Most Cancer Drug Side Effects May Not Be Preventable, Vietnamese Database Reveals

When a patient with cancer begins chemotherapy, the question that haunts every oncologist, nurse, and pharmacist is not only whether the drug will work, but what harm it might cause along the way. Adverse drug reactions, or ADRs, are among the most common and feared events in cancer therapy. They can stretch hospital stays, derail treatment schedules, and in the worst cases kill. Yet a deceptively simple question has long gone underexplored in many parts of the world: how many of these harmful reactions could actually have been prevented? A new study from Vietnam, drawing on a full year of reports from the country’s regional pharmacovigilance system, offers one of the most detailed answers yet, and its conclusion is striking: of hundreds of confirmed reactions to anticancer drugs, only two were judged preventable.

The research, published in BMC Cancer, was conducted by a team led by Thi Thu Thuy Nguyen of Hong Bang International University in Ho Chi Minh City, working with colleagues from Cho Ray Hospital and the Health Technology Assessment & Application Research Institute. The investigators performed a retrospective, cross-sectional analysis of all ADR reports related to anticancer drugs submitted to the Ho Chi Minh City Regional Drug Information and Pharmacovigilance Center between January 1 and December 31, 2023. Reports concerning drug quality problems and those not associated with anticancer drugs were excluded, leaving 291 reports for analysis. In Vietnam, where studies assessing the severity and preventability of ADRs using regional pharmacovigilance data have been scarce, the study fills an important evidence gap.

The technical backbone of the study rests on two internationally recognized assessment systems. First, the researchers used the World Health Organization–Uppsala Monitoring Centre (WHO-UMC) causality assessment system, a scale that classifies the likelihood that a suspected drug truly caused a reported reaction, ranging from unlikely through possible, probable, and certain. Only reports with a confirmed causal relationship were eligible for the next step. Second, the team applied the WHO Preventability method, known as the WHO P-method, which determines whether a reaction could have been avoided through actions such as respecting a documented drug history, avoiding therapeutic duplication, or adjusting doses for patient characteristics. This two-stage filtering is what allows pharmacovigilance scientists to separate genuine, unavoidable toxicities of powerful drugs from errors that better clinical practice might have caught.

The demographic picture that emerged is telling. Female patients experienced ADRs far more frequently than male patients, accounting for 212 of the 291 reports, or 72.9 percent, compared with 76 reports, or 26.1 percent, among men. The 40 to 64 age group dominated, representing 54.3 percent of cases, or 158 of 291 reports. This skew toward middle-aged women reflects both the epidemiology of cancer in Vietnam and the biology of the drugs involved. Agents such as carboplatin and paclitaxel, the two most frequently implicated drugs in the study, are mainstays of breast and gynecologic cancer treatment, and paclitaxel in particular is notorious for hypersensitivity reactions that occur more readily in female patients.

Of the 291 reports, causality assessment using the WHO scale confirmed a causal relationship between the suspected drug and the reaction in 253 reports, corresponding to 433 distinct drug–ADR pairs. These confirmed pairs were then evaluated for preventability, and the result was remarkable: only 2 reports, or 0.8 percent, were determined to be preventable. The vast majority of the confirmed reactions were classified as non-serious, accounting for 70.8 percent, or 179 of 253 reports. Nevertheless, a considerable proportion carried real clinical weight: 18.2 percent, or 46 reports, required hospitalization or prolonged a hospital stay, and 6.3 percent, or 16 reports, resulted in death. These figures underscore that even reactions deemed non-preventable can be devastating when the drugs involved are cytotoxic agents designed to kill dividing cells.

