Dopamine D2 Receptor Gene Polymorphisms and Response to Cabergoline Therapy in Patients with Prolactin-Secreting Pituitary Adenomas

Home , Raclopride

Dopamine D2 receptor gene polymorphisms and response to cabergoline therapy in patients with prolactin-secreting pituitary adenomas

M Filopanti1, AM Barbieri1, Dopamine-agonist cabergoline (CB) reduces prolactin (PRL) secretion and 1 2 3 tumor size in 80% of patients with prolactin-secreting adenomas (PRL-omas) AR Angioni ,AColao , V Gasco , by binding type 2 dopamine receptor (DRD2). The mechanisms responsible 3 4 4 S Grottoli , A Peri , S Baglioni , for resistance to CB remain largely unknown. To assess the association of MF Fustini5, F Pigliaru6, DRD2 with sensitivity to CB, TaqI-A1/A2, TaqI-B1/B2, HphI-G/T and NcoI-C/T PD Monte7, G Borretta8, genotypes were determined in a cross-sectional retrospective study, B Ambrosi9, ML Jaffrain-Rea10, including 203 patients with PRL-oma. DRD2 alleles frequencies did not differ 11 12 between patients and 212 healthy subjects. Conversely, NcoI-T allele M Gasperi , S Brogioni , frequency was higher in resistant rather than responsive patients, considering S Cannavo` 13, G Mantovani1, both PRL normalization (56.6 vs 45.3%, P ¼ 0.038) and tumor shrinkage P Beck-Peccoz1, A Lania1 and (70.4 vs 41.4%, P ¼ 0.006). Finally, [TaqIA1À/TaqIB1À/HphITÀ/NcoITÀ] A Spada1 haplotype was found in 34.5% of patients normalizing PRL with p3 mg/ week of CB vs 11.3% of resistants (P ¼ 0.021). In conclusion, resistance to CB 1Department of Medical Sciences, University of Milan, was associated with DRD2 NcoI-T þ allele, consistent with evidence Endocrine and Diabetes Unit Fondazione Ospedale suggesting that this variant may lead to reduction and instability of DRD2 Maggiore IRCCS, Milan, Italy; 2Department of mRNA or protein. Molecular and Clinical Endocrinology and Oncology, The Pharmacogenomics Journal (2008) 8, 357–363; doi:10.1038/tpj.2008.1; Federico II University of Naples, Naples, Italy; 3Division of Endocrinology and Metabolism, published online 11 March 2008 Department of Internal Medicine, University of Turin, Turin, Italy; 4Endocrine Unit, Department of Clinical Keywords: prolactinomas; cabergoline; polymorphisms; DRD2 Physiopathology, Center for Research, Transfer and High Education DENOThe, University of Florence, Florence, Italy; 5Endocrinology Unit, Department of Medicine, Ospedale Bellaria, Bologna, Italy; 6Endocrinology, Department of Medical Sciences ‘M. Aresu’, University of Cagliari, Cagliari, Italy; 7Unit of Endocrinology, Galliera Hospital, Genoa, Italy; Introduction 8Endocrinology Unit, Santa Croce Hospital, Cuneo, Italy; 9Endocrinology Unit, Department of Medical Prolactin-secreting pituitary adenomas (PRL-omas) are rare and well-differen- and Surgical Sciences, University of Milan, IRCCS 10 tiated tumors constituted by lactotroph cells with an estimated incidence of 6–10 Policlinico San Donato, Milan, Italy; Department of 1 Experimental Medicine, University of L’Aquila and cases per million per year. The first-choice treatment of patients with PRL-omas Neuromed IRCCS, Pozzilli, Italy; 11Chair of is the dopamine agonist cabergoline (CB), a lysergic acid amide derivative, which Endocrinology, University of Molise, Campobasso, is a potent type 2 dopamine receptor (DRD2) agonist and causes less side effects 12 Italy; Department of Endocrinology, University of and a better pharmacological profile than other dopamine agonists. Resistance to Pisa, Pisa, Italy and 13Department of Medicine and Pharmacology, University of Messina, Messina, Italy CB, defined as failure to normalize PRL levels and/or to reduce tumor size, is reported in 10–15% of patients with PRL-omas.2 Reduction of dopaminergic Correspondence: binding sites, decreased expression of DRD2 mRNA and a relative decrease in Gai Professor A Spada, University of Milan, expression have been described in resistant PRL-omas, as well as reduced levels of Department of Medical Sciences, Endocrine the short receptor isoform.3–5 Finally, no mutations in the DRD2 gene have been and Diabetes Unit, Pad. Granelli, Via Francesco 6 Sforza 35, Milan 20122, Italy; so far described in PRL-omas resistant to medical treatment. E-mail: [email protected] Multiple alternative splicing, including that occurring in exon 6 and generating the two major long and short isoforms, have been described for the Received 18 July 2007; revised 12 December 7,8 2007; accepted 20 December 2007; published DRD2. Moreover, there are several single nucleotide polymorphisms of the online 11 March 2008 DRD2 gene that have mostly synonymous changes. Due to the critical role of DRD2 polymorphisms and response to cabergoline in PRL-omas M Filopanti et al 358

