Abstract
Previous studies have found an association between elevated circulating prolactin levels and increased risk of breast cancer. Prolactin stimulates breast cancer cell proliferation, migration, and survival via binding to the cell-surface prolactin receptor. The association of prolactin receptor expression with breast tumorigenesis remains unclear as studies that have focused on this association have had limited sample size and/or information about tumor characteristics. Here, we examined the association of prolactin expression with tumor characteristics among 736 cases, from a large population-based case–control study of breast cancer conducted in Poland (2000–2003), with detailed risk factor and pathology data. Tumors were centrally reviewed and prepared as tissue microarrays for immunohistochemical analysis of prolactin receptor expression. Association of prolactin receptor expression across strata of tumor characteristics was evaluated using χ 2 analysis and logistic regression. Prolactin receptor expression did not vary by menopausal status; therefore, data from pre- and post-menopausal women were combined in the analyses. Approximately 83 % of breast cancers were categorized as strong prolactin receptor staining. Negative/low prolactin receptor expression was independently associated with poorly differentiated (p = 1.2 × 10−08) and larger tumors (p = 0.0005). These associations were independent of estrogen receptor expression. This is the largest study to date in which the association of prolactin receptor expression with tumor characteristics has been evaluated. These data provide new avenues from which to explore the associations of the prolactin/prolactin receptor signaling network with breast tumorigenesis.
Electronic supplementary material
The online version of this article (doi:10.1007/s12672-013-0165-7) contains supplementary material, which is available to authorized users.
Keywords: Breast Cancer, Progesterone Receptor, Estrogen Receptor Status, Breast Cancer Risk Factor, Menopausal Hormone Therapy
Background
Prolactin (PRL) and its receptor (PRLR) are necessary for the development and differentiation of the normal breast [1–5]. When PRL binds to PRLR, the Janus kinase 2 (Jak2)/signal transducers and activators of transcription 5 (Stat5), phosphoinositide 3-kinase (PI3K), and mitogen-activated protein kinases (MAPK) signaling pathways are activated, resulting in the enhanced survival, proliferation, and differentiation of breast epithelial cells [1, 2, 5]. These signaling pathways lead to the expansion of the epithelial cell population in the breast during pregnancy and promote differentiation of the cells that synthesize and secrete milk during lactation.
Evidence from both laboratory and epidemiology studies support a role for PRL and PRLR signaling in breast tumorigenesis. PRL can stimulate cell proliferation, migration, and survival of human breast cancer cells in culture [2, 5–7]. In vivo, high PRL levels are significantly associated with increased breast cancer risk in both large prospective and case–control studies [5, 8–13], independent of established breast cancer risk factors. These results are consistent across both pre- and post-menopausal women. A recent analysis of women from Nurses’ Health Study I and II found that higher PRL levels from samples collected less than 10 years prior to diagnosis were associated with breast cancer risk, irrespective of menopausal status (relative risk [RR] = 1.20, 95 % confidence interval [CI] = 1.03–1.40) [13]. When this analysis was restricted to only post-menopausal women with estrogen receptor (ER)-positive disease, the risk estimate increased to RR = 1.52 (95 % CI 1.19–1.93) [13]. In a large case–control study, we found that among post-menopausal women, higher PRL levels were associated with increased risk of breast cancer (odds ratio (OR) = 1.76, 95 % CI 1.21–2.57, p = 0.003) [12]. Among controls in both studies cited above, circulating PRL levels have been associated with select breast cancer risk factors (e.g., parity, body mass index, menopausal hormone therapy use, and mammographic density). In contrast, data relating PRL levels to tumor characteristics are limited and conflicting [5, 8, 10–12].
Epidemiologic studies of PRL and PRLR in association with human breast cancer have predominantly focused on circulating PRL levels. Analyses of PRLR expression in breast cancer have been limited by small sample sizes (<70 cases), no or partial risk factor information from the patients who contributed tissue to these studies, and use of varying methods for assessing PRLR, including in situ hybridization and reverse-transcriptase PCR for assessing mRNA expression and immunohistochemistry employing different reagent antibodies and staining protocols to assess protein levels [5, 14–19]. These analyses have generally not found significant relationships between PRLR protein expression and histological type, grade, size, or node status [14, 15, 18, 19]. These reports also yielded mixed results regarding the association of PRLR expression with estrogen receptor expression [14, 15, 18].
With the emergence of elevated circulating PRL levels as an independent breast cancer risk factor, it is important to know whether PRLR expression is associated with specific tumor characteristics. Here, we examine the associations of PRLR protein expression with tumor characteristics among cases in a large population-based case–control study conducted in Poland. Given that associations between PRLR status and other tumor characteristics could be modified by circulating PRL levels, we repeated these analyses for tumors cross-classified by PRL levels and PRLR expression when possible because this has not been well explored.
