TRIPLE POSITIVE GROUP

Tumor markers..

http://www.cancer.gov/cancertopics/factsheet/detection/tumor-markers?cid=sf13020968&cid=sf14120874

Hi everyone,

I received several private messages asking me to cut and paste the article that I described on page 606 of this forum.  I have done so, and here it is. It is lengthy, but very interesting, informative:

Impact of Metabolizing Enzymes on Drug Response of Endocrine Therapy in Breast Cancer

Abstract

Estrogen-receptor positive breast cancer accounts for 75% of diagnosed breast cancers worldwide. There are currently two major options for adjuvant treatment: tamoxifen and aromatase inhibitors. Variability in metabolizing enzymes determines their pharmacokinetic profile, possibly affecting treatment response. Therefore, prediction of therapy outcome based on genotypes would enable a more personalized medicine approach, providing optimal therapy for each patient. In this review, the authors will discuss the current evidence on the most important metabolizing enzymes in endocrine therapy, with a special focus on CYP2D6 and its role in tamoxifen metabolism.

Introduction

Breast cancer has the highest incidence of cancer in women worldwide. According to statistics reported by the WHO, over 1.38 million women are affected with breast cancer and of those, 458,000 die each year.^[201] A majority of these women are found to have hormone receptor-positive breast cancer, specifically classified with estrogen receptor (ER)- or progesterone receptor-positive status or both. Approximately 75% of diagnosed breast cancers occur in postmenopausal women and within that group, more than two-thirds are characterized as having ER-positive status. These affected women are considered candidates for endocrine therapies, which include either tamoxifen or aromatase inhibitors (AIs) as standard adjuvant therapy for early breast cancer.

Tamoxifen was first approved in 1977 by the US FDA in the adjuvant setting and thereafter was approved in the chemopreventive setting.^[202] Adjuvant tamoxifen is considered the gold standard of care in antiestrogenic treatment of ER-positive breast cancer in premenopausal women and remains a major option of treatment in postmenopausal women. Tamoxifen therapy for 5 years substantially lowers the yearly relapse rates and mortality in primary breast cancer,^[1] and this clinical benefit is further maintained when tamoxifen is continued for 10 years rather than stopping at 5 years, as shown by the ATLAS trial.^[2] Notably, 10 years of tamoxifen can halve breast cancer mortality during the second decade after diagnosis. Pharmacologically, tamoxifen is a competitive partial agonist inhibitor of ER and belongs to a class of drugs known as selective ER modulators, which exhibit either agonistic or antagonistic effects at the ER depending on the location of action in the body.^[3] For instance, tamoxifen acts antagonistically in mammary tissue by blocking the ER from its endogenous ligand, estradiol, but acts agonistically in the bone.^[3] Thus, the drug provides a further benefit by increasing bone density and preventing osteoporosis in breast cancer patients.

Since the 1990s, direct hormone suppression has been another option for antihormonal therapy to prevent growth of estrogen-dependent breast cancer cells. As the enzyme aromatase (CYP19A1) catalyzes the crucial last step of estradiol biosynthesis, it is an efficient target to lower estrogen levels by inhibition with AIs. Indeed, the currently used third generation AIs provide efficient reduction of estradiol levels and have been proven to be effective in the adjuvant endocrine treatment of postmenopausal breast cancer. Clinical studies have shown that AIs are superior to tamoxifen with regard to disease-free survival^[4–7] and are deemed to be the preferred treatment for postmenopausal ER-positive breast cancer, either as up-front monotherapy or sequential to tamoxifen.^[8] In contrast to earlier generation AIs, the nonsteroidal compounds anastrozole and letrozole, as well as the steroidal exemestane, provide the highest target specificity and inhibitory potential.^[9–11] Although they are proven to be safe, a substantial proportion of patients suffers from severe side effects including arthralgia, myalgia and loss of bone density.^[12,13]

This review addresses the role of drug metabolizing enzymes (DMEs) that potentially influence pharmacokinetics and drug efficiency of both tamoxifen and AIs. Importantly, both drug classes target estrogen-mediated growth of breast cancer cells. Therefore, metabolizing enzymes controlling the supply of circulating estrogens are probably relevant for drug efficiency; however, with the exception of CYP19A1, other enzymes' contribution or confounding impact is beyond the scope of this review.

Tamoxifen

The Phase I metabolism of tamoxifen is extensive and involves several CYP450 DMEs including, but not limited to: CYP2D6, CYP2C9, CYP2C19, CYP2B6, CYP3A4 and CYP3A5 (Figure 1). As a prodrug, tamoxifen requires conversion into its primary metabolites, (Z)-4-hydroxytamoxifen ([Z]-4-OH-Tam) and N-desmethyltamoxifen (NDM-Tam). These metabolites are further converted into several secondary metabolites, including (Z)-4-hydroxy-N-desmethyltamoxifen, also known as endoxifen. Both (Z)-4-OH-Tam and endoxifen are shown to have significantly higher efficacy at the ER compared with the parent drug.^[14,15] Furthermore, in vitro studies revealed that endoxifen inhibits estrogen-induced breast cancer cell proliferation depending on metabolite concentration.^[16] CYP2D6 is involved in several steps of the metabolic pathway but is the only enzyme responsible for the specific conversion of NDM-Tam to endoxifen.^[17,18] As steady-state plasma levels of endoxifen are approximately five-times higher than those of (Z)-4-OH-Tam and as endoxifen is the metabolite with lowest IC₅₀ at the ER, it is currently perceived as the major active metabolite.^[16,19]

(Enlarge Image)

Figure 1.

Major pathways of tamoxifen metabolism including the active metabolites and Phase II metabolism of (Z)-4-OH-Tam.

Tamoxifen and its metabolites are inactivated via glucuronidation and sulfation by UDP-glucuronosyltransferases (UGTs) and sulfotransferases (SULTs), respectively, and are eliminated mostly by biliary/fecal excretion.^[20] Some of the major enzymes responsible for these inactivation steps include UGT1A4, UGT2B7, UGT2B15 and SULT1A1.

