Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jul 3.
Published in final edited form as: Lancet Glob Health. 2025 Jul;13(7):e1250–e1257. doi: 10.1016/S2214-109X(25)00138-X

Substandard anticancer medications in clinical care settings and private pharmacies in sub-Saharan Africa: a systematic pharmaceutical investigation

Maximilian J Wilfinger 1, Jack Doohan 1, Ekezie Okorigwe 1, Ayenew Ashenef 1, Atalay Mulu Fentie 1, Ibrahim Chikowe 1, Hanna Stambuli Kumwenda 1, Paul Ndom 1, Yauba Saidu 1, Jesse Opakas 1, Phelix Makoto Were 1, Sachiko Ozawa 1, Benyam Muluneh 1, Marya Lieberman 1
PMCID: PMC12224173  NIHMSID: NIHMS2093084  PMID: 40580990

Summary

Background

The quality of anticancer drugs is crucial for good patient outcomes, but quality surveillance in low-income and middle-income countries (LMICs) has been deterred by the high toxicity of the drugs. Despite worrisome reports about substandard or falsified products, no systematic studies of anticancer drug quality across multiple LMICs have been reported.

Methods

Between April 6, 2023, and Feb 12, 2024, cisplatin, oxaliplatin, methotrexate, doxorubicin, cyclophosphamide, ifosfamide, and leucovorin dosage forms were collected both covertly and overtly from 12 hospitals and 25 private or community pharmacies in Ethiopia, Kenya, Malawi, and Cameroon, with the goal of obtaining ten different brands and lot numbers of each type of active pharmaceutical ingredient (API). Each product was visually inspected. The percentage of active pharmaceutical ingredient relative to the stated API content was assayed with high-performance liquid chromatography (HPLC). Assay values were compared with US Pharmacopoeia acceptance criteria for different APIs and dosage forms. Samples with assay values that failed to meet the appropriate acceptance criteria were categorised as having failed HPLC assay, samples with assay values that fell within the allowed acceptance criteria were categorised as having passed HPLC assay, and samples with assay values that fell within the 2% margin of error of the acceptance criteria were categorised as inconclusive. Critical failure rates were calculated with 95% CIs and significance testing was done for differences between failure rates. For the comparison of visual inspection with HPLC results, sensitivity was calculated as the number of lots that failed both HPLC assay and visual inspection divided by the total number of lots that failed HPLC assay. Specificity was calculated as the number of lots that passed both HPLC assay and visual inspection divided by the total number of lots that passed HPLC assay.

Findings

251 samples of chemotherapy drugs (dosage forms) were collected between April 6, 2023, and Feb 12, 2024, and 191 unique brands and lot numbers were obtained. Products from eight of 191 (4%) unique lot numbers (collected in countries coded as W, X, and Y) failed visual inspection. Active pharmaceutical ingredient contents ranged from 28% to 120% of stated contents, and failure rates ranged from 14% to 24% across the different countries; these rates were not significantly different at the 95% CI. Nearly a quarter of the products (59 [24%] of 251) had expired before analysis, some by nearly a year, but the expired products did not fail HPLC assay at a higher rate than the non-expired products. Ten of the 59 post-expiry products failed assay (ie, a 17% failure rate), whereas 38 of the 189 pre-expiry samples failed assay (ie, a 20% failure rate); these rates were not different at the 95% CI. Failing products were found in all four countries and in both major hospitals and private pharmacies (with no difference in failure rates at the 95% CI). The sensitivity of visual inspection for the detection of products failing HPLC assay was 9% (three of 32 lots) and the specificity was 97% (155 of 159 lots). The sensitivity of visual inspection is low because many quality defects, such as a shortage of an uncoloured active pharmaceutical ingredient, are not visually apparent.

Interpretation

Oncology practitioners and health systems in sub-Saharan Africa need to be aware of the possible presence of substandard anticancer products when designing care protocols and evaluating patient outcomes, and regulatory system strengthening is needed to provide better surveillance of this crucial class of medicines.

Introduction

Oncology drugs are widely suspected of being vulnerable to pharmaceutical crime, particularly in low-income and middle-income countries (LMICs).1 This vulnerability is driven by an exponential growth in cancer treatment in these countries. For instance, 10 years ago in Ethiopia2 and Kenya,3 cancer care was available to only a few thousand patients per year in a few hospitals. Today, over 75 000 people receive cancer treatment each year in each of these countries. This rising trend has also been observed in Cameroon and Malawi, where cancer care is provided to over 20 000 patients each year.4,5 The required anticancer drugs are expensive. In Cameroon for instance, where the average gross domestic product per capita in 2023 was US$1674, the cost of six cycles of chemotherapy is $1986.4 To maximise access to these life-saving treatments for their populations, ministries of health with constrained budgets are under pressure to buy the lowest cost products on the market, in turn exerting pressure on manufacturers to cut corners in the manufacturing process. At the same time, medicine regulatory agencies in LMICs do not have the capacity to measure the quality of anticancer drugs.6 This combination of high demand but low capacity for regulatory oversight in a market renders it vulnerable to incursion by substandard and falsified medical products.7

