Understanding variability in treatment responses among patients

# Understanding Variability in Treatment Responses Among Patients

The phenomenon of differential treatment responses represents one of medicine’s most persistent challenges. Two patients presenting with identical diagnoses, receiving the same medication at equivalent doses, can experience dramatically different outcomes—one achieving complete remission whilst the other shows no improvement or develops severe adverse reactions. This therapeutic lottery affects millions globally, driving healthcare costs upward whilst compromising patient outcomes. Modern precision medicine seeks to unravel the complex biological, genetic, and environmental factors underpinning this variability, transforming treatment from a one-size-fits-all approach into individualised therapeutic strategies that maximise efficacy whilst minimising harm.

Pharmacogenomic determinants of drug response heterogeneity

The genetic blueprint inherited from parents profoundly influences how your body processes medications. Pharmacogenomics—the study of genetic variations affecting drug response—has identified numerous polymorphisms that account for substantial inter-individual treatment variability. These genetic differences can alter drug absorption, distribution, metabolism, and elimination, creating a spectrum of responses ranging from therapeutic failure to life-threatening toxicity.

Cytochrome P450 enzyme polymorphisms and metaboliser phenotypes

The cytochrome P450 (CYP) enzyme superfamily metabolises approximately 75% of clinically prescribed medications. Genetic polymorphisms in CYP genes classify individuals into distinct metaboliser phenotypes: poor, intermediate, extensive, and ultra-rapid metabolisers. A poor metaboliser carrying two non-functional CYP2D6 alleles cannot effectively convert codeine into its active morphine form, rendering the medication ineffective for pain relief. Conversely, ultra-rapid metabolisers possess multiple functional gene copies, producing morphine at dangerous rates and risking fatal respiratory depression at standard doses.

CYP2C19 polymorphisms similarly affect proton pump inhibitor efficacy in treating gastro-oesophageal reflux disease. Poor metabolisers achieve superior acid suppression compared to extensive metabolisers, explaining why approximately 30% of patients experience inadequate symptom control with standard dosing regimens. Recent estimates suggest pharmacogenomic testing could prevent over 85,000 adverse drug reactions annually in the United Kingdom alone.

Human leukocyte antigen (HLA) allelic variations in adverse drug reactions

Human leukocyte antigen genes encode cell surface proteins crucial for immune system function. Specific HLA alleles dramatically increase susceptibility to severe, potentially fatal adverse drug reactions. The HLA-B*57:01 allele elevates abacavir hypersensitivity risk approximately 117-fold in HIV patients, whilst HLA-B*15:02 increases carbamazepine-induced Stevens-Johnson syndrome and toxic epidermal necrolysis risk over 1,000-fold in Asian populations.

Regulatory agencies now mandate HLA-B*57:01 screening before abacavir initiation, demonstrating pharmacogenomics’ clinical implementation. This pre-emptive testing has virtually eliminated abacavir hypersensitivity in screened populations, showcasing precision medicine’s protective potential. The cost-effectiveness of such screening becomes evident when considering that preventing a single case of Stevens-Johnson syndrome saves healthcare systems approximately £80,000 in treatment costs.

TPMT and DPYD genetic testing in oncological therapeutics

Thiopurine methyltransferase (TPMT) and dihydropyrimidine dehydrogenase (DPYD) enzymes metabolise critical chemotherapeutic agents. TPMT-deficient patients receiving standard azathioprine or mercaptopurine doses experience severe, potentially lethal myelosuppression, as approximately 0.3% of Caucasians carry two non-functional TPMT alleles. Intermediate metabolisers, comprising roughly 10% of the population, require dose reductions to avoid toxicity.

