Prevalence of Potential Pharmacological Interactions in Patients Undergoing Systemic Chemotherapy in a Tertiary Hospital

Abstract

Introduction: Pharmacological interactions are frequently observed in patients with chronic diseases, and their occurrence is proportional to the amount of medication used daily. Patients undergoing chemotherapy treatment commonly have comorbidities, which favor a greater prevalence of polypharmacy, increasing the risk of drug interactions. Therefore, the aim of this study was to estimate the prevalence of drug interactions in patients undergoing intravenous chemotherapy treated at a hospital oncology service in southern Brazil. Methods: This was an observational study with a cross-sectional design that was carried out with the analysis of secondary data obtained through the review of medical records. The population assessed consisted of all cancer patients who received intravenous chemotherapy from October to December 2020. Results: Out of the 297 patients included in the study, 231 (77.8%) had at least 1 potential pharmacological interaction. In total, 1044 drug interactions were found that were classified according to severity, resulting in 18 (1.7%) contraindicated drug-drug interactions (DDI), 699 (67%) severe, 281 (26.9%) moderate, and 46 (4.4%) minor interactions. There was an association between polypharmacy and the prevalence of drug interactions. Conclusion: The results demonstrate that a large percentage of patients undergoing chemotherapy are susceptible to drug interactions. Thus, it is necessary that prescribers consider all drugs used by patients and, when possible, prescribe alternative drugs that have less potential for interaction in order to prevent drug interactions adverse effects and provide a better prognosis for patients.

Introduction

Pharmacological interactions are observed in 19% to 33% of patients seen in primary and secondary care services^1-3; their occurrence is proportional to the number of drugs used daily.^3-5 Patients with chronic diseases commonly have a higher incidence of drug-drug interactions (DDI),^6,7 especially those affected with more than 1 chronic disease.^8,9

Patients diagnosed with cancer often have comorbidities,^8,9 such as diabetes mellitus, dyslipidemia, and cardiovascular diseases which are the most common chronic diseases. ⁸ This fact contributes to the increased prevalence of polypharmacy among cancer patients,^10,11 favoring DDI occurrence both due to the increased number of drugs used to treat cancer and comorbidities and also to the support drugs prescribed, thus increasing the susceptibility of those patients to DDI.^12-14

A drug-drug interaction is considered when the concomitant use of 2 or more medications alters the effect of 1 or more medications used, modifying the patient’s response to therapy, which may cause toxicity or reduced effectiveness.^15,16 In patients undergoing chemotherapy, DDIs affect more than two-thirds of patients,^13,17 and its diagnosis is made difficult by the fact that the effects of these interactions can be mistakenly attributed to chemotherapy. ¹⁸

The high prevalence of DDI in cancer patients is of concern due to its potential serious consequences for patients, ¹⁹ including cardiac arrhythmias due to increased QT interval,^20,21 which can lead to permanent injury ⁹ or even fatal events. ²² DDI is estimated to be the cause of up to 4% of deaths among hospitalized cancer patients^22,23; however, even when DDI does not result in fatal events, DDI classified as severe and moderate have a high prevalence in cancer patients, and can generate serious adverse reactions in those patients.^9,12

Studies show that the main DDIs found in cancer patients were due to the interaction between chemotherapeutic agents and antihypertensive and antithrombotic drugs,^9,24 in addition to the high prevalence of interaction between chemotherapy and anticonvulsant drugs^9,13 as well as diuretics.^17,22 These interactions can occur through pharmacodynamic and/or pharmacokinetic mechanisms.^13,25 Pharmacodynamic interactions result in an agonist or antagonist effect, while pharmacokinetics interactions affect drug’s absorption, distribution, metabolism and excretion.^12,25

Considering the large number of drugs used by patients undergoing intravenous chemotherapy and the increased risk of drug interactions in these patients, the aim of this study was to estimate the prevalence of potential pharmacological interactions in patients undergoing intravenous chemotherapy treated at an oncology service hospital in southern Brazil.