Anticancer products must meet compendial requirements for licensing or regulatory approval. The assay value (ie, the quantity of the active pharmaceutical ingredient provided in the dosage form compared with the quantity stated on the packaging materials) is of particular importance for ensuring that patients receive the correct dose. Regulatory agencies in LMICs often do not have the necessary reference materials for conducting chemical assays of anticancer drugs, and might not have the personal protective equipment and chemical fume hoods or laminar flow cabinets needed to handle anticancer drugs safely.

Substandard or falsified anticancer drugs have already harmed many people. Around 2012, 119 individuals received a falsified bevacizumab chemotherapy product in a state-run hospital in Veracruz, Mexico.8 Fake bevacizumab was also found in multiple sites in the USA around 2012; the manufacturer reported that the product they were given to examine was falsified and contained only water.9 WHO later issued a rapid alert for falsified bevacizumab and sunitinib in east Africa.10 A recent retrospective review of case reports in Kenya identified 551 cases of suspected therapeutic failure associated with the antineoplastic agent imatinib.11 Multiple substandard asparaginase products were identified in 2018–22,12 and their harm to patients was documented in hospitals in Brazil.13–15 Journalists found that batches of these asparaginase products had been imported 167 times by countries in sub-Saharan Africa.16 In 2018, Eberle and colleagues17 reported three lots of substandard cisplatin in a government hospital in Ethiopia. In 2019, WHO issued a global alert warning patients, doctors, and pharmacies about a fake anticancer drug packaged to look like the cancer drug Iclusig (ponatinib); the pills contained only paracetamol.18

Given the number of known cases in which patients have been harmed by substandard or falsified anticancer medicines, the dearth of publicly available quality data is surprising. The US Pharmacopoeia Medicines Quality Database contains records for over 17 170 medical products from 25 LMICs that were screened or tested by compendial methods between 2003 and 2017; however, it has no entries for any chemotherapy, oncological, antineoplastic, or anticancer products.19,20 In this study, we aimed to investigate the quality of seven anticancer drugs (cisplatin, oxaliplatin, methotrexate, doxorubicin, cyclophosphamide, ifosfamide, and leucovorin) collected from the public and private sectors of health-care systems in Ethiopia, Kenya, Malawi, and Cameroon in 2023–24.

Methods

Sample collection, storage, and shipping

Sample collectors in Ethiopia, Kenya, Malawi, and Cameroon set out to obtain ten different brands and lot numbers for each of seven different active pharmaceutical ingredients (APIs; cisplatin, oxaliplatin, methotrexate, doxorubicin, cyclophosphamide, ifosfamide, and leucovorin) between April 6, 2023, and Feb 12, 2024. The country in which each unique product was obtained was coded as country W, country X, country Y, or country Z in order to not hinder ongoing regulatory activities related to our results. For the same reason, we did not attempt to authenticate products by contacting manufacturers. Samples were first collected from major national hospitals that provide centralised cancer care. This phase was done overtly in countries X, Y, and Z with full cooperation from the hospitals and covertly in country W. Overt sampling meant anticancer medicine samples were collected after disclosing the purpose of the collection to the pharmacy or health-care workers at the sampling sites (appendix 1 p 15).21 No hospital provided ten unique lot numbers of any active pharmaceutical ingredient, and every site experienced stockouts of multiple active pharmaceutical ingredients during the sample collection period. In the following phase of collection, collectors identified private or community pharmacies that supplied anticancer products. For these sites, collection was done by study staff who presented medical prescriptions, in accordance with standard practice. This phase was done overtly in country Y and covertly in countries W, X, and Z. In country X, clinics or pharmacies that offer particular anticancer medicines (ie, products that are approved or registered by national regulators and imported by legally established or licensed importers) to patients but are not legally authorised to do so are termed grey markets.

None of the samples were cold chain products. At each main collection site (ie, the major national hospitals, which can safely store anticancer drugs; appendix 1 pp 14–15), samples were stored separately from other medications per manufacturer recommendations, generally at 20–25°C. The sample metadata were entered into a secure database system (Artifacts VERIFY, Artifactsofresearch, San Diego, CA, USA). Samples were packaged for shipping following a standard operating procedure (appendix 1 pp 20–21) to minimise risks in case of breakage. The samples and temperature logging devices were sent via courier (DHL or FedEx) to the University of Notre Dame, Notre Dame, IN, USA, where they were stored per manufacturer recommendations. Samples were typically in transit for 6–8 days.