CYP2D6 Pharmacogenetics

Tamoxifen provides a significant clinical benefit in endocrine therapy; however, approximately 30–50% of the patients relapse or eventually die from the disease.^[1,21] The causes for breast cancer relapse are attributed to a variety of factors, one of which may be traced to interindividual variability in DME, particularly CYP2D6 and its role in the formation of endoxifen. CYP2D6 is one of the most studied CYP450 enzymes and has been extensively reviewed.^[22–25] The CYP2D6 gene is located on chromosome 22q13.1 and its sequence is characterized by numerous variants. Currently, 105 distinct alleles are known, listed on the CYP allele nomenclature website.^[203] Many of these alleles lead to absent, decreased or increased CYP2D6 expression. Comprehensive genotype–phenotype correlation studies in Caucasians using debrisoquine or sparteine as probe-drug clearly indicate the impact of CYP2D6 polymorphisms on CYP2D6 function in vivo ^[26–28] and result in the discrimination of the poor (PM), intermediate (IM), extensive (EM) and ultrarapid (UM) metabolizer phenotypes (Figure 2). The EM status is determined by the presence of two normal functioning alleles (e.g., *1, *2, *35). The PM status is defined by a lack of CYP2D6 enzyme function and is determined by the presence of two nonfunctional alleles (so-called null alleles), of which more than 20 have been identified.^[203] Approximately 99.9% of CYP2D6 PM metabolizers can be predicted by CYP2D6 null alleles either in a homozygous variant or a compound heterozygous genotype.^[25,29] Subjects carrying one partially defective CYP2D6 allele (e.g., *41, *9, *10) together with another partially defective or null allele are classified as IM (e.g., *9/*41 or *4/*41), which is a distinct phenotypic CYP2D6 subgroup (Figure 2). Moreover, heterozygous carriers of one defective and one normal CYP2D6 allele (e.g., *1/*4) tend to also have a lower median CYP2D6 enzyme activity compared with wild-type carriers (e.g., *1/*1).^[19,30] As this group is not as clearly defined as EMs and PMs but represents a rather heterogenous group, many of the current studies combine heterozygous carriers together with classically defined IMs into one decreased function phenotype. The UM phenotype can be inferred from CYP2D6 gene duplications/amplifications of fully functional alleles (e.g., *1xN/*2) resulting in excessive enzymatic function.^[31,32] Since not all CYP2D6 duplications comprise functional genes, the prediction of the correct phenotype may be difficult.^[33,34] Only up to 30% of the UM phenotype can be attributed to gene amplification.^{[26,27,35,36]}

(Enlarge Image)

Figure 2.

Sparteine oxidation phenotypes and distribution in a German population. MR_S: Urinary metabolic ratio for sparteine.Adapted with permission from [25].

In the European population, approximately 5–10% are PMs, 10–15% are IMs and approximately 5–10% are UMs (Figure 2). Furthermore, ethnic variations in reduced or null function alleles must also be considered as several studies have reported on the rarity of CYP2D6 PM alleles in individuals of African and Asian descent.^[37] Table 1 highlights the major CYP2D6 variants with reported allele frequencies categorized by the main population groups. CYP2D6*17 is the most common reduced function allele in African populations while the IM *10 allele is more prevalent in Asian populations ranging from a frequency of 38 to 70%.^[37–39]

Several CYP2D6 genotyping strategies are employed to determine pharmacogenetic DME variations from various DNA sources. Table 2 summarizes the currently available methods. There are two FDA-approved methods used in the molecular diagnostic setting: the AmpliChip® CYP450 test system (Roche, CA, USA) and the xTAG® assay CYP2D6 kit (Luminex). An advantage of the AmpliChip system is the high-density allele coverage of 33 variant CYP2D6 alleles including allele-specific gene duplication. The DNA source is critical for accurate genotyping. Because formalin-fixed paraffin-embedded (FFPE) material contains highly fragmented DNA, this source is not suitable for the long-range PCR-based AmpliChip CYP450 test.^[40] Other multiplexing technologies include the MassARRAY® system (Sequenom) that incorporates MALDI-TOF MS, which allows for the use of archived materials, such as FFPE, as a source of DNA.^[41] As an extension of validated CYP2D6 TaqMan® genotyping assays (Applied Biosystems), the Fluidigm platform may process several individual TaqMan assays simultaneously in a medium-throughput setting.^[42]

Tamoxifen Pharmacokinetics & CYP2D6

Within the past decade, there have been numerous pharmacokinetic studies stressing the link between patients' CYP2D6 genotype and its prediction of enzyme function, and plasma levels of active tamoxifen metabolites. Table 3 summarizes such pharmacokinetic studies, which report the correlation between CYP2D6 genotype and plasma endoxifen levels of patients under tamoxifen treatment. Most of these studies were conducted in women with different menopausal statuses. A recent study in postmenopausal women showed a CYP2D6 gene-dose effect in which EMs and UMs had the highest endoxifen concentrations while PMs had the lowest.^[19] There is first evidence in premenopausal women that *1/*1 carriers had endoxifen concentrations that were up to 2.4- and 2.6-fold higher than those of *10/*10 and *5/*10 carriers, respectively.^[43] Taken together, the correlation between CYP2D6 genetics and plasma endoxifen levels in individuals undergoing tamoxifen therapy is well established (Table 3). By contrast, there is a lack of studies addressing a direct link between tamoxifen metabolite levels and effect on clinical outcome. A first study by Madlensky et al. shows that patients in the lowest quintile of endoxifen plasma concentrations were associated with increased risk of recurrence.^[44] This needs further confirmation and there are studies currently underway. Indeed, there is a need to investigate whether monitoring plasma levels under therapy may complement the prediction of CYP2D6 enzyme function and whether this approach may be used as a surrogate for clinical efficiency.

