Pharmaceutical and Biomedical Sciences Journal Volume 7.
, 2025, 192-201
________________________________________________________________
QSAR Analysis and ADMET Prediction of Tamoxifen Derivatives Using LFER Hansch Model Andzar Fikranus Shofa*.
Rawwdha Luthfi Nur Muhammad.
Norma Adityaningsih Prameswari.
Sendi Handika Putra.
Aisyah Afifah Setiawan.
Dhia Salsabila Rizki Departement of Pharmacy.
Faculty of Health and Sciences.
Syarif Hidayatullah State Islamic University.
South Tanggerang.
Indonesia 15412 *Corresponding author: andzar@apps.
Received: 02 December 2025.
Accepted 26 December 2025 Abstract: Tamoxifen is a pharmaceutical compound that can be widely used in the therapeutic regimen for breast cancer, particularly for postmenopausal women.
Nevertheless, this clinical efficacy is frequently diminished and limited due to the emergence of drug resistance and heterogeneous therapeutic impacts with different degrees of This has led to the discovery of structurally connected compounds that are more effective than tamoxifen.
The current research study describes QSAR modeling and predicting the ADMET parameters of tamoxifen derivatives for the treatment of breast cancer.
SPSS software was used for analysis of multiple linear regression and found the pIC50 = -0.
059CLogP 0.
759LUMO - 0.
011MR 3.
Model validation yielded R = 0.
= 0.
848 and QA = 0.
651, which suggests high predictability.
The most important characteristic being LUMO energy, the second most important descriptor was followed by CLogP and MR.
The ADMET prediction showed high intestinal absorption values (HIA > 90%) and satisfactory permeability over skin.
Water solubility was impaired but also low.
The metabolism of compounds seemed to predominantly occur via CYP3A4 enzyme.
However.
LD50 values were of acceptable size, ranging from 324 to 3000 mg/kg .
ithin the safety profil.
The findings of this work will thus show help in designing anticancer tamoxifen derivative products with the least Additional structural optimization is advocated to achieve maximal therapeutic benefit and least toxicity.
Keywords: ADMET, breast cancer.
drug design.
QSAR.
Tamoxifen DOI: https://doi.
org/10.
15408/pbsj.
INTRODUCTION
Breast cancer remains one of the most prevalent malignancies affecting women globally, accounting for a significant proportion of cancer-related morbidity and mortality (Mulyani & Nasution, 2.
Despite advancements in early detection and therapeutic interventions, resistance to conventional treatment options, including chemotherapy and hormone therapy, continues to pose a critical challenge (Chen et al.
, 2.
Among the therapeutic agents used for breast cancer.
Tamoxifen, a selective estrogen receptor modulator (SERM), has been extensively employed due to its ability to modulate estrogen receptor activity (Shagufta & Ahmad, 2.
However, variations in individual metabolic response to Tamoxifen, primarily influenced by the CYP2D6 enzyme, result in diverse therapeutic outcomes (Chen et al.
, 2.
The need to optimize the therapeutic efficacy of Tamoxifen has driven researchers to explore its structural derivatives, aiming to enhance anti-cancer potency while minimizing adverse effects.
Quantitative Structure-Activity Relationship (QSAR) is considered to be a key method in predicting chemical compounds biological activities from molecular descriptors (Zekri et al.
, 2.
The computational approach can be used to predict potential drug candidates initially at the early stage of drug discovery, thus minimizing the time- and resource-consuming experimental validation process (Lipinski, 2.
QSAR is used for screening Tamoxifen and its derivatives to elucidate correlation between chemical structure and biological activity and guide synthesis of more effective analogs (Shagufta & Ahmad, 2.
However, the accuracy of the QSAR models is affected by the quality and size of training data and complexity of biological endpoints, which may result in conflicting predictions 192 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
from different QSAR methods (Zekri et al.
, 2.
As a solution to these limitations, several validation strategies such as internal validation, external validation, and Y-randomization are suggested to make the QSAR models more robust and powerful for predictive purposes (Lipinski, 2.
Although QSAR models can be of great importance, they do face many challenges like the inherent variation of biological systems, especially in predicting complex endpoints like pharmacokineticstoxicity, and offer a comprehensive strategy to evaluate the safety and efficacy of Tamoxifen derivatives (Zekri et al.
, 2.
As we discussed earlier.
QSAR was utilized for dose optimization for Tamoxifen and its analogs .
uch as Z-endoxife.
, a metabolite of this compound, which also plays a key role in therapeutics (Chen et al.