DPYD deficiency affects fluoropyrimidine metabolism, including 5-fluorouracil and capecitabine—cornerstones of colorectal cancer treatment. Complete DPYD deficiency occurs in 0.5% of patients, causing life-threatening toxicities including severe mucositis, neutropenia, and neurotoxicity. The European Medicines Agency now recommends

pre-treatment screening for common DPYD variants to identify high-risk individuals. Dose reductions of 50–75% for heterozygous carriers, or complete avoidance of fluoropyrimidines in severely deficient patients, can reduce the incidence of grade 3–4 toxicities by up to 60%. As pharmacogenomic testing panels become cheaper and faster, integrating TPMT and DPYD genotyping into routine oncology workflows represents a pragmatic step towards safer, more predictable cancer treatment responses.

Single nucleotide polymorphisms affecting warfarin and clopidogrel efficacy

Warfarin and clopidogrel are classic examples of how single nucleotide polymorphisms (SNPs) can dramatically alter drug response. Variants in CYP2C9 and VKORC1 explain up to 50% of inter-individual variability in warfarin dose requirements. Patients with reduced-function CYP2C9 alleles clear warfarin more slowly, predisposing them to bleeding if standard initiation protocols are used. Meanwhile, certain VKORC1 haplotypes decrease vitamin K epoxide reductase activity, meaning lower warfarin doses are sufficient to achieve therapeutic anticoagulation.

Clopidogrel, a pro-drug requiring metabolic activation, is heavily influenced by CYP2C19 polymorphisms. Carriers of loss-of-function alleles such as CYP2C19*2 have reduced formation of the active metabolite, resulting in suboptimal platelet inhibition and a higher risk of stent thrombosis or recurrent myocardial infarction. In contrast, ultra-rapid metabolisers may face increased bleeding risks. Clinical guidelines now recommend alternative antiplatelet agents like prasugrel or ticagrelor for poor metabolisers, illustrating how genotyping can refine cardiovascular treatment strategies and reduce avoidable complications.

Disease pathophysiology and molecular subtyping in treatment outcomes

Beyond genetics related to drug metabolism, the intrinsic biology of a disease itself plays a decisive role in treatment variability. Two patients may share the same diagnostic label on paper, yet harbour cancers or inflammatory conditions driven by entirely different molecular pathways. As we dissect diseases into finer molecular and cellular subtypes, it becomes clear that traditional classifications often obscure crucial heterogeneity that determines whether a therapy will work. Understanding this disease-level variability is essential if we want treatment decisions to move from broad protocols to precise, patient-specific pathways.

Modern diagnostics now combine histopathology with next-generation sequencing, proteomics, and immune profiling to characterise tumours and chronic diseases at unprecedented depth. This molecular subtyping helps predict who is more likely to respond to targeted therapies, immunotherapies, or conventional cytotoxic drugs. It also explains why some patients relapse quickly despite receiving guideline-concordant care. In clinics around the world, we are already seeing how integrating detailed pathophysiology into decision-making can turn an apparently “average” responder into a long-term survivor.

Tumour mutational burden and immunotherapy response rates

Tumour mutational burden (TMB) refers to the total number of somatic mutations per megabase of tumour DNA. High TMB cancers, such as melanoma and some lung cancers, generate more abnormal proteins that can act as neoantigens, making them more visible to the immune system. Immune checkpoint inhibitors like anti-PD-1 and anti-CTLA-4 therapies exploit this vulnerability, but their success is far from uniform. Studies have shown that patients with high TMB may experience response rates exceeding 40%, compared with under 10% in low TMB tumours receiving the same immunotherapy regimen.

However, TMB is not a perfect predictor; some high TMB tumours remain “cold” because of immunosuppressive microenvironments or defective antigen presentation. Conversely, certain low TMB cancers still respond well due to strong pre-existing immune infiltration. For clinicians, measuring TMB is akin to reading a weather forecast: it does not guarantee an outcome, but it markedly improves the odds of making the right choice. As sequencing becomes more accessible, integrating TMB with PD-L1 expression, gene expression signatures, and microsatellite instability status will refine patient selection and reduce the number of individuals exposed to ineffective and costly immunotherapies.