The two preventable cases deserve close attention because each represents a different, well-known category of avoidable harm. One involved cisplatin, a platinum-based chemotherapy drug, in a patient with a documented history of drug hypersensitivity. Administering a drug to a patient with a known allergy to it is precisely the kind of error that careful medication reconciliation and allergy documentation are designed to prevent. The other case involved epirubicin, an anthracycline agent, and was attributed to therapeutic duplication, meaning the patient received overlapping therapy that should not have been co-administered. Both scenarios fall squarely within the WHO P-method’s definition of preventability, which considers whether a reaction could have been avoided given the information available before the drug was given.

The most common anticancer drugs causing ADRs were carboplatin, implicated in 92 of the 433 drug–ADR pairs, or 21.2 percent, and paclitaxel, implicated in 87 pairs, or 20.1 percent. Carboplatin is a second-generation platinum compound whose dose is typically calculated using a formula based on kidney function, and its characteristic toxicities include myelosuppression, nausea, and thrombocytopenia. Paclitaxel, a taxane derived originally from the bark of the Pacific yew, is famous for infusion-related hypersensitivity reactions driven by the solvent Cremophor EL, as well as peripheral neuropathy and neutropenia. Both drugs are frequently combined in regimens for ovarian, breast, and lung cancers, which helps explain their prominence in the pharmacovigilance database.

Why does the preventability figure matter so much? In pharmacovigilance research, preventability estimates vary widely across settings and drug classes. Studies in general hospital populations have sometimes found that a third or more of ADRs are preventable, particularly those arising from prescribing errors, drug interactions, or failure to respect documented allergies. The very low preventability rate of 0.8 percent in this anticancer drug database suggests that most reactions to cytotoxic agents are intrinsic toxicities, predictable consequences of the drugs’ mechanisms of action rather than lapses in clinical judgment. This does not mean such reactions are unimportant; rather, it reframes the challenge. If most harm from chemotherapy is inherent to the drugs, then prevention must come from better supportive care, dose optimization, premedication, and monitoring, not simply from avoiding prescribing mistakes.

The study’s findings carry implications for how pharmacovigilance systems in low- and middle-income countries can inform patient safety. Vietnam’s regional drug information and pharmacovigilance centers collect spontaneous reports from hospitals across their catchment areas, and this analysis demonstrates that such databases can yield actionable characterizations of drug safety. The identification of preventable ADRs, even if rare, highlights the need for ongoing monitoring and evaluation to inform appropriate preventive strategies. The authors note that most anticancer drug–related ADRs in the study were non-serious, although a considerable proportion still resulted in hospitalization or death, and that the identification of preventable cases underscores the value of continuous surveillance in cancer treatment.

For patients and clinicians, the message is nuanced but ultimately reassuring in one respect and sobering in another. The reassuring part is that the overwhelming majority of confirmed reactions to anticancer drugs in this Vietnamese database were not the result of preventable errors; they were the expected, if unwelcome, consequences of drugs whose therapeutic power and toxic potential are two sides of the same coin. The sobering part is that a meaningful fraction of these reactions, nearly one in five, still sent patients to the hospital or extended their stays, and one in sixteen proved fatal. As cancer therapy evolves toward targeted agents and immunotherapies with their own distinctive safety profiles, the kind of systematic, database-driven preventability assessment exemplified by this study will remain an essential tool for distinguishing avoidable harm from the inherent price of treating cancer with drugs that attack the very machinery of cell division.

Subject of Research: Pharmacovigilance

Article Title: Preventability of adverse drug reactions to anticancer drugs: insights from the database of the Regional Drug Information and Pharmacovigilance Center

Article References: Nguyen, T. T. T., Nguyen, Q. B., Nguyen, T. K. U., & Do, H. Q. T. (2026). Preventability of adverse drug reactions to anticancer drugs: insights from the database of the Regional Drug Information and Pharmacovigilance Center. BMC Cancer. https://doi.org/10.1186/s12885-026-16459-4

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16459-4

Keywords: adverse drug reactions, anticancer drugs, pharmacovigilance, drug safety, cancer therapy, WHO-UMC causality assessment, preventability, adverse, drug, reactions, anticancer, drugs