Patterns of platinum drug use in an acute care setting: a retrospective study

Abstract

Purpose

Platinum drugs have been in use in cancer treatment for more than 40 years, but little is known about the pattern of their use. The aim of this study was to examine the patterns of platinum drug use, with a secondary aim to describe the occurrence of dose reductions.

Methods

A retrospective analysis was conducted of oncology pharmacy dispensing records from a single hospital in Australia. Data related to drug choice, regimen and dose reductions were included in this study if the patient had received their last round of chemotherapy between November 2014 and July 2015.

Results

Of the 156 patients included in the study, 46% were dispensed a platinum drug during their treatment. The most commonly dispensed drugs were cisplatin (40%), carboplatin (40%) and oxaliplatin (15%), while some patients (5%) received more than one platinum drug. Dose reductions were more common in patients who were treated with a platinum drug (73%) compared with patients treated with non-platinum drugs (55%). The most common reason for a dose reduction was cytopenia.

Conclusions

The findings suggest that platinum drugs remain one of the most commonly dispensed drugs to treat cancer patients and most patients receive a dose reduction during treatment.

Introduction

Cancer is the second leading cause of death and economic disease burden throughout the world, with the World Health Organization estimating that the one in six deaths is due to the disease (Stewart and Wild 2014). While there is a range of treatment options for cancer, including surgery, radiotherapy, and immunotherapy, more than 65% of hospitalised patients receive some form of chemotherapy as part of their anticancer therapy (Australian Institute of Health and Welfare 2017).

Platinum drugs (cisplatin, carboplatin and oxaliplatin) are highly effective chemotherapy agents that have been in use for over 40 years (Muggia et al. 2015; Wheate et al. 2010). These medicines are indicated for more than 25 different cancer types, including: colorectal, gastrointestinal, gynaecological, head and neck, multiple myeloma, lung and urogenital tumours (Greystoke et al. 2017; Muggia et al. 2015), with high testicular cancer cure rates which exceed 90% (Bagheri-Sereshki et al. 2016; Bosl et al. 2001; Einhorn and Foster 2006; Husain et al. 2002; Kopp et al. 2006).

The use of platinum drugs is, however, limited by their considerable, dose-limiting side effects, such as: nephrotoxicity, neurotoxicity, ototoxicity, emetogenicity and a range of cytopenias (Apps et al. 2015; Wang and Lippard 2005; Yanagimoto et al. 2016). Additionally, cancers may either develop, or be inherently resistant to, platinum drugs (Giaccone 2000; McEvoy et al. 2015; Rabik and Dolan 2007; Stewart 2007). The three platinum drugs with worldwide marketing approval: cisplatin, carboplatin and oxaliplatin, share similar pharmacological properties but differ in their side effect profiles and their anticancer activity (Rixe et al. 1996; Wheate et al. 2010).

Platinum drugs are listed in approximately 20–50% of the chemotherapy protocols described in various national chemotherapy regimen databases, such as the EviQ guidelines in Australia, the NICE guidelines in the United Kingdom, and in scholarly pharmaceutical texts such as Martindale (Sweetman 2017), and are often co-administered with other chemotherapy agents including antimetabolites, topoisomerase inhibitors, nitrogen mustards, and vinca alkaloids (Cancer Institute 2011).

Despite extensive research into the various mechanisms of action of these types of drugs (Gibson 2016) and the development of new agents in this class, knowledge on the patterns of clinical use of platinum drugs is sparse. Studies have reported that between 50–70% of all patients receive a platinum drug as part of their cancer treatment, but there are no robust data to support these claims (Dyson and Sava 2006; Hannon 2007; Harper et al. 2010; Johnstone et al. 2014; Zalba and Garrido 2013). In fact, while there are limited studies into the chemotherapy prescribing and dispensing for specific cancer types, (Langer et al. 2005; Sacher et al. 2015) to the best of our knowledge no study to date specifically examined the patterns of platinum drug use. This is important because there is real demand to better understand how these medicines are used to drive future healthcare forecasting. It is also vital to understand the potential harm to patients given the severity of the side effects patients may experience.