Phenocopying of CYP2D6

Currently, the American Society of Clinical Oncology (ASCO) recommends using caution when CYP2D6 inhibitors are coadministered throughout the duration of tamoxifen treatment.^[8] Potent inhibitors of CYP2D6 reduce its enzymatic function, resulting in a change of phenotype of EM to PM in the most extreme case, known as phenocopying.^[45,46] Thus, drug interactions of potent inhibitors are a major confounding factor of the CYP2D6 phenotype. Known CYP2D6 inhibitors predominantly used in breast cancer patients are selective serotonin reuptake inhibitors (SSRIs) given to overcome tamoxifen side effects like hot flashes as a consequence of antiestrogen action.^[47] The potency of SSRIs to inhibit CYP2D6-catalyzed oxidation of sparteine has been analyzed using human liver microsomes.^[48] Inhibition constants (K_i) for various SSRIs are in descending order: fluvoxamine (8.2 µM), citalopram (5.1 µM), sertraline (0.70 µM), fluoxetine (0.60 µM) and paroxetine (0.15 µM), indicating that paroxetine is the strongest inhibitor. Several pharmacokinetic studies evaluated the effect of SSRIs on tamoxifen metabolism and showed that coadministration of paroxetine substantially reduced the level of plasma endoxifen concentrations.^[30,47,49] CYP2D6 inhibitor use impairs tamoxifen metabolism and potentially diminishes the clinical efficacy toward an inferior outcome.^[50] Notably, comedication with paroxetine during tamoxifen treatment was shown to raise the risk of death from breast cancer.^[51] By contrast, the antidepressant venlafaxine, a serotonin–norepinephrine reuptake inhibitor, is a very weak CYP2D6 inhibitor.^[52] The K_i value of venlafaxine, as determined by dextromethorphan O-demethylation, is 33 µM, which is one to three orders of magnitude larger compared with SSRIs including paroxetine, norfluoxetine, sertraline and fluvoxamine.^[53] Interestingly, venlafaxine use did not change breast cancer mortality with concomitant tamoxifen treatment.^[51] Altogether, these findings have strengthened the argument against the practice of coadministration of potent CYP2D6 inhibitors with tamoxifen treatment. The concept of drug–drug interaction between potent CYP2D6 inhibitors and tamoxifen resulting in worse outcome in postmenopausal women strongly supports an important role of CYP2D6 on tamoxifen efficacy.

Impact of CYP2D6 in Endocrine Therapy

The overwhelming evidence of the pharmacokinetic association between CYP2D6 genotype and plasma endoxifen levels has led to the focus of how CYP2D6 genotype translates to tamoxifen outcome. The current understanding is that CYP2D6 matters^[54] based on a large study conducted in Germany and the USA.^[41] The study included 1325 of mostly postmenopausal patients with ER-positive breast cancer who underwent tamoxifen monotherapy. Patients carrying at least one PM or IM allele had higher rates of recurrence, worse event-free survival and worse disease-free survival compared with those carrying two copies of EM alleles. Overall survival, however, was not statistically different when classified by CYP2D6 genotype. This retrospective study has recently been challenged by pharmacogenetic analyses performed in patients previously investigated for the superiority of tamoxifen versus AI, the ATAC and the BIG 1-98 trials.^[55,56] Although these new pharmacogenetic studies did not observe a correlation between CYP2D6 and tamoxifen outcome, it has become clear that their data cannot be interpreted owing to serious methodological issues pointing to the critical issue of DNA source – that is, germline versus tumor genome.^[54,57–59] The genomic region surrounding the CYP2D6 locus at chromosome 22q13 is frequently deleted in breast cancer, causing loss of heterozygosity.^[60] As the BIG 1-98 and ATAC studies used DNA derived from FFPE,^[55,56] the strong departure from Hardy–Weinberg equilibrium of CYP2D6*4 genotypes in their pharmacogenetic studies suggests that their genotyping most likely reflected the tumor and not the constitutional genotype, leading to errors in the prediction of CYP2D6 phenotype and consequently, to false-negative interpretation of study results. As it stands, subsequent studies must now be more carefully designed with regard to the source of DNA and CYP2D6 genotyping methods in order to produce reliable data and interpretation.

Another important issue within the context of molecular CYP2D6 diagnostics for tamoxifen metabolizer phenotype is the possibility of the misclassification of CYP2D6 phenotypes due to insufficient coverage of CYP2D6 alleles by genotyping. It has been clearly shown that it is not sufficient to genotype the *4 allele alone because this will either only detect 70% of PM or misclassify a substantial proportion of subjects (up to 30%) as the wrong phenotype.^[40,61]

As discussed in the previous section, phenocopying is a major confounder of the CYP2D6 phenotype as comedication with inhibitors like SSRIs may result in diminished tamoxifen efficacy and decreased clinical benefit. Altogether, Table 4 summarizes the major confounders based on CYP2D6 genotype-predicted phenotypes that may impact tamoxifen outcome and have so far been discussed in this review. In addition, it is important to note that a single factor alone is not sufficient in explaining the frequently observed heterogeneity in CYP2D6 outcome studies. In fact, clinical outcome is influenced not only by genetic factors, but also by a combination of several nongenetic factors: combination therapies, nonadherence, increased tamoxifen dosing, AI switch during endocrine therapy and so on.^[62]

While the technical issues require thorough attention in CYP2D6 genotyping, it is of interest that a most recent study showed that PMs derived a slight clinical benefit when switching from tamoxifen to AIs during their 5-year treatment compared with those who remained on tamoxifen for the entire duration of therapy.^[63] This crossover study calls attention to the fact that AI use may be clinically advantageous, which is in line with the notion that CYP2D6 PM patients have less benefit from tamoxifen as compared with EM patients.^[41] Therefore, CYP2D6 molecular diagnostics can provide a means to identify patients poorly responding to tamoxifen and allows for evidence-based assignments of AI treatment, thereby improving clinical outcome.