Her2-positive versus triple-negative breast cancer treatment paradigms

Breast cancer provides a clear illustration of how molecular subtyping reshapes treatment paradigms. HER2-positive tumours overexpress the human epidermal growth factor receptor 2, driving aggressive growth but also creating a therapeutic vulnerability. Targeted agents such as trastuzumab, pertuzumab, and newer antibody–drug conjugates have transformed HER2-positive disease from one of the most lethal subtypes into a far more manageable condition, with significant improvements in overall survival. Without HER2 testing, many women would either miss out on these life-saving drugs or receive them without clear benefit.

Triple-negative breast cancer (TNBC), defined by the absence of oestrogen receptor, progesterone receptor, and HER2 expression, is more heterogeneous and generally associated with poorer outcomes. Chemotherapy remains the backbone of treatment, but not all TNBCs respond equally. Emerging molecular classifications identify subgroups such as basal-like, mesenchymal, and immune-enriched TNBC, each with distinct vulnerabilities to PARP inhibitors, immunotherapies, or targeted agents. For patients and oncologists, this means that a single label—“breast cancer”—now covers multiple biologically distinct diseases, and aligning therapy with subtype is crucial to avoid both undertreatment and unnecessary toxicity.

Asthma endotypes and biologic therapy selection

Asthma was once treated as a monolithic disease, managed mainly with inhaled corticosteroids and bronchodilators. We now recognise multiple asthma endotypes—biologically distinct forms of the condition—driven by different inflammatory pathways. Type 2 (T2) high asthma, characterised by eosinophilia and elevated biomarkers like FeNO and periostin, responds particularly well to biologics targeting IL-5, IL-4/IL-13, or IgE. In contrast, T2-low asthma, which may be neutrophilic or paucigranulocytic, often responds poorly to standard steroids and currently lacks equally effective targeted options.

Biologic treatments such as mepolizumab, benralizumab, dupilumab, and omalizumab can dramatically reduce exacerbation rates and oral steroid requirements in appropriately selected patients. Yet not every individual with severe asthma benefits to the same degree, even within the same biomarker-defined endotype. This is where careful phenotyping—combining blood eosinophil counts, allergic status, exacerbation history, and lung function—helps guide biologic selection. For someone struggling with uncontrolled symptoms despite maximal inhaled therapy, understanding their specific asthma endotype can be the difference between years of hospital admissions and a return to near-normal daily life.

Crohn’s disease phenotypic classifications and anti-TNF response

Crohn’s disease is another condition where outwardly similar symptoms conceal profound biological diversity. The Montreal classification separates Crohn’s into phenotypes based on disease location (ileal, colonic, ileocolonic) and behaviour (inflammatory, stricturing, penetrating). These phenotypes influence not only natural history but also response to therapies such as anti-TNF agents. For example, patients with predominantly inflammatory disease often exhibit better initial responses to infliximab or adalimumab than those with long-standing stricturing disease, where fibrosis rather than active inflammation predominates.

Early introduction of biologics in high-risk phenotypes—such as young patients with extensive ileocolonic involvement and deep ulcers—can reduce the need for surgery and hospitalisation. Yet up to 30–40% of patients exhibit primary non-response to anti-TNF therapy, and another subgroup loses response over time. Factors such as high baseline inflammatory burden, immunogenicity leading to anti-drug antibodies, and specific genetic variants (for instance in the NOD2 gene) all contribute. Recognising which Crohn’s phenotypes are likely to benefit from anti-TNF therapy versus those needing alternative mechanisms of action (e.g. anti-integrin or anti-IL-12/23 agents) is an evolving but essential element of personalised care.

