A Decade of Decline: The Decreasing Use of Positive Controls Threatens the Reliability of In vitro Cancer Research

Abstract

Positive controls are indispensable for validating in vitro cancer bioassays and ensuring reproducibility in cancer drug discovery research. However, evidence suggests their inclusion in in vitro bioassays during drug discovery is declining, undermining assay reliability, reproducibility, and scientific integrity. This mini-review analysed 150 peer-reviewed studies published in 2015, 2020, and 2025 (50 from each year) to evaluate trends in positive control usage within cytotoxicity assays targeting cancer cell lines. Results indicate that only 52% (2015) and 54% (2020) of studies included appropriate positive controls; however, this dropped to a worrying 32% in 2025. The omission was most pronounced in natural product research and studies using high-throughput assays, such as MTT. In addition, reporting of dose-response parameters (e.g., IC₅₀) decreased from 26% to only 16% across the decade. These findings highlight a systemic erosion of methodological rigor, likely driven by publication pressure, limited resources, and a lack of standardised protocols. We recommend the mandatory inclusion of validated positive controls, transparent reporting of preparation and storage conditions, and journal-level enforcement of experimental standards. Reinforcing these practices will improve reproducibility, comparability, and ultimately the translatability of preclinical cancer research outcomes.

Keywords

assay standardisation - cancer - cytotoxicity - in vitro bioassays - preclinical validation - positive controls - reproducibility

Introduction

Cancer incidence is projected to rise substantially, particularly in lower-income countries, potentially reaching 32 million annual cases by 2050. Correspondingly, global cancer-related expenditure reached $223 billion in 2023 and is projected to exceed $400 billion by 2028 [1]. Within this, preclinical drug research and development (R&D) costs were estimated at approximately $18 billion between 2016 and 2020 [2]. However, true global preclinical R&D expenses are challenging to quantify, as this figure excludes researcher salaries, capital investments, and laboratory maintenance, likely amounting to tens of billions annually [3]. Drug R&D encompasses several stages, including target selection and validation, in vitro bioactivity screening of chemical entities, chemical optimisation of promising leads, and clinical trials (phases 1 – 4) [4]. Preclinical processes represent a substantial portion of total R&D budgets and are pivotal in developing new treatments [3], [5].

In cancer research, cytotoxicity assays are indispensable for high-throughput screening of potential anticancer agents, evaluating their impact on cell viability and informing preclinical safety and efficacy [6]. However, result reliability hinges on robust experimental design, particularly the inclusion of valid positive controls–known cytotoxic agents like doxorubicin, cisplatin, or 5-fluorouracil specific to applicable cancer cell lines to validate assay performance. Positive controls confirm that observed effects stem from the test substance, not procedural errors such as inconsistent pipetting or incubator temperature variations [7]. They are crucial for verifying system functionality amid cell culture variability, enabling interlaboratory comparisons, providing context for test substance activity, and validating dose-response curves [8]. Many journals reject manuscripts lacking positive controls [9], as Butterweck and Nahrstedt (2012) [10] noted: “If no positive control was used in a standard assay, it is impossible to interpret the resulting data with regard to their scientific content”.

[Fig. 1] provides various theoretical scenarios to illustrate how positive controls provide ʼcontextʼ when conducting preclinical drug discovery and development experiments using in vitro bioassays. At its core, the issue relies on a basic experimental principle: positive controls are essential for distinguishing between precision and accuracy. While increasing sample replicates leads to improved statistical scrutiny regarding consistency (precision), only a positive control verifies that results reflect the true biological effects (accuracy). Without this benchmark, even highly precise results may be systematically misleading and cause type I (false positive) or type II (false negative) errors due to assay failure, solvent interference, human error, etc., ultimately undermining the validity of scientific conclusions. [Fig. 1] also includes the US National Cancer Instituteʼs guidelines regarding IC₅₀ values for plant extracts that should be considered active or inactive [11]. Ideally, an accurate and precise result is required, which can only be achieved by including a proper number of sample replicates and positive controls. Failure to conduct a properly designed study is believed to play a role in the so-called ʼreproducibility crisisʼ, where it was found that the probability that a positive research finding is true is likely below 50% in many research fields [12].