CYP2D6 Genotype-Driven Dosing of Tamoxifen

In postmenopausal women, AI can be given as another valid treatment option, which is especially important for patients who are CYP2D6 PMs. Nevertheless, the idea arises that it should be considered whether at least in IM women a higher dosage of tamoxifen may be an appropriate adjustment. Important clues came from recent prospective studies of CYP2D6 genotype-guided dosing of tamoxifen (Table 3). An increase in the standard dose of tamoxifen from 20 to 40 mg/day in IM patients whose baseline endoxifen levels were 46% lower compared with EM patients resulted in an increase in endoxifen concentrations close to those of EM patients independent of menopausal status. However, PMs displayed no significant increase of endoxifen levels after dose increase, in line with the fact that null alleles for CYP2D6 result in nonexpression of CYP2D6 protein.^[64] It follows that at least one functional or partially functional allele, in the case of IMs, is necessary to reach endoxifen levels that are not statistically different from EMs upon dose adjustment.^[64] Supportive evidence came from another study that increased tamoxifen dose from 20 to 30 mg/day in patients with baseline endoxifen levels below 40 nmol/l.^[65] An increase to clinically meaningful endoxifen concentrations was observed;^[16] however, this was dependent on CYP2D6 genotype, with PMs having lower rates of increase.^[65] Moreover, after tamoxifen dose adjustments to 30 or 40 mg/day based on CYP2D6 genotype, Japanese patients with one copy of reduced functional allele reached endoxifen levels comparable with the patients with two functional alleles.^[66] Collectively, these studies suggest that CYP2D6 molecular diagnostics could play an important role in the case of planned tamoxifen dose adjustment in patients with moderately impaired metabolizer statuses such as IM or heterozygous EM patients.

Other Enzymes Involved in Tamoxifen Metabolism

Plasma metabolite analytics in patients receiving tamoxifen showed that despite a CYP2D6 PM status, patients can have high levels of endoxifen,^[19] with 24% of PM patients even reaching therapeutic levels.^[44] It follows that other DMEs are likely to influence endoxifen formation and therefore the formation of endoxifen cannot be predicted solely by CYP2D6 genotype.

Phase I

Reduced CYP2C9 enzyme function variants are encoded by the *2 and *3 alleles.^[67] They are frequent in populations of European descent (35%) but are less common in Asians and Africans.^[68,69] German individuals genotyped as carriers of CYP2C9 reduced-activity alleles had lower plasma levels of endoxifen compared with those who were homozygous for the normal function alleles.^[19] No such association was observed in Asian patients, most likely due to the lower frequency of these variants in that population.^[43] Given that CYP2C9 contributes significantly to the formation of 4-OH-Tam, it can be concluded that a proportion of endoxifen derives from demethylation of (Z)-4-OH-Tam.^[70]

The major variants of CYP2C19 include the *2 and *3 alleles, both considered null, and the promoter variant *17 resulting in increased expression.^[25] Studies in The Netherlands linked the *2 allele with increased survival rates in women undergoing tamoxifen treatment;^[71,72] however, this was not confirmed in a German study.^[73] Rather, the UM allele *17,^[74] which is frequent in Europeans and North Africans but not in Asians, has been linked to improved benefit with respect to breast cancer recurrences and relapse-free survival.^[73]

The most common CYP2B6 variant allele is *6, which is present in various populations with frequencies ranging from 15 to 60%.^[75] The role of the *6 allele was investigated specifically with regard to the hydroxylation of tamoxifen to (Z)-4-OH-Tam, with no correlation being observed between the CYP2B6 variant and the metabolite.^[19]

The CYP3A isoforms CYP3A4 and CYP3A5 contribute to tamoxifen metabolism, particularly during the conversion of the parent drug to both NDM-Tam and (Z)-4-OH-Tam. CYP3A4 shows remarkable interindividual variability attributed to environmental factors including medication and food, but its variability is also influenced by gender and age. For instance, CYP3A4 expression was found to be twofold higher in liver samples from females compared with males, and was also considerably increased after treatment with carbamazepine or St John's wort.^[76] CYP3A4 activity, like other CYPs, was shown to increase with age.^[77] The influence of grapefruit juice on CYP3A4 was discovered by chance and has since been extensively investigated.^[78,79] Notably, potent CYP3A4-inducing compounds (e.g., rifampin, aminoglutethimide, tamoxifen and 4-OH-Tam)^[80–82] may cause drug–drug interactions modulating the levels of parent drug tamoxifen and its metabolites. Known variants of CYP3A4 include the *1B allele, a promoter variant linked with altered enzyme activity; however, this finding has been controversially debated.^[83] CYP3A4*22 has a frequency of 5–7% in Europeans and this intron 6 polymorphism results in a reduced mRNA expression and decreased enzyme activity in human liver.^[84] While cis-acting genetic variants may play a minor role in CYP3A4 variability, it is of note that a recently identified SNP (rs4253728) in the gene coding for PPAR-α, a transcription factor that binds to the CYP3A4 promoter, was shown to be associated with the CYP3A4 phenotype.^[85,86] Such trans-acting regulatory factors could be of potential interest for genetic testing; however, their contribution to the variability in tamoxifen response awaits further clarification.

The common variant CYP3A5*3 leads to reduced enzyme function. The frequency is higher in populations of European descent (73–93%) compared with those of Asian (74–77%) and African descent (27–50%).^[87,88] A slight increase in endoxifen levels in individuals carrying at least one copy of the normal function allele has been reported,^[47] but subsequent studies did not confirm this.^[19,43] It is likely, however, that the CYP3A5*3 deficiency allele contributes to variation in the levels of the primary metabolite NDM-Tam.