Methods

This was an observational study with a cross-sectional design reviewing the secondary data obtained from the assessment of electronic medical records. The population assessed consisted of all the cancer patients who received intravenous chemotherapy treatment between the period of October 1st and December 31st, 2020 in an oncology service in southern Brazil that is part of the National Network of Clinical Research, and is a reference high complexity oncology hospital for the Ministry of Health.

The oncology service in which the investigation was carried out provides health care to cancer patients through the Sistema Único de Saúde (SUS, Unified Health System), through private health plans and to private individuals, providing chemotherapy for all types of solid tumors.

Inclusion criteria covered patients receiving at least one intravenous chemotherapy drug and being at least 18 years old. Patients which pharmacological data had not been entered in their medical records were excluded.

Procedures

Data were obtained from a report that had been requested to the Information Technology Department of the aforementioned hospital, which contained all the records of care entered during the study period, in order to retrieve the medical records for review. Based on the visit number the electronic medical records were reviewed for sociodemographic variables, and for anamnesis, evolution and prescribed medications. The following social and clinical data were retrieved: age, gender, education, alcohol and smoking habits, medications being used, disease under treatment and other comorbidities, chemotherapy medications and other drugs used by the patient.

Case Definition

After obtaining data on the drugs used by patients, drug interactions were identified from the Medscape Drug Interaction Check ²⁶ and Micromedex^®27 databases. The first classifies interactions according to the mode of action, while the second classifies interactions according to clinical risk and mode of action. Interactions were classified as contraindicated when the concomitant use of drugs is contraindicated; severe, when the interaction could be fatal and/or could require medical intervention to prevent or minimize serious adverse effects; moderate, when the interaction could exacerbate the patient’s condition and/or would require a change in therapy and; minimal, when the interaction would be likely to have limited clinical effects. ²⁷

Regarding the mode of action, interactions were classified into pharmacokinetics and pharmacodynamics. Pharmacokinetics interactions occur when the interaction interferes with the drug absorption, distribution, metabolism, and/or excretion. When the interaction results in synergism or antagonism, it corresponds to a pharmacodynamic interaction.^28,29 Polypharmacy users were considered to be patients who used a number equal or greater than 5 medications daily.^30,31 When a medication contained 2 or more active pharmaceutical ingredients these were counted individually. Adverse drug reactions were considered when there was a warning sign in the electronic medical record, reporting reactions from drug interactions or proven drug allergies.

Data Analysis

The data collected were entered into the Microsoft Office Excel software (version 16.45) and statistical analysis was performed using SPSS Statistics^® version 26 (IBM, Armonk, New York, USA). Mean and standard deviation for continuous variables and percentages for categorical variables were calculated. An exploratory analysis was performed to assess the normality of continuous variables distribution using the Kolmogorov–Smirnov test, and a parametric analysis was conducted for variables with normal distribution and the Mann–Whitney test (U) was performed for non-parametric variables. In addition, Pearson’s chi-square test was used to test the association between the variables of interest. The level of statistical significance adopted was 5%.

This study was conducted after approval by the Research Ethics Committee of the University of Southern Santa Catarina (CEP-Unisul), opinion number 4 821 670. The patients’ identification data were kept confidential as recommended by Resolution No. 466/2012, of the Conselho Nacional de Saúde (CNS, National Health Council).

Results

During the period assessed, 1097 patients were seen at the Oncology Department of the study site. A total of 800 patients were excluded from the investigation for not meeting the pre-established inclusion criteria. Thus, data from 297 patients were included in the survey, among which 231 (77.8%) had potential drug-drug interactions. In total, 1044 DDI were counted and classified according to severity criteria. These data are reported in Figure 1.

Figure 1.

Flowchart indicating the inclusion and exclusion criteria and the pharmacological interactions found. The medical records of patients who did not undergo chemotherapy during the study period were excluded from the study.