Fig. 1 Theoretical scenarios are illustrated to indicate how positive controls provide ʼcontextʼ when conducting in vitro preclinical drug discovery experiments. While increasing sample replicates leads to improved statistical scrutiny regarding consistency (precision), only a positive control verifies that results reflect the true biological effects (accuracy). Also included are the US National Cancer Instituteʼs guidelines regarding IC₅₀ values for plant extracts that should be considered active or inactive.

A recent review of in vitro bioactivity studies on the ʼCancer Bushʼ (Sutherlandia frutescens (syn. Lessertia frutescens) (L.) R.Br. ex W. T.Aiton (Fabaceae)) found that only 5 of 15 studies (33%) used positive controls, rendering most results unreliable [10]. Considering the costs involved, the time-consuming nature of the drug discovery process, and importantly, the severity and impact of cancer on society, we found this to be unconscionable, which prompted us to investigate whether this failure to include positive controls is a widespread phenomenon or limited to research conducted on the Cancer Bush. This mini-review aims to survey recent publications on positive control usage in preclinical drug discovery. By promoting standardised positive control inclusion and detailed methodological reporting, we seek to elevate preclinical study quality and expedite effective therapy development.

Search Strategy and Selection Criteria

Using the search term “in vitro cytotoxic assay” on PubMed (15 May 2025), over 50 articles with free full-text links were identified. Each was evaluated against two inclusion criteria: (1) investigation of activity against cancer cell lines via in vitro bioassays and (2) assessment of bioactivity of any substances (synthetic molecules, natural products, extracts, or fractions), excluding established treatments. The 50 most recent qualifying publications were reviewed for positive control usage, cancer cell lines investigated, bioassay types, and related metadata. To identify chronological trends, the process was repeated for publications emanating from 2015 and 2020, yielding a total of 150 articles.

Results and Discussion

Tables 1S–3S (Supporting Information) provide comprehensive data on positive controls (including instances where none were used), reported IC₅₀ values for positive controls, cancer cell lines investigated, bioassay types, and compound classes studied. They also include the DOIs of assessed references and journal impact factors.

The data reveal a concerning trend: in the 2025 dataset, only 16 of 50 studies (32%) employed appropriate positive controls, such as standard chemotherapeutic drugs (Table 1S, Supporting Information). In contrast, the 2015 and 2020 datasets (Tables 2S and 3S, Supporting Information) showed positive control usage in 52% and 54% of studies, respectively. This represents a 38% and 41% decline in positive control utilisation in 2025 relative to 2015 and 2020, indicating a downward trajectory from roughly a half to a third over the past decade ([Fig. 2]). These findings align with our prior review of Cancer Bush (Sutherlandia frutescens) research, where only 5 of 15 studies (33.3%) incorporated proper positive controls [10]. Such inconsistencies undermine result trustworthiness and comparability, as positive controls are essential for detecting assay issues and ensuring interlaboratory reproducibility. Beyond the underutilisation of positive controls, several other trends emerged from the investigated studies.

Fig. 2 Over a decade, two key indicators of methodological rigour declined sharply: the proportion of studies using positive controls fell from 52% to 32%, and among those, IC₅₀ reporting dropped from 32% to 18%. Meanwhile, the mean journal impact factor rose from 2.78 to 4.59. Data from 150 peer-reviewed studies (n = 50/year).

Inappropriate selection of positive controls. Although 16 studies in the 2025 dataset reported using positive controls, four (articles 2, 7, 13, and 32; Table 1S, Supporting Information) employed unsuitable agents. Studies 2 and 7 claimed that positive controls were used, but used DMSO, which functions as a negative (vehicle) control or, at high concentrations, a ʼdeadʼ control to establish maximum cell death baselines. Study 13 utilised Triton X, similarly misclassified and better suited as a negative or dead control. Study 32 used an ethanolic propolis extract (Kleeva tinktura, Higytest) as a positive control, yet this acts as a comparator for assessing relative efficacy between substances. This mislabelling was absent in the 2015 and 2020 datasets (Tables 2S and 3S, Supporting Information), where positive controls were correctly identified.

