Original Article · Originalarbeit Onkologie 2013;36:637–640 DOI: 10.1159/000355663
Published online: October 14, 2013
Clinical Significance of Lymph Node Ratio in Locally Advanced Breast Cancer Molecular Subtypes Fatih Demircioglua Umut Demircib Diclehan Kilicc Secil Ozkand Eray Karahaciogluc a
Rize Recep Tayyip Erdogan University Hospital, Department of Radiation Oncology, Rize, Dr.Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital, Department of Medical Oncology, Ankara, c Department of Radiation Oncology, d Department of Public Health, Gazi University Faculty of Medicine, Ankara, Turkey
Keywords Breast cancer · Lymph node ratio · Molecular subtypes · Prognostic factors
Schlüsselwörter Mammakarzinom · Lymphknotenratio · Molekulare Subtypen · Prognosefaktoren
Summary Background: The ratio of metastatic to dissected lymph nodes (lymph node ratio; LNR) is a sensitive and supe rior prognostic factor for lymph node evaluation, but its relationship to cancer subtypes is unclear. Patients and Methods: Data from 469 patients with axillary lymph node metastasis out of 640 early breast cancer cases were retrospectively analyzed. They were classified into 4 molecular subtypes; luminal A, luminal B HER2(+), HER2 overexpression, basal-like. LNRs were compared between groups and with other prognostic factors. Results: The distribution of LNRs was 35.2% in luminal A, 43.2% in luminal B HER2(+), 46.9% in HER2 over expression, and 39.1% in basal-like. A significant differ ence was found between luminal A and HER2 over expression subtypes (p = 0.023). LNR was significantly correlated with tumor size and lymphovascular invasion, but not with other prognostic factors including meno pausal status, laterality, grade, and perineural invasion. An LNR of 29.8% was defined as the cut-off value, and significant differences in survival rates were identified accordingly between basal-like and both luminal A (p = 0.003) and luminal B HER2(+) (p = 0.04). Conclusion: The LNR differs between some molecular subtypes of breast cancer, and it correlates with certain prognostic factors and survival. These data support using the LNR to assess breast cancer patients.
Zusammenfassung Hintergrund: Das Zahlenverhältnis zwischen befallenen und entfernten Lymphknoten (Lymphknotenratio) ist ein sensibler und überlegener Prognosefaktor für die Lymph knotenbewertung. Das Verhältnis zwischen Lymphkno tenratio und Tumorsubtypen ist jedoch unklar. Patienten und Methoden: Die Daten von 469 Patientinnen mit axil lären Lymphknotenmetastasen (von insgesamt 640 Pati entinnen mit frühem Mammakarzinom) wurden retros pektiv analysiert. 4 molekulare Subtypen wurden unter schieden: Luminal A, Luminal B HER2(+), HER2-Über expression und basal-like. Die Lymphknotenratios wurden innerhalb der Gruppen sowie mit anderen Prog nosefaktoren verglichen. Ergebnisse: Die Distribution der Lymphknotenratios war 35,2% für Luminal A, 43,2% für Luminal B HER2(+), 46,9% für HER2-Überexpression und 39,1% für den basal-like Zelltyp. Ein signifikanter Unterschied bestand zwischen den Subtypen Luminal A und HER2-Überexpression (p = 0,023). Die Lymphknoten ratio war signifikant mit Tumorgröße und lymphovas kulärer Invasion korreliert, nicht jedoch mit anderen Pro gnosefaktoren wie menopausalem Status, Lateralisation, Grad und perineuraler Invasion. Eine Lymphknotenratio von 29,8% wurde als Cut-Off-Wert definiert, und signi fikante Unterschiede in der Überlebensrate wurden ent sprechend zwischen dem basal-like Subtyp und sowohl Luminal A (p = 0,003) als auch Luminal B HER2(+) (p = 0,04) identifiziert. Schlussfolgerung: Die Lymphknotenratio ist bei gewissen molekularen Subtypen des Mammakarzi noms unterschiedlich und mit bestimmten Prognose faktoren und Überleben korreliert. Die vorliegenden Daten unterstützen den Einsatz der Lymphknotenratio bei der Bewertung von Mammakarzinompatientinnen.