Methods
Study Population
The population and design of the Polish case–control study has been reported elsewhere [20]. Participants provided written informed consent and the study protocol was approved by ethical review boards in Poland and the USA. Briefly, eligible cases were women ages 20–74 years diagnosed with pathologically confirmed breast cancer from 2000 to 2003 while living in Warsaw or Lodz, Poland. The Polish Electronic System, a database with demographic information of Polish citizens was used to randomly select controls, defined as women without breast cancer, frequency matched to cases on city and age in 5-year categories. In total, 2,386 cases (79 % of eligible) and 2,502 controls (69 % of eligible) consented to participate in an interview regarding breast cancer risk factors. Of the 2,386 cases identified, 2,143 were found to have an invasive diagnosis. Paraffin-embedded tumor tissue from 1,437 (67 % of invasive cases) was prepared as tissue microarrays (TMAs). Of these 1,437 cases, 336 cases had missing cores leaving 1,101 cases for PRLR analyses.
Pathology and Immunohistochemical Staining
Histopathological features, including histology, grade (based on modified Elston grading criteria that include architecture, nuclear atypia, and mitotic rate), tumor size, and axillary lymph node metastases were assessed using surgical pathology reports and independent review (M.E.S.). Invasive breast cancers were prepared as TMAs with duplicate representation as 0.6-mm diameter cores (Beecher Instruments). The methods and results for immunohistochemical stains for ER, progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) were reported previously [21–23]. In brief, immunohistochemistry (IHC) staining was done with antigen retrieval prior to antibody incubation according to established protocols for ER-α (clone 6 F11, 1:200; Novocastra, Newcastle upon Tyne, UK), PR (clone PgR636, 1:1,000; Dako), and HER2 (polyclonal, 1:2,000; DakoCytomation, Glostrup, Denmark). The percentage (0–100 %) and intensity (0 = negative, 1 = weak, 2 = intermediate, and 3 = strong) of tumor cells stained were recorded for each marker. ER and PR were classified as positive if the product of intensity and percentage was >10. HER2 was considered positive if intermediate or strong staining was identified in at least 20 % of tumor cells. In a prior study, ER, PR, and HER2 status also were assessed using automated quantitative analysis (AQUA) method [23]. In this prior publication, the AQUA and semi-quantitative immunohistochemistry methods were highly correlated for ER (r s 2 = 0.80), PR (r s 2 = 0.86), and HER2 (r s 2 = 0.87). The AQUA data were utilized here to examine ER, PR, and HER2 status using a continuous measure and compare these results to those determined from microscopic assessment of semi-quantitative immunohistochemistry.
PRLR IHC staining was performed using a mouse anti-PRLR extracellular domain (ECD) antibody (Cat# 35-9200; Invitrogen, Camarillo, CA). This antibody has been shown to be the most specific for the PRLR, out of a panel of six putative PRLR antibodies tested, and demonstrated potential immunoreactivity in human breast carcinomas when incubated with slides overnight at 4 °C, at a concentration of 3 ug/ml [24]. Here, the anti-PRLR ECD antibody was applied at 10 ug/ml for 1 h at RT, followed by detection using the anti-Mouse VECTASTAIN Elite ABC kit protocol (Vector Laboratories, Burlingame, CA). This kit utilizes diaminobenzidine as a substrate for the peroxidase enzyme. The resulting staining is a brown precipitate which permits categorizing the staining as negative, equivocal, weakly positive, and strongly positive (Supplementary Data–Fig. 1). A pathologist (M.D.) scored all tumor samples using a digital imaging system. Scores were based on intensity only, as this was the predominant factor that differed across the categories. For tumors with multiple cores on the TMAs, scoring was based on simultaneous review of staining patterns across all cores. Staining patterns were largely homogenous across the cores; therefore, scoring was assigned by the pathologist based on overall impression of the staining across the cores. Intra-rater agreement on 23 tumors with duplicate representation on the TMA was 74 %. The category of “weakly positive” was the most inconsistent across the duplicate readings; therefore, PRLR staining was characterized as a dichotomous variable–strongly positive (referred to as “positive”) compared to weakly positive/equivocal/negative (referred to as “negative”) for subsequent analyses. A categorical variable was created that combined PRL levels and PRLR expression into four categories: low PRL/negative PRLR; low PRL/positive PRLR; high PRL/negative PRLR; and high PRL/positive PRLR.
Statistical Analysis
Tumor tissue staining of PRLR could not be assessed as determined by a pathologist (MD) for 365 of the 1,101 cases, due to inadequate tumor tissue representation or digital image quality. In total, 735 invasive cases with PRLR data were available for analyses. The χ 2 analyses were used to determine if tumor characteristics varied between those tumors for which there was PRLR reading and those for which a PRLR reading was not available. In comparing tumors with PRLR expression data to those without, the tumors not included in the analyses were more likely to be of lobular histology (p = 0.02), less than 2 cm (p = 0.03), and HER2 negative (p = 0.03). All other tumor characteristics reported here were not statistically different across the two groups.
Comparisons across strata of tumor characteristics by PRLR status were also assessed using χ2 analyses. Tumor characteristics significantly associated with PRLR staining in univariate analyses were included in multivariate logistic regression analyses with the relevant tumor characteristics as the predictors and PRLR staining (negative vs. positive) as the outcome variable. Exploratory case–case comparisons examining associations between risk factors and PRLR staining were also conducted using χ 2 analyses. Statistical significance was defined at two-sided p < 0.05 and all analyses were completed with SAS (version 9.0, SAS Institute, Inc., Cary, NC) or STATA (version 9.0) software.