Predominant use of broad-spectrum positive controls. The most common positive controls in the 16 studies from the 2025 dataset were broad-spectrum antineoplastic agents, rather than those specific to particular cancer types. These included cisplatin, doxorubicin, 5-fluorouracil, and paclitaxel, which are primarily employed as non-selective chemotherapeutic drugs. Of these 16 studies, 13 (81%) utilised one or more of these broad-spectrum agents. In the 2015 dataset, 13 of 22 studies with positive controls (59%) adopted similar broad-spectrum drugs. By 2020, this rose slightly, with 15 of 24 such studies (63%) doing so. This pattern indicates a growing reliance on less targeted controls over the decade, potentially reflecting a shift towards broader applicability in experimental design.

However, this trend raises concerns regarding drug selectivity and technical deficiencies of the MTT assay resulting in the elevated IC₅₀ values. Conversely, broad-spectrum agents may yield artificially inflated IC₅₀ values (less active), as their lack of specificity for the investigated cancer cell lines could make test substances appear comparably active – and thus more promising – than justified [13]. This could skew interpretations and limit insights into targeted therapies. Nonetheless, these drugsʼ widespread clinical use underscores their reliability and broad efficacy, making them suitable for general cytotoxicity assessments across diverse cell lines.

Omission of dose-response studies and IC₅₀ values. Another notable trend is the frequent omission of IC₅₀ values (derived from dose-response studies) for positive controls or test substances. This pattern persists across all reviewed years among studies that included positive controls, as follows: in 2015, 16 of 26 studies (62%) reported IC₅₀ values; in 2020, 12 of 27 (44%); and in 2025, 9 of 16 (56%). Overall, the proportion of studies incorporating both positive controls and IC₅₀ values decreased from 32% in 2015 to 24% in 2020 and 18% in 2025, indicating a declining trend of overall IC₅₀ reported values ([Fig. 2]) (Tables 1S–3S, Supporting Information). In studies lacking IC₅₀ values, results were often presented solely as percentage cell viability, which offers basic insights but lacks the precision needed for robust comparisons. Such approaches typically rely on simple graphs, limiting quantitative analysis.

Comparison of synthetic versus natural compounds. The 2015 dataset showed 22 studies (44%) were conducted on naturally derived products and 28 studies (56%) on synthetic compounds. By 2020, natural products were investigated in 42% of studies, and by 2025 they dominated at 72% (with synthetics at 28%). This shift reflects a growing emphasis on natural sources in drug discovery, necessitating aligned research protocols to maintain rigour [14], [15].

In the 26 studies that included positive controls in 2015, only 35% were conducted on natural products, and in the 27 studies in 2020 that included positive controls, 52% were conducted on natural products. Of the 16 studies that included positive controls in 2025, 69% were conducted on natural products, which indicates a complete shift in the percentage of positive controls used for natural products vs. synthetic research. This is a promising trend for natural product research, despite the declining use of positive controls in studies.

Inadequate methodological details, cell line usage patterns, and clinical relevance of positive controls. Many studies fail to provide detailed methodologies for the preparation and storage of test substances or positive controls, raising substantial concerns about reproducibility. In the 2025 dataset, 37 of 50 studies (Table 1S, Supporting Information) offered insufficient details. A substanceʼs solubility can profoundly affect assay outcomes, and without clear preparation methods, interpreting results becomes challenging. For example, doxorubicin is often used in a hydrochloride salt, liposomal, or non-liposomal form to improve water solubility and ionisation at specific pH levels [16], [17]; however, these critical specifics are frequently omitted, hindering result replication. Similarly, paclitaxelʼs limited water solubility leads to variable IC₅₀ values against identical cell lines [18], [19]. Excluding such information impedes follow-up and comparative studies.