© 2013 S. Karger GmbH, Freiburg 0378-584X/13/3611-0637$38.00/0 Fax +49 761 4 52 07 14 [email protected] www.karger.com
Accessible online at: www.karger.com/onk
Fatih Demircioglu, MD Department of Radiation Oncology Rize Recep Tayyip Erdogan University Hospital Șehitler st. 74 5300 Rize, Turkey [email protected]
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
b
Breast cancer (BC) is the most frequent type of cancer and the second most common cause of cancer-related deaths in women [1, 2]. Hormone receptor (HR) status, human epidermal growth factor receptor 2 (HER2) expression, and axillary lymph node involvement are the most important prognostic factors [3]. The heterogeneity of disease makes identification of molecular ‘intrinsic’ subtypes by gene expression profiling important for prognosis and for guiding molecular therapies [4, 5]. Due to the high correspondence between molecular subtypes and clinicopathological features, HR status, HER2 overexpression, and Ki-67, as well as grading are important when classifying BC [6]. Axillary lymph node status is classified only by the number of positive lymph nodes according to the American Joint Committee on Cancer (AJCC) staging system [7]. 10 or more excised lymph nodes is considered sufficient for axillary dissections [8]. However, this number varies for each patient, and there is a close relationship between the number of metastatic lymph nodes and the total number dissected. Some studies have suggested that each involved lymph node affects the prognosis of BC [9, 10]. In this situation, the AJCC lymph node staging system can lead to heterogeneous results. To avoid this confusion, research efforts have focused on implementing another metric, the ratio of metastatic to total dissected lymph nodes (lymph node ratio; LNR), which has proven to be a more sensitive and superior prognostic factor than the total number of involved lymph nodes [11–15]. To our knowledge, no studies have evaluated the relationship between LNR and molecular subtypes in BC. To address this, we examined whether LNR differs between molecular subtypes and sought to identify markers that may contribute to any differences. In addition, the effect of LNR on survival and its correlation with other prognostic factors was investigated.
Patients and Methods Patients were classified into 4 molecular subtypes based on histological grade and immunohistochemical staining for estrogen receptor (ER), progesterone receptor (PR), and HER2. Grading was performed according to the Bloom-Richardson grading system. ER, PR, and HER2 status were determined by immunohistochemical evaluation. 1 (+)HER2 was defined as negative, 3 (+)HER2 was defined as positive, and fluorescence in situ hybridization was performed for 2 (+)HER2. Accordingly, the luminal A group was defined as ER(+) and/or PR(+), HER2(–), and grade 1; the luminal B HER2(+) group as ER(+) and/or PR(+), and HER2(+); the HER2 overexpression group as ER(–), PR(–), and HER2(+); and the basal-like group as those with triple-negative molecular indicators. All patients in the study had complete axillary dissection. LNR was defined as the ratio of metastatic to total dissected lymph nodes, calculated using the formula of (metastatic lymph node number/total dissected lymph node number) × 100, and given as percentages. Patients without lymph node metastasis were excluded because LNRs were not calculated. Descriptive statistics are presented as frequencies, percent, and median-mean (min, max) values. Continuous data were tested for normality with the Kolmogorov-Smirnov test. Since the variables are not nor-
638
Onkologie 2013;36:637–640
mally distributed, the continuous variables were compared by MannWhitney U test between 2 groups and Kruskal-Wallis test of variance between more than 2 groups. Correlation coefficients and their significance were calculated using the Spearman test. The capacity of LNR in predicting overall survival was analyzed by ROC (receiver operating characteristic) curve analysis. When a significant cut-off value was observed, sensitivity and specificity values were presented. The Kaplan-Meier survival estimates were calculated. A separate log-rank test was used to identify the independent effect of the various prognostic factors. The association between clinicopathological factors, biological subtype, and outcome was evaluated with univariate analysis. Statistical analyses were performed using SPSS v16.0 (IBM SPSS Inc., Chicago, IL, USA), and values of p < 0.05 were considered statistically significant.