Safety precautions for the handling of all cytotoxic anticancer drugs included the use of a fume hood or laminar flow hood, designated laboratory space, and personal protective equipment (ie, double chemo-rated gloves, splash-proof goggles, chemo-rated gowns, and face masks; appendix 1 p 22). Ethical approval for this research was obtained from hospital and university institutional review boards, independent research ethics committees, or ethical review boards in each of the four countries in which samples were collected and in the USA where the samples were analysed (appendix 1 p 16). We shared and discussed the findings of this study with pharmacovigilance or regulatory contacts in all four countries and WHO.

Visual inspection and high-performance liquid chromatography analysis

Visual inspection is the only quality control method for anticancer drugs practised in all four countries included in this study. In this study, inspection included scrutinising products’ inner and outer packaging, labelling, and dosage forms for signs of falsification, bad handling, deterioration, or improper manufacture, and evaluating reconstituted product.22 Products that were deemed as having failed visual inspection might, for example, not have had a labelled expiration date, have arrived in damaged packaging, or have turned an unexpected colour when reconstituted.

System suitability was established for each active pharmaceutical ingredient on one or more high-performance liquid chromatography (HPLC) instruments following US Pharmacopoeia methods.23 The instruments included an e2695 separation module coupled with a 2487 dual-wavelength absorbance detector (Waters, Milford, MA, USA), an UltiMate DAD-3000 diode array detector (Thermo Fisher Scientific, Waltham, MA, USA), and a 1100 series G1315B diode array detector (Agilent, Santa Clara, CA, USA). Certified reference standards from US Pharmacopoeia and other analytical grade materials and chemicals from Sigma Aldrich (St Louis, MO, USA) were used. The percentage of active pharmaceutical ingredients in each product was calculated by comparing the actual concentration of each HPLC sample with the concentration expected from the stated active pharmaceutical ingredient content. The US Pharmacopoeia acceptance criteria for different active pharmaceutical ingredients and dosage forms are shown in appendix 1 (p 235); typically, products must contain 90–110% of the stated amount of active pharmaceutical ingredient to meet the assay standard. Samples with assay values that failed to meet the appropriate acceptance criteria were categorised as having failed, samples with assay values that fell within the allowed acceptance criteria were categorised as having passed, and samples with assay values that fell within the 2% margin of error of the acceptance criteria were categorised as inconclusive. A failure could be the result of several possible issues, including bad manufacturing practices, falsification, and improper shipping or storage. The full details of the materials, instruments, standard and sample preparation, system suitability procedures, analytical metrics, and calculations related to the analysis are provided in appendix 1 (pp 12–22, 25)

For a group of n samples with F samples failing assay, critical failure rates were calculated as (F/n) ± C, where C is the 95% binomial CI. Significance testing for differences between failure rates was done with the normal distribution method and a 95% CI. For the comparison of visual inspection to so-called gold standard HPLC results, sensitivity was calculated as the number of lots that failed both HPLC assay and visual inspection divided by the total number of lots that failed HPLC assay (reported as a percentage), and specificity was calculated as the number of lots that passed both HPLC assay and visual inspection divided by the total number of lots that passed HPLC assay (reported as a percentage).

Role of the funding source

The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.

Results

251 samples of chemotherapy drugs (dosage forms) were collected between April 6, 2023, and Feb 12, 2024, in Ethiopia, Kenya, Malawi, and Cameroon (appendix 1 pp 4–11; appendix 2 pp 1–2) and shipped to the University of Notre Dame in seven batches. Of these 251 samples, 61 (24%) were obtained from hospital pharmacies; the remaining 190 (76%) samples were obtained from private pharmacies and so-called grey market sources. Overall, 191 unique brands and lot numbers were obtained from country W (n=28), country X (n=57), country Y (n=25), and country Z (n=81; appendix 1 pp 4–11; appendix 2 pp 3–4). For 163 (85%) of the 191 unique lots from countries X, Y, and Z, we were able to check registration status. 95 (58%) of these 163 unique lots could be verified with relevant national regulatory agencies, but there were wide differences in regulatory practices. Registered lots constituted 0% (0 of 28) of the lots collected in country W, 28% (16 of 57) of lots collected in country X, 0% (0 of 25) of lots collected in country Y, and 98% (79 of 81) of lots collected in country Z. Samples collected across all four sites included 191 unique lot numbers, comprising cisplatin (30 unique lots), oxaliplatin (26 unique lots), methotrexate (44 unique lots, and three repackaged products without lot information), doxorubicin (35 unique lots), cyclophosphamide (31 unique lots), ifosfamide (four unique lots), and leucovorin (18 unique lots).