Therefore, this study aimed to investigate the patterns of platinum drug use, including their common cancer indications, treatment protocols, doses and reported adverse effects, in the Australian acute care setting. The secondary objective was to examine the level of, and reasons for, platinum drug–dose reductions.

Methods

Study design and population

This was a retrospective analysis of oncology pharmacy dispensing records of patients treated at Concord Repatriation General Hospital (CRGH) (Sydney, Australia). Oncology pharmacy dispensing records were filed under the date patients received their last cycle of chemotherapy. Potentially eligible participants were assessed by reviewing their dispensing history using iPharmacy dispensing software to ensure they did not receive further chemotherapy treatment after this 9-month period.

Patients treated for any type of cancer (excluding haematological cancers), with any stage of cancer, at the CRGH Cancer Centre were included. There were no age restrictions. Exclusion criteria included any patients who continued to undergo chemotherapy treatment beyond the specified time (November 2014–July 2015) or any patient who was diagnosed with a haematological cancer, as these records could not be accessed.

Data collection

Data were extracted from oncology pharmacy records by a single, trained pharmacy researcher and stored securely in a RedCap database with the server located at the University of Sydney. A detailed data collection sheet was used that included age, height, weight, body surface area, cancer diagnosis, chemotherapy treatment protocols (both platinum- and non-platinum-based chemotherapy), the specific chemotherapeutic drugs and doses used, dose reductions, whether the drugs were used as first or second-line treatment, any changes to prescribed drugs during treatment, the name of the chemotherapy protocol used, and the time since last round of chemotherapy. Patient outcomes were not evaluated as part of this study.

For the purposes of this study a dose reduction is defined as either a reduction in the dose given to the patient with respect to their first cycle of chemotherapy or where a patient was prescribed a dose lower than that recommended in the protocol before their first cycle of chemotherapy, because of their underlying health.

A cycle of chemotherapy was defined as the administration of a treatment repeated on a regular basis with a specified rest interval. A round of chemotherapy was defined as the commencement and cessation of a chemotherapy protocol that comprised one or more chemotherapy cycles.

Statistical methods

Statistical analysis was undertaken using IBM’s Statistical Package for the Social Sciences software, version 24 (New York, USA). Descriptive statistics were used to describe patient characteristics. The normality of each study variable was assessed using the Kolmogorov–Smirnov statistic. For data that were normally distributed, results are presented as mean $\pm$ standard deviation, and for data that were not normally distributed, the median and interquartile range (IQR) are reported. Independent samples t-tests were carried out for normally distributed data with continuous dependent variables, and Mann–Whitney U tests were performed on non-normally distributed data to determine if any characteristics were significantly different between those patients that received a platinum drug and those that did not. P values lower than 0.05 were considered to be statistically significant. Dose reductions $\ge$ 10% were included in this analysis.

Results

Baseline patient characteristics

The data on 156 patients were included in this study. Of these, 71 patients (46%) were dispensed at least one platinum drug during their cancer treatment compared with 85 patients (54%) who were not treated with a regimen that included a platinum.

From Table 1, there were no significant differences between the groups, except with respect to the number of chemotherapy rounds administered and the types of cancers treated with each drug. Patients who were administered a platinum drug tended to undergo more rounds of chemotherapy (median 2, IQR 1–4) when compared with the non-platinum-treated patients (median 1, IQR 1–2).

Table 1.

Characteristics of study population

Characteristic	Patients treated with a platinum drug	Patients not treated with a platinum drug	P value
N (n = 156)	71 (46%)	85 (54%)	–
Sex (M/F) (n = 156)*	31/40	28/57	0.169
Mean age (years) (n = 152)**	61 $\pm$ 11	65 $\pm$ 13	0.056
Height (cm) (n = 132)**	165 $\pm$ 9	162 $\pm$ 9	0.061
Weight (kg) (n = 134)***	69.1 (50.8–87.3)	64.0 (42.0–86.0)	0.205
BMI (kg/m2) (n = 131)***	25.5 (19.5–31.5)	24.7 (19–30.4)	0.511
Total number of chemotherapy rounds received (n = 156)***	2 (IQR 1–4)	1 (IQR 1–2)	0.027

BMI body-mass index

*$\mathit{\chi }$²-test performed

**Independent samples t test performed

***Mann–Whitney U test performed

Among patients treated with a platinum drug, the most common cancers were gastrointestinal (32%), lung (27%) and gynaecological (16%) (Table 2). In this study cohort patients treated with a chemotherapy regimen that did not include a platinum drug were more likely to be diagnosed with either breast (48%) or colorectal (27%) cancer.