Trends in cancer cell line usage. In 2015, MCF-7 (breast cancer) was the most common, appearing in 16 studies, followed by A549 (lung cancer) in 10, and HeLa (cervical cancer) in 8. In 2020, MCF-7 remained prominent in 22 studies, with A549 and MDA-MB-231 (breast cancer) in 9 studies each. By 2025, A549 led with 9 studies, MCF-7 in 7, MDA-MB-231 in 6, and HeLa in 5. The minimal use of Caco-2 (colorectal cancer) cell lines – only one study in 2025 – contrasts with the rising incidence of colorectal cancer, the fastest-growing cancer among those aged 50 or younger [20]. This suggests a potential lag in research focus despite epidemiological shifts. Breast cancer cell lines also showed a decline: MCF-7 and MDA-MB-231 combined appeared in 20 studies in 2015, in 31 in 2020, but only in 13 studies in 2025. This reduction is concerning, given the ongoing global rise in breast cancer incidence, underscoring the need for sustained research efforts. The enduring popularity of cell lines like A549, MCF-7, and HeLa stems from their status as immortalised cultures modelling prevalent cancers, along with their ease of proliferation, cost-effectiveness, accessibility, and suitability for diverse applications. Commercially available lines are typically more economical and faster-growing than patient- or animal-derived alternatives, which are costlier, slower growing, and non-immortalised [21], [22].

Evaluating these cell lines revealed inconsistencies in matching positive controls to clinical treatments. For A549 (non-small cell lung cancer, specifically adenocarcinoma), all five positive controls used were clinically validated for NSCLC, demonstrating appropriate selection. In contrast, for MCF-7 (luminal A subtype: ER+, PR+, HER2-, hormone-sensitive breast cancer), only 2 of 5 positive controls aligned with standard clinical options. For Caco-2 (colorectal cancer), one of two positive controls was oxaliplatin – a standard treatment – while the other was doxorubicin, which is not commonly used for this cancer. These examples illustrate varying diligence in positive control selection, with some studies exhibiting neglect of clinical relevance.

Additionally, studies employing multiple cell lines often treat positive controls as adjuncts rather than logically selected agents. For instance, article 34 in the 2025 dataset (Table 1S, Supporting Information) tested over 60 cell lines with just two positive controls (gefitinib and erlotinib). These tyrosine kinase inhibitors are suitable for NSCLC but become inadequate when applied broadly across diverse lines, effectively serving as comparators rather than true validators [23], [24]. Such approaches risk overstating similarities in cell viability or IC₅₀ values, emphasising the need for tailored, multiple controls in multi-line experiments to ensure relevance and accuracy.

Relationship between journal impact factors and positive control usage. Journal impact factors were evaluated as a proxy for study quality, with higher values generally indicating superior standards [25], [26]. In the 2025 dataset, the 50 studies appeared in 38 journals. For the 16 studies using positive controls, impact factors ranged from 0.8 to 11.3 (mean: 4.69). The remaining 34 studies had impact factors from 1.9 to 12.7 (mean: 4.55), showing a negligible difference between groups. Similar patterns emerged in 2015, where overall impact factors averaged 2.78 (range: 0.2 – 7.6). Studies with positive controls averaged 2.80 (range: 0.8 – 5.6), compared to 2.77 (range: 0.2 – 7.6) for those without. In contrast, the 2020 data revealed a counterintuitive trend, with impact factors ranging from 0.7 to 15.3. Studies employing positive controls averaged 2.96 (range: 0.7 – 6.2), while those omitting them averaged 4.3 (range: 0.7 – 15.3) with a statistically significant difference of − 1.34 (p = 0.042) ([Fig. 2]). This suggests that even high-impact journals frequently overlook positive controls, underscoring that impact factors alone are insufficient as quality metrics for scientific publications.

Inclusion of positive controls in society-related journals vs. commercial-based journals. A comparison of different journal types was essential for ensuring methodological rigour in our study and to investigate possible selection bias. To do this, we compared the studies in 2025 that were society-related vs. commercial-based, where society-related is defined as journals that are owned by or are the designated official publications of a learned society, professional association, or academic institution. The society-related journals yielded 16 studies, of which only 5 included positive controls (31.25%). By comparison, the remaining 34 studies published in commercial-based journals were found to have 11 uses of positive controls (32.35%). These findings are consistent with broader literature observations, showing that the quality of in vitro anticancer screening methods (such as including positive controls) is shaped more by field-specific standards and reviewer expectations than solely by the publisher model.