Products from eight (3%) of 251 products (collected in countries W, X, and Y) failed visual inspection. Three unique lots of 205 injectable products collected in countries X and Y failed because they produced unexpected colours when dissolved. However, only two (67%) of these three lots failed HPLC assay (figure 1A–C). Five (63%) of eight products failed visual inspection because they had serious labelling defects, including one calcium leucovorinate product provided in a vial without a brand name or manufacturer information and one lot of a doxorubicin product that was labelled only in Turkish but was obtained in a country that requires labels in English (appendix 1 p 23). Three sets of tablets collected in country W were provided in paper envelopes (figure 1D) that did not include the brand name, manufacturer information, lot number, or expiration date (these were grouped as one of the eight products). A blister pack of tablets collected in country W with a particular manufacturer’s information arrived inside a box marked with another manufacturer’s information (appendix 1 p 24). Finally, the packaging of one lot of methotrexate collected in country W contained substantial differences from the packaging of the same product collected in country Z (figure 1E). Of the five unique lots associated with labelling defects, just one (20%) failed HPLC assay. The sensitivity of visual inspection for the detection of products failing HPLC assay was 9% (three of 32 unique lots) and the specificity was 97% (155 of 159 unique lots). The sensitivity of visual inspection is low because many quality defects, such as a shortage of an uncoloured active pharmaceutical ingredient, are not visually apparent. An example is provided in appendix 1 (p 25).

Figure 1: Anticancer products that failed visual inspection.

Figure 1:

Open in a new tab

The percentages on each photograph show the corresponding HPLC assay result. The products are ifosfamide (A), injectable methotrexate (B; packaged in amber vials so transferred to Eppendorf tubes for this photo), cyclophosphamide (C), methotrexate tablets in secondary packaging (D; this practice was banned in 2022, but some vendors still have bulk supplies and are allowed to dispense in this way), and injectable methotrexate (E) that contains some interesting packaging variations between failing and passing samples. The substitution of grey printed dots for embossed Braille dots and the fact that only the sample from country Z matches the product pictured on the manufacturer’s website raises the suspicion that the samples collected in country W might be falsified. HPLC=high-pressure liquid chromatography.

HPLC assay results for the percentage of active pharmaceutical ingredients in the 251 dosage forms ranged from 28% to 120% of the stated active pharmaceutical ingredient content. Figure 2 summarises the assay results for the 191 products with unique lot numbers collected in this study. If multiple dosage forms of a lot number were analysed, the average assay result is shown in figure 2, whereas all results are included in figure 3. 32 (17%) of the 191 unique lot numbers were categorised as failures. The total number of products failing HPLC assay was 48 (19%) of 251. Failure rates ranged from 14% to 24% across the different countries. For country W, nine (14% [95% CI 5–22]) of 65 products failed. For country X, 18 (24% [15–34]) of 74 products failed. For country Y, 6 (21% [6–35]) of 29 products failed. For country Z, 15 (18% [10–26]) of 83 products failed (appendix 2 p 2). There were 11 unique lot numbers (6%) of 191 products that contained 75% or less of the stated active pharmaceutical ingredient. Nearly a quarter (59 [24%]) of 251 products had expired before analysis, some by nearly a year, but the expired products did not fail HPLC assay at a higher rate than the non-expired products. A total of 248 (99%) of 251 assayed products had expiration dates that could be identified. Ten of the 59 post-expiry products failed assay (ie, a 17% failure rate), whereas 38 of the 189 pre-expiry samples failed assay (ie, a 20% failure rate); these rates were not different at the 95% CI (appendix 2 p 7).

Figure 2: Active pharmaceutical ingredient content of 191 unique anticancer product lot numbers as assessed by HPLC.

Figure 2:

Open in a new tab

The HPLC assay result is expressed as a percentage of the stated active pharmaceutical ingredient content of each product. The green bar illustrates the 90–110% active pharmaceutical ingredient content range, which is the acceptance standard for all of the drugs except methotrexate (powder for injection), which has an acceptance standard of 95–115%; cyclophosphamide (sterile powder for injection), which has an acceptance standard of 90–105%); and calcium leucovorin (predissolved), which has an acceptance standard of 90–120%. Purple dots (n=133) represent products that confidently meet the assay standard. Blue dots (n=26) lie within the 2% margin of error of the HPLC assay. Red dots represent products that failed assay and are either subpotent (n=25) or superpotent (n=7). HPLC=high-pressure liquid chromatography.

Figure 3: HPLC assay result variability for every anticancer product lot number for which multiple dosage forms were assayed (90 samples; 30 unique lot numbers).

Figure 3:

Open in a new tab

The HPLC assay result is expressed as a percentage of the stated active pharmaceutical ingredient content of each product. The mean for each lot number is shown as a circle (injectable vials) or triangle (tablets). The error bars represent the standard deviation. HPLC=high-pressure liquid chromatography.