Table 2.

Distribution of cancer types in comparison to recommended Australian chemotherapy protocols that contain a platinum drug

Cancer type	Patients treated with a platinum drug (n = 71)	Patients not treated with a platinum drug (n = 85)
Breast	4 (5%)	41 (48%)
Colorectal	6 (9%)	23 (27%)
Gastrointestinal	23 (32%)	5 (6%)
Gynaecological	11 (16%)	3 (4%)
Lung	19 (27%)	2 (2%)
Urogenital	5 (7%)	5 (6%)
Not documented	3 (4%)	6 (7%)

Patterns of platinum drug use

Among patients prescribed a platinum drug (n = 71), 32 patients (45%) received at least one additional type of non-platinum drug during the course of their cancer treatment. For the remaining 39 patients (55%), their chemotherapy treatment consisted solely of a platinum drug.

Cisplatin and carboplatin were used for an identical number of patients, 28 of 71 patients each (40% each) while oxaliplatin was administered to only 11 patients in the platinum group (15%). A minority of patients received more than one platinum drug during their treatment (4 patients, 5%). Two patients initially commenced treatment with carboplatin which was later changed to cisplatin. One patient was changed from carboplatin to oxaliplatin, while another patient was treated with all three platinum drugs; initially with carboplatin, which was then changed to cisplatin, and later, oxaliplatin.

For 66 of the 71 patients in the platinum-treated group (93%), the platinum treatment was dispensed as first-line therapy. A small minority of patients received a platinum drug as second-line therapy (5 patients).

Gemcitabine was the most commonly co-administered chemotherapy drug and was dispensed on 33 occasions in combination with a platinum drug (Table 3). Epirubicin and capecitabine were co-dispensed on 17 and 16 occasions, respectively. A variety of other drugs were co-administered with platinum-based agents including antitumour antibiotics (bleomycin, doxorubicin, epirubicin), antimetabolites (5-fluorouracil, capecitabine, gemcitabine, pemetrexed), taxanes (paclitaxel, docetaxel), topoisomerase inhibitors (etoposide, irinotecan) and monoclonal antibodies (trastuzumab, panitumumab, bevacizumab).

Table 3.

Chemotherapy drugs co-prescribed with platinums

Co-prescribed non-platinum drug	Frequency prescribed (n = 109)
Antimetabolites
Gemcitabine	33
Capecitabine	17
Fluorouracil	12
Pemetrexed	3
Antitumour antibiotics
Bleomycin	2
Doxorubicin	4
Epirubicin	16
Monoclonal antibodies
Bevacizumab	1
Panitumumab	1
Trastuzumab	2
Taxanes
Docetaxel	2
Paclitaxel	8
Topoisomerase inhibitors
Etoposide	6
Irinotecan	2

The most common chemotherapy protocol prescribed differed depending on the cancer type. For people undergoing treatment for colorectal cancer, there was a preference for the use of the XELOX regimen (oxaliplatin/capecitabine) (5/8 patients) over the FOLFOX (fluorouracil/oxaliplatin) regimen (2/8 patients). One patient received XELOX in combination with the monoclonal antibody bevacizumab (1/8 patients).