Assay selection and positive control incorporation. In the 2025 dataset, assay choice correlated with positive control usage. MTT assays featured in 28 of 50 studies, but only 11 (39%) included positive controls. In contrast, all three studies using sulphorhodamine B (SRB) assays incorporated positive controls. ATP CellTiter-Glo assays appeared in three studies, none with positive controls. Annexin V-FITC/PI assays showed 40% positive control inclusion, akin to MTT rates. This suggests a general disregard for positive controls, irrespective of assay simplicity or complexity. Similar patterns emerged in 2015: of 30 MTT studies, 13 (43%) used positive controls – comparable to 2025′s 39% and 2020′s 53%. SRB assays in seven studies had positive controls in five (71%), a 65% increase over MTT. No statistical difference existed in positive control usage between MTT (43%) and non-MTT (57%) studies.

In 2020, 18 of 34 MTT studies (53%) included positive controls. SRB assays, used twice, both incorporated them; though limited, this indicates a higher frequency versus MTT. MTT remained the dominant method, comprising 56% of studies in 2025, 60% in 2015, and 68% in 2020, due to its high-throughput, cost-effective, rapid nature and minimal expertise requirements compared to advanced techniques like Annexin V-FITC/PI [27], [28], [29].

Positive control suitability. In this study, we observed limited use of positive controls, even in high-impact factor journals. It is important to recognise that not all positive controls are equally appropriate for in vitro cytotoxic assays. Established chemotherapeutic agents serve as essential benchmarks, validating assay performance, ensuring reproducibility, and facilitating cross-study comparisons by confirming that the experimental system responds as expected to known cytotoxic compounds. Nevertheless, the choice of positive control can impact interpretation, particularly when compounds vary in potency due to experimental conditions.

Using appropriate controls in the study must be mechanistically or clinically appropriate. An example of this is the use of a platinum-resistant cancer cell lines in a study that includes a positive control with a platinum-derived chemotherapeutic drug (cisplatin, oxaliplatin, carboplatin, etc.). These studies often present a false comparison, as the tested treatment option can outperform the positive control and thus is not mechanistically efficient; in such cases, a non-platinum-derived drug must be used to fairly evaluate the assay.

Overall, cisplatin is used as a common reference standard in cytotoxicity assays. However, a 2023 study found that discrepancies in IC₅₀ values were observed across multiple cell lines when tested with cisplatin, often resulting in significant differences within the same cell lines [30]. This is not isolated, and non-platinum-derived drugs undergo the same problem when it comes to consistency. A study analysing doxifluridine showed that endpoint in vitro analysis techniques, such as WST-8 and MTT assays, significantly underreport IC₅₀ values, whereas more advanced techniques, such as real-time cycle analysis (RTCA), yield significantly lower IC₅₀ values [31]. Because endpoint analysis assays are preliminary, drugs that are specific for cell cycle inhibition, such as paclitaxel or docetaxel, will usually have higher IC₅₀ values in these endpoint assays.

Conclusions

The data indicate persistently low positive control usage in in vitro cancer bioassays over the past decade, with rates of 52% in 2015, 54% in 2020, and a concerning decline to 32% in 2025. This suggests that only about one-third of recent preclinical cancer studies include this essential validation, compromising result reliability, interlaboratory comparability, and overall research integrity. Such omissions exacerbate the reproducibility crisis in biomedical research, where variability in experimental design hinders the translation of findings into clinical applications [12], [32], [33].

Several factors likely contribute to this trend. Factors such as publication bias, where journals favour studies with positive or statistically significant outcomes, leading to an overrepresentation of such results in the literature [12], [34]. Between 1990 and 2007, the proportion of published positive results rose by 22%, reflecting systemic pressures that discourage reporting of null or negative findings [35], [36]. In this context, omitting positive controls may enable researchers to present unvalidated data as promising, avoiding comparisons that could reveal weak efficacy relative to established agents. Without controls, results are more susceptible to manipulation or misinterpretation, as there is no benchmark to confirm assay functionality or contextualise activity levels.