, 2.
It has been found that the efficacy of Tamoxifen may be highly dependent on the patient's genetic polymorphism in CYP2D6, and treatment must therefore be adjusted according to dose (Shagufta & Ahmad, 2.
Moreover, the hepatotoxic potential and mutagenicity of Tamoxifen derivatives require meticulous characterization of their ADMET (Absorption.
Distribution.
Metabolism.
Excretion, and Toxicit.
profiles, which is where QSAR modeling is useful (Lipinski.
Previous studies have explored the use of various Tamoxifen derivatives, including Raloxifene.
Droloxifene, and Idoxifene, which have shown potential anti-cancer activities with different pharmacological profiles (Shagufta & Ahmad, 2.
These derivatives are often evaluated using QSAR techniques to understand their interaction with estrogen receptors and predict their pharmacokinetic behavior.
However, conflicting results regarding their efficacy and toxicity profiles persist, primarily due to differences in experimental conditions, computational methods, and sample sizes (Lipinski, 2.
Addressing these discrepancies requires further research, especially in validating QSAR models across diverse datasets and experimental frameworks to achieve more consistent and reliable predictions (Zekri et al.
, 2.
Given these research gaps, this study aims to further explore the potential of QSAR modeling in predicting the pharmacokinetic and toxicity profiles of Tamoxifen derivatives.
By analyzing a broader set of molecular descriptors and integrating comprehensive ADMET predictions, this research seeks to identify novel derivatives with enhanced therapeutic potential and minimized adverse effects.
The novelty of this study lies in its approach to combine computational modeling with extensive dataset validation, aiming to refine the predictive power of QSAR models and contribute to the advancement of precision medicine in breast cancer treatment to provide a more accurate, efficient, and personalized approach to breast cancer therapy through the development of optimized Tamoxifen analogs.
MATERIAL AND METHODS
1 Materials Table 1.
Chemical Structures and ICCICA Values of Tamoxifen and Its Derivatives Againts Breast Cancer
Compoud
Tamoxifen
Chemical Formula
C26H29NO
IC50 (AAM)
Raloxifene C28H27NO4S
Tamoxifene A C23H23NO
Tamoxifene B
C27H29NO3
Idoxifene C28H30INO
Toremifene C26H28ClNO
Ospemifene C24H23ClO2
4-hydroxy tamoxifene C26H29NO2
Lasofoxifene C28H31NO2
Endoxifen
C25H27NO2
The research utilized the following tools and materials: a laptop with an Asus brand, a Windows 10 operating system, a 64-bit AMD Quad Core A8-7410 processor, a CPU @ 2.
5 GHz, and 4.
00 GB of 193 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
RAM.
The software employed included Chem Draw Ultra 12.
0 and Chem Bio 3D Ultra (CambridgeSof.
, as well as IBM SPSS Statistics 27.
The data on tamoxifen-derived compounds and the IC50 values of breast anticancer activity were obtained from research conducted by Brauch and Jordan .
, as shown in Table 1 .
2 Methods Research Design The study performs a computational and in silico study to characterize structure-activity relationships and pharmacokinetic properties of tamoxifen derivatives.
Computational methods are favored due to their cost-efficiency, reduced experimental hazards, and the ability to screen multiple compounds simultaneously (Zekri et al.
, 2.
This design is exploratory in nature by correlating molecular descriptors to biological activities.
The QSAR analysis follows a supervised learning approach through multilinear regression (MLR) which is suitable for the analysis of complex data involving multiple predictors (Frimayanti et al.
, 2.
ADMET predictions necessary to understand drug-like properties were performed with pkCSM and Protox i and were accessible through their respective online platforms (Pires et al.
, 2.
Data Collection The first phase of the study encompassed collecting applicable chemical and biological information from PubChem, a trusted, open-access chemical database.
A text notation of chemical structures known as SMILES was obtained for each tamoxifen derivative.
These SMILES notations were entered into ChemDraw and Chem3D to compute molecular descriptors: lipophilicity (CLogP), electronic properties (LUMO and HOMO), steric parameters .
olecular refractivity.
MR), polar surface area (TPSA).
Furthermore, pkCSM was also used for ADMET property prediction .
bsorption, distribution, metabolism, excretion, and toxicit.
(Pires et al.
, 2.
The retrieved data were collated into organized datasets for statistical analysis to ensure data integrity and completeness.
Descriptors were standardized and normalized to mitigate bias caused by differences in scale in order to enhance reliability of the data.