Similarly, in people undergoing treatment for lung cancer, there was a clear preference for the carboplatin/gemcitabine regimen (14/23 patients) over various combinations of pemetrexed, etoposide or paclitaxel with either cisplatin or carboplatin (9/23 patients).The two most common chemotherapy regimens for people receiving treatment for gynaecological cancer both utilised carboplatin, in combination with either paclitaxel (7/18 patients) or doxorubicin (5/18 patients). For those patients with either breast or gastrointestinal cancers there was no chemotherapy regimen that was used more often than any other. For the gastrointestinal cancer group, both ECX (epirubicin/cisplatin/capecitabine) and the cisplatin/gemcitabine regimens were commonly prescribed, 11/32 patients (34%) and 10/32 patients (31%), respectively. For breast cancer treatment, two patients received a gemcitabine/carboplatin combination (2/4 patients) while two others received docetaxel/carboplatin/trastuzumab (2/4 patients).

Dose reductions

For the 156 patients included in the study, there were a total of 1592 doses of chemotherapy and 322 dose reduction events recorded in the clinical notes (Table 4). The highest number of dose reductions were reported for people dispensed carboplatin (33/322 patients, 10%), followed by cisplatin (29/322 patients, 9%) and finally, oxaliplatin (16/322 patients, 5%). One patient received dose reductions in two platinum drugs sequentially. Overall, 52 patients (73%, 52/71) who were treated using a protocol that included a platinum received a dose reduction for at least one of their platinum drugs.

Table 4.

Dose reductions in the platinum- and non-platinum-based patient groups

Dose reductions	Received a platinum drug				Did not receive a platinum drug
	Cisplatin (n = 30)	Carboplatin (n = 32)	Oxaliplatin (n = 13)	Non-platinum chemotherapy agents (used during or after platinum chemotherapy) (n = 71)	Non-platinum chemotherapy agents (n = 85)
1 dose reduction	15	17	6	18	20
2 dose reductions	4	3	5	11	15
3 dose reductions	0	2	0	10	6
4 dose reductions	0	1	0	6	4
$\ge$ 5 dose reductions	1	0	0	9	2
Cumulative number of dose reduction events	29	33	16	148	96
Individuals who received a dose reduction	67% (20/30)	72% (23/32)	85% (11/13)	76% (54/71)	55% (47/85)

Twenty-two patients were prescribed a reduction in their first cycle of chemotherapy (when compared to protocol-based empiric dosing guidelines) with a platinum drug (22/71 patients). These dose reductions ranged in size from 10 to 60% of the protocol recommended dose. The most frequent initial dose reduction was 25% (5/22 patients) followed by 15% (3/22 patients).

All six patients undergoing platinum-based treatment for colorectal cancer received a dose reduction. Dose reductions were also common for those treated with a platinum drug for their urogenital and breast cancers; 4/5 patients (80%) and 3/4 patients (75%), respectively.

Reasons for dose reductions

Reductions in platinum-based chemotherapy dose were prescribed for a variety of reasons; however, the reason for the decrease in dose was often not documented in the dispensing records for the patients receiving a platinum drug (26/52 patients, 50%). For the other patients where a reason was recorded in the dispensing report, neutropenia was recorded as the cause of a dose reduction on seven separate occasions (7/52 patients, 13%) followed by myelosuppression (4/52). Pancytopenia, was recorded as a reason for dose reductions for 5 of 52 patients (10%). There were only two recorded occasions where a dose reduction was given due to extreme gastrointestinal adverse effects including nausea/vomiting or diarrhoea. Other reasons for dose reductions included hypersensitivity reactions (3/52 patients), age and frailty (2/52 patients), increased creatinine levels (2/52 patients) and weight loss (1/52 patients).

Two patients underwent platinum desensitisation cycles for oxaliplatin and carboplatin where the drugs were given at 1/1000th of the recommended dose, then 1/100th of the dose, before it was increased to 1/10th of the recommended dose on day one of treatment. The full recommended dose was then administered on day two of the cycle. The patient who received a carboplatin desensitisation protocol later went on to receive cisplatin desensitisation treatment as well.

Discussion

Prevalence of platinum drug use

To our knowledge, this is the first study to investigate the use of platinum drugs across the different cancer types in the acute care setting. Our main findings were that platinum drugs were commonly prescribed (46%) among cancer patients undergoing chemotherapy, and dose reductions were common for patients who received one or more platinum drugs. Patients treated with a platinum drug were more likely to receive additional rounds of chemotherapy when compared to those not treated with a platinum.