Phase II

Tamoxifen metabolites are inactivated prior to elimination through conjugation with a glucuronide or sulfate group by UGTs and SULTs, respectively. Some of the major enzymes include, but are not limited to, UGT1A4, UGT2B7, UGT2B15 and SULT1A1 (Figure 1). It is known that these enzymes, similar to CYPs, have functionally relevant genetic variants potentially influencing the efficacy of tamoxifen. For instance, in vitro studies showed that the UGT1A4*3 (rs2011425, 142T>G, L48V)^[204] polymorphism increases glucuronidation of tamoxifen and 4-OH-Tam.^[89] Plasma metabolite studies showed a significant decrease in the metabolic ratio of tamoxifen/tamoxifen-N-glucuronide in patients homozygous for the *3 allele; however, an association with 4-OH-Tam and its glucuronidation reaction was not apparent in vivo.^[19] Currently, the clinical relevance of the UGT1A4*3 variant is unknown.

The UGT2B7 gene has a common polymorphism referred to as UGT2B7*2 ^[90] with a frequency of approximately 50% in individuals of European descent.^[91] In vitro studies indicated that the variant displayed significantly reduced glucuronidation activities against the trans isomers of 4-OH-Tam and endoxifen compared with the wild-type UGT2B7 allele.^[92] However, no association between UGT2B7*2 and tamoxifen and its metabolites was observed.^[19] Regarding tamoxifen outcome, no correlation with UGT2B7*2 was reported.^[56,93]

The UGT2B15 gene harbors a nonsynonymous polymorphism which is designated the *2 allele, resulting in twofold increased activity.^[94] A trend of increased breast cancer recurrence for carriers of UGT2B15*2 alleles was observed for patients treated with tamoxifen;^[95] however, no association with either metabolite levels or outcome has been confirmed in other studies.^[19,93,96]

SULT1A1 is the most abundant hepatic SULT enzyme and is mainly present in breast cancer cells. Moreover, it is predominantly involved in the sulfation of 4-OH-Tam. Because the SULT1A1*2 allele has been linked with reduced enzyme activity,^[97] several studies addressed the possible influence of SULT1A1*2 on tamoxifen metabolism and outcome. Results of these studies were inconsistent, showing either a lack of association with regard to metabolite levels^[47] or nonfavorable outcome,^[98] but most showed a lack of association with clinical outcome of tamoxifen therapy.^[96,99,100] SULT1A1 copy number variations seem to best explain variable activity in vitro, questioning the relevance of other polymorphisms.^[101] No correlation between SULT1A1 copy number and disease-free survival in patients receiving tamoxifen has been observed.^[102]

The genetic polymorphisms of other DMEs involved in tamoxifen metabolism may be important but their value in predicting treatment response is presently unknown due to the lack of evidence.

Consequences for Molecular Diagnostics

The current evidence highlights CYP2D6 to be the most relevant molecular diagnostic factor to be utilized for stratified use of tamoxifen in the endocrine treatment setting of postmenopausal breast cancer. The highest standard in CYP2D6 genetic testing equates to comprehensive allele coverage by using the FDA-approved AmpliChip CYP450 test system in order to accurately predict the metabolizer phenotype.^[40,61] However, a limited analysis of the most common CYP2D6 variants listed in Table 1 also seems appropriate for most molecular diagnostic purposes^[45] and represents an acceptable compromise if AmpliChip testing is deemed impractical for routine testing. Suitable options are the FDA-approved xTAG® CYP2D6 kit and the INFINITI® CYP450 2D6I assay, which both simultaneously interrogate 15 genetic variants.^[103] It must be stressed that testing for only one variant, such as *4, is highly unacceptable for research purposes and unethical for clinical practice. This may not only be relevant for tamoxifen, but also for other CYP2D6 substrates.

Another critical requirement is obtaining the most appropriate source of DNA for genetic testing. Because DNA from tumor samples is susceptible to somatic gene deletions, a blood sample that provides constitutional DNA must be the requirement for diagnostic genetic testing.

In addition to high standard diagnostic test criteria, it is critical to consider that testing results should be accompanied with an expert interpretation.^[104] This should include suggestions for other valid treatment options, the avoidance of known CYP2D6 inhibitors and dosing adjustments where appropriate, thereby effectively conveying the clinical ramifications that are critical to treatment decision-making.^[104]

Diagnostic testing relies on the professional experience associated with patient management and treatment, which cannot be sufficiently covered by specialist molecular knowledge and analytical expertise alone. In fact, genetic testing laboratories require interdisciplinary cooperation involving clinicians, clinical pharmacologists and molecular scientists with appropriate training in molecular diagnostics, to produce testing reports of the highest standard of care.

Aromatase Inhibitors

The third-generation AIs letrozole, anastrozole and exemestane are frequently used as adjuvant therapy for ER-positive breast cancer in the postmenopausal setting. Large clinical studies have shown an improved disease-free survival for AI therapy in comparison to tamoxifen,^[5–7] a reason why AI use has been on the rise. The current ASCO guidelines recommend the use of an AI as upfront therapy or sequential to tamoxifen. Still, optimal timing and duration of AI therapy remains unclear.^[8] The patients' individual risk profile, including familial predisposition to particular side effects (e.g., thromboembolism or osteoporosis) and the adverse effect profiles of the drugs are considered in the treatment decision between tamoxifen and AI.

CYP19A1 Genetics

The drug target of AIs is the enzyme aromatase, which belongs to the superfamily of CYP450 enzymes. Aromatase converts androgens (testosterone, androstenedione) to estrogens and its inhibition results in estrogen depletion, which prevents growth of estrogen-dependent breast tumor cells. This is the therapeutic principle used in the antihormonal AI treatment of ER-positive postmenopausal breast cancer.