Results
Characteristics of Population
The demographics of the 142 pre- and 594 post-menopausal women included in this analysis are reported in Table 1. Using the dichotomous variable for PRLR staining that was based on grouping the categories of negative, equivocal, and weak together as “low/negative” PRLR expression, 83 % of tumors in this sample were scored as a strong PRLR staining (Table 2). The proportion of tumors staining positive was similar across both pre- and post-menopausal cases; therefore, data for these groups were combined.
Table 1.
Characteristics of 736 cases by menopausal status from the Polish Breast Cancer Study
| Premenopausal cases N = 142 | Post-menopausal cases N = 594 | |||
|---|---|---|---|---|
| Mean | SD | Mean | SD | |
| Age, years | 44.9 | 5.7 | 61.2 | 7.9 |
| Age at menarche, years | 13.0 | 1.7 | 13.6 | 1.7 |
| Age at first birth, years | 24.3 | 4.8 | 24.1 | 4.6 |
| Age at menopause, years | – | – | 49.7 | 4.7 |
| Parous, frequency† | 128 | 90.1 | 501 | 84.3 |
| Family history of breast cancer, frequency† | 14 | 9.9 | 69 | 11.6 |
| History of benign breast disease, frequency† | 35 | 24.7 | 105 | 17.7 |
| HRT use, frequency† | – | – | 148 | 24.9 |
†Frequencies represented as N and percent instead of mean and SD
Table 2.
Association of tumor characteristics with prolactin receptor expression among 736 women with breast cancer (pre- and post-menopausal) from the Polish Breast Cancer Study
| PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value* | |
|---|---|---|---|
| Menopausal status N = 736 | |||
| Pre-menopausal | 29 (20.4) | 113 (79.6) | |
| Post-menopausal | 98 (16.5) | 496 (83.5) | 0.27 |
| PRL serum levels** N = 349 | |||
| Low | 30 (17.9) | 138 (82.1) | |
| High | 30 (16.9) | 148 (83.2) | 0.81 |
| Histological type N = 713 | |||
| Ductal | 106 (17.6) | 488 (82.2) | |
| Lobular | 17 (14.3) | 102 (85.7) | 0.35 |
| Tumor size N = 715 | |||
| ≤2 cm | 42 (11.9) | 310 (88.1) | |
| >2 cm | 79 (21.8) | 284 (78.2) | 0.0005 |
| Tumor grade N = 595 | |||
| Well/moderately differentiated | 35 (10.8) | 288 (89.2) | |
| Poorly differentiated | 80 (29.4) | 192 (70.6) | 1.2x10-08 |
| Axillary node metastasis N = 716 | |||
| Node negative | 62 (15.7) | 333 (84.3) | |
| Node positive | 60 (18.7) | 261 (81.3) | 0.29 |
| ER N = 732 | |||
| Negative | 50 (20.1) | 199 (79.9) | |
| Positive | 76 (15.7) | 407 (84.3) | 0.14 |
| PR N = 731 | |||
| Negative | 70 (18.4) | 311 (81.6) | |
| Positive | 56 (16.0) | 294 (84.0) | 0.40 |
| HER2 N = 721 | |||
| Negative | 104 (16.3) | 534 (83.7) | |
| Positive | 20 (24.1) | 63 (75.9) | 0.08 |
| ER/PR N = 730 | |||
| ER+ PR+ | 49 (15.6) | 265 (84.4) | |
| ER+ PR− | 27 (16.1) | 141 (83.9) | |
| ER− PR+ | 7 (19.4) | 29 (80.6) | |
| ER− PR− | 43 (20.3) | 169 (79.7) | 0.15 |
| ER/PR/HER2 N = 721 | |||
| ER+ or PR+ and HER2− | 78 (15.5) | 425 (84.5) | |
| ER+ or PR+ and HER2+ | 5 (35.7) | 9 (64.3) | |
| ER− and PR− and HER2+ | 9 (17.3) | 43 (82.7) | |
| ER− and PR− and HER2− | 32 (21.1) | 120 (78.9) | 0.11 |
Not all strata sums to 100 % due to rounding
*p Values from χ 2 test
**Category for PRL levels assigned using median PRL levels for the subset of cases with this measurement
Relationships between Prolactin Receptor Status and Pathological Characteristics
PRLR-negative status was significantly associated with larger tumor size (p = 0.0005) and poorly differentiated histological grade (p = 1.2 × 10−08) (Table 2). A non-significant but suggested relationship was found between PRLR-negative and HER2-positive status (p = 0.08). Other pathological factors were unrelated to PRLR status both in the overall population (Table 2) and when stratified by menopausal status (Table 3). The associations of PRLR expression with size and differentiation were similar for tumors cross-classified by menopausal status and ER status, except for the small strata of ER negative premenopausal women, which was limited to 44 cases (Table 4).
Table 3.