Data preprocessing also took into account missing values, the detection of outliers, and the consistency of descriptor calculation.
The descriptors with high multicollinearity (VIF > .
were removed in order to avoid redundancy and improve interpretation of models (Nystrom & Sanchez, 2.
QSAR Model Development The QSAR modeling process follows these steps: .
data acquisition and preprocessing, .
descriptor calculation using ChemDraw and Chem3D, .
statistical analysis using SPSS, and .
model validation through internal and external techniques.
Internal validation employs leave-one-out (LOO) crossvalidation, while external validation uses independent datasets (Frimayanti et al.
, 2011.
Putra et al.
The evaluation parameters include correlation coefficient (R), determination coefficient (RA), and cross-validated RA (QA), adhering to the threshold values specified in previous QSAR studies.
ADMET Predictions ADMET predictions are critical for assessing the pharmacokinetic properties of the compounds.
Absorption metrics include solubility.
Caco-2 permeability, human intestinal absorption (HIA), and skin permeability.
Distribution is assessed via VDss .
olume of distributio.
and b .
lood-brain barrie.
Metabolism focuses on enzyme interaction, specifically with CYP450 enzymes (CYP2D6 and CYP3A.
Excretion evaluates total clearance and renal OCT2 substrates, while toxicity considers hepatotoxicity and carcinogenicity using Protox i (Pires et al.
, 2.
Statistical Analysis Statistical analysis was performed to measure the relationships between the molecular characteristics and biological activity of tamoxifen derivatives in this study.
Multilinear regression (MLR) and Pearson's correlation were used to conduct this analysis .
sing IBM SPSS Statistics .
MLR
194 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
performed well in estimating how multiple independent variables contributed to a dependent variable, allowing a comprehensive understanding of how molecular descriptors influence biological activity (Nystrom & Sanchez, 2.
To examine the strength and direction of the relationship across molecular descriptors.
Pearson's correlation was employed.
A correlation matrix was created to identify the linear relationships among the variables, ensuring that multicollinearity did not influence the regression The degree of multicollinearity was determined with the variance inflation factor (VIF).
a VIF value above 5 indicates multicollinearity issues that potentially disrupt the model's accuracy (Zekri et , 2.
The significance parameters of predictor variables were evaluated by p-values (<0.
indicating that only significant descriptors were included.
Using coefficient of determination (RA) and adjusted RA to model multiple predictors, goodness of fit of the regression model was analyzed.
Finally, standardized coefficients () were examined in order to understand the relative importance of each descriptor in predicting biological activity.
Residual analysis including inspection of standardized residuals and Cook's distance was performed for potential outliers or influential data points.
Model Validation Model validation is crucial to ensure the credibility of QSAR predictions.
Internal validation through LOO cross-validation was used to evaluate the model's stability.
Additionally, external validation was performed using independent datasets to verify the model's predictability.
The criteria for model acceptability included RA > 0.
QA > 0.
5, and F-statistics surpassing the threshold (Hadaji et al.
, 2.
RESULTS AND DISCUSSION
The QSAR analysis was performed on ten tamoxifen derivatives to establish the correlation between molecular descriptors and anticancer activity against breast cancer.
The QSAR model was constructed using pIC50 as the dependent variable, while descriptors such as lipophilicity (CLogP.
LogS), electronic parameters (HOMO.
LUMO, pKa, tPSA), and steric factors (MR.
MW.
MV) were considered as independent variables (Table .
Table 2.
Molecular Descriptors of Lipophilic.
Electronic, and Steric Properties of Selected Tamoxifen Derivatives Compoud
pIC50
CLogP
LogS
HOMO
LUMO
tPSA Tamoxifen Raloxifene Tamoxifene A Tamoxifene B Idoxifene Toremifene Ospemifene 4-hydroxy Lasofoxifene Endoxifen To refine the QSAR model, a Pearson correlation analysis was performed (Table .
A significant positive correlation between LUMO and pIC50 .
= 0.
indicates that molecules with higher LUMO energy exhibit increased anticancer potency.
Conversely.
CLogP exhibited a weak negative correlation .
= -0.
, suggesting that excessive lipophilicity could reduce biological activity.
The MR descriptor also showed a minor positive correlation .
= 0.
, indicating its minor influence on potency.
Among the molecular descriptors.
LUMO (Lowest Unoccupied Molecular Orbita.
demonstrates the strongest positive correlation .
with pIC50, emphasizing its significant role in predicting anticancer activity.
A higher LUMO value suggests increased electron affinity, enhancing the potential for nucleophilic attacks, which is critical for interaction with estrogen receptors (Zhang et al.