Aromatase is encoded by the CYP19A1 gene, which is located on chromosome 15q21.2, spanning 123 kb. The gene contains nine protein-coding exons and ten noncoding upstream exons controlled by tissue-specific promoters.^[105] CYP19A1 is polymorphic and therefore genetic variants may influence gene expression or enzyme activity, which might alter AI efficacy.^[106,107] The impact of CYP19A1 variants on AI outcome has been thoroughly reviewed by Del Re et al..^[108]

Using a CYP19A1 resequencing approach, Ma et al. identified 88 polymorphisms resulting in 44 haplotypes and four nonsynonymous coding SNPs associated with aromatase activity.^[106] A study investigating the prognosis of invasive primary breast cancer linked a 3'-UTR polymorphism, rs10046 (C>T), with a favorable 5-year disease-free survival in premenopausal women.^[109] Prolonged time to progression in letrozole-treated patients has been linked with the 3'-UTR polymorphism rs4646 (C>A);^[110] however, another study reported an association with reduced time to progression.^[111] Wang et al. investigated the changes of aromatase activity in AI-treated postmenopausal breast cancer patients in relation to two tightly linked SNPs located in the 5'-flanking region of exon 1.1 (rs6493497 and rs7176005). Notably, both SNPs resulted in greater change in aromatase activity before and after AI treatment in tumor samples, indicating that patients carrying these two SNPs had greater inhibition of aromatase activity. Moreover, a moderate association of these two SNPs with baseline aromatase activity and their association with higher estradiol levels in tumor samples was found.^[107] Women carrying the eight-repeat allele (TTTA₈) of the tetranucleotide repeat polymorphism in intron 4 (rs60271534) suffered less frequently from AI-associated arthralgia.^[112] In healthy individuals, the homozygous CC genotype of a SNP in exon 1.2 (rs1062033) was linked to lower spine and hip bone density.^[113] Although the CYP19A1 polymorphisms rs10046 C>T, rs4646 G>T, rs749292 C>T and rs727479 T>G have been suggested to be related to circulating estrogen levels, no association with plasma estrone sulfate levels was reported in a recent study.^[114] Despite numerous studies, as of yet there is questionable evidence whether CYP19A1 polymorphisms may impact AI drug response, currently not justifying any genetic testing in breast cancer patients and therefore further comprehensive studies are warranted.

AI Metabolism & the Role of Polymorphisms in DMEs

Although the role of DMEs for AI outcome is less investigated compared with tamoxifen, initial studies addressed pharmacogenetic relationships that may indicate potential clinical implications. Inspired by the CYP2D6 tamoxifen pharmacogenetic paradigm, it is of high interest to understand the possible role of DME genetic variants involved in AI metabolism. However, this is by far a more complex issue because letrozole, anastrozole and exemestane follow different metabolic pathways.

Letrozole

The nonsteroidal AI letrozole is metabolized to the pharmacologically inactive 4,4'-(hydroxymethylene)dibenzonitrile (carbinol), which is further glucuronidated to bis(4-cyanophenyl)methyl hexopyranosiduronic acid (carbinol-gluc).^[115–117] Carbinol-gluc is the major metabolite of letrozole and approximately 65% of the oral letrozole dose is excreted as carbinol-gluc in the urine.^[118] The conversion of letrozole to carbinol is catalyzed by CYP2A6 and CYP3A4.^[116] The gene CYP2A6, located on chromosome 19q13.2,^[119] is mainly expressed in human liver and contributes less than 5% to the total liver P450 pool.^[25] CYP2A6 expression and function appear to be sex-dependent with higher levels in females.^[120,121] Moreover, interindividual hepatic expression and function are highly variable.^[122] In addition to transcriptional regulation of CYP2A6 by pregnane X receptor and constitutive androstane receptor activators, like rifampicin and phenobarbital^[123] and the anti-inflammatory drug dexamethasone,^[124] at least 38 CYP2A6 alleles have been identified so far,^[205] in part with functional consequences.^{[122,125,126]} For instance, the null alleles CYP2A6*2 and CYP2A6*4 (gene deletion) predispose to a PM phenotype. Of note, the CYP2A6*4 allele frequency differs widely between Europeans (1.2%) and Asians (10–20%).^[127] Other impaired function alleles include *9B, *12B, *7, *10, *17 and *35. Due to the relevance of CYP2A6 in nicotine metabolism, several pharmacogenetic investigations were focused on smoking behavior, nicotine withdrawal symptoms and lung cancer risk.^[128] With regard to letrozole, CYP2A6 genotype-predicted phenotypes showed a correlation to plasma levels of 261 postmenopausal breast cancer patients.^[129] A Japanese clinical study investigated the influence of CYP2A6 polymorphisms as well as nongenetic factors on the population pharmacokinetics of letrozole in 25 healthy postmenopausal women.^[130] The apparent systemic clearance (CL/F) was significantly altered by CYP2A6 genotype in that carriers of one or two variant alleles had reduced clearance of 84.3 and 44.8%, respectively. With respect to CYP3A5 genotypes, no association with letrozole plasma levels has been reported.^[129]

Anastrozole

Anastrozole metabolism consists mainly of hydroxylation to OH-anastrozole by CYP3A4 with subsequent glucuronidation to glucuronide-OH-anastrozole. Another metabolic pathway is the direct N-glucuronidation via UGT1A4.^[131] Triazol, the major cleavage product, is pharmacologically inactive.^[132] Plasma levels of anastrozole and its metabolite levels show substantial interindividual variability as demonstrated in 191 postmenopausal women with early breast cancer.^[133] Whether variability of anastrozole pharmacokinetics is related to incomplete aromatase inhibition and subsequent residual estrogen levels under therapy is currently unknown. UGT1A4 is polymorphically expressed^[134,135] and therefore the impact of UGT1A4 variants including the alleles *2 (rs6755571, 70C>A, P24T) and *3 (rs2011425, 142T>G, L48V) on the in vitro metabolism of anastrozole has been investigated. No significant effects were found,^[131] which has been recently confirmed.^[136] Of interest, three polymorphisms in the UGT1A4 promoter region (-163G>A, -217T>G, -219C>T) were shown to influence anastrozole N-glucuronidation activity in vitro.^[136]