dopaminergic activity in the central nervous system, several heterozygotes, forming TaqIA1þ , TaqIB1þ , HphITþ and studies examined the association between common poly- NcoITþ groups, respectively. morphisms of the DRD2 gene and behavioral traits, No association between DRD2 polymorphisms and sex, psychiatric and neurological diseases, such as schizophrenia, age at diagnosis and PRL levels was observed in patients with alcoholism, smoking, opioids and polysubstance abuse, and PRL-oma (Table 2). PRL Z-scores did not significantly differ susceptibility to migraine.9–14 Accordingly, in vivo evidence between genotypes (TaqIA1À 275.4±440.0 vs A1 þ supporting the potential functional significance of DRD2 550.0±280.5, P ¼ 0.187; TaqIB1À 550.3±486.7 vs B1 þ variants has been accumulating. In particular, previous 267.4±442.3, P ¼ 0.176; HphITÀ 529.5±292.2 vs studies reported the association of DRD2 TaqIA,TaqIB TÀ 1171.7±1235.4, P ¼ 0.262; NcoITÀ 506.5±432.3 vs and NcoI polymorphisms with modifications in D2 receptor T þ 1328.0±753.6, P ¼ 0.592). Conversely, patients with binding and expression in human central nervous system, microadenoma showed a strong linkage disequilibrium particularly in the striatum.15–18 between TaqIA1À and HphITÀ polymorphisms (D0 ¼ 1.0, The aim of this study was to assess the possible association r2 ¼ 0.879), which was not present in patients with macro- between four common DRD2 gene polymorphisms (that is, adenoma. TaqI A1/A2, TaqI B1/B2, HphI G/T and NcoI C/T) and the efficacy of CB treatment in patients with PRL-oma. Genotype analysis in patients with PRL-omas and sensitivity to CB Results treatment When considering patients who achieved PRL normaliza- Clinical data tion and tumor-size reduction under CB treatment (from 0.5 Two hundred and three patients with PRL-omas (98 male to 3 mg/week), no association between DRD2 polymorph- and 105 female, mean age at diagnosis 36.6±15.0 years) isms and the effective CB dosage was observed. Similarly, no were enrolled in this retrospective cross-sectional study. As it association between TaqIA,TaqI B and HphI polymorphisms is known that the size of the tumor is one of the major and resistance to CB in terms of PRL and tumor size was determinants of PRL levels, the study included a similar found (Table 3). By contrast, the prevalence of NcoI number of patients with either micro- or macro-PRL-oma genotype was different in patients resistant to CB treatment (adenoma size o10 mm or X10 mm, respectively) consecu- in comparison with those who were responsive. In fact, NcoI tively seen by the different centers. Subsequently, about half T þ allele was identified in 56.6% (12/22) of patients (50.3%) of the patients had a macroadenoma. All patients resistant to CB and 45.3% (84/181) of those normalizing at had a clinical and hormonal reevaluation every 3–4 months CB p3 mg/week (Fisher’s exact test, P ¼ 0.038) (Table 3). In and magnetic resonance imaging every 6–9 months. During the group of patients with NcoI T allele, TT homozygotes follow-up (range: 18–38 months), the mean lowest dosage of were 36% in responsive subjects vs 50% in resistants and TC CB required for PRL normalization and reduction of the heterozygotes were 9 vs 4%, respectively. tumor size was 1.40±1.13 mg/week, 45.3% patients requir- This association was also found, considering the tumor ing o1 mg/week. Twenty-two (10.8%) patients failed to shrinkage in response to CB, as NcoITþ variant was present normalize PRL levels with CB at the dosage of 43 mg/week in 70.4% of patients (12/17) who did not reach a 430% after 3–4 months follow-up and were considered resistant. In tumor-size reduction at CB p3 mg/week vs 41.4% (77/186) this subset, 54.6% of patients were males, 68.2% had a in the responsive group (Fisher’s exact test, P ¼ 0.006) (Table macroadenoma and the median PRL was 452 ng/ml (inter- 3). In the set of NcoI T allele carriers, TT homozygotes were quartile range 257–12 230). Seventeen (8.4%) patients did 35% in responsive patients vs 65% in resistants, and TC not reduce tumor size at CB 43 mg/week. A discrepancy heterozygotes were 7 vs 6%, respectively. between hormonal and tumor size responses was observed A multivariate logistic regression analysis has been carried in five patients with macroadenoma (three males and two out using backward method, considering failure to normal- females) in whom 43 mg/week of CB induced a tumor-size ize PRL levels with CB at the dosage of 43 mg/week as reduction from 60 to 75% without normalizing PRL levels. dependent variable and with the following predictors: TaqI The increase of CB dosage to 4, 4.5, 5 and 9 mg/week in 17, A þ , TaqIBþ , Hph T þ, NcoITþ, PRL levels, presence of 2, 2 and 1 patient, respectively, did not overcome the macroadenoma and age at diagnosis. Only NcoITþ resistance in terms of PRL normalization and tumor-size genotype was a significant predictor of the unsuccessful reduction in any patient. normalization of PRL by CB therapy (odds ratio 5.181, 95% confidence interval 1.369–19.1645, P ¼ 0.015). As far as the Genotype analysis in normal subjects and in patients with failure in tumor shrinkage was considered, none of the PRL-omas above predictors were retained in the residual model. The four DRD2 polymorphisms were genotyped in 203 Finally, haplotype analysis of DRD2 polymorphisms patients with PRL-oma and 212 healthy subjects. No showed that [TaqIA1À/TaqIB1À/HphITÀ/NcoITÀ] combi- statistically significant differences in genotypes and alleles nation was found in 34.5% of patients taking p3 mg/week proportions were found between the two groups (Table 1). to normalize PRL levels vs 11.3% of resistant patients Due to the low prevalence of TaqI A1, TaqI B1, HphI T and (Fisher’s exact test, P ¼ 0.021). Such association was not NcoI T alleles, their homozygotes were grouped with relative present when the effect on tumor shrinkage was evaluated.