Products obtained from hospitals were not of better quality than products purchased from private pharmacies or grey market or unknown sources. 50 (26%) of the 191 unique lot numbers were sourced overtly from hospital pharmacies. Of these 50 products, nine (18%) failed assay, with six (12%) of the products containing less than 75% of the stated active pharmaceutical ingredient content. Of the 251 products, 100 (40%) were identified as registered in the countries where they were collected; 16 (16% [95% CI 9–23]) of these 100 products failed assay. 84 (33%) of the 251 products were identified as not registered; 23 (27% [95% CI 18–37]) of these 84 products failed assay. 67 (27%) of the 251 products were of unknown registration status, and nine (13% [5–22]) of these 67 products failed assay. These failure rates are not significantly different at the 95% confidence level (appendix 2 p 3). All registered products that failed assay were associated with country Z, which reported 98% of products as registered.

We did not have sufficient dosage units of most products to perform the US Pharmacopoeia content uniformity test, so we estimated product variability across multiple dosage units with the size of the standard deviations for samples with the same brand and lot number. Figure 3 shows assay values obtained for the 30 unique lot numbers (90 total samples) for which 2–6 replicate dosage units were available for assay.

The 191 unique lot number samples came from 50 unique manufacturers with 11 stated countries of origin. All manufacturer names and countries of origin are those printed on the product packaging. It has not been established that any of the products assayed were not fraudulently labelled or counterfeit. 31 (62%) of the 50 stated manufacturers did not have any products fail assay. The overall rate of lot failure was evaluated for all manufacturers for which we had three or more lot samples. Four (2%) of 251 samples did not include a manufacturer name (three groups of methotrexate tablets were purchased in secondary packaging that did not include the manufacturer’s name and one vial of calcium leucovorinate did not include a manufacturer’s name on the label). None of these four samples failed assay. The average failure rate seen across all 247 products with known manufacturers was 19% (48 of 247 [95% CI 14–24]). Two stated manufacturers had failure rates that were statistically larger: Zee Laboratories (five of seven lots failed; 71% [95% CI 38–10] failure rate) and Venus Remedies (nine of 11 lots failed; 82% [59–105] failure rate). Three other stated manufacturers had failure lots that were statistically higher than zero: Namen Pharma (four of eight lots failed; 50% [95% CI 15–85] failure rate), Beta Drugs (six of 20 lots failed; 30% [10–50] failure rate), and Zuvius Lifesciences (three of eight lots failed; 38% [4–72] failure rate). The full data for all products are provided in appendix 1 (pp 4–11) and appendix 2 (p 6).

Discussion

In this study, 32 (17%) of 191 unique lots of seven anticancer products collected in four countries in sub-Saharan Africa in 2023–24 did not contain the correct amount of active pharmaceutical ingredient. Substandard or falsified products were present in major cancer hospitals and in the private market in all four countries. Given the wide reach of global supply chains for pharmaceuticals, substandard products are probably also present in other LMICs, meaning that health-care professionals providing cancer care in these settings might be treating patients with drugs that are substandard or falsified. We suggest four main strategies to address this issue.

First, in locations where compendial testing is not yet available, hospitals and regulators must make decisions about what products to administer on the basis of visual inspection. Only three of the 32 products that failed the HPLC assay in this study could be identified by visual inspection. Increased use of chemical screening technologies such as the GPHF Minilab (Global Pharma Health Fund Minilab, Giessen, Germany),24 chemoPADs,25 and portable near-infrared or Raman spectrometers could identify larger numbers of substandard or falsified products.26 Some of these devices are not yet validated for use with chemotherapy drugs, and manufacturers will need to develop new methods, reference materials, and library entries. Hospitals and regulatory agencies should formulate policies for responding to anticancer products that fail screening tests, just as they already have policies for responding to products that fail visual inspection.

Second, increased post-market surveillance of anticancer products, careful supplier selection or prequalification, root cause investigations of products that fail compendial testing, swift and stringent regulatory actions enabled by laboratory data, and information sharing will all contribute to sustainable quality improvement.27 Poor access to compendial anticancer drug testing in LMICs must be addressed. In the short term, analysis could be provided out of country, but in the long term, regulatory and academic laboratories in LMICs need funding to train technicians, buy equipment and reference standards, and obtain chemo-rated personal protective equipment and hoods or laminar flow cabinets so that staff can work safely with these cytotoxic drugs.

Third, problems with anticancer medicine quality coexist at each site with stockouts, unsafe dispensing practices, and medicines that are unavailable or unaffordable for patients.28,29 Cost–benefit analysis of interventions that address all these problems should be performed to help policy makers and funders ensure the best results for patient outcomes with their available resources.