Association of tumor characteristics with prolactin receptor expression among 736 women (stratified by menopausal status) with breast cancer from the Polish Breast Cancer Study
| Premenopausal N = 142 | Postmenopausal N = 594 | |||||
|---|---|---|---|---|---|---|
| PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value | PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value* | |
| PRL serum levels** N = 349 | ||||||
| Low | 5 (17.9) | 23 (82.1) | 25 (17.9) | 115 (82.1) | ||
| High | 7 (23.3) | 23 (76.7) | 0.61 | 23 (15.5) | 125 (84.5) | 0.60 |
| Histological type N = 713 | ||||||
| Ductal | 27 (22.3) | 94 (77.7) | 79 (16.7) | 394 (83.3) | ||
| Lobular | 1 (7.1) | 13 (92.7) | 0.30 | 16 (15.2) | 89 (84.8) | 0.71 |
| Axillary node metastasis N = 716 | ||||||
| Node negative | 13 (19.7) | 53 (80.3) | 49 (14.9) | 280 (85.1) | ||
| Node positive | 15 (20.6) | 58 (79.5) | 0.90 | 45 (18.2) | 203 (81.9) | 0.30 |
| ER N = 732 | ||||||
| Negative | 12 (19.7) | 49 (80.3) | 38 (20.2) | 150 (79.8) | ||
| Positive | 17 (21.3) | 63 (78.8) | 0.82 | 59 (14.6) | 344 (85.4) | 0.09 |
| PR N = 731 | ||||||
| Negative | 14 (22.2) | 49 (77.8) | 56 (17.6) | 262 (82.4) | ||
| Positive | 15 (19.2) | 63 (80.8) | 0.66 | 41 (15.1) | 231 (84.9) | 0.41 |
| HER2 N = 721 | ||||||
| Negative | 25 (20.5) | 97 (79.5) | 79 (15.3) | 437 (84.7) | ||
| Positive | 4 (22.2) | 14 (77.8) | 1.0 | 16 (24.6) | 49 (75.4) | 0.06 |
| ER/PR N = 730 | ||||||
| ER+ PR+ | 13 (20.3) | 51 (79.7) | 36 (14.4) | 214 (85.6) | ||
| ER+ PR− | 4 (25.0) | 12 (75.0) | 23 (15.1) | 129 (84.9) | ||
| ER− PR+ | 2 (14.3) | 12 (85.7) | 5 (22.7) | 17 (77.3) | ||
| ER− PR− | 10 (21.3) | 37 (78.7) | 0.99 | 33 (20.0) | 132 (80.0) | 0.11 |
| ER/PR/HER2 N = 721 | ||||||
| ER+ or PR+ and HER2− | 18 (19.6) | 74 (80.4) | 60 (14.6) | 351 (85.4) | ||
| ER+ or PR+ and HER2+ | 1 (50.0) | 1 (50.0) | 4 (33.3) | 8 (66.7) | ||
| ER− and PR− and HER2+ | 2 (16.7) | 10 (83.3) | 7 (17.5) | 33 (82.5) | ||
| ER− and PR− and HER2− | 8 (24.2) | 25 (75.8) | 0.64 | 24 (20.2) | 95 (79.8) | 0.13 |
Not all strata sum to 100 % due to rounding
* p Values from χ 2 test
**Category for PRL levels assigned using median PRL levels for the subset of cases with this measurement
Table 4.
Association of tumor size and grade with prolactin receptor expression stratified by ER status among 736 women with breast cancer from the Polish Breast Cancer Study
| Overall Population N = 736 | Premenopausal N = 142 | Postmenopausal N = 594 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value* | PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value* | PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value* | |
| ER+ tumors | |||||||||
| Tumor size | |||||||||
| ≤2 cm | 31 (11.9) | 230 (88.1) | 0.02 | 7 (16.7) | 35 (83.3) | 0.33 | 24 (11.0) | 195 (89.0) | 0.04 |
| >2 cm | 41 (19.5) | 169 (80.5) | 9 (25.7) | 26 (74.3) | 32 (18.3) | 143 (81.7) | |||
| Tumor grade | |||||||||
| Well/moderately differentiated | 26 (10.2) | 230 (89.8) | 2.2 × 10−7 | 5 (12.2) | 36 (87.8) | 0.007 | 21 (9.8) | 194 (90.2) | 9.3 × 10−6 |
| Poorly differentiated | 45 (31.7) | 97 (68.3) | 11 (40.7) | 16 (59.3) | 34 (29.6) | 81 (70.4) | |||
| ER− tumors | |||||||||
| Tumor size | |||||||||
| ≤2 cm | 10 (11.4) | 78 (88.6) | 0.01 | 2 (8.0) | 23 (92.0) | 0.04 | 8 (12.7) | 55 (87.3) | 0.08 |
| >2 cm | 38 (25.0) | 114 (75.0) | 10 (29.4) | 24 (70.6) | 28 (23.7) | 90 (76.3) | |||
| Tumor grade | |||||||||
| Well/moderately differentiated | 8 (12.3) | 57 (87.7) | 0.02 | 4 (25.0) | 12 (75.0) | 1.0 | 4 (8.2) | 45 (91.8) | 0.007 |
| Poorly differentiated | 35 (26.9) | 95 (73.1) | 7 (25.0) | 21 (75.0) | 28 (27.5) | 74 (72.6) | |||
*p Values from χ 2 test
Additional analyses of 579 women with data for size, grade, ER status, and PRLR using multivariate logistic regression demonstrated that PRLR-negative status was associated with larger tumors (p = 0.01, OR = 1.81, 95 % CI 1.14–2.86) and poor differentiation (p = 0 < 0.0001, OR = 3.13, 95 % CI 1.94–5.03), independent of ER status. In these models, PRLR staining was not associated with ER status (p = 0.51, OR = 1.18, 95 % CI 0.74–1.88). Stratified analyses of data for premenopausal women was limited by sample size (n = 109).