Although it has been shown that concurrent tamoxifen therapy reduces anastrozole plasma levels by 27%, putatively by CYP3A4 induction,^[82,137] no change of plasma levels of anastrozole and hydroxyanastrozole was observed in a small study of nine women with concomitant intake of simvastatin, another known CYP3A4 substrate.^[138] Thus, it is presently unclear whether a CYP3A4-related drug–drug interaction affects anastrozole metabolism and efficacy. Altogether, the available knowledge on anastrozole's DME pharmacogenetics is quite limited and awaits further investigation

Exemestane

The steroidal AI exemestane undergoes Phase I metabolism to the active metabolite 17-dihydroexemestane^[139] by CYP1A1/2 and possibly CYP4A11. The oxidation to 6-hydroxymethylexemestane is mediated by CYP3A4, CYP3A5 and CYP2B6.^[140] 17-dihydroexemestane is subsequently glucuronidated to exemestane-17-O-glucuronide, predominantly by UGT2B17, but also by UGT1A4 as shown in vitro.^[141] Although CYP3A4, CYP1A1 and CYP4A11 are polymorphic, there are currently no data on a potential impact of genetic variants on exemestane metabolism.^[142] The oxidative CYP3A4-mediated pathway is of minor importance as concomitant medication with strong CYP3A4 inhibitors did not alter exemestane pharmacokinetics.^[206] UGT2B17 pharmacogenetics have been described to impact Phase II metabolism of exemestane.^[141]

The UGT2 locus is located on chromosome 4 at position 4q13.^[143] The UGT2B17 gene shares 95% sequence homology with UGT2B15 ^[144] and is expressed in the liver, intestine and antigen-presenting cells. Spielman et al. reported a remarkable variability in UGT2B17 gene expression with a 22-fold higher expression level in lymphoblastoid cell lines of individuals of European descent compared with those of Asians.^[145] A frequent variation of UGT2B17 is a gene deletion (*2) with a prevalence of approximately 30% in Europeans.^[144] As a consequence, approximately 11% of Europeans are UGT2B17 deficient. An enzyme-selective probe drug is currently not available for UGT2B17 phenotyping.^[143] Because of the predominant role of UGT2B17 in exemestane metabolism, the impact of the UGT2B17 gene deletion has been investigated and a 14-fold decrease of exemestane-17-O-glucuronide formation in human liver microsomes homozygous for *2 compared with wild-type has been observed.^[141] No information on the clinical consequences of this polymorphism is available. There was no influence of the UGT1A4 alleles *2 and *3 on exemestane glucuronidation activity in vitro.

Finally, a recent genome-wide association study (GWAS) provided initial evidence for the possible role of variants in target genes other than DMEs for AI-related adverse side effects. This case–control GWAS (551,395 SNPs investigated in 878 women) identified four SNPs in close proximity to the TCL1A gene that were associated with musculoskeletal adverse events during AI therapy of early breast cancer.^[146] Additionally, Ingle et al. identified a relationship between TCL1A expression and IL-17, a cytokine involved in inflammatory processes.^[146] Indeed, an estradiol-induced TCL1A expression and subsequently altered cytokine expression (including IL-17) has been observed in in vitro studies.^[147]

Conclusion of DME Impact on Clinical Outcome of AI Therapy

So far, evidence supporting an impact of genetic variation of DMEs on AI therapy outcome in breast cancer patients is limited. Further studies on AI pharmacogenomics regarding efficacy as well as AI-related adverse drug reactions are required to better understand interindividual variability of drug response and to identify patients at risk.

Expert Commentary

The evidence of CYP2D6 genotype–phenotype correlation associated with tamoxifen pharmacokinetics in early breast cancer is compelling. However, the scientific community is currently divided on the clinical relevance of CYP2D6 genotype predicting tamoxifen response. Conflicting outcome studies have led to confusion about its utility in treatment decision making. The major aim of molecular diagnostics is to provide an accurate determination of phenotype in relation to outcome; however, outcome may be confounded by several factors that include inaccurate phenotype, CYP2D6 phenocopying and unsuitable sources of DNA for genetic testing. In addition, other DMEs involved in tamoxifen biotransformation may also be influential in prediction of treatment response but further studies are warranted to understand their clinical value. It must be noted that there is a subset of patients, mainly postmenopausal women, who may benefit from genetic testing if found to have impaired CYP2D6 metabolism. So far, most pharmacogenetic results have been based on retrospective studies with differing study designs; therefore, a final decision must await appropriately designed prospective clinical trials.

The current evidence suggests different roles and strengths for pharmacogenetic diagnostics within the context of tamoxifen and AI treatment. CYP2D6 was shown to be a promising predictor for tamoxifen outcome while AI pharmacogenetics is still in its infancy, suggesting a possible role for the prediction of the onset of adverse drug reactions.

Five-Year View

The controversy surrounding the clinical validity of CYP2D6 genetic testing in tamoxifen therapy will most likely not abate in the near future. However, as it stands, the current evidence suggest that CYP2D6 matters and CYP2D6 genetics may provide useful information to support the choice of endocrine treatment for postmenopausal breast cancer patients, in that CYP2D6 PM patients may benefit more from AI therapy, which is a valid treatment option to tamoxifen. Whether implemented in routine testing or not, the challenge in this is to conduct the genetic testing at the highest standards to avoid confounders. Future research should address those factors that influence CYP2D6 phenotype prediction and this will include other metabolizing enzymes for their contribution to tamoxifen efficacy. Current results of the ATLAS trial could influence tamoxifen therapy recommendations in prolonging duration of treatment up to 10 years. Therefore, predicting the drug's efficacy before the start of a long-term treatment will be of greater importance to aid both the patients and the healthcare system. Regarding AIs, future research will contribute to the understanding of a possible role of DMEs in therapy outcome.