The Pharmacogenomics Journal DRD2 polymorphisms and response to cabergoline in PRL-omas M Filopanti et al 359

Table 1 Allele frequencies of TaqIA,TaqIB,HphI and NcoI DRD2 polymorphisms in 203 patients affected with PRL-oma and 212 healthy subjects

PRL-omas Controls P

TaqIA A1–A1 7, 0.03 (0.01–0.07) 3, 0.01 (0.00–0.04) 0.412 A1–A2 58, 0.29 (0.23–0.35) 61, 0.29 (0.23–0.35) A2–A2 138, 0.68 (0.61–0.74) 148 0.70 (0.63–0.76)

Allele A1+ 65, 0.32 (0.26–0.39) 64, 0.30 (0.24–0.37) 0.774 Allele A1À 196, 0.97 (0.93–0.98) 209, 0.99 (0.96–1.00)

TaqIB B1–B1 7, 0.03 (0.01–0.07) 1, 0.00 (0.00–0.03) 0.051 B1–B2 59, 0.29 (0.23–0.36) 52, 0.24 (0.19–0.31) B2–B2 137, 0.67 (0.61–0.74) 159, 0.75 (0.69–0.80)

Allele B1+ 66, 0.32 (0.27–039) 53, 0.25 (0.20–0.31) 0.137 Allele B1À 196, 0.97 (0.93–0.99) 211, 0.99 (0.97–1.00)

HphI T–T 7, 0.34 (0.01–0.07) 2, 0.01 (0.00–0.04) 0.253 T–G 51, 0.25 (0.20–0.31) 52, 0.24 (0.19–0.31) G–G 145, 0.71 (0.65–0.77) 158, 0.74 (0.68–0.80)

Allele T+ 58, 0.29 (0.23–0.35) 54, 0.25 (0.20–0.32) 0.542 Allele TÀ 196, 0.97 (0.93–0.98) 210, 0.99 (0.96–1.00)

NcoI T–T 18, 0.09 (0.06–0.14) 22, 0.10 (0.07–0.15) 0.540 T–C 78, 0.38 (0.32–0.45) 90, 0.42 (0.36–0.49) C–C 107, 0.53 (0.46–0.59) 100, 0.47 (0.41–0.54)

Allele T+ 96, 0.47 (0.40–0.54) 105, 0.49 (0.43–0.56) 0.746 Allele TÀ 186, 0.92 (0.87–0.95) 175, 0.90 (0.77–0.87)

Entries are expressed as number of cases, frequencies (95% confidence interval). Variables were compared using w2 test.

Discussion lactotroph proliferation,2 the possible association of DRD2 polymorphisms with PRL-omas has been so far disregarded. This study first investigated the presence of DRD2 poly- Indeed, the few studies investigating the possible relation- morphisms in patients with PRL-omas and provided evi- ships between PRL secretion and DRD2 variants focused on dence for the association of the NcoI C/T polymorphism the impact of DRD2 polymorphisms on PRL levels in with resistance to CB in these patients. Emerging evidence patients with schizophrenia during treatment with a variety suggests that, although most polymorphisms in the coding of antipsychotic medications.27 and noncoding regions are functionally neutral, some We focused our analysis on four frequent synonymous variations, even when synonymous, may have functional variations, TaqI A1/A2, TaqI B1/B2, HphI G/T and NcoI C/T, consequences.19–21 Multiple polymorphisms as well as which have been thought to be associated with neuropsy- alternative splicing have been described in DRD2.As chiatric diseases, eating disorders and personality traits, dopamine D2 receptor is involved in motor control, although studies attempting to replicate these data fre- cognition and drug-related reward and reinforcement, quently gave discordant results. In particular, Taq1 A1/A2 DRD2 polymorphic variants have been extensively analyzed (located at 30 of DRD2 and recently reported to be a for disease-association studies, particularly focused on polymorphism of ankyrin repeat and kinase domain con- psychiatric/neurological phenotypes.22–26 Moreover, some taining 1 ANKK1 gene) has been found to be associated with of these synonymous variations have been demonstrated to schizophrenia, severe alcoholism and other substance-use affect D2 receptor expression in central nervous system, disorders as well as obesity. Similarly, TaqI B1/B2 (located in particularly in the striatum and caudate nucleus.15–18 By intron 1, 882 bp upstream exon 2) and HphI G/T (located contrast, although it is well established that defective in intron 6, 83 bp upstream exon 7) have been thought to dopamine signaling may result in PRL hypersecretion and be associated with alcohol and drug dependence, whereas