In addition to data from standard microscopic assessment of semi-quantitative immunohistochemistry methods, almost all the 579 women also had data from the AQUA method for ER (n = 564), PR (n = 561), and HER2 (n = 557) [23]. Including these continuous data in the regression models did not change the outcome of the association of PRLR-negative status with larger tumors (p = 0 < 0.0001, OR = 1.81, 95 % CI 1.14–2.89) and poor differentiation (p = 0 < 0.0001, OR = 2.89, 95 % CI 1.78–4.63), independent of ER status.
Approximately 349 cases in this analysis were included in a prior report assessing associations with circulating PRL levels [8]. The current data do not show an association between PRL levels and PRLR expression (Table 2). The cross-classification of circulating PRL levels with PRLR staining did not reveal stronger associations than found for analyses of PRLR staining alone.
Association between Breast Cancer Risk Factors and Tumor PRLR Status among Cases
Increasing age at first birth was associated with positive PRLR staining (p = 0.03) and increasing BMI was associated with negative PRLR status (p = 0.002) (Table 5). These results were similar among post-menopausal women (n = 467) but were weaker among premenopausal women (n = 109) (data not shown). Other risk factors were unrelated. When BMI was added to the multivariate model containing tumor size, grade, and ER status, BMI remained independently associated with PRLR expression. Therefore, the associations of PRLR with tumor size, grade, and patient BMI are independent of each other and ER status.
Table 5.
Association of risk factors with prolactin receptor expression among 736 women with breast cancer (pre- and post-menopausal) from the Polish Breast Cancer Study
| PRLR staining (negative or low) N (%) | PRLR staining (strong) N (%) | p Value** | |
|---|---|---|---|
| Parity N = 736 | |||
| Nulliparous | 19 (17.8) | 88 (82.8) | |
| Parous | 108 (17.2) | 521 (82.8) | 0.88 |
| Age at first birth* N = 628 | |||
| <20 years | 16 (21.3) | 59 (78.7) | |
| 20–24 | 63 (19.7) | 257 (80.3) | |
| 25–30 | 19 (12.5) | 133 (87.5) | |
| 30+ | 10 (12.4) | 71 (87.7) | 0.03 |
| Number of full-term births* N = 629 | |||
| 1 | 42 (16.2) | 217 (83.8) | |
| 2 | 52 (18.0) | 237 (82.0) | |
| 3+ | 14 (17.3) | 67 (82.7) | 0.69 |
| Age at last birth* N = 629 | |||
| <24 years | 30 (20.3) | 118 (79.7) | |
| 24–27 | 25 (18.4) | 111 (81.6) | |
| 27–31 | 26 (16.2) | 135 (83.9) | |
| 31+ | 27 (14.7) | 157 (85.3) | 0.15 |
| Years since last birth* N = 629 | |||
| <22.5 years | 30 (19.0) | 128 (81.0) | |
| 22.5–31.0 | 28 (19.3) | 117 (80.7) | |
| 31.0–40.0 | 22 (14.0) | 135 (86.0) | |
| 40.0+ | 28 (16.6) | 141 (83.4) | 0.36 |
| Number of months breastfeeding* N = 629 | |||
| Never | 30 (22.1) | 106 (77.9) | |
| <6 months | 46 (13.8) | 287 (86.2) | |
| 6–11 | 17 (18.1) | 77 (81.9) | |
| 12–23 | 15 (22.7) | 51 (77.3) | 0.85 |
| BMI category N = 733 | |||
| <25 | 29 (12.4) | 205 (87.6) | |
| 25 ≤ 30 | 34 (13.4) | 220 (86.6) | |
| 30+ | 62 (25.3) | 183 (74.7) | 0.002 |
Not all strata sum to 100 % due to rounding
*Analyses conducted among only parous women
**p Values from χ 2 test
Discussion
In this study, breast cancers negative for PRLR expression were larger and more poorly differentiated than breast cancers with positive PRLR expression, independent of ER status, other tumor characteristics, and such patient characteristics as BMI. Similarly, previous studies have found that PRLR expression is independent of ER status [14–19]. Here, ER status was reported using microscopic-assisted semi-quantitative immunohistochemistry and a score of positive or negative determined from methods outlined in prior publications based on this population [21–23]. The results of the associations of PRLR-negative staining with larger and more poorly differentiated tumors, independent of ER status, also were examined using the continuous AQUA data. Including the continuous measures of ER, PR, or HER2 staining in the multivariate models here did not change the outcome.
The associations between high levels of PRLR, smaller tumor size, and lower grade suggests that PRLR may be a marker of better differentiation and may be involved in this phenotype. These results are consistent with data demonstrating breast cancer xenografts grown in immunocompromised mice upregulated expression of genes involved in differentiation pathways when stimulated in vivo with prolactin [25]. Similarly, the phosphorylation of a key signaling component of the prolactin receptor pathway, Stat5, is associated with more favorable prognosis with regard to breast cancer outcomes [26–30]. We are unable to examine the association of PRLR expression with breast cancer outcomes because of limited number of deaths in follow-up. Therefore, it remains unclear whether PRLR staining provides additional prognostic information beyond that contributed by other factors. In addition, examining markers of PRLR activation, such as Stat5 activation, may permit further refinement of the strong PRLR staining category by separating tumors that highly express PRLR from tumors that both highly express PRLR and exhibit characteristics associated with PRLR activation.