Finally, constructive engagement of care providers is needed to formulate site-specific response policies, develop messaging for patients, and engage regulators, donors, and other resources. Small changes could have substantial effects. For example, electronically capturing the brand names and lot numbers of products administered to patients could retrospectively identify products that are associated with reports of adverse effects or poor patient outcomes.11

Our findings shed light on the complexity of global supply chains and the factors that give rise to substandard or falsified anticancer drugs. The impacts of these substandard or falsified products on patient care outcomes need to be measured. Comprehensive, affordable, and innovative surveillance solutions could be a crucial step towards addressing existing gaps in medicine quality.

Additional sample collection and analysis will be needed to better understand the quality of anticancer medicines in LMIC settings. There are 73 antineoplastic and supportive care drugs in the WHO Model List of Essential Medicines, with hundreds of manufacturers and global scales of distribution, but this study included only seven active pharmaceutical ingredients collected from four countries in sub-Saharan Africa. We did not sample cold chain drugs that are sensitive to poor storage or shipping conditions; Barr and colleagues12 review serious quality problems that have been identified for one such sensitive anticancer drug, asparaginase. Our overt sampling methodology could also have biased sample collection away from products that sellers in the private sector thought were of bad quality. We also did not have much information about private sector shipping and storage conditions, which could have contributed to the degradation of products post manufacturing. We hope to collaborate with regulatory authorities in the future to extend the scope of this study and contribute to root cause investigations.

Supplementary Material

1
2

Research in context.

Evidence before this study

Information on the prevalence of substandard or falsified anticancer medicines is almost completely missing from medicine quality databases, primary research reports, and reviews about substandard or falsified pharmaceuticals. We reviewed the US Pharmacopoeia medicine quality database (which covers the 2003–17 period). This important resource summarises screening and compendial test results from regulatory agencies in multiple low-income and middle-income countries (LMICs) that participate in the Promoting the Quality of Medicines programme, including non-English speaking countries. 17 170 reports are indexed, but there were no results for any of the seven chemotherapy products we aimed to study (cisplatin, oxaliplatin, methotrexate, doxorubicin, cyclophosphamide, ifosfamide, and leucovorin). The Pharmaceutical Security Institute (PSI), an industry group made up of major pharmaceutical manufacturers, includes some reports of counterfeit cytostatic products, but their database does not include analytical results or brand, batch, or manufacturer information, and the manufacturers who belong to the PSI are not major suppliers of anticancer medicines in sub-Saharan Africa. In 2017, WHO published a structured literature review on the evidence basis for substandard or falsified pharmaceuticals, which included published papers and data from WHO’s rapid alert system (eg, medicine quality alerts). The published papers in the WHO report did not include any that performed high-performance liquid chromatography analysis of anticancer drugs. The WHO alerts were based on incidents in which patients were harmed by particular products or in which products failed visual inspection. We searched the PubMed and Web of Science databases on Sept 30, 2024, for articles and reports published after 2000. The search terms used were [substandard OR falsified OR counterfeit OR SF], either with [anticancer OR oncology OR chemotherapy] or with the names of the seven medicines we studied. We found several commentaries related to WHO or US Food and Drug Administration alerts about specific products such as Genentech’s Avastin (bevacizumab), a cluster of published papers about substandard or falsified asparaginase products discovered in Brazil around 2019 (these products were later found to be sold in over 100 LMICs), and two papers published by the Lieberman group that include data from a number of anticancer drug samples collected in Ethiopia in 2017–18. Media reports usually described police raids of illegal distributors, mentioned oncology products tangentially, and did not include data about the brands, lot numbers, chemical analysis, or quality of the products in question.

Added value of this study

This study is the first investigation of anticancer drug quality in sub-Saharan Africa with good geographical representation of the eastern (ie, Ethiopia and Kenya), southeastern (ie, Malawi), and western (ie, Cameroon) parts of Africa. In this study, active pharmaceutical ingredient contents for seven different anticancer products ranged from 28% to 120% of the stated contents, and failure rates ranged from 14% to 24% across the four countries sampled. Failing products were found in all four countries and in both major hospitals and private or community pharmacies (with no difference in failure rates at the 95% CI). The sensitivity of visual inspection for the detection of products failing high-pressure liquid chromatography assay was low at 9% (three of 32 lot numbers) because many quality defects, such as a shortage of an uncoloured active pharmaceutical ingredient, are not visually apparent. Since there is almost no previous evidence about the quality of anticancer products sold in sub-Saharan Africa, or in other LMICs, and since correct dosing of these drugs is crucial for good therapeutic outcomes, the added scientific value of this study is high.

Implications of all the available evidence

The confirmed presence of substandard or falsified anticancer medicines in sub-Saharan Africa will complicate cancer care, and the situation is very likely to be similar in other LMICs as the production lot sizes of the cancer products we investigated are much larger than the numbers bought by small countries. Future research building on this study should include studies designed to probe the market share and prevalence of substandard or falsified anticancer products in other regions, epidemiological modelling or retrospective case analysis to evaluate the effects on patient care outcomes, and economic modelling of the costs of substandard or falsified anticancer products. We hope the global research community can reveal the true quality of this class of medicines, as was done for antimalarials in the 2000s after Paul Newton published a series of papers about bad quality artesunate in southeastern Asia.