, 2.
This supports the hypothesis that tamoxifen derivatives with enhanced LUMO energy may achieve stronger binding to receptor sites, thus exhibiting superior anticancer activity (Chen et al.
, 2.
195 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
Table 3.
Pearson Correlation Matrix between pICCICA and Selected Physicochemical and Electronic Descriptors of Tamoxifen Derivatives.
QSAR Descriptor
pIC50
CLogP
LogS
HOMO
LUMO
tPSA
pIC50
CLogP
LogS
HOMO
LUMO
tPSA Conversely, the negative correlation of CLogP (-0.
suggests that excessive lipophilicity may compromise drug bioavailability and increase toxicity risks.
An optimal balance of hydrophilicity and lipophilicity is crucial for effective membrane permeability and pharmacokinetic performance (Megawati et al.
, 2.
These findings are consistent with previous research indicating that overlipophilicity often leads to poor pharmacokinetics and off-target interactions, potentially elevating toxicity (Mulyani & Nasution, 2.
Additionally.
MR (Molar Refractivit.
demonstrates a weaker correlation .
, indicating a less prominent impact of steric hindrance in influencing biological However.
MR's contribution cannot be disregarded, as slight modifications in steric properties can still impact receptor binding and overall molecular stability (Putra et al.
, 2.
The QSAR model was validated using the Leave-One-Out (LOO) cross-validation technique, yielding a QA value of 0.
651, meeting the acceptance threshold (QA > 0.
Additionally, the statistical significance of the model was verified with an F-value of 11.
137, surpassing the F-critical value of 4.
at a 99% confidence level.
The strong correlation and predictive reliability of the model confirm its applicability in designing more effective tamoxifen derivatives.
The validated QSAR equation obtained was:
pIC50 = -0.
059CLogP 0.
759LUMO - 0.
011MR 3.
n = 10, r = 0.
921, ycI2 = 0.
F= 11.
137, ycE2 = 0.
The validation through the Leave-One-Out (LOO) cross-validation method, yielding QA = 0.
demonstrates the model's reliability while acknowledging its limitations.
The moderate QA value implies that other influential factors, potentially unaddressed molecular descriptors or complex biological interactions, may contribute to the variations in anticancer activity.
These findings align with existing literature highlighting the intricate nature of QSAR modeling for complex biological responses (Frimayanti et al.
, 2011.
Male et al.
, 2.
Future studies should consider integrating additional descriptors, such as quantum chemical or three-dimensional structural descriptors, to refine predictive The model showed strong statistical validation with n = 10, r = 0.
RA = 0.
F = 11.
137, and QA = 651, indicating high predictive power and stability.
Among the descriptors.
LUMO energy demonstrated the highest correlation with pIC50 .
= 0.
, followed by CLogP (-0.
and MR .
indicates the complex interplay between lipophilicity (CLogP), electronic properties (LUMO), and steric factors (MR) in determining biological activity.
The correlation coefficient (RA = 0.
suggests a robust model, capable of predicting the anticancer efficacy of tamoxifen derivatives with considerable accuracy (Arba et al.
, 2.
These findings suggest that increasing LUMO energy while reducing CLogP and MR values may enhance anticancer activity.
196 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
The LUMO energy has the largest positive coefficient, indicating that the electronic features of the molecules are central in determining inhibitory potency.
Higher LUMO energy, reflecting a lower tendency of the molecule to accept electrons, may lead to enhanced specificity and stability in the ligandAereceptor interactions, especially in targets where charge transfer or electrostatic complementarity is crucial.
This is in line with earlier QSAR and quantum chemical analyses indicating that the frontier molecular orbitals, particularly LUMO, are a major determinant in the ligand reactivity and binding affinity in the enzymeAeinhibitor and receptorAeligand systems (Parr & Yang, 1989.
Cherkasov et al.
, 2.
The CLogP reflects the lipophilicity and enters activity negatively, as evidenced from the negative regression coefficient.
Lipophilicity is traditionally a major driver of membrane permeability and target however, excessive hydrophobicity can be deleterious by promoting nonspecific interactions with either biological membranes or plasma proteins, ultimately decreasing the effective concentration of free ligand at the target site.
This negative influence of CLogP in this model suggests that a ligand with moderate or low hydrophobic character may be preferred in the binding site of the biological target, possibly because of the presence of polar or charged amino acid residues within the binding pocket.