Results This study was a retrospective analysis of data from 469 patients with axillary lymph node metastasis out of 640 early BC cases followed at the Gazi University Faculty of Medicine, Ankara, Turkey, between 2001 and 2008. 171 patients without lymph node metastasis were excluded because LNRs were not calculated. Patient characteristics are summarized in table 1. The LNRs calculated for each molecular subtype were: 35.2% in luminal A, 43.2% in luminal B HER2(+), 46.9% in HER2 overexpression, and 39.1% in basal-like. Kruskal- Wallis analysis revealed a statistically significant difference between the groups (p = 0.012). The Spearman test was used Table 1. Patient characteristics Age, median (range), years Menopausal status, n Premenopausal Postmenopausal ER status, n Positive Negative PR status, n Positive Negative HER2 status, n Positive Negative Molecular subtypes, n (%) Basal-like HER2+ Luminal A Luminal B HER2(+) Laterality, n Right Left Grade, n (%) 1 2 3 NA BCS/MRM Tumor size, median (range), cm Patients with metastatic LN, n (%) Median dissected LN, n (%) Median metastatic LN, n (%) Median LNR, %
56 (23–87) 184 285 291 178 291 178 166 303 72 (15.3) 62 (13.2) 231 (49.3) 104 (22.2) 233 246 242 (51.6) 96 (20.5) 75 (16) 56 (11.9) 37/432 3.2 (0.4–12.5) 469 (73) 17 (2.64) 5 (1–44) 31.8 (95% CI 2.1–100)
ER = Estrogen receptor; PR = progesteron receptor; HER2 = human epidermal growth factor receptor 2; NA = not available; BCS/MRM = breast conserving surgery/modified radical mastectomy; LN = lymph nodes; LNR = lymph node ratio.
Demircioglu/Demirci/Kilic/ Ozkan/Karahacioglu
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
Introduction
to evaluate the significance of the correlation for each subtype (table 2), which revealed that the LNR differed significantly between the luminal A and HER2 overexpression groups (p = 0.023). The relationship between LNR and other prognostic factors was examined separately. A statistically significant association was found between LNR and both tumor size and lymphovascular invasion (LVI), but not between other prognostic factors including menopausal status, tumor location, tumor grade, and perineural invasion (PNI) (table 3). The survival times of patients with different molecular subtypes of BC were 112.8 months in luminal A, 104.2 months in luminal B HER2(+), 95.9 months in HER2 overexpression, and 93.4 months in the basal-like group. Statistically significant difference was found between survival times according to molecular subtype (p = 0.024). The impact of prognostic factors on survival rates was analyzed, and only molecular subtypes Table 2. Molecular subtypes according to lymph node ratio (Kruskal-Wallis test)
Discussion
p
Standard error
0.39 0.75 0.78
4.9523 3.8581 4.3822
0.39 0.02 0.86
4.9523 4.0884 4.5863
0.75 0.02 0.08
3.8581 4.0884 3.3754
0.78 0.86 0.08
4.3822 4.5863 3.3754
Table 3. The lymph node ratio (LNR) and survival differences according to prognostic factors Prognostic factors Menopausal status Premenopausal Postmenopausal Tumor location Right Left Tumor grade I II III Tumor size, cm ≤5 >5 PNI Positive Negative LVI Positive Negative
LNR, %
p
Survival
p
0.796
0.271
0.187
0.45
0.286
0.471
0.027
0.059
0.11
0.039
0.014
0.048
38.07 39.81 37.76 40.37 35.93 38.64 41.19 37.94 44.74 39.21 36.71 42.99 35.75
PNI = Perineural invasion; LVI = lymphovascular invasion.