A limitation of this study is that it was only possible to score PRLR dichotomously, which, in combination with the low percentage of negative cases, limited statistical power. Our estimate of PRLR-negative cases was conservative whereby PRLR-negative scores were restricted to cancers in which both TMA cores were scored negative. Prior publications by Camp and colleagues demonstrated that immunohistochemical staining scores assigned based on duplicate representation in a tissue microarray correlate with staining assigned from review of whole sections in >95 % of breast cancer cases [31].
Another factor in this analysis was the use of a PRLR reagent antibody for immunohistochemistry that binds to the extracellular domain of the receptor and detects multiple forms of the PRLR. There are six different PRL receptors that arise from alternative transcripts of the PRLR gene, which have different functions. [5, 17, 32–35]. The most well characterized isoform is the long-form (i.e., 90 kDa) of the receptor, which initiates the Jak2/Stat5, PI3K, and MAPK signaling pathways, resulting in enhanced survival, proliferation, and differentiation of breast epithelial cells [5]. Other isoforms of the PRLR may be negative inhibitors of signaling [17, 32–35].
The results presented here, in combination with those from prior studies demonstrating associations between circulating PRL (i.e., ligand) levels and selected breast cancer risk factors [5, 8–13], provide a basis for additional epidemiologic studies extending these results to other aspects of the PRL/PRLR signaling network. While several large studies have now examined circulating PRL levels, this report is the first to look at the association of circulating PRL levels with PRLR expression but could do so only in a subset of cases. These associations should be examined in larger, prospective studies. There is also the unresolved question of autocrine PRL production, including whether this could result in locally higher concentrations of the hormone and have implications for PRLR signaling [36].
Conclusions
The study reported here is the largest to evaluate the associations of PRLR expression with tumor characteristics and the findings of associations between negative PRLR staining with larger tumor size and poorly differentiated histological grade are interesting. However, replication and extension of these results to include other relevant markers, including expression and/or activation of specific prolactin receptor isoforms or downstream signaling molecules, and follow-up would be useful to understand whether PRLR expression has clinical importance beyond established factors such as tumor size and grade. Future studies using assays that enable more detailed characterization of the PRLR may increase our knowledge of the importance of this receptor in breast cancer.
Electronic Supplementary Material
Below is the link to the electronic supplementary material.
(DOC 256 kb)
Acknowledgments
We thank Pei Chao and Michael Stagner from Information Management Services (Sliver Spring, MD, USA), for their valuable contributions to the data management aspects of the study. We also thank the participants, physicians, pathologists, nurses, and interviewers from participating centers in Poland for their efforts during the field-work. The study was funded by Intramural Research Funds of the National Cancer Institute, Department of Health and Human Services, USA. This research was supported by the Intramural Research Programs of the Division of Cancer Epidemiology and Genetics and Center for Cancer Research of the National Cancer Institute. Dr Faupel-Badger’s research was also supported by the Cancer Prevention Fellowship Program, Center for Cancer Training, NCI.
Conflict of Interest
The authors declare that they have no conflict of interest.
References
- 1.Binart N, Ormandy CJ, Kelly PA. Mammary gland development and the prolactin receptor. Adv Exp Med Biol. 2000;480:85–92. doi: 10.1007/0-306-46832-8_10. [DOI] [PubMed] [Google Scholar]
- 2.Das R, Vonderhaar BK. Prolactin as a mitogen in mammary cells. J Mammary Gland Biol Neoplasia. 1997;2:29–39. doi: 10.1023/A:1026369412612. [DOI] [PubMed] [Google Scholar]
- 3.Hovey RC, Trott JF, Vonderhaar BK. Establishing a framework for the functional mammary gland: from endocrinology to morphology. J Mammary Gland Biol Neoplasia. 2002;7:17–38. doi: 10.1023/A:1015766322258. [DOI] [PubMed] [Google Scholar]
- 4.Ormandy CJ, Binart N, Kelly PA. Mammary gland development in prolactin receptor knockout mice. J Mammary Gland Biol Neoplasia. 1997;2:355–364. doi: 10.1023/A:1026395229025. [DOI] [PubMed] [Google Scholar]