Acknowledgments

Research reported in this Article was supported by the US National Cancer Institute of the National Institutes of Health under award number U01CA269195. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Some of the HPLC instrumentation used was provided by the Center for Environmental Science and Technology at the University of Notre Dame. We would like to thank Alyssa Wicks for writing the R code for the dot plot in figure 2.

Funding

US National Cancer Institute of the National Institutes of Health.

Footnotes

Declaration of interests

ML serves on the board of Artifacts VERIFY (an unpaid position). ML also serves as a paid advisory council member for the Massachusetts Drug Supply Data Stream. BM serves as a consultant at Servier Pharmaceuticals. BM’s spouse is an employee and stockholder at Novartis Pharmaceuticals. All other authors declare no competing interests.

See Online for appendix 1

See Online for appendix 2

For more on the WHO essential medicines see https://list.essentialmeds.org/

Data sharing

Data collected for this study will be available after publication, including sample metadata (ie, active pharmaceutical ingredient, stated dose and dosage form, product name, manufacturer name, lot number, and expiration date), the complete methods, and analytical metrics for the high-performance liquid chromatography (HPLC) assay (ie, monograph methods, HPLC settings, system suitability test results, information about the reference materials, chromatograms and peak integration for all controls, calibration checks, experimental samples, and spreadsheets showing calculations of assay values from the raw data). Data are provided in the appendix. Readers who need access to the country data can contact the corresponding author, who will transmit the request to the project steering council; access will require the signing of a data use agreement.