The participants’ mean age was 59.3 (SD = 11.4) years; over two thirds were female and most patients (51.9%) were elderly. The majority of patients did not have smoking (44.1%) or drinking (57.6%) habits; however, this information was not included in approximately one third of the medical records. The independence test showed that there is an association between the variable polypharmacy and the prevalence of drug interactions, with no association between the other patient’s clinical and social variables. The clinical and social characteristics, as well as the results of the independence test are shown in Table 1.

Table 1.

Clinical and Social Variables and Their Statistical Associations with DDI n = 297.

Variables	n (%)	Interaction		P-value
		Yes n (%)	No n (%)
Gender				.814
Female	206 (69.4)	161 (54.2)	45 (15.1)
Male	91 (30.6)	70 (23.6)	21 (7.1)
Elderly a				.951
Yes	154 (51.9)	120 (40.4)	34 (11.4)
No	143 (48.1)	111 (37.4)	32 (10.8)
Age (years)				.982
21-40	13 (4.4)	10 (3.4)	3 (1.0)
41-60	138 (46.5)	108 (36.4)	30 (10.1)
>60	146 (49.2)	113 (38.0)	33 (11.1)
Education				.973
Low education level b	202 (68.0)	157 (52.9)	45 (15.1)
High education level c	95 (32.0)	74 (24.9)	21 (7.1)
Smoking habit				.804
Yes	32 (10.8)	24 (8.1)	8 (2.7)
No	131 (44.1)	103 (34.7)	28 (9.4)
Ex-smoker	40 (13.5)	33 (11.1)	7 (2.4)
NA	94 (31.6)	71 (23.9)	23 (7.7)
Alcohol consumption				.848
Yes	26 (8.8)	20 (6.7)	6 (2.0)
No	171 (57.6)	135 (45.5)	36 (12.1
NA	100 (33.6)	76 (25.6)	24 (8.1)
Polypharmacy d				.001*
Yes	125 (42.1)	120 (40.4)	5 (1.7)
No	172 (57.9)	111 (37.4)	61 (20.5)
Adverse drug reaction				.161
Yes	32 (10.8)	28 (9.4)	4 (1.3)
No	265 (89.2)	203 (68.4)	62 (20.9)
Comorbidities				.065
Yes	182 (61.3)	148 (49.8)	34 (11.4)
No	115 (38.7)	83 (28.0)	32 (10.8)

Note. Values are n (%). NA = Not available. ^a≥60 years old. *P < .05.

^bLess than 10 years of schooling; ^cEqual or above 10 years of schooling; ^dUse of 5 or more medications daily/

In addition to cancer disease under treatment, approximately two thirds of the patients included in the study had comorbidities. The main comorbidity found was systemic arterial hypertension (SAH), corresponding to 114 (31.2%) of the total of 365 chronic diseases described, followed by diabetes mellitus (n = 54; 14.8%) and major depressive disorder (n = 33; 9%). The main patients’ clinical features are described in Table 2.

Table 2.

Clinical characteristics of patients (n = 297).

Variables	n (%)
Site of the neoplasia
Breast	145 (48.7)
Gastrointestinal	81 (27.3)
Genitourinary	21 (7.1)
Gynecologic	19(6.4)
Head and neck	18(6.1)
Lung	8 (2.7)
Others	5 (1.7)
Comorbidities
Cardiovascular diseases	133 (36.5)
Endocrine diseases	103 (28.3)
Psychological illnesses	60 (16.4
Musculoeskeletal diseases	20 (5.5)
Respiratory diseases	14 (3.8)
Kidney diseases	10 (2.7)
Others	25 (6.8)

The analysis of the pharmacological profile shows that patients used an average 4.96 (SD = 4.55) drugs per day, reaching 20 different medications daily. Furthermore, almost half of the patients were considered polypharmacy users. About two thirds of the patients were exclusively submitted to chemotherapy treatment, while the others were concomitantly undergoing radiotherapy, hormone therapy or both. The main medications used, as well as the other patients’ pharmacological characteristics are displayed in Table 3.

Table 3.

Pharmacological Characteristics of Patients (n = 297).