This behavior is in good agreement with the classical Hansch theory, which advocates an optimum lipophilicity window beyond which the biological activity decreases (Hansch & Fujita.
Kubinyi, 1.
The negative contribution to activity from the steric and polarizability effects is captured by the MR descriptor with a relatively small coefficient.
MR reflects both molecular volume and electronic thus, bulkier and more polarizable molecules tend to show slightly reduced biological This may imply that steric hindrance could restrict the ability of larger ligands to assume an optimal binding conformation within the active site, assuming that the pocket of the target is spatially Similar observations have been reported in QSAR studies where increases in molecular size adversely affected binding efficiency due to unfavorable steric clashes or reduced conformational flexibility (Hansch et al.
, 1995.
Todeschini and Consonni, 2.
Taking into consideration this negative MR contribution supports the idea of compact molecular frameworks with controlled polarizability being preferable for achieving higher potency within this ligand series (OECD, 2014.
Roy et al.
, 2.
The ten tamoxifen derivatives studied based on ADMET properties: absorption, distribution, metabolism, excretion and toxicity have the best absorption of ten tamoxifen derivatives.
Most of the compounds had good skin permeability (Log KP > -2.
, high intestinal absorption (HIA > 90%) and good skin permeability based on absorption analysis.
Meanwhile all substances exhibited a low solubility aqueously, but the better solubility was achieved by raloxifene (-3.
716 log mol/L).
Most of the included compounds also showed high Human Intestinal Absorption (HIA > 90%), suggesting good absorption and possibility of effective oral bioavailability.
Still, poor aqueous solubility continues as an important barrier, which may block systemic circulation and requirement for modifications for formulations (Pires et al.
, 2.
Notwithstanding effective skin penetration (Log KP > Oe2.
transdermal administration might not be the ideal method for the systemic anticancer therapy (Nystrom & Sanchez, 2.
The Caco-2 permeability values varied greatly across the derivatives, with ospemifene providing the highest permeability .
178 x 10AA cm/.
The derivatives identified as Pglycoprotein (P-g.
substrates and inhibitors and multidrug resistance mechanisms.
Distribution of VDss varied widely with idoxifene with the largest VDss .
056 log L/k.
, confirming wide tissue The b permeability .
og BB) ranged from -1.
indicating that certain derivatives were sufficient to cross the blood-brain barrier, as per the latter of which the CNS may take influence.
Yet.
CNS permeability .
og PS) values were all < -2, suggesting that none of the compounds can penetrate the CNS effectively.
All derivatives were metabolized by CYP3A4, and most were CYP1A2 inhibitors, suggesting possible drug-drug interactions.
Interestingly, raloxifene and ospemifene were not CYP2D6 inhibitors, which suggests a lower likely effect of CYP2D6-metabolized medications on CYP2D6-metabolized drug.
The total values of clearance ranged 192 .
81 log ml/min/kg .
, suggesting differential elimination rates.
None of these compounds were kidney OCT2 substrates, indicating that they showed a slight interaction 197 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
with renal excretion.
Metabolically, most tamoxifen derivatives serve as substrates for CYP3A4, underlining the enzyme's pivotal role in their metabolic clearance.
Notably.
Raloxifen and Ospemifen, which do not inhibit CYP2D6, present a reduced risk of adverse drug interactions, highlighting their potential for safer use in polypharmacy settings (Chen et al.
, 2.
These findings are consistent with prior studies that emphasize the role of CYP3A4 and CYP2D6 in the metabolism of selective estrogen receptor modulators (SERM.
(Brysen & Naranjo, 2.
The varying clearance rates, particularly Lasofoxifen's high clearance .
81 log ml/min/k.
, imply a narrower therapeutic window, necessitating careful therapeutic monitoring and dosage adjustments (Krihariyani et al.
, 2.
Toxicity predictions revealed that all derivatives were non-carcinogenic, but some exhibited hepatotoxicity .
amoxifen B, toremifene, 4-hydroxy tamoxifen, and tamoxifen A).
The LD50 values ranged from 324 mg/kg .
to 3000 mg/kg .
amoxifen B), indicating a generally acceptable safety profile.
Most compounds fell within toxicity level IV, with tamoxifen B categorized as level V, suggesting the highest safety margin.
Toxicological assessments indicate minimal carcinogenic risks across the compounds, aligning with their current clinical applications.
However, hepatotoxicity observed in Toremifen and 4-hydroxy tamoxifen necessitates hepatic function monitoring during therapeutic use.
The wide LD50 range .
Ae3000 mg/k.
implies an overall acceptable safety profile, yet further in vivo validation is essential for clinical translation (Banerjee et al.