The Pharmacogenomics Journal DRD2 polymorphisms and response to cabergoline in PRL-omas M Filopanti et al 360

Table 2 DRD2 gene polymorphisms and clinical features in 203 patients with prolactin-secreting pituitary adenoma

Female gender Pa Macroadenoma Pb PRL levels (ng/ml) Pa

TaqIA A2–A2 (138) A1À (138) 0.59 0.51 43±154 A1–A2 (58) [0.57] 1.0 [0.50] 0.77 [63±55] 0.75 A1+ (65) 0.59 0.53 86±43 A1-A1 (7) [0.71] [0.86] [103±39]

TaqIB B2–B2 (137) B1À (137) 0.58 0.51 86±73 B1–B2 (59) [0.56] 0.88 [0.53] 0.66 [65±91] 0.42 B1+ (66) 0.59 0.54 42±155 B1–B1 (7) [0.71] [0.71] [39±77]

HphI G–G (145) TÀ (145) 0.57 0.51 80±45 T–G (51) [0.57] 1.0 [0.53] 0.64 [240±189] 0.11 T+ (58) 0.57 0.55 236±224 T–T (7) [0.71] [0.71] [142±146]

NcoI C–C (107) TÀ (107) 0.57 0.50 80±62 T–C (78) [0.60] 0.67 [0.54] 0.89 [248±139] 0.55 T+ (96) 0.60 0.61 244±182 T–T (18) [0.61] [0.44] [139±75]

Abbreviation: PRL, prolactin. Due to the low prevalence of TaqI A1, TaqIB1,HphI T and NcoI T alleles, their homozygotes were grouped with relative heterozygote, forming TaqI A1+, TaqI B1+, HphI T+ and NcoI T+ groups, respectively. Statistical tests were performed for each polymorphism, between two groups of genotypes with (+) or without (À) the specified allele. The number of patients is indicated in parenthesis. Female gender and macroadenoma are expressed as frequencies. Data in square brackets refer to homozygous or heterozygous for polymorphic variants. aFisher’s exact test. bMann–Whitney test.

NcoI C/T (a synonymous cytosine to thymine transition at according to the size of the adenoma, a strong linkage position 957) has been thought to be associated with disequilibrium between TaqIA1À and HphITÀ polymorph- schizophrenia and susceptibility to migraine.9–14,22–26 isms was found only in patients bearing microadenomas. By genotyping about 200 controls and patients with Taking into account the association between TaqI A2 and prolactinomas, we found similar allelic frequencies of TaqI HphI G alleles and high D2 dopamine receptor density in A1/A2, TaqI B1/B2, HphI G/T and NcoI C/T in the two groups caudate nuclei samples reported previously,29 these data are con- that were comparable to those of European and American sistent with the notion that microadenomas are generally populations (NCBI, Entrez SNP database, TaqI A: http:// well responsive to CB therapy,2 as also shown in this study. www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?rs ¼ 1800497; TaqIB: We then investigated the association of DRD2 polymorph- http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?rs ¼ 1079597; isms with the responsiveness to cabergoline, a potent HphI: http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?rs ¼ dopamine D2 receptor agonist that is currently the first- 1076560; NcoI: http://www.ncbi.nlm.nih.gov/SNP/snp_ref. choice treatment for patients with PRL-oma. As tumor size is cgi?rs ¼ 6275). We then investigated the possible associa- one of the most important determinants of resistance to tions between the different genotypes and clinical and dopaminergic drugs, a similar number of patients with hormonal characteristics of patients. No statistically sig- micro- and macroprolactinomas were genotyped according nificant differences in genotype and allele proportions were to the study protocol. Criteria for the definition of resistance found between the two groups, thus ruling out a role of to dopaminergic agents as well as the standard CB dose receptor variants in prolactinoma development. thresholds required for the assessment of this status are not As far as the clinical phenotype of patients was concerned, clearly established.2 For the purpose of this genotype/ no association between DRD2 polymorphisms and sex, age phenotype analysis, patients were defined as resistant when at diagnosis and PRL levels was observed in patients with they failed to normalize PRL levels taking 43 mg/week CB. PRL-oma. These data are at variance with a previous study In agreement with previous studies,2 resistance to CB was investigating the impact of variants of somatostatin receptor observed in about 10% of patients and was associated with a isoforms and the clinical characteristics of patients with lack of tumor-size reduction. Among resistant patients, acromegaly. Indeed, in this cohort of patients, polymorphic patients with DRD2 NcoITþ genotype, that is, homozygotes variants of somatostatin receptor type 5 were associated or heterozygotes for 957T allele, were significantly more with growth hormone and insulin like growth factor-I frequent than patients with DRD2 NcoITÀ, that is, levels.28 When we analyzed patients with prolactinoma homozygotes for C957. This association was further con-