Pharmacokinetic and pharmacodynamic individual variability

Even when underlying disease biology and pharmacogenomic profile are similar, patients can differ markedly in how their bodies handle drugs. Pharmacokinetics describes what the body does to a drug—absorption, distribution, metabolism, and excretion—while pharmacodynamics focuses on what the drug does to the body at its sites of action. Imagine two people drinking the same amount of alcohol: one feels only mildly relaxed, the other becomes incapacitated. Similar principles apply to prescribed medications, where organ function, body composition, and age can turn a “standard” dose into either too much or too little.

Clinicians routinely adjust dosing based on renal and hepatic function tests, body weight, and age, but residual variability remains substantial. Therapeutic drug monitoring, clinical observation, and adverse event surveillance help fine-tune treatment over time. For you as a patient, understanding these factors can demystify why your doctor might prescribe a lower dose than a friend’s or insist on more frequent blood tests. Ultimately, appreciating pharmacokinetic and pharmacodynamic variability reinforces a key message: optimal dosing is personal, not average.

Renal and hepatic impairment effects on drug clearance

The kidneys and liver are the main exit routes for many drugs, so impairment in either organ can significantly prolong drug exposure. In chronic kidney disease, reduced glomerular filtration leads to accumulation of renally cleared medications such as certain antibiotics, digoxin, and low-molecular-weight heparins. Without dose reductions guided by estimated glomerular filtration rate (eGFR), patients face heightened risks of toxicity, from bleeding complications to arrhythmias. This is why prescribing information often includes detailed renal dosing tables that clinicians must follow.

Similarly, hepatic impairment alters metabolism and biliary excretion, particularly for drugs undergoing extensive first-pass metabolism. Cirrhosis can reduce the clearance of opioids, benzodiazepines, and many psychotropics, increasing the likelihood of sedation, confusion, or respiratory depression. Standard liver function tests do not always correlate perfectly with metabolic capacity, making careful titration and clinical vigilance essential. For high-risk therapies, such as certain tyrosine kinase inhibitors or immunosuppressants, combining pharmacokinetic modelling with therapeutic drug monitoring can help maintain efficacy while avoiding dangerous accumulation in patients with compromised organs.

Body mass index and volume of distribution considerations

Body mass index (BMI) and overall body composition shape how widely and where a drug distributes in the body. Lipophilic medications such as benzodiazepines and some anaesthetic agents tend to accumulate in adipose tissue, potentially prolonging their effects in individuals with obesity. In contrast, hydrophilic drugs distribute mainly in lean body mass and extracellular fluid, meaning that simple weight-based dosing may overestimate the appropriate dose in those with high BMI. This mismatch can contribute to both underdosing and overdosing if not carefully accounted for.

For certain therapies, including chemotherapy and anticoagulants, dosing strategies are evolving to consider factors beyond total body weight. Concepts like adjusted body weight and lean body mass are increasingly used to calculate more accurate starting doses. At the same time, obesity-related physiological changes—such as altered cardiac output and glomerular filtration—further complicate pharmacokinetics. For patients at either end of the weight spectrum, a “standard” tablet or infusion rate may therefore be anything but standard, underscoring the need for bespoke dosing plans informed by BMI and body composition.

Age-related alterations in drug absorption and metabolism

Age may be one of the most visible yet underappreciated drivers of variability in treatment responses. In older adults, gastric pH tends to increase and gastric emptying slows, subtly altering the absorption of some oral medications. More significantly, age-related declines in hepatic blood flow and renal function reduce the clearance of many drugs, even when laboratory values appear within the “normal” range. Polypharmacy, common in this group, introduces additional risks of drug–drug interactions and cumulative toxicity.

In children, developmental pharmacology presents the opposite challenge: enzyme systems and organ functions mature over time, so neonates, infants, and adolescents each require different dosing strategies. A dose that is safe and effective for a 10-year-old might be dangerous for a newborn or inadequate for a rapidly growing teenager. As a result, paediatric dosing often relies on weight- or surface area-based calculations, combined with age-specific adjustments. Across the lifespan, respecting how age reshapes both pharmacokinetics and pharmacodynamics is key to avoiding avoidable harm whilst still achieving the desired therapeutic effect.