The proportion of patients administered a platinum drug is reasonably high considering the known limitations of platinum drugs that include their severe side effects, the development of drug resistance, and the availability of modern, targeted, less toxic drugs. There may be several explanations for the high rates of platinum drug use including the very high costs of new chemotherapy drugs.

For example, with regards to cost, for advanced or metastatic non-small cell lung cancer (NSCLC), there are a total of 20 chemotherapy protocols listed in the Australian EviQ chemotherapy protocol database (http://www.eviq.org.au/medical-oncology). Of these protocols, nine contain a platinum drug, four involve small molecule inhibitors and one utilises a monoclonal antibody drug (nivolumab). Looking at these protocols comparatively, a regimen that contains a combination of carboplatin and docetaxel has a weekly cost in the EviQ database of just $57 AUD per cycle. In contrast, current pricing for 1 week’s supply of nivolumab, a monoclonal antibody is $2360 AUD. Similar cost differences are likely in other nations, reflecting the fact that older drugs (like platinum drugs) are off-patent, while the newer drugs are significantly higher cost.

A second consideration is the advice and recommendations made to clinicians when they consider prescribing the newer chemotherapy drugs. The evidence for gefitinib in a first line setting for epidermal growth factor receptor-positive NSCLC is significant for improving progression-free survival (median 10.8 vs 5.4 months in chemotherapy patients) and has demonstrated an increased objective response rate (73.7 vs 30.7%, P < 0.001) when compared to standard chemotherapy (Yang et al. 2010). As such, eviQ recommends the use of gefitinib, but only when a patient has failed previous treatment. Similarly, crizotinib, which is another small molecule inhibitor, is only indicated for patients who previously failed to respond to platinum-based chemotherapy. As the protocol guidelines do not advise clinicians to use drugs like gefitinib and crizotinib as first line treatment, then this is an additional factor that results in the high rate of platinum drug prescribing.

Finally, prescriber familiarity and confidence may be a factor in the high rates of platinum prescribing. In Australia, there are currently no requirements for clinicians to adhere to eviQ guidelines; clinicians be hesitant to prescribe novel chemotherapy agents, and instead opt for familiar chemotherapy regimens which are likely to contain a platinum drug. Generally, there is more established evidence available for the older chemotherapy regimens regarding their success rates and possible adverse events. This factor, however, may play a relatively minor role in chemotherapy decision-making, as various reports indicate fast adoption of new therapeutics in the treatment of breast, prostate and lung cancers (Vijayvergia et al. 2016).

The distribution of cancer types among the patients who received platinum drugs was not wholly unexpected as the trends displayed in Table 2 approximately correspond to the cancer types known to be sensitive to platinum drugs. In this cohort of patients, platinum-based drugs were more commonly prescribed to patients diagnosed with gastrointestinal (32%), gynaecological (16%) and lung (27%) cancers. Similarly, in the eviQ guidelines, gastrointestinal, gynaecological and lung cancers have the highest proportion of protocols that recommend a platinum drug. To truly determine whether these observed trends are significant and reflective of oncological practice, a larger sample size covering all types of cancer would be required.

Rates and reasons for dose reductions

Overall, dose reductions were common for patients who received one or more platinum drugs. Of the 71 patients who received a platinum as part of their treatment, 52 (73%) received a dose reduction for the platinum drug, and 54 patients (76%) received a dose reduction for one of their co-administered non-platinum chemotherapy drugs. Patients were more likely to be prescribed a dose reduction in a non-platinum chemotherapy agent if they were currently receiving, or had a history of receiving, a platinum drug when compared with those patients who never received a platinum (76 vs 55%, respectively). Unfortunately, the reasons behind the dose reductions were not clearly documented in some patients’ dispensing records which made it difficult to draw firm conclusions as to why patients treated with a platinum had such high rates of dose reductions in comparison to the patients not treated with a platinum.