Variables	n (%)
Current conventional therapies
Chemotherapy	191 (64.3)
Chemotherapy + Radiotherapy	55 (18.5)
Chemotherapy + Radiotherapy + Hormone therapy	32 (10.8)
Chemotherapy + Hormone therapy	19 (6.4)
Main anticancer drugs associated with interactions
Doxorubicin	222 (35.8)
Cyclophosphamide	133 (21.4)
Paclitaxel	97 (15.7)
Fluorouracil	75 (12.1)
Docetaxel	40 (6.5)
Oxaliplatin	32 (5.2)
Cisplatin	13 (2.1)
Capecitabine	6 (1.0)
Irinotecan	1 (0.2)
Main supportive drugs involved in interactions
Psychotropic drugs	642 (23.9)
Antiemetic agents	385 (14.3)
Opiods	209 (7.79)
Antihipertensive drugs	193 (7.19)
Non opiods analgesic	160 (5.96)
Diuretics	147 (5.48)
Nonsteroidal anti-inflammatory drugs	131 (4.88)
Antiarrhthimics	105 (3.91)
Anticoagulants/Antiplatelets	94 (3.50)
Antidiabetics (except insulin)	89 (3.32)
Estatinas	75 (2.79)
Others	453 (16.9)

According to the criteria established in the research protocol, 1044 potential DDIs were identified in the medical records assessed, totaling 197 (18.9%) DDI involving 2 chemotherapy agents, 227 (21.7%) involving chemotherapy and supportive drugs and 620 (59.4%) involving other medicines. Classification of interactions according to the pharmacological mechanism resulted in 865 (82.9%) pharmacokinetic interactions, 93 (8.9%) pharmacodynamic interactions, and 86 (8.2%) mixed interactions. In addition, 11% of the electronic medical records contained warning signs of adverse drug reaction

The categorization of DDIs according to severity criteria shows that the interactions classified as contraindicated resulted from the simultaneous use of selective serotonin reuptake inhibitors (SSRIs) with dopaminergic D2 blockers, while the most prevalent severe interactions were due to the use of dopaminergic antagonists with opioid/or hypnotics or SSRI with opioids. On the other hand, moderate DDIs involved mainly thiazide diuretics with chemotherapy or with non-steroidal anti-inflammatory drugs (NSAIDs). DDIs description is reported in Table 4.

Table 4.

Drug-drug interactions in cancer patients (n = 1044 drug-drug interactions).