, 2.
The identification of hepatotoxic risk factors also encourages structural modifications to reduce off-target toxicity while maintaining therapeutic efficacy (Widiyanti et al.
, 2.
Table 4.
In Silico Prediction of Absorption-Distribution Profiles for Tamoxifen and Its Derivatives Absorption Water Solubi .
og Caco2 Permea .
og Papp in cm/.
Tamoxif Raloxife Distribution Skin Permea (Log KP) P-gp P-gp I Inhibi P-gp Inhibi .
og L/k Unbo Fracti .
b Permea .
og BB) CNS Permea .
og PS) Ospemif Lasofoxi Endoxif Compo Tamoxif Tamoxif Idoxifen Toremif The rational design of next-generation tamoxifen analogs necessitates an early, integrated ADMET strategy to balance absorption, distribution, metabolism, excretion, and toxicity.
This strategy involves fine-tuning key physicochemical and electronic properties to maximize efficacy and safety (Xiao et al.
198 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
Wang.
Clark & MaAoayan, 2.
In accordance with Lipinski's rule of five, it is imperative to maintain calculated lipophilicity .
LogP) below 5 to ensure adequate membrane permeability and oral bioavailability (Lipinski et al.
, 2.
Furthermore, it is essential to optimize molar refractivity (MR) between 40 and 130 to support favorable van der Waals interactions without exacerbating hydrophobic The hepatotoxicity of tamoxifen, which has been extensively documented, encompasses a range of adverse effects, including steatosis, cholestasis, and, in rare instances, peliosis hepatis.
Table 5.
In silico prediction of metabolism-excretion profiles for tamoxifen and its derivatives Metabolism Excretion
CYP3A4
Inhibitor Total Clearence .
og ml/min/k.
Renal
Substrate OCT2
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Ospemifene 4-hydroxy Lasofoxifene Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Endoxifen
Yes
Yes
Yes
Compound
CYP2D6
Substrate CYP3A4
Substrate CYP1A2
Inhibitor
Tamoxifen
Yes
Raloxifene Tamoxifene Tamoxifene Idoxifene Yes
CYP2C9
Inhibitor
CYP2D6
Inhibitor Yes Yes Yes Yes Yes Yes Toremifene The mechanism underlying this toxicity is primarily attributed to the generation of reactive metabolites by CYP2D6, which induce mitochondrial dysfunction, oxidative stress, and hepatocyte apoptosis (Chitturi & Farrell, 2013.
Zhao et al.
, 2.
At the molecular-orbital level, the energy of the lowest unoccupied molecular orbital (LUMO) is reduced by electron-withdrawing substituents, thereby enhancing the stacking of pi-pi bonds and the formation of hydrogen bonds with aromatic residues in the ER binding pocket.
This, in turn, strengthens the binding of antagonists while preserving cLogP and MR within optimal ranges through iterative in silico docking and ADMET modeling (Zhang et al.
Yuan et al.
, 2.
Table 6.
In silico toxicity assessment of tamoxifen and its Derivatives Tamoxifen LD50 Prediction .
g/k.
Toxicity Toxicity level Raloxifene Tamoxifene A Yes
Tamoxifene B
Yes
Idoxifene Yes
Toremifene Yes
Ospemifene 4-hydroxy tamoxifene Yes
Lasofoxifene Endoxifen
Compoud
Carcinogenicity Hepatotoxicity Yes
CONCLUSION
The findings of this study reveal a significant relationship between molecular parameters and the anticancer activity of tamoxifen derivatives against breast cancer.
The validated QSAR equation, pIC50 199 | S h o f a e t a l Pharmaceutical and Biomedical Sciences Journal, 2025 Vol.
= -0.
059CLogP 0.
759LUMO - 0.
011MR 3.
444, highlights the influence of LUMO energy as the most impactful molecular descriptor, followed by CLogP and MR.
Increasing LUMO while reducing CLogP and MR values may enhance the anticancer potential of these derivatives.
ADMET predictions indicate that most derivatives exhibit high intestinal absorption (HIA > 90%), low water solubility, and adequate skin permeability.
While metabolism primarily involves the CYP3A4 enzyme, raloxifene and ospemifene do not affect CYP2D6, minimizing potential drug interactions.
The hepatotoxic potential of certain derivatives suggests the need for careful consideration in further drug development.
Despite these risks, the LD50 values remain within a safe range .
Ae3000 mg/k.
, suggesting acceptable safety REFERENCES