The Pharmacogenomics Journal DRD2 polymorphisms and response to cabergoline in PRL-omas M Filopanti et al 361

Table 3 TaqIA,TaqIB,HphI and NcoI DRD2 polymorphisms in patients with PRL-oma responsive or resistant to cabergoline treatment in terms of prolactin normalization and tumor-size reduction

PRL normalization (% patients) Tumor-size reduction (% patients)

Responsive a Resistant Responsive b Resistant

TaqI A1+ 33.0 (60/181) 37.5 (8/22) 33.3 (62/186) 35.3 (6/17) TaqIA1À 67.0 (121/181) 62.5 (14/22) 66.7 (124/186) 64.7 (11/17)

TaqI B1+ 30.4 (55/181) 43.8 (10/22) 31.7 (59/186) 35.3 (6/17) TaqIB1À 69.6 (126/181) 56.3 (12/22) 68.3 (127/186) 64.7 (11/17)

HphI T+ 28.2 (51/181) 37.5 (8/22) 29.6 (55/186) 23.5 (4/17) HphITÀ 71.8 (130/181) 62.5 (14/22) 70.4 (121/186) 76.5 (13/17)

NcoI T+ 45.3 (84/181) 56.6c (12/22) 41.4 (77/186) 70.4d (12/17) NcoITÀ 54.7 (97/181) 43.4 (10/22) 58.7 (102/186) 29.6 (5/17)

Abbreviation: PRL, prolactin. Due to the low prevalence of TaqI A1, TaqI B1, HphI T and NcoI T alleles, their heterozygotes were grouped with relative homozygotes, forming TaqI A1+, TaqI B1+, HphI T+ and NcoI T+ groups, respectively. aPatients who normalized PRL levels taking CB at the dosage of p3 mg/week were considered responsive. bPatients who showed a 430% reduction of tumor size taking CB at the dosage of p3 mg/week were considered responsive. cP ¼ 0.038 vs responsive patients. dP ¼ 0.006 vs responsive patients. Fisher’s exact test. firmed by the observation that the haplotype negative for that subjects carrying the 957T allele had markedly lower NcoITþ allele, that is, [TaqIA1À/TaqIB1À/HphITÀ/NcoI striatal D2 receptor availability than those with C975.17 It is TÀ], was associated with a greater sensitivity to CB therefore tempting to speculate that a similar reduction of treatment in terms of PRL normalization. This is consistent D2 receptors in subjects with 957T allele may also occur at with the notion that normalization of PRL levels and the pituitary level. In this respect, it would be of interest to reduction in tumor mass are frequently concomitant events. test whether the only changes so far identified in PRL-omas DRD2 NcoITþ genotype was also significantly associated removed from resistant patients, that is, reduction and with resistance to the antitumoral action of CB. instability of DRD2 mRNA or protein,3–5 were associated The polymorphism examined by the restriction enzyme with the presence of 957T variant. NcoI is a common synonymous cytosine to thymine In conclusion, this study suggests that 957T variant of transition at position 957 that has been associated pre- DRD2 was associated with resistance to dopaminergic drugs viously with susceptibility to migraine and schizo- in patients with PRL-oma. This in vivo neurobiological phrenia.12,30,31 Interestingly, this polymorphism has been correlate further supports the view that C957T may reported to have an endocrine correlate.32 In fact, a recent have functional consequences and suggests that this study carried out on a cohort of subjects from the Sweden polymorphism might be considered an interesting candi- National Population Registry reported a high prevalence of date for genetic predisposition studies in DRD2-associated 957T genotype in subjects with high systolic and diastolic neuroendocrinologic disorders. Further studies are needed blood pressure. On the basis of this association, the authors to assess the clinical implication of genotyping patients with suggested that this variant might cause defective dopamine prolactinoma for DRD2 polymorphisms. signaling, resulting in elevated blood pressure due to deficient inhibition of noradrenaline release from sympa- thetic nerve endings and aldosterone secretion from the Materials and methods adrenals.32 Indeed, increasing in vitro and in vivo evidence indicates that this change in DRD2, though synonymous, Subjects has marked functional consequences. In particular, it has A computerized database was developed using Microsoft been demonstrated that 957T, rather than being silent, Access 2000 software (Microsoft Corporation, 1999), as altered the predicted mRNA folding, leading to a decrease in described previously.33 The database was divided into two mRNA stability and translation.18 In fact, in DRD2-trans- parts: a principal table, including information about the fected CHO-K1 cells, the DRD2 mRNA carrying the 957T patient at the time of recruitment, and several follow-up allele decayed significantly faster than mRNA carrying C957. records. Each patient received an identification number, and This event resulted in a 50% decrease in the receptor personal information was available only to principal synthesis.18 Moreover, by investigating in vivo DRD2 bind- investigators or coinvestigators. Patients (n ¼ 203) with ing in healthy subjects using [11C]raclopride and PET PRL-oma were collected in 11 Italian Endocrine and (positron emission tomography), it has been demonstrated Neurosurgery Departments between 1998 and 2005. In