Interactions	n	MA	Interactions	n	MA	Interactions	n	MA
Doxorubicin + Cyclophosphamide †††	76	K	Metoprolol + Metformin ††	8	D/K	Enalapril + Glyburide ††	4	K
Doxorubicin + Paclitaxel †††	72	K	Morphine + Ondansetron †††	8	K	Enalapril + Hydrochlorothiazide ††	4	K
Fluorouracil + Leucovorin Calcium ††	49	K	Paclitaxel + Simvastatin ††	8	K	Methadone + Ondansetron †††	4	D/K
Doxorubicin + Docetaxel †	37	K	Cisplatin + Paclitaxel †††	7	K	Metoprolol + Aspirin ††	4	K
Cyclophosphamide + Tamoxifen †††	26	K	Enalapril + Metformin ††	7	K	Salsalate + Losartan ††	4	K
Dipyrone + Dexamethasone †††	15	K	Methadone + Morphine †††	7	K	Amitriptyline + Aspirin †††	3	K
Cyclophosphamide + Hydrochlorothiazide ††	15	K	Citalopram + Codeine †††	6	K	Amlodipine + Simvastatin †††	3	D/K
Oxaliplatin + Ondansetron †††	13	D	Clonazepam + Zolpidem †††	6	K	Atenolol + Clonidine †††	3	D
Fluorouracil + Hydrochlorothiazide ††	12	K	Hydrochlorothiazide + Aspirin †††	6	K	Atenolol + Glyburide ††	3	K
Codeine + Ondansetron ††	11	K	Levothyroxine + Simvastatin ††	6	K	Atorvastatin + Ciprofibrate †††	3	D
Doxorubicin + Trastuzumab †††	11	K	Bromazepam + Zolpidem †††	5	K	Bromopride + Morphine †††	3	D
Atenolol + Metformin ††	10	K	Citalopram + Rivaroxaban †††	5	K	Capecitabine + Omeprazole †††	3	K
Metoclopramide + Morphine †††	10	K	Clonazepam + Metoclopramide †††	5	K	Citalopram + Metoprolol ††	3	K
Codeine + Metoclopramide †††	9	K	Fluorouracil + Methotrexate †††	5	K	Clonazepam + Codeine †††	3	K
Domperidone + Ondasetron †††	9	D	Hydrochlorothiazide + Salsalate †††	5	K	Clopidogrel + Aspirin †††	3	K
Losartan + Aspirin ††	9	D/K	Levothyroxine + Pantoprazole ††	5	K	Codeine + Morphine †††	3	K
Metoclopramide + Zolpidem †††	9	K	Methadone + Metoclopramide †††	5	K	Codeine + Ondasetron †††	3	K
Citalopram + Metoclopramide ††††	8	K	Amitriptyline + Carbamazepine ††	4	K	Codeine + Zolpidem †††	3	K
Citalopram + Omeprazole †††	8	D	Atenolol + Aspirin ††	4	K	Cyclophosphamide + Carbamazepine †††	3	K
Citalopram + Ondansetron †††	8	D/K	Bromopride + Codeine †††	4	D	Dipyrone + Fluoxetine †††	3	K
Cyclophosphamide + Ondansetron ††	8	K	Citalopram + Dipyrone †††	4	K	Dipyrone + Rivaroxaban †††	3	K
Dipyrone + Losartan ††	8	K	Codeine + Metoclopramide †††	4	K	Oxaliplatin + Domperidone †††	3	D
Doxorubicin + Amitriptyline †††	8	K	Doxorubicin + Dexamethasone †††	4	K	Other drug interactions	404	D/K

Note. n = number of prescriptions; MA = mechanism of action; D = Pharmacodynamic; K = Pharmacokinetics.

Severity: ^†Minor, ^††Moderate, ^†††Major, ^††††Contraindicated.

Discussion

This study evaluated the prevalence of drug interactions, their associated factors and their degree of severity in patients undergoing intravenous chemotherapy. Interactions are shown according to the clinical risk to the patient and the mode of action involved in each interaction. It was observed that approximately 78% of the patients assessed had a potential risk for DDI, corroborating previous research that estimated between 60% and 100% the occurrence of DDI in cancer patients.^12,13,32

Classification of interactions according to the mode of action resulted in more than 80% of pharmacokinetic interactions. These interactions may involve the induction or inhibition of cytochrome P450 enzymes (CYP450).^33,34 The induction of CYP450 can result in an increase in the metabolic rate of a drug or a reduction in its serum concentration, which in turn causes an increase in the bioavailability of the active metabolite, potentially enhancing its therapeutic effect and possibly causing toxicity or causes the loss of therapeutic efficacy. On the other hand, enzyme inhibition often reduces the metabolic rate, increasing the serum concentration of the drug, which may cause an increase in therapeutic response or toxicity.^19,34

A high percentage of interactions involving 2 chemotherapeutic drugs (18.1%) was found; however, these interactions are already recognized in the literature for being part of cancer treatment protocols.^35-37 Other very prevalent DDIs involved the use of a chemotherapy agent with supportive medication, especially the use of fluorouracil with calcium folinate and cyclophosphamide with tamoxifen. It is known that calcium folinate, an active metabolite of folic acid, is used to potentiate the effect of fluorouracil,^38,39 while tamoxifen is used as an adjuvant of the treatment with cyclophosphamide.^40-42

The data obtained demonstrate a high prevalence of DDI in the patients evaluated, showing a positive association with the variable polypharmacy. It has already been established in the literature that the risk of drug interactions is proportional to the number of medications used by patients^9,31; polypharmacy is a predictor of hospitalization, adverse reactions and increased mortality risk.^31,43