Microbiome composition and therapeutic efficacy

The trillions of microbes that inhabit the human gut collectively form a metabolic organ as influential as the liver in some contexts. This gut microbiome not only helps digest food but also modifies bile acids, produces vitamins, and interacts closely with the immune and endocrine systems. It is therefore unsurprising that microbiome composition can profoundly affect how patients respond to treatments ranging from cancer immunotherapy to common metabolic drugs. Yet, because these microbes are invisible to the naked eye, their contribution to treatment variability is often overlooked.

Emerging research suggests that differences in microbial diversity, dominant bacterial species, and metabolite profiles can help explain why one person responds dramatically to a therapy while another sees little benefit. Antibiotic exposures, diet, age, geography, and even birth mode shape the microbiome over a lifetime. As we move deeper into the era of precision medicine, accounting for microbiome-driven variability may become as routine as checking kidney function or performing genetic tests. For now, understanding these interactions gives us a glimpse of why lifestyle factors can so powerfully modulate treatment outcomes.

Gut dysbiosis impact on immunotherapy response in melanoma patients

Some of the most striking evidence for microbiome-driven variability comes from melanoma patients treated with immune checkpoint inhibitors. Several landmark studies have shown that responders and non-responders possess distinct gut microbial signatures before therapy begins. Responders often exhibit higher abundances of bacteria such as Akkermansia muciniphila and certain Ruminococcaceae species, which appear to promote a more robust anti-tumour immune response. In contrast, dysbiosis characterised by low microbial diversity or overgrowth of potentially pathogenic species is associated with poorer outcomes.

Intriguingly, faecal microbiota transplantation (FMT) from immunotherapy responders into non-responders has, in small trials, converted some previously resistant patients into partial or complete responders. This suggests that, at least in some cases, modifying the microbiome can reshape the immune system’s capacity to recognise and attack cancer. For clinicians and patients, these findings raise practical questions: should we avoid unnecessary antibiotics during immunotherapy, and could targeted probiotics or dietary interventions enhance response rates? While definitive answers are still emerging, the evidence strongly supports considering the gut microbiome as a key determinant of immunotherapy success.

Bacteroides and prevotella ratios affecting metformin action

Metformin, a first-line treatment for type 2 diabetes, offers another example of how gut bacteria influence drug efficacy. Not all patients experience the same degree of glycaemic control or weight stabilisation despite similar doses and adherence. Research has implicated the balance between bacterial genera such as Bacteroides and Prevotella in modulating metformin’s metabolic effects. Individuals with a higher relative abundance of specific Bacteroides species, for instance, may exhibit more pronounced improvements in insulin sensitivity and glucose tolerance.

Metformin itself alters the gut microbiome, increasing certain short-chain fatty acid–producing bacteria that enhance intestinal hormone secretion and glucose homeostasis. Yet, in patients with marked baseline dysbiosis, these beneficial shifts may be blunted, contributing to variable treatment responses and gastrointestinal side effects. For you as a patient, this highlights why dietary patterns rich in fibre and plant diversity—foods that nourish beneficial microbes—might indirectly enhance how well metformin works. In time, personalised nutrition and microbiome profiling could become routine companions to standard diabetes pharmacotherapy.

Antibiotic-induced microbiome depletion in cancer treatment

Antibiotics are essential, life-saving drugs, but they can inadvertently sabotage other therapies by disrupting the gut microbiome. In oncology, multiple studies have linked broad-spectrum antibiotic exposure before or during immune checkpoint blockade with reduced response rates and shorter progression-free survival. By depleting commensal bacteria that support immune priming and anti-tumour activity, antibiotics may tilt the balance in favour of the cancer. This effect is particularly concerning in patients already at high risk of infection, where prophylactic antibiotic use is common.