Equally surprising were the number of dose reductions experienced by patients treated with carboplatin (23/32 patients, 72%) or oxaliplatin (11/13 patients, 85%). This is of particular interest as carboplatin was designed as a less toxic alternative to cisplatin (Kelland 2007; Wheate et al. 2010). Carboplatin does, however, lead to the greatest risk for myelosuppression, (Egorin et al. 1984; Gutierrez et al. 2016) which is consistent with the high rates of myelosuppression (21%, 11/52) observed in this cohort of patients.

Strengths and limitations of the study

This is the first study to investigate patterns of platinum drug use in Australia across cancer types. Our study provides important data to inform the judicious prescribing of platinum drugs in the acute care setting; however, there are important limitations to this study including the relatively small sample size (n = 156), which may limit the applicability of the study findings. The patients included in this study are not necessarily indicative of current cancer trends as our data did not include any patients with neurological, blood and bone marrow, sarcoma, head and neck or melanoma-related cancers. This factor limits the generalisability of this study. In saying this, platinum drugs are not normally prescribed to patients with neurological, sarcoma or melanoma-related cancers, as these are commonly treated using a combination of surgery and radiation as first-line therapy. Patients with head and neck cancers were not included in this study, as these cancer types are not treated at CRGH. In addition, this study did not evaluate patient outcomes based on the different chemotherapy drugs.

This study was a retrospective analysis, and as such, is limited by the availability of data. For example, for half the patients the clinical reasons for some dose reductions were not retrievable as they were not documented in the dispensing records. This data may be of particular clinical significance (e.g. patient renal function) and may have contributed significantly to the rate of dose reductions in both groups of patients.

This study’s findings are significant in the field of oncology pharmacy and platinum chemotherapy research as this is the first data available regarding platinum drug use in an acute care setting across cancer types. Despite being in use for over 40 years, platinum drugs continue to be used regularly to treat many types of human cancers. Their clinical use is, however, limited by the high prevalence of dose reductions. This study validates and endorses current research focused on reducing the toxicity of platinum chemotherapy agents.

Future research may include a more comprehensive examination into the reasons behind the dose reductions for platinum drugs. This could include whether patient co-morbidities and co-administered medications are correlated to dose reductions. Additionally, further research may focus on collecting data across multiple sites including different countries and healthcare settings, and with attention to regional and rural oncology care settings. Further studies could also include a larger and wider cancer population rather than focusing on cancer patients who received chemotherapy as their main form of treatment.

Conclusions

This study found that 46% of patients undergoing chemotherapy were administered a platinum drug at some stage of their cancer treatment. Cisplatin and carboplatin were administered an equal number of times (40%) with oxaliplatin being used less frequently (15%). Five percent of patients received more than one platinum drug. Dose reductions were common in patients receiving platinum drugs with 73% of this cohort experiencing a decrease in the dose of at least one of their platinum drugs. Patients were also more likely to receive a dose reduction in a non-platinum chemotherapy drug if they were currently taking, or had a history of receiving a platinum drug, in comparison to the group of patients that were never administered a platinum drug (76 vs 55%, respectively). The most common reason for a dose reduction was cytopenia (31%). Further studies should investigate the prevalence of platinum drug use across the wider cancer population and the factors that may contribute to dose reductions for patients treated with a platinum drug, including co-morbidities. Future investigation into platinum drug use in a clinical setting is justified in order to drive future healthcare forecasting and maintain patient safety.

Abbreviations

AUD

Australian dollar

BMI

Body-mass index

CRGH

Concord Repatriation General Hospital

EGFR

Epidermal growth factor receptor

IQR

Interquartile range

NSCLC

Non-small cell lung cancer

Compliance with ethical standards

Conflict of interest

Author A declares that he/she has no conflict of interest. Author B declares that he/she has no conflict of interest. Author C declares that he/she has no conflict of interest. Author D declares that he/she has no conflict of interest. Author E declares that he/she has no conflict of interest. Author F declares that he/she has no conflict of interest. Author G declares that he/she has no conflict of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by the Sydney Local Health District Concord Hospital Human Research Ethics Committee (HREC Reference Number: LNR/17/CRGH/196).