The Pharmacogenomics Journal DRD2 polymorphisms and response to cabergoline in PRL-omas M Filopanti et al 362

particular, the inclusion criteria were the following: (1) containing 1 (ANKK1) gene; TaqI B1/B2 (rs1079597), confirmed hyperprolactinemia by blood sample collections localized in intron 1, 882 bp upstream of exon 2, causing in supine patients during a slow saline infusion; (2) an adenine to guanine transition; HphI G/T (rs1076560) exclusion of macroprolactin by polyethylene glycol pre- located in intron 6, 83 bp upstream of exon 7, causing a cipitation test; (3) no hyperprolactinemic drugs, pregnancy guanine-to-thymine transversion; NcoI C/T (C957T, rs6275), or major psychiatric illness; in particular, in all women causing a synonymous cytosine-to-thymine transition at the estrogen was withdrawn before diagnosis and eventually third nucleotide of the codon 313. The primers used to restarted after PRL normalization during CB treatment in amplify TaqI A polymorphism were 50-CCGTCGACGGCT post-menopausal women; (4) evidence of pituitary adenoma GGCCAAGTTGTCTA-30 (forward) and 50-CCGTCGACCCTT at magnetic resonance imaging. All patients had a clinical CCTGAGTGTCATCA-30 (reverse), whereas for TaqI B poly- and hormonal reevaluation every 3–4 months and magnetic morphism the primers were 50-GATACCCACTTCAG resonance imaging every 6–9 months. CB treatment was GAAGTC-30 (forward) and 50-GATGTGTAGGAATTAGCCA started with 0.25–0.5 mg/week and the dosage was subse- GG-30 (reverse). HphI and NcoI polymorphisms were en- quently titrated up to 3 mg/week according to biochemical closed in one polymerase chain reaction (PCR) product by and tumor-size response, during periodical follow-up exam- this primers pair: 50-GGCATGGTCTGGATCTCA-30 (forward) inations. The criteria for responsiveness to the therapy were and 50-TGACCCTGTGGTGTTTGC-30 (reverse). PCRs were the following: (1) normalization of PRL levels at CB dosage carried out in a 50-ml reaction mix with 0.8 U Taq DNA of p3 mg/week; (2) reduction of the tumor size more than polymerase DyNAzyme (Finnzymes, Espoo, Finland) and 30% at CB dosage of p3 mg/week, an arbitrary cutoff in the subjected to denaturation at 95 1C for 5 min, followed by 35 range of the proposed cutoffs (20–80) used in literature.2 cycles of 95 1C for 30 s, 58 1C for 30 s and 72 1C for 30 s. A Finally, as it is known that the size of the tumor is one of final cycle at 72 1C for 5 min was carried out to allow the major determinants of PRL levels, the study included a complete extension of the amplified fragments. TaqI similar number of patients with either micro- or macro-PRL- enzyme digestions were performed in 50 ml of reaction mix oma consecutively seen by the different centers. As CB with 10 U of TaqI restriction enzyme (New England Biolabs, dosages used in these patients were definitely lower than Ipswich, MA, USA) and 17.5 ml of PCR product. The reactions 2–5 mg/day, a dosage reported to be responsible for were incubated at 65 1C overnight. NcoI and HphI enzyme cardiac valvulopathy in subjects affected with Parkinson digestion were performed together in 50 ml of reaction mix disease, periodical ultrasound heart examination was not with 2 U of NcoI and 2.5 U of HphI (New England Biolabs) carried out.34 and 17.5 ml of PCR product. The reactions were incubated at Genotyping was performed on blood DNA from 212 sex- 37 1C overnight. Digestions were resolved on a 3% agarose gel. matched healthy subjects as controls. Written informed Finally, 50 randomly selected samples underwent direct consent was obtained from all the participants in the study, sequencing (primers and conditions used available upon and the study was approved by the Ethics Committee of request). In all cases, direct sequencing data confirmed the Fondazione Ospedale Maggiore. results of RFLP analysis. Statistics Prolactin assay Quantitative variables were reported as mean±s.d. or as In the participating centers, immunoreactive PRL levels medians and interquartile ranges if variables were not were measured in duplicate using the commercial assay kits normally distributed, unless otherwise indicated. Differ- reported below: PRL-CTK IRMA (Sorin Biomedica, Saluggia, ences among means were evaluated by t-test, in agreement Italy) with normality range 3–20 ng/ml in females and 3– with normal distribution. Non-Gaussian variables were 13 ng/ml in males; PRL 1-step immunoluminometric sand- analyzed using Mann–Whitney test. w2 analysis or Fisher’s wich assay using directly coated magnetic microparticles exact test was used to test the difference between categorical (DiaSorin, Saluggia, Italy) with normality range 5–25 ng/ml variables. Two-tailed P-value less than 0.05 was considered in females and 4–20 ng/ml in males; Delfia (PerkinElmer statistically significant. Multivariate logistic regression ana- Inc., Waltham, MA, USA) with normality range 3–15 ng/ml lysis was carried out with backward elimination method in females and 2–10 ng/ml in males and RAD Prolactin based on the likelihood-ratio statistic, using X0.10 as (IRMA, Radim, Pomezia, Italy) with normality range 5– removal threshold. Po0.05 was considered to be a residual 25 ng/ml in females and 5–15 ng/ml in males. Z-scores were statistical significance. Calculations were performed by SPSS calculated for each prolactin assay. 11.0 software (SPSS, Paris, France).