Less-intentional reasons include resource constraints: positive controls add direct costs (e.g., purchasing agents like doxorubicin) and indirect expenses (e.g., additional assays), which may deter underfunded labs. Availability issues, such as limited access to suitable controls for specific cell lines, or entrenched protocols from prior training that normalise their omission, further perpetuate the practice. Ignorance of best practices – stemming from inadequate education on experimental design – may also play a role, particularly in high-throughput settings where simplicity (e.g., MTT assays) prioritises speed over rigour. The “publish or perish” culture amplifies these issues, prioritising quantity over quality and fostering unethical shortcuts. Industry funding can also introduce biases, as seen in the 2012 curcumin fraud case, where manipulated data on its anticancer properties enriched promoters while wasting resources and distorting the field [37].

This is not an isolated or unsubstantiated problem; it reflects deeper epistemological weaknesses in bioactivity screening, as incisively critiqued in the review article, “How scientific is the science in ethnopharmacology? Historical perspectives and epistemological problems” [38]. Gertsch traces the evolution of ethnopharmacology from colonial bioprospecting to modern high-throughput assays, questioning the fieldʼs scientific rigour amid a surge in medicinal plant studies. He highlights that many in vitro assays, particularly those evaluating natural extracts, are conducted without sufficient validation or appropriate reference standards, yielding results that are difficult to compare, reproduce, or pharmacologically contextualise. This directly parallels the observed absence of positive controls in over two-thirds of recent cytotoxicity studies, which undermines the reliability of reported bioactivities and obscures true pharmacological relevance. Gertsch advocates for hypothesis-driven experimentation, anchored in proper controls, full dose-response evaluation, and in vivo correlation standards that are conspicuously absent in much of todayʼs preclinical cancer literature.

At its core, the issue relies on a basic experimental principle: positive controls are essential for distinguishing between precision and accuracy. While increasing sample replicates improves consistency (precision), only a positive control verifies that results truly reflect biological effects (accuracy). Without this benchmark, even highly reproducible data may be systematically misleading due to assay failure, solvent interference, or undetected variability, ultimately undermining the validity of scientific conclusions. Addressing this requires standardised methodologies, including the mandatory use of positive controls with detailed reporting of preparation (e.g., solubility and storage conditions). Journals should enforce these guidelines through pre-registration of protocols, incentives to publish negative results, and other measures to reduce bias and improve reproducibility. Ultimately, raising the use of positive controls to a non-negotiable standard will enhance the credibility of preclinical cancer research and accelerate therapeutic progress.

Data availability

Data are available on request from the corresponding author.

Contributorsʼ Statement

Conceptualisation, HvdB; FvdK; data curation, HvdB; FvdK; formal analysis, HvdB; FvdK; SvN; HM investigation, HvdB; FvdK methodology, HvdB; FvdK project administration, FvdK software, HvdB supervision, FvdK; SvN; HM validation, HvdB writing–original draft, HvdB writing–review and editing, FvdK; SvN; HM. All authors have read and agreed to the published version of the manuscript.

Conflict of Interest

The authors declare that they have no conflict of interest.