- 5.Clevenger CV, Furth PA, Hankinson SE, Schuler LA. The role of prolactin in mammary carcinoma. Endocr Rev. 2003;24:1–27. doi: 10.1210/er.2001-0036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Maus MV, Reilly SC, Clevenger CV. Prolactin as a chemoattractant for human breast carcinoma. Endocrinology. 1999;140:5447–5450. doi: 10.1210/endo.140.11.7245. [DOI] [PubMed] [Google Scholar]
- 7.Vonderhaar BK. Prolactin involvement in breast cancer. Endocr Relat Cancer. 1999;6:389–404. doi: 10.1677/erc.0.0060389. [DOI] [PubMed] [Google Scholar]
- 8.Faupel-Badger JM, Sherman ME, Garcia-Closas M, Gaudet MM, Falk RT, Andaya A, Pfeiffer RM, Yang XR, Lissowska J, Brinton LA, et al. Prolactin serum levels and breast cancer: relationships with risk factors and tumour characteristics among pre- and postmenopausal women in a population-based case-control study from Poland. Br J Cancer. 2010;103:1097–1102. doi: 10.1038/sj.bjc.6605844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hankinson SE, Willett WC, Michaud DS, Manson JE, Colditz GA, Longcope C, Rosner B, Speizer FE. Plasma prolactin levels and subsequent risk of breast cancer in postmenopausal women. J Natl Cancer Inst. 1999;91:629–634. doi: 10.1093/jnci/91.7.629. [DOI] [PubMed] [Google Scholar]
- 10.Tworoger SS, Eliassen AH, Rosner B, Sluss P, Hankinson SE. Plasma prolactin concentrations and risk of postmenopausal breast cancer. Cancer Res. 2004;64:6814–6819. doi: 10.1158/0008-5472.CAN-04-1870. [DOI] [PubMed] [Google Scholar]
- 11.Tworoger SS, Eliassen AH, Sluss P, Hankinson SE. A prospective study of plasma prolactin concentrations and risk of premenopausal and postmenopausal breast cancer. J Clin Oncol. 2007;25:1482–1488. doi: 10.1200/JCO.2006.07.6356. [DOI] [PubMed] [Google Scholar]
- 12.Tworoger SS, Sluss P, Hankinson SE. Association between plasma prolactin concentrations and risk of breast cancer among predominately premenopausal women. Cancer Res. 2006;66:2476–2482. doi: 10.1158/0008-5472.CAN-05-3369. [DOI] [PubMed] [Google Scholar]
- 13.Tworoger SS, Eliassen AH, Zhang X, Qian J, Sluss PM, Rosner BA, Hankinson SE. A 20-year prospective study of plasma prolactin as a risk marker of breast cancer development. Cancer Res. 2013;73:4810–4819. doi: 10.1158/0008-5472.CAN-13-0665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gill S, Peston D, Vonderhaar BK, Shousha S. Expression of prolactin receptors in normal, benign, and malignant breast tissue: an immunohistological study. J Clin Pathol. 2001;54:956–960. doi: 10.1136/jcp.54.12.956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Reynolds C, Montone KT, Powell CM, Tomaszewski JE, Clevenger CV. Expression of prolactin and its receptor in human breast carcinoma. Endocrinology. 1997;138:5555–5560. doi: 10.1210/endo.138.12.5605. [DOI] [PubMed] [Google Scholar]
- 16.Touraine P, Martini JF, Zafrani B, Durand JC, Labaille F, Malet C, Nicolas A, Trivin C, Postel-Vinay MC, Kuttenn F, et al. Increased expression of prolactin receptor gene assessed by quantitative polymerase chain reaction in human breast tumors versus normal breast tissues. J Clin Endocrinol Metab. 1998;83:667–674. doi: 10.1210/jcem.83.2.4564. [DOI] [PubMed] [Google Scholar]
- 17.Trott JF, Hovey RC, Koduri S, Vonderhaar BK. Alternative splicing to exon 11 of human prolactin receptor gene results in multiple isoforms including a secreted prolactin-binding protein. J Mol Endocrinol. 2003;30:31–47. doi: 10.1677/jme.0.0300031. [DOI] [PubMed] [Google Scholar]
- 18.Glasow A, Haidan A, Gillespie J, Kelly PA, Chrousos GP, Bornstein SR. Differential expression of prolactin receptor (PRLR) in normal and tumorous adrenal tissues: separation of cellular endocrine compartments by laser capture microdissection (LCM) Endocr Res. 1998;24:857–862. doi: 10.3109/07435809809032697. [DOI] [PubMed] [Google Scholar]
- 19.Mertani HC, Garcia-Caballero T, Lambert A, Gerard F, Palayer C, Boutin JM, Vonderhaar BK, Waters MJ, Lobie PE, Morel G. Cellular expression of growth hormone and prolactin receptors in human breast disorders. Int J Cancer. 1998;79:202–211. doi: 10.1002/(SICI)1097-0215(19980417)79:2<202::AID-IJC17>3.0.CO;2-B. [DOI] [PubMed] [Google Scholar]
- 20.Garcia-Closas M, Brinton LA, Lissowska J, Chatterjee N, Peplonska B, Anderson WF, Szeszenia-Dabrowska N, Bardin-Mikolajczak A, Zatonski W, Blair A, et al. Established breast cancer risk factors by clinically important tumour characteristics. Br J Cancer. 2006;95:123–129. doi: 10.1038/sj.bjc.6603207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yang XR, Pfeiffer RM, Garcia-Closas M, Rimm DL, Lissowska J, Brinton LA, Peplonska B, Hewitt SM, Cartun RW, Mandich D, et al. Hormonal markers in breast cancer: coexpression, relationship with pathologic characteristics, and risk factor associations in a population-based study. Cancer Res. 2007;67:10608–10617. doi: 10.1158/0008-5472.CAN-07-2142. [DOI] [PubMed] [Google Scholar]