Genetic analysis Blood samples were collected from the patients, and leukocyte DNA was extracted by Nucleon BACC2 genomic Abbreviations DNA purification kit (GE Healthcare, Piscataway, NJ, USA) in compliance with the manufacturer’s instructions. CB cabergoline Taq DRD2 type 2 dopamine receptor The following polymorphisms were analyzed: I A1/A2 PRL prolactin 0 (rs1800497), located at the 3 of DRD2 and recently reported PRL-oma PRL-secreting pituitary adenoma to be a polymorphism of ankyrin repeat and kinase domain

The Pharmacogenomics Journal DRD2 polymorphisms and response to cabergoline in PRL-omas M Filopanti et al 363

Acknowledgments predicts low D2 receptor availability in healthy volunteers. Mol Psychiatry 1998; 3: 256–260. This work was partially supported by FIRST, Fondazione Ospedale 17 Jonsson EG, Nothen MM, Grunhage F, Farde L, Nakashima Y, Propping P et al. Polymorphisms in the dopamine D2 receptor gene and their Maggiore IRCCS and Associazione Italiana per la Ricerca sul Cancro relationships to striatal dopamine receptor density of health volunteers. AIRC (Milan). We are grateful to Drs EP Noble and T Ritchie for Mol Psychiatry 1999; 4: 290–296. advice concerning analysis of NcoI and HphI polymorphisms. 18 Duan J, Wainwright MS, Comeron JM, Saitou N, Sanders AR, Gelernter J et al. Synonymous mutations in the human dopamine receptor D2 (DRD2) affect mRNA stability and synthesis of the receptor. Hum Mol Duality of interest Genetics 2003; 12: 205–216. 19 Akashi H. Gene expression and molecular evolution. Curr Opin Genet None declared. 2001; 11: 660–666. 20 Iida K, Akashi H. A test of translational selection at ‘silent’ sites in the human genome: base composition comparisons in alternatively spliced References genes. Gene 2000; 261: 93–105. 21 Carlini DB, Chen Y, Stephan W. The relationship between third-codon 1 Ezzat S, Asa SL, Couldwell WT, Barr CE, Dodge WE, Vance ML et al. The position nucleotide content, codon bias, mRNA secondary structure prevalence of pituitary adenomas: a systematic review. Cancer 2004; and gene expression in the drosophilid alcohol dehydrogenase genes 101: 613–619. Adh and Adhr. Genetics 2001; 159: 623–633. 2 Gillam MP, Molitch ME, Lombardi G, Colao A. Advances in the 22 Noble EP. D2 dopamine receptor gene in psychiatric and neurologic treatment of prolactinomas. Endocr Rev 2006; 27: 485–534. disorders and its phenotypes. Am J Med Genet B Neuropsychiatr Genet 3 Pellegrini I, Rasolonjanahary R, Gunz G, Bertrand P, Delivet S, Jedynak 2003; 116: 103–125. CP et al. Resistance to bromocriptine in prolactinomas. J Clin Endocrinol 23 Sharfetter J. Pharmacogenetics of dopamine receptors and response to Metab 1989; 69: 500–509. antipsychotic drugs in schizophrenia—an update. Pharmacogenomics 4 Caccavelli L, Feron F, Morange I, Rouer E, Benarous R, Dewailly D et al. 2004; 5: 691–698. Decreased expression of the two D2 dopamine receptor isoforms in 24 Sakumoto N, Kondo T, Mihara K, Suzuki A, Yasui-Furukori N. Dopamine bromocriptine resistant prolactinomas. Neuroendocrinol 1994; 60: D2 receptor gene polymorphisms predict well the response to 314–322. dopamine antagonists at therapeutic dosages in patients with schizo- 5 Caccavelli L, Morange-Ramos I, Kordon C, Jaquet P, Enjalbert A. phrenia. Psychiatry Clin Neurosci 2007; 61: 174–180. Alteration of Ga subunits mRNA levels in bromocriptine resistant 25 Robinson JD, Lam CY, Minnix JA, Wetter DW, Tomlinson GE, Minna JD adenomas. J Neuroendocrinol 1996; 8: 737–746. et al. The DRD2 TaqI-B polymorphism and its relationship to smoking 6 Friedman E, Adams EF, Hoog A, Geyman PV, Carson E, Larsson C et al. abstinence and withdrawal symptoms. Pharmacogenomics J 2007; 7: Normal structural dopamine type 2 receptor gene in prolactin secreting 266–274. and other pituitary tumors. J Clin Endocrinol Metab 1994; 78: 568–574. 