References

  • 1.Venhuis BJ, Oostlander AE, Giorgio DD, Mosimann R, du Plessis I. Oncology drugs in the crosshairs of pharmaceutical crime. Lancet Oncol 2018; 19: e209–17. [DOI] [PubMed] [Google Scholar]
  • 2.Wondimagegnehu A, Negash Bereded F, Assefa M, et al. Burden of cancer and utilization of local surgical treatment services in rural hospitals of Ethiopia: a retrospective assessment from 2014 to 2019. Oncologist 2022; 27: e889–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Nyangasi MF, McLigeyo AA, Kariuki D, Mithe S, Orwa A, Mwenda V. Decentralizing cancer care in sub-Saharan Africa through an integrated regional cancer centre model: the case of Kenya. PLOS Glob Public Health 2023; 3: e0002402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Mapoko BSE, Frambo A, Saidu Y, et al. Assessment of barriers to optimal cancer control in adult cancer treatment centres in Cameroon. Ecancermedicalscience 2023; 17: 1601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ferlay J, Ervik M, Lam F, et al. Global Cancer Observatory: cancer today. International Agency for Research on Cancer, 2024. https://gco.iarc.who.int/today (accessed May 15, 2025).
  • 6.WHO. 4 substandard and falsified medical products: the causes. In: Pisani E, ed. WHO Global Surveillance and Monitoring System for substandard and falsified medical products. World Health Organization, 2017. [Google Scholar]
  • 7.Pisani E, Nistor AL, Hasnida A, Parmaksiz K, Xu J, Kok MO. Identifying market risk for substandard and falsified medicines: an analytic framework based on qualitative research in China, Indonesia, Turkey and Romania. Wellcome Open Res 2019; 4: 70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Mexican Secretary of Health. Results of the investigation of the case of Veracruz. Government of Mexico, 2017. https://www.gob.mx/salud/prensa/resultado-de-la-investigacion-del-caso-veracruz-96471?idiom=es (accessed Dec 17, 2017). [Google Scholar]
  • 9.Genetech. Genentech Statement on Counterfeit Drug Labeled as Avastin® (Bevacizumab) in the United States. Genetech, Feb 14, 2012. https://www.gene.com/media/statements/ps_021412 (accessed May 15, 2025). [Google Scholar]
  • 10.WHO. Medical product alert n°3/2017 falsified avastin (bevacizumab) and sutent (sunitinib malate) circulating in east Africa. World Health Organization, 2017. [Google Scholar]
  • 11.Toroitich AM, Armitage R, Tanna S. Suspected poor-quality medicines in Kenya: a retrospective descriptive study of medicine quality-related complaints reports in Kenya’s pharmacovigilance database. BMC Public Health 2024; 24: 2561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Barr RD, Furneaux R, Margottini L, Eden TOB. The international scandal of defective asparaginase: a blight on children with cancer. Pediatr Blood Cancer 2023; 70: e30403. [DOI] [PubMed] [Google Scholar]
  • 13.Michalowski MB, Cecconello DK, Lins MM, et al. Influence of different asparaginase formulations in the prognosis of children with acute lymphocytic leukaemia in Brazil: a multicentre, retrospective controlled study. Br J Haematol 2021; 194: 168–73. [DOI] [PubMed] [Google Scholar]
  • 14.Zenatti PP, Migita NA, Cury NM, et al. Low bioavailability and high immunogenicity of a new brand of E coli L-asparaginase with active host contaminating proteins. EBioMedicine 2018; 30: 158–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Sidhu J, Gogoi MP, Agarwal P, et al. Unsatisfactory quality of E coli asparaginase biogenerics in India: implications for clinical outcomes in acute lymphoblastic leukaemia. Pediatr Blood Cancer 2021; 68: e29046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Furneaux R, Margottini L. The drug was meant to save children’s lives. Instead, they’re dying. The Bureau of Investigative Journalism, Jan 25, 2023. https://www.thebureauinvestigates.com/stories/2023-01-25/the-drug-was-meant-to-save-childrens-lives-instead-theyre-dying/ (accessed Jan 14, 2024). [Google Scholar]
  • 17.Eberle MS, Ashenef A, Gerba h, Loehrer PJ Sr, Lieberman M. Substandard cisplatin found while screening the quality of anticancer drugs from Addis Ababa, Ethiopia. J Glob Oncol 2020; 6: 407–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.WHO. Medical product alert n°2/2019: falsified ICLUSIG traded globally. World Health Organization, 2019. [Google Scholar]
  • 19.US Pharmacopoeia. Medicines quality database. https://apps.usp.org/app/worldwide/medQualityDatabase/terms.html (accessed May 15, 2025).
  • 20.WHO. WHO Global Surveillance and Monitoring System for substandard and falsified medical products. World Health Organization, 2017. [Google Scholar]
  • 21.Kaur H, Allan EL, Mamadu I, et al. Quality of artemisinin-based combination formulations for malaria treatment: prevalence and risk factors for poor quality medicines in public facilities and private sector drug outlets in Enugu, Nigeria. PLoS One 2015; 10: e0125577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Schiavetti B, Wynendaele E, Melotte V, Van der Elst J, De Spiegeleer B, Ravinetto R. A simplified checklist for the visual inspection of finished pharmaceutical products: a way to empower frontline health workers in the fight against poor-quality medicines. J Pharm Policy Pract 2020; 13: 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.US Pharmacopoeia. (1225) Validation of compendial procedures. US Pharmacopeia, 2024. https://doi.usp.org/USPNF/USPNF_M99945_04_01.html (accessed May 15, 2025). [Google Scholar]
  • 24.Gnegel G, Häfele-Abah C, Neci R, Heide L, Difäm-EPN Minilab Network. Surveillance for substandard and falsified medicines by local faith-based organizations in 13 low- and middle-income countries using the GPHF Minilab. Sci Rep 2022; 12: 13095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Smith M, Ashenef A, Lieberman M. Paper analytic device to detect the presence of four chemotherapy drugs. J Glob Oncol 2018; 4: 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Caillet C, Vickers S, Zambrzycki S, et al. Multiphase evaluation of portable medicines quality screening devices. PLoS Negl Trop Dis 2021; 15: e0009287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Hamilton WL, Doyle C, Halliwell-Ewen M, Lambert G. Public health interventions to protect against falsified medicines: a systematic review of international, national and local policies. Health Policy Plan 2016; 31: 1448–66. [DOI] [PubMed] [Google Scholar]
  • 28.Stocker KJ, Tiemann A, Brunk KM, et al. Processes and perceptions of chemotherapy supply chain in Ethiopia: a mixed-method study. J Oncol Pharm Pract 2023; 29: 1555–64. [DOI] [PubMed] [Google Scholar]
  • 29.Fentie AM, Mekonen ZT, Gizachew Z, et al. Chemotherapy supply chain management, safe-handling and disposal in Ethiopia: the case of Tikur Anbessa specialized hospital. Pediatr Hematol Oncol 2023; 40: 258–66. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2093084-supplement-1.pdf (77.8MB, pdf)
2
NIHMS2093084-supplement-2.xlsx (196.6KB, xlsx)

Data Availability Statement

Data collected for this study will be available after publication, including sample metadata (ie, active pharmaceutical ingredient, stated dose and dosage form, product name, manufacturer name, lot number, and expiration date), the complete methods, and analytical metrics for the high-performance liquid chromatography (HPLC) assay (ie, monograph methods, HPLC settings, system suitability test results, information about the reference materials, chromatograms and peak integration for all controls, calibration checks, experimental samples, and spreadsheets showing calculations of assay values from the raw data). Data are provided in the appendix. Readers who need access to the country data can contact the corresponding author, who will transmit the request to the project steering council; access will require the signing of a data use agreement.

RESOURCES