Classification of DDI according to the clinical risk evidenced a high prevalence of contraindicated, severe and moderate interactions, as described in previous studies.^17,44,45 In connection with these interactions, it is known that the parallel use of SSRIs with dopaminergic antagonists^46-48 or with opioids^49,50 can result in serotonergic syndrome and neuroleptic-malignant syndrome. In addition, it is possible that the use of opioid dopaminergic antagonists may cause respiratory depression. ⁵¹ Other interactions often observed in this investigation were due to the use of thiazide diuretics with NSAIDs and chemotherapy. The concomitant use of diuretics with NSAIDs has been described and can result in a reduction in the diuretic effect, increasing the chances of toxicity.^52,53 On the other hand, the use of those diuretics with chemotherapeutics is associated with an increased risk of granulocytopenia and myelosuppression.^54,55

Data from this study show a high prevalence of DDI involving supportive drugs. Among the supportive drugs, psychotropic medications were the most involved in DDI, followed by antiemetics and opioids. Psychotropic drugs are frequently used in cancer patients, either because of pre-existing psychological disorders, or because these patients develop symptoms such as insomnia and anxiety as a result of cancer therapy.^33,56 Likewise, antiemetics and opioid analgesics are frequently prescribed to cancer patients with the aim of reducing the symptoms of nausea and vomiting caused by chemotherapy^57,58 and for pain control,^59,60 respectively.

The high incidence of DDI in cancer patients has already been demonstrated in previous studies,^{13,17,25,45,61} which can be explained by the polymedication to which many of these patients are exposed.^10,24,62 However, the fact that cancer patients commonly have comorbidities^8,11 makes the concomitant use of supportive medications critical, ⁵⁷ increasing the susceptibility of these patients to DDI.^13,14 Thus, it is necessary that, especially for polypharmacy users, the prescription of a drug be made after careful review of all drugs used by the patient in order to avoid potential DDI. ⁵⁷

In this study, adverse drug reactions were documented in approximately 11% of patients. This value is divergent from previously published studies.^63,64 However, it is important to emphasize that in patients undergoing cancer treatment, the effects resulting from adverse reactions can be mistakenly attributed to chemotherapy, making early recognition a challenge in medical practice. ¹⁸ In addition, the fact that patients commonly receive prescriptions by more than one physician is associated with a greater chance of developing DDI and, consequently, adverse drug reactions.^65,66

Taking these facts into account, drug reconciliation is considered a potential alternative for the prevention of DDI^67,68 which is the process in which a team of health professionals work to obtain information about the drugs used by patients, ensuring that this information obtained is passed on in health care transitions, increasing the safety of patient treatment. ⁶⁹ Thus, the importance of communication within the interdisciplinary team for the effectiveness of the patient’s treatment is evident. ⁷⁰

There are some limitations in this study, such as: only patients undergoing outpatient treatment were included, that is, patients with less severe conditions. In addition, the survey was carried out at only one center. Only 2 sources of research on DDI were used and only potential or probable DDI were determined. However, the results are consistent with other previously published studies on polypharmacy and drug interaction in cancer patients.

Conclusion

Patients undergoing chemotherapy treatment are often exposed to drug interactions, many of which are considered to be of clinical relevance and are even contraindicated. The results of this study show a large percentage of patients at potential risk of DDI, mainly classified as severe. Therefore, these patients are continuously at risk of developing adverse drug reactions, either by reduction of the effect of some medication or the toxicity resulting from those interactions.

In order to avoid DDI and potential harmful reactions resulting from them, it is necessary that the oncologist and other prescribers consider all medications used by patients and, whenever possible, indicate alternative drugs with minor interaction potential in order to prevent the harmful effects of DDI and provide a better prognosis for patients. When it is not possible to avoid the use of the drugs involved in DDI, caution, treatment delay and continuous monitoring of these patients are suggested.