26 Southon A, Walder K, Sanigorski AM, Zimmet P, Nicholson GC, 7 Grandy DK, Marchionni MA, Makam H, Stofko RE, Alfano M, Kotowicz MA et al. The Taq IA and Ser311 Cys polymorphisms in the Frothingham L et al. Cloning of the cDNA and gene for a human D2 dopamine D2 receptor gene and obesity. Diabetes Nutr Metab 2003; dopamine receptor. Proc Natl Acad Sci USA 1989; 86: 9762–9766. 16: 72–76. 8 Usiello A, Baik JH, Rouge-Pont F, Picetti R, Dierich A, LeMeur M et al. 27 Young R, Lawford BR, Barnes M, Burton SC, Ritchie T, Ward WK et al. Distinct functions of the two isoforms of dopamine D2 receptors. Nature Prolactin levels in antipsychotic treatment of patients with 2000; 408: 199–203. schizophrenia carrying the DRD2+A1 allele. Br J Psychiatry 2004; 185: 9 Parsons MJ, Mata I, Beperet M, Iribarren-Iriso F, Arroyo B, Sainz R et al. A 147–151. dopamine D2 receptor gene-related polymorphism is associated with 28 Filopanti M, Ronchi C, Ballare E, Bondioni S, Lania AG, Losa M et al. schizophrenia in a Spanish population isolate. Psychiatr Genet 2007; 17: Analysis of somatostatin receptors 2 and 5 polymorphisms in patients 159–163. with acromegaly. J Clin Endocrinol Metab 2005; 90: 4824–4828. 10 Itokawa M, Arinami T, Futamura N, Hamaguchi H, Toru M. A structural 29 Ritchie T, Noble EP. Association of seven polymorphisms of the D2 polymorphism of human dopamine D2 receptor, D2 (Ser311Cys). dopamine receptor gene with brain receptor-binding characteristics. Biochem Biophys Res Commun 1993; 196: 1369–1375. Neurochem Res 2003; 28: 73–82. 11 Seeman P, Ohara K, Ulpian C, Seeman MV, Jellinger K, Van Tol HH et al. 30 Lawford B, Young R, Swagell C, Barnes M, Burton S, Ward W et al. The Schizophrenia: normal sequence in the dopamine D2 receptor region C/C genotype of the C957T polymorphism of the dopamine D2 that couples to G-proteins DNA polymorphisms in D2. Neuropsycho- receptor is associated with schizophrenia. Schizophr Res 2005; 73: pharmacology 1993; 8: 137–142. 31–37. 12 Peroutka SJ, Price SC, Wilhoit TL, Jones KW. Comorbid migraine with 31 Hoenicka J, Araguë´sM,Rodrı´guez-Jimeńez R, Ponce G, Martıńez I, Rubio aura, anxiety, and depression is associated with dopamine D2 receptor G et al. C957T DRD2 polymorphism is associated with schizophrenia in (DRD2) NcoI alleles. Mol Med 1998; 4: 14–21. Spanish patients. Acta Psychiatr Scand 2006; 114: 435–438. 13 Wong AH, Buckle CE, Van Tol HH. Polymorphisms in dopamine 32 Rosmond R, Rankinen T, Chagnon M, Pe´russe L, Chagnon YC, Bouchard receptors: what do they tell us? Eur J Pharmac 2000; 410: 183–203. C et al. Polymorphism in exon 6 of the dopamine D(2) receptor gene 14 Munafo` MR, Johnstone EC, Welsh KI, Walton RT. Association between (DRD2) is associated with elevated blood pressure and personality the DRD2 gene Taq1A (C32806T) polymorphism and alcohol con- disorders in men. J Hum Hypertens 2001; 15: 553–558. sumption in social drinkers. Pharmacogenomics J 2005; 5: 96–101. 33 Ferrante E, Ferrarono M, Castrignano` T, Menicatti L, Anagni M, 15 Thompson J, Thomas N, Singleton A, Piggott M, Lloyd S, Perry EK et al. Reimondo G et al. Non-functioning pituitary adenoma database: a D2 dopamine receptor gene (DRD2) TaqIA polymorphism: reduced useful resource to improve the clinical management of pituitary tumors. dopamine D2 receptor binding in the human striatum associated with Eur J Endocrinol 2005; 155: 823–829. the A1 allele. Pharmacogenetics 1997; 7: 479–484. 34 Zanettini R, Antonini A, Gatto G, Gentile R, Tesei S, Pezzoli G. Valvular 16 Pohjalainen T, Rinne JO, Nagren K, Lehikoinen P, Anttila K, Syvalahti heart disease and the use of dopamine agonists for Parkinson’s disease. EKG et al. The A1 allele of the human D2 dopamine receptor gene N Engl J Med 2007; 356: 39–46.

The Pharmacogenomics Journal