• References

• 1 IQVIA Institute for Human Data Science. Aitken M, Kleinrock M, Pritchett J. Global Oncology Trends 2024: Outlook to 2028. Accessed September 2, 2025 at: https://www.iqvia.com/insights/the-iqvia-institute/reports-and-publications/reports/global-oncology-trends-2024
• 2 McIntosh SA, Alam F, Adams L, Boon IS, Callaghan J, Conti I, Copson E, Carson V, Davidson M, Fitzgerald H, Gautam A, Jones CM, Kargbo S, Lakshmipathy G, Maguire H, McFerran K, Mirandari A, Moore N, Moore R, Murray A, Newman L, Robinson SD, Segaran A, Soong CN, Walker A, Wijayaweera K, Atun R, Cutress RI, Head MG. Global funding for cancer research between 2016 and 2020: A content analysis of public and philanthropic investments. Lancet Oncol 2023; 24: 636-645
• 3 Schlander M, Hernandez-Villafuerte K, Cheng CY, Mestre-Ferrandiz J, Baumann M. How much does it cost to research and develop a new drug? A Systematic Review and Assessment. Pharmacoecon 2021; 39: 1243-1269
• 4 Hoelder S, Clarke PA, Workman P. Discovery of small molecule cancer drugs: Successes, challenges and opportunities. Mol Oncol 2012; 6 (02) 155-176
• 5 Schuhmacher A, Gassmann O, Hinder M. Changing R&D models in research-based pharmaceutical companies. J Transl Med 2016; 14: 105
• 6 Lee OW, Austin S, Gamma M, Cheff DM, Lee TD, Wilson KM, Johnson J, Travers J, Braisted JC, Guha R, Klumpp-Thomas C, Shen M, Hall MD. Cytotoxic profiling of annotated and diverse chemical libraries using quantitative High-Throughput Screening. SLAS Discov 2020; 25: 9-20
• 7 Riss TL, Moravec RA, Niles AL, Duellman S, Benink HA, Worzella TJ, Minor L. Cell Viability Assays. In: Markossian S, Grossman A, Baskir H, Arkin M, Auld D, Austin C, Baell J, Brimacombe K, Chung TDY, Coussens NP, Dahlin JL, Devanarayan V, Foley TL, Glicksman M, Gorshkov L, Grotegut S, Hall MD, Hoare S, Inglese J, Iversen PW, Lal-Nag M, Li Z, Manro JR, McGee J, Norvil A, Pearson M, Riss T, Saradjian P, Sittampalam GS, Tarselli MA, Trask jr. OJ, Weidner JR, Wildey MJ, Wilson K, Xia M. and Xu. eds. Assay Guidance Manual [Internet]. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2013. Available at https://www.ncbi.nlm.nih.gov/books/NBK144065/ Accessed September 4, 2025
• 8 Petersen EJ, Nguyen A, Brown J, Elliott JT, Clippinger A, Gordon J, Kleinstreuer N, Roesslein M. Characteristics to consider when selecting a positive control material for an in vitro assay. ALTEX 2021; 38: 365-376
• 9 Gernandt N, van der Kooy F. The “Cancer Bush” Sutherlandia (syn. Lessertia) frutescens. An example of promotional research, or is there scientific merit?. Planta Med 2025; 91: 730-744
• 10 Butterweck V, Nahrstedt A. What is the best strategy for preclinical testing of botanicals? A critical perspective. Planta Med 2012; 78: 747-754
• 11 Vogel D, Gouws C, van der Kooy F. Antineoplastic activity of selected cytotoxic plants from the Kalahari. Rev Bras Farmacogn 2024; 34: 899-909
• 12 Ioannidis JPA. Why most published research findings are false. PLoS Med 2005; 2: e124
• 13 He Y, Zhu Q, Chen M, Huang Q, Wang W, Li Q, Huang Y, Di W. The changing 50 % inhibitory concentration (IC₅₀) of cisplatin: a pilot study on the artifacts of the MTT assay and the precise measurement of density-dependent chemoresistance in ovarian cancer. Oncotarget 2016; 7: 70803-70821
• 14 Newman DJ, Cragg GM. Natural products as sources of new drugs over the nearly four decades from 01/1981 to 09/2019. J Nat Prod 2020; 83: 770-803
• 15 Cordell GA. Evolving paradigms for natural-product drug discovery. Natl Sci Rev 2022; 9: nwac075