It is Canada Day weekend this weekend, so off to my brother-in-law's cottage I, DH and my two sons go. Finished #14 of Herceptin today, yahoo!!! Welcome to the new thread members. May you find, solace, knowledge and compassion as you communicate with us. To all the American gals, in advance, Happy 4th of July!!

Aw, Pbrain! STOP COPYING ME! LOL. I hope your path diverges from mine, in that they will let you go back on herceptin. But, if not, like my MO said, um, you have to be alive to continue tx! So it is what it is. I would rather have continued herceptin too, but I at least got what I figured 6 months worth. With my weekly H, with H with each tx, I got about half, as I was able to get two more tx before my EF tanked. I did ask my MO if I ever recurred if I could go back on it, he said if we stopped it now, I probably could. Sorry you had to take that break, but perhaps yours is temporary. Much love.

Hi

I am new to this forum and was diagnosed last year as having grade 2 triple positive.  I am actually in the UK rather than US, but I decided to join this forum as well as UK ones as you have a triple positive group.

There are a few differences between the UK and US: our treatment is completely free so we don't have to worry about insurance issues, but perhaps patients in the US may get new drugs a little sooner.

Part of my treatment is at St Bartholomews (founded in 1123) located near St Pauls in the City of London (small area corresponding to medieval London), so it is an interesting area to wonder round after treatment.

I hope to find out more about triple positive cancer from your collective knowledge and experience.

Welcome Bluefox..

We've had a few Brits in this forum and others from all over the world. I've discovered that there are many different protocols that are used around the world and indeed in different regions, states, hospitals etc right here in the USA.

But most of the conversations we have here in this forum are universal to this experience. Ask questions, vent , cry....we've all been there . Someone will usually show up to help you shortly.

You are not alone.

Aw nuts pbrain!

I'll try to keep up with posts over the next few days, but keep falling asleep in mid post. Valium and percoset will do that I guess.

2 days post op and all is well.

You guys crack me up!  And thanks for the support.  Cami, Chandler dances like a white boy 

Bluefox, I have worked for the company that makes herceptin for 11 years (Roche) and I was horrified to find out I was Her2+.  I was good with everything I'd learned about my disease except that.  So then I started meeting with my doctors and nurses to get my treatment plan started I was stunned to see that they were actually pleased with the Her2 status because they have a targetted therapy.  I actually had one nurse give me the thumbs up when I told her I was triple positive.  I was shockered!

I can't answer you on the neoadjuvant chemo and your tumor response, but it sounds good to me.  I so wish I had some way of knowing I responded to all the junk they poured into me over the 15 weeks, but heck, I've survived and am here and feel good.  Really, I feel good finally!  YAY!!  I'm doing the Chandler dance!!!

Pbrain so sorry that you had to take a break from herceptin.

Hello BF's! (Breastie Friends),

I have been away a bit selling houses, working like crazy. I have not been able to work out for a couple weeks and it shows. But I felt llike I needed to make hay while the sun shines, lol.

I just glanced through the last few pages...PBrain...that sucks...can you go to your company and ask them what else they have in the pipeline? Tell them you need some juice!

On the port...my PS took mine out on one of my fat grafting trips. BS would have but she said since I was already having the surgery to just have him do it.

On uterine issues...my GYN has me on a 6 month ultrasound plan. I have some thickening. I did have an endometrial biopsy once several years ago, that was during my regular exam, so no pain meds. Hurt like hell. If I have one again, I'm taking drugs.

I agree with everyone on the fatigue. there has got to be something to that. I think we should all be getting shots of b vitamins or something for a boost. I ran myself ragged planning the trip to thailand and then was wiped out part of the time.

We are taking an Alaskan cruise in August with three other couples. I have been trying to plan and develop an agenda of what I want and somehow have become the tour coordinator. It is stressing me out. And like someone else noted, I worry about putting too much in, and then I worry that if I am only going once, then I need to pack it all in. Just trying to coordinate the logistics is exhausting.

Sometimes, it isn't always that I am tired, as in sleepy, but I just feel like I have no energy, or just achy. 

Welcome BlueFox--I really think u'll like it here, lots of info.

And pbrain that stupid dance picture took me forever to figure out how to post it, I just wanted to make someone smile cuz I don't know much at all so I do what I do best. Nothing much.

 BTW when I heard +++ I was so happy, not because of herceptin (I never heard of it) I just thought all positive is a good work and I liked it. My Dr. started to explain, that's when I started I don't want to know. I stuck with the good thought of guess what I'm positive.Sounds really positive to me.

How is everyone's weather make sure u keep hydrated--it's super hot in some places and cool here--i'm so glad I like to weat a sweter or a blanket around here. As far as I'm concerned it can stay this way.

Just popping in. Last week was a busy one. Managed to scan several pages.

Welcome BlueFox & DXat32

PBrain, sorry you are joining us on the Herceptin Holiday.  Are you being referred to a cardiologist?  I have been seeing one since December, when they pulled me off Herceptin.  I get echos every six weeks now. My last one showed a drop but my cardiologist says my heart is strong and pumping well and allowed me to stay on.  I had a 3 month break, and I've been on since January. New target date to finish is October for me.  I'm also on a low-dose ACE inhibitor and am taking COQ10.  I think seeing a cardiologist should be standard of care, but that's just my opinion.

Had my gall bladder out last Monday. Yay for four more scars on my abdomen.  I look like a human pin cushion. This was not the piece-of-cake, easy surgery people told me that it would be.  The pain was ridiculous and I spent more time being mad that I was so misinformed. I'm feeling better and will head to work today.

I'm still on my quest to find a local DIEP surgeon that I have confidence in.

Vballmom are u sure u'r ready to go back to work? I hope u don't rush it.

OK Vballmom as long as u can pick what and when u want then u have my OK--but still don't overdo, u might feel fine but in the long run u won't be, Did u have lapriscopy GB or cut into GB>