012 recommending the use of actual body weight for BSA calculation in obese adult patients with cancer, especially when the treatment intent is cure.11

Maintaining protocol‐recommended doses and dose intensity is imperative in the curative setting where reductions in dose delivery may compromise survival benefits.12,13,14,15,16,17 Dose intensity is defined as the total amount of drug delivered over time, expressed as mg/m²/week. Relative dose intensity (RDI) is the ratio of delivered to planned dose intensity, optimally aiming to achieve an RDI of 100%. A retrospective analysis of adjuvant chemotherapy in breast cancer demonstrated a clear dose–response effect where patients receiving less than 85% of the total planned dose had worse survival outcomes compared with those receiving 85% or greater of the planned dose.12 With long term follow‐up, the former group did not appear to derive any benefit from adjuvant chemotherapy.16 A meta‐analysis of randomised trials demonstrated improved survival outcomes in patients with breast cancer when adjuvant chemotherapy was delivered weekly or fortnightly in a dose‐dense schedule compared with previously recommended three‐weekly regimens.14 For germ cell testis cancers, delivery of higher total doses of etoposide and bleomycin in cisplatin‐based combination regimens improved survival outcomes.15 Outcomes were improved when RDI was maintained when treating aggressive lymphomas.17 Therefore, chemotherapy dose and RDI should be maintained in curative and adjuvant treatment settings, achievable with the addition of granulocyte colony‐stimulating factor (GCSF) support, approved in Australia for specific curative and adjuvant indications including aggressive lymphomas, paediatric malignancies, germ cell testis, and breast cancers but not colorectal or lung cancer.

Patients recruited to clinical trials need to meet strict eligibility criteria, often excluding patients who are elderly and frail, with multiple medical comorbidities, or with suboptimal performance status. It is increasingly difficult to translate chemotherapy dosing from trials to real‐world patients. In a nationwide study of United States community practices of adjuvant chemotherapy prescribing for patients with breast cancer, less than half received an RDI of at least 85%.18 Similar prescribing across six different types of cancers was described in another US community oncology practice review.19 In a trial comparing cabazitaxel with mitoxantrone for metastatic castration‐resistant prostate cancer, febrile neutropenia occurred at a higher rate in cabazitaxel‐treated patients.20 This study allowed GCSF prophylaxis for treatment‐induced neutropenia, which is not funded by the Australian Pharmaceutical Benefits Scheme for this indication. These examples highlight that in real‐world clinical practice, the delivery of full doses or optimal RDI may be challenging for reasons including fear of toxicity in more frail and aged patients and restricted access to GCSF. Patients with organ dysfunction are frequently excluded from participating in clinical trials, on the basis of strict eligibility criteria. Uncertainty remains regarding the risk of increased toxicity from altered pharmacokinetics or potential exacerbation of organ dysfunction in this population.

Regarding delivered RDI and total dose in the palliative treatment setting, key to any discussion is the understanding that with higher doses (and RDI) there is the risk of greater toxicity and compromised quality of life balanced against survival benefit. While palliative chemotherapy may not always confer significant survival advantage, the palliation–toxicity trade‐off should favour quality of life. In metastatic breast cancer, patients treated with standard dose chemotherapy had improved overall survival and quality of life, albeit with higher toxicity, compared with those receiving a reduced dose regimen.21 Other evidence has indicated that in advanced breast and ovarian cancer, reduced RDI has been associated with worse survival,22,23 although these publications did not report on toxicity or quality of life outcomes. When discussing palliative chemotherapy with patients with good performance status and minimal comorbidities, the intention should be to commence treatment at standard doses, aiming to optimise symptom control and survival benefit, with the understanding that dose modification may be indicated if there is significant toxicity or compromised quality of life. Conducting prospective randomised clinical trials comparing standard dose versus reduced dose regimens in the palliative setting will be difficult owing to patient heterogeneity, and patient expectations and consent.

Stakeholders in patient and medical communities may hold different definitions and perceptions regarding off‐protocol prescribing. Off‐protocol prescribing relates to use of a drug that is not consistent with its original intended use or indication. The Medical Oncology Group of Australia states that off‐protocol prescribing should not be confused with individualised dose adjustment, where the dose is adjusted for an individual patient based on performance status, comorbidities and the burden of the cancer.24 The geriatric oncology community divides patients into “fit”, “vulnerable” and “frail” and has published extensively regarding the importance of individualised dose adjustments for the vulnerable subset, considered at higher risk of treatment‐related adverse outcomes.25 Individualised dose adjustment is an intrinsic part of good practice, learnt through years of patient care, experience and continuing education, while off‐protocol prescribing is inconsistent with good practice when it is not evidence‐based.24

A National Quality Use of Medicines Indicator for Australian Hospitals is the “percentage of patients receiving cytotoxic chemotherapy whose treatment is guided by a hospital approved chemotherapy treatment protocol”, which aims to standardise care and reduce non‐evidence‐based variation.26 It is crucial to understand that treatment protocols and guideline‐based dosing should not be upheld as absolutes, as these do not factor individualised dose adjustment based on complex patient factors and may not reflect new trial data suggesting better patient outcomes. Above all, informed consent and good documentation lie at the crux of the issue. If a cancer treatment is not yet standard practice but there is appropriate evidence that it may offer superior outcomes, open discussion with the patient is important to better inform and guide treatment decisions. Documenting the exchange in patient records and in written communication with the treating teams should be a matter of good practice. Protocol and practice review should be implemented at all levels, from hospital and local health districts through to national specialty groups such as the Medical Oncology Group of Australia, and the specialty tumour stream groups. Lastly, governance structures should recognise that a one‐size‐fits‐all approach may not be appropriate for all patients.

Cancer treatment is complex and dynamic. Protocols and guidelines are often based on clinical trials on patients with minimal comorbidities. Real‐world patients often have multiple comorbidities, which must be considered when prescribing chemotherapy. Guidelines on chemotherapy protocols and dosing are guides and not absolutes. Rather, dosing based on treatment intent and patient factors, open communication, documentation and peer review are key to good practice. It is an issue that we should discuss, not only with patients but among ourselves, to ensure that peer review and quality assurance processes are in place for appropriate chemotherapy dosing and treatment.