• 16 Gabizon A, Martin F. Polyethylene glycol-coated (pegylated) liposomal doxorubicin: Rationale for use in solid tumours. Drugs 1997; 54: 15-21
• 17 Alley MC, Scudiero DA, Monks A, Hursey ML, Czerwinski MJ, Fine DL, Abbott BJ, Mayo JG, Shoemaker RH, Boyd MR. Feasibility of drug screening with panels of human tumor cell lines using a microculture tetrazolium assay. Cancer Res 1988; 48: 589-601
• 18 Rowinsky EK, Donehower RC. Paclitaxel (Taxol). N Engl J Med 1995; 332: 1004-1014
• 19 Sparreboom A, van Zuylen L, Brouwer E, Loos WJ, de Bruijn P, Gelderblom H, Pillay M, Nooter K, Stoter G, Verweij J. Cremophor EL–mediated alteration of paclitaxel distribution in human blood: Clinical pharmacokinetic implications. Cancer Res 1996; 56: 2112-2115
• 20 Berrington de Gonzalez A, Brayley M, Frost R, Freedman N, Gunter MJ, Jackson I, Lapitan P, Shiels MS, García-Closas M. Trends in cancer incidence in younger and older adults: An international comparative analysis. Ann Intern Med 2025; 178: 1677-1687 Accessed September 5, 2025 at: https://www.acpjournals.org/doi/10.7326/ANNALS-24–02718
• 21 Masters JR. Human cancer cell lines: Fact and fantasy. Nat Rev Mol Cell Biol 2000; 1: 233-236
• 22 Freshney RI. Culture of Animal Cells: A Manual of Basic and Specialized Techniques. 7th ed. Hoboken, NJ: Wiley-Blackwell; 2016
• 23 Zhang J, Yang PL, Gray NS. Targeting cancer with small molecule kinase inhibitors. Nat Rev Cancer 2009; 9: 28-39
• 24 Gazdar AF, Girard L, Lockwood WW. Lung cancer cell lines as tools for biomedical discovery and research. J Natl Cancer Inst 2010; 102: 1310-1321
• 25 Seglen PO. Why the impact factor of journals should not be used for evaluating research. BMJ 1997; 314: 498-502
• 26 Garfield E. The history and meaning of the journal impact factor. JAMA 2006; 295: 90-93
• 27 Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays. J Immunol Methods 1983; 65: 55-63
• 28 van Meerloo J, Kaspers GJL, Cloos J. Cell sensitivity assays: The MTT assay. Methods Mol Biol 2011; 731: 237-245
• 29 Telford WG. Multiparametric analysis of apoptosis by Flow Cytometry. Methods Mol Biol 2018; 1678: 167-202
• 30 Ćwiklińska-Jurkowska M, Wiese-Szadkowska M, Janciauskiene S, Paprocka R. Disparities in cisplatin-induced cytotoxicity-A meta-analysis of selected cancer cell lines. Molecules 2023; 28: 5761
• 31 Hazekawa M, Nishinakagawa T, Kawakubo-Yasukochi T, Nakashima M. Evaluation of IC50 levels immediately after treatment with anticancer reagents using a real-time cell monitoring device. Exp Ther Med 2019; 18: 3197-3205
• 32 Begley CG, Ellis LM. Raise standards for preclinical cancer research. Nature 2012; 483: 531-533
• 33 Baker M. 1, 500 scientists lift the lid on reproducibility. Nature 2016; 533: 452-454
• 34 Dickersin K. The existence of publication bias and risk factors for its occurrence. JAMA 1990; 263: 1385-1389
• 35 Duyx B, Urlings MJ, Swaen GM, Bouter LM, Zeegers MP. Scientific citations favor positive results: A systematic review and meta-analysis. J Clin Epidemiol 2017; 88: 92-101
• 36 Mlinarić A, Horvat M, Smolčić VŠ. Dealing with the positive publication bias: Why you should really publish your negative results. Biochem Med 2017; 27: 030201
• 37 Harvey L. Research fraud: a long-term problem exacerbated by the clamour for research grants. Quality in Higher Education 2020; 26: 243-261
• 38 Gertsch J. How scientific is the science in ethnopharmacology? Historical perspectives and epistemological problems. J Ethnopharmacol 2009; 122: 177-183

Publication History

Received: 19 January 2026

Accepted after revision: 17 February 2026

Article published online:
02 March 2026