- 22.Yang XR, Sherman ME, Rimm DL, Lissowska J, Brinton LA, Peplonska B, Hewitt SM, Anderson WF, Szeszenia-Dabrowska N, Bardin-Mikolajczak A, et al. Differences in risk factors for breast cancer molecular subtypes in a population-based study. Cancer Epidemiol Biomarkers Prev. 2007;16:439–443. doi: 10.1158/1055-9965.EPI-06-0806. [DOI] [PubMed] [Google Scholar]
- 23.Sherman ME, Rimm DL, Yang XR, Chatterjee N, Brinton LA, Lissowska J, Peplonska B, Szeszenia-Dabrowska N, Zatonski W, Cartun R, et al. Variation in breast cancer hormone receptor and HER2 levels by etiologic factors: a population-based analysis. Int J Cancer. 2007;121:1079–1085. doi: 10.1002/ijc.22812. [DOI] [PubMed] [Google Scholar]
- 24.Galsgaard ED, Rasmussen BB, Folkesson CG, Rasmussen LM, Berchtold MW, Christensen L, Panina S. Re-evaluation of the prolactin receptor expression in human breast cancer. J Endocrinol. 2009;201:115–128. doi: 10.1677/JOE-08-0479. [DOI] [PubMed] [Google Scholar]
- 25.Sato T, Tran TH, Peck AR, Liu C, Ertel A, Lin J, Neilson LM, Rui H. Global profiling of prolactin-modulated transcripts in breast cancer in vivo. Mol Cancer. 2013;12:59. doi: 10.1186/1476-4598-12-59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cotarla I, Ren S, Zhang Y, Gehan E, Singh B, Furth PA. Stat5a is tyrosine phosphorylated and nuclear localized in a high proportion of human breast cancers. Int J Cancer. 2004;108:665–671. doi: 10.1002/ijc.11619. [DOI] [PubMed] [Google Scholar]
- 27.Nevalainen MT, Xie J, Torhorst J, Bubendorf L, Haas P, Kononen J, Sauter G, Rui H. Signal transducer and activator of transcription-5 activation and breast cancer prognosis. J Clin Oncol. 2004;22:2053–2060. doi: 10.1200/JCO.2004.11.046. [DOI] [PubMed] [Google Scholar]
- 28.Peck AR, Witkiewicz AK, Liu C, Klimowicz AC, Stringer GA, Pequignot E, Freydin B, Yang N, Ertel A, Tran TH, et al. Low levels of Stat5a protein in breast cancer are associated with tumor progression and unfavorable clinical outcomes. Breast Cancer Res : BCR. 2012;14:R130. doi: 10.1186/bcr3328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Peck AR, Witkiewicz AK, Liu C, Stringer GA, Klimowicz AC, Pequignot E, Freydin B, Tran TH, Yang N, Rosenberg AL, et al. Loss of nuclear localized and tyrosine phosphorylated Stat5 in breast cancer predicts poor clinical outcome and increased risk of antiestrogen therapy failure. J Clin Oncol. 2011;29:2448–2458. doi: 10.1200/JCO.2010.30.3552. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yamashita H, Nishio M, Ando Y, Zhang Z, Hamaguchi M, Mita K, Kobayashi S, Fujii Y, Iwase H. Stat5 expression predicts response to endocrine therapy and improves survival in estrogen receptor-positive breast cancer. Endocr Relat Cancer. 2006;13:885–893. doi: 10.1677/erc.1.01095. [DOI] [PubMed] [Google Scholar]
- 31.Camp RL, Charette LA, Rimm DL. Validation of tissue microarray technology in breast carcinoma. Lab Invest. 2000;80:1943–1949. doi: 10.1038/labinvest.3780204. [DOI] [PubMed] [Google Scholar]
- 32.Hu ZZ, Meng J, Dufau ML. Isolation and characterization of two novel forms of the human prolactin receptor generated by alternative splicing of a newly identified exon 11. J Biol Chem. 2001;276:41086–41094. doi: 10.1074/jbc.M102109200. [DOI] [PubMed] [Google Scholar]
- 33.Kline JB, Roehrs H, Clevenger CV. Functional characterization of the intermediate isoform of the human prolactin receptor. J Biol Chem. 1999;274:35461–35468. doi: 10.1074/jbc.274.50.35461. [DOI] [PubMed] [Google Scholar]
- 34.Kline JB, Rycyzyn MA, Clevenger CV. Characterization of a novel and functional human prolactin receptor isoform (deltaS1PRLr) containing only one extracellular fibronectin-like domain. Mol Endocrinol. 2002;16:2310–2322. doi: 10.1210/me.2001-0033. [DOI] [PubMed] [Google Scholar]
- 35.Trott JF, Hovey RC, Koduri S, Vonderhaar BK. Multiple new isoforms of the human prolactin receptor gene. Adv Exp Med Biol. 2004;554:495–499. doi: 10.1007/978-1-4757-4242-8_71. [DOI] [PubMed] [Google Scholar]
- 36.Nitze LM, Galsgaard ED, Din N, Lund VL, Rasmussen BB, Berchtold MW, Christensen L, Panina S (2013) Reevaluation of the proposed autocrine proliferative function of prolactin in breast cancer. Breast Cancer Res Treat. DOI 10.1007/s10549-013-2731-7. In press. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOC 256 kb)