Beyond immunotherapy, chemotherapy-induced mucositis and barrier disruption can interact with antibiotic-driven dysbiosis to increase the risk of bloodstream infections and sepsis. Clinicians therefore face a delicate balancing act: protecting patients from dangerous infections while preserving as much microbiome diversity as possible to maintain treatment efficacy. Strategies under investigation include narrower-spectrum antibiotics, microbiome-sparing regimens, and post-antibiotic microbiota restoration via diet, prebiotics, or FMT. Recognising the microbiome as a collateral victim of aggressive anti-cancer regimens is the first step towards designing protocols that protect both the patient and their microbial allies.

Epigenetic modifications and treatment resistance mechanisms

Epigenetics refers to reversible chemical changes to DNA and histone proteins that regulate gene expression without altering the underlying genetic code. These modifications—such as DNA methylation and histone acetylation—act like software instructions running on the hardware of our genes. In cancer and other chronic diseases, epigenetic reprogramming can switch critical genes on or off, driving disease progression and influencing how cells respond to therapy. Crucially, epigenetic states are dynamic and can evolve under treatment pressure, fostering drug resistance.

For example, tumour cells exposed to targeted therapies or chemotherapy may gradually silence pro-apoptotic genes via hypermethylation or upregulate drug-efflux pumps through histone modifications. Over time, this creates a population of cells less sensitive to the original treatment, even though their DNA sequence remains largely unchanged. In haematological malignancies, epigenetic dysregulation is a recognised driver of resistance to hypomethylating agents and other targeted drugs. Understanding these mechanisms helps explain why initial responses can be followed by relapse, and why simply increasing the dose rarely overcomes resistance.

The therapeutic flip side is that epigenetic changes are, at least in principle, reversible. Drugs such as DNA methyltransferase inhibitors (e.g. azacitidine, decitabine) and histone deacetylase inhibitors aim to reset aberrant epigenetic marks, restoring sensitivity to other treatments or directly inducing tumour cell death. Combination regimens that pair epigenetic therapies with immunotherapy, chemotherapy, or targeted agents are actively being explored to prevent or overcome resistance. From a clinical perspective, integrating epigenetic profiling into routine diagnostics may one day allow us to anticipate resistance trajectories and adapt treatment plans before clinical failure becomes evident.

Therapeutic drug monitoring and dose individualisation strategies

Given the multiple layers of variability—genetic, physiological, microbial, and epigenetic—it is clear that fixed, one-size-fits-all dosing is often suboptimal. Therapeutic drug monitoring (TDM) offers a practical way to tailor treatment by directly measuring drug concentrations in blood or plasma and adjusting doses accordingly. This approach is well established for medications with narrow therapeutic windows and high inter-individual variability, such as lithium, vancomycin, aminoglycosides, antiepileptic drugs, and certain biologics. By aligning drug exposure with known therapeutic ranges, TDM helps maximise efficacy while minimising toxicity.

Modern dose individualisation strategies increasingly combine TDM with population pharmacokinetic models and Bayesian forecasting. Rather than waiting for overt toxicity or treatment failure, clinicians can use early concentration measurements and patient-specific covariates—such as age, weight, organ function, and comorbidities—to predict the optimal maintenance dose. For example, in oncology, model-informed precision dosing of drugs like methotrexate or busulfan has been shown to reduce serious adverse events and improve outcomes. For patients, this can translate into fewer hospitalisations, shorter treatment delays, and a more predictable therapeutic journey.

Implementing TDM and personalised dosing in routine care does pose challenges, including access to timely assays, integration with electronic health records, and the need for specialised expertise. However, as laboratory technologies improve and decision-support tools become more user-friendly, we can expect wider adoption across multiple therapeutic areas. Combining pharmacogenomics, disease subtyping, microbiome insights, and TDM represents a comprehensive framework for understanding variability in treatment responses among patients. In practice, this means that the future of medicine will look less like following a rigid recipe and more like adjusting a complex, finely tuned instrument for each individual sitting in front of us.