Lymph Node Ratio in Locally Advanced Breast Cancer Subtypes
BC is a heterogeneous disease that is associated with several prognostic factors. The ratio of metastatic to dissected lymph nodes was shown to be a more sensitive and superior prognostic factor than the classical AJCC staging system [11– 15]. In the present study, the LNR and other prognostic factors were examined with respect to tumor molecular subtype, and the relationship of these factors to survival rates were investigated in BC patients. The lowest LNR was found in the luminal A group (35.2%), while the highest was found in the HER2 overexpression group (46.9%); this difference was statistically significant. In agreement with the literature [16–20], the LNR was as high as 43.2% in the luminal B HER2(+) group; however, this did not distinguish it from other subtypes as the differences were not statistically significant. The basallike group had a lower LNR of 39.1% that may be contrary to expectations. Crabb et al. [21] investigated axillary nodal involvement in 4,444 early-stage BC cases and showed that the basal-like subtype predicts a lower incidence of axillary nodal involvement, which is in agreement with our study. In addition, studies of triple-negative BC revealed marked heterogeneity within this group, and gene expression profiling is important for inter-group discrimination [22–24]. Despite the poor survival data, the lower LNR may have arisen from this heterogeneity, a fact that should be considered in future genetic studies. Moreover, hematogenous tumor spread can be independent from lymphatic tumor spread and may be responsible for this situation [25]. Hartkopf et al. [25] investigated the relationship between disseminated tumor cells in bone marrow and lymph node involvement, and found no association of lymph node status and molecular subtype with bone marrow involvement. LNR cut-off values have been defined by others with the aim of staging. These levels were 20 and 65% in the study of Vinh-Hung et al. [11] which is the most comprehensive to date. In another study, a value of 25% was correlated with a statistically significant impact on survival rate [13]. We found Onkologie 2013;36:637–640
639
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
Basal-like HER2+ Luminal A Luminal HER2(+) HER2+ Basal-like Luminal A Luminal B HER2(+) Luminal A Basal-like HER2+ Luminal B HER2(+) Luminal B HER2(+) Basal-like HER2+ Luminal A
and LNR showed a significant association. There were no significant differences in survival according to menopausal status, tumor location, or grade (table 3). The significant cut-off value was 29.8% in ROC curve analysis, sensitivity and specificity values were 70.9 and 59.6%, respectively. Separation of patients below and above this value corresponded to survival rates of 82.9 and 57.3%, respectively (p = 0.0001). A KaplanMeier analysis based on this cut-off value revealed statistically significant differences between basal-like subtype and both the luminal A (p = 0.003) and luminal B HER2(+) groups (p = 0.04). Each group was also compared in pairs by KaplanMeier analysis, and no statistically significant values were found between other groups (luminal A and luminal B HER2(+) (p = 0.658); luminal A and HER2 overexpression (p = 0.159); luminal B HER2(+) and HER2 overexpression (p = 0.428); and HER2 overexpression and basal-like (p = 0.249)).
the most specific and sensitive LNR value was 29.8%. According to this cut-off, significant differences were found between the molecular subtypes by Kaplan-Meier analysis (basal-like and both luminal A (p = 0.003) and luminal B HER2(+) (p = 0.04)). Although a statistically significant difference in LNR was found between luminal A and HER2 overexpression subtypes, this did not extend to the basal-like group as was the case for survival rate. This should be pursued in future studies taking into consideration the heterogeneity of the basal-like subtype. In addition, although our study showed the LNR differed according to molecular subtype, all groups were classified as medium according to the cut-off values defined by Vinh-Hung et al. [11]. For this reason, further studies will be required to refine the classification based on LNR to narrower intervals. This study explored the relationship between LNR and various BC prognostic factors. No association was found with menopausal status, tumor grade or location, and PNI. However, statistically significant associations between LNR and both tumor size and LVI were identified. Analysis revealed that only PNI and LVI were of prognostic value for estimating survival, and the latter was also associated with LNR. There
was no significant impact of menopausal status, tumor grade, and tumor location on survival rate, as was the case for LNR. These findings are compatible with each other and demonstrate the prognostic reliability of LNR. However, the dis advantages of the study were its retrospective structure and the fact that there was no evaluation of BC histopathology or Ki-67 status. In summary, the LNR differed significantly between the luminal A and HER2 overexpression subtypes. Evaluation of their respective survival rates revealed luminal A has a more favorable prognosis, and this corresponds to the lowest LNR. In addition, similar relationships were identified between LNR and survival rates based on other prognostic factors. However, the basal-like subtype must be examined in more detail because of its heterogeneity and low survival rate despite having a low LNR. For this reason, more precise LNR intervals are required that may come from studies with larger BC cohorts.
Disclosure Statement The authors declare that they have no conflict of interest.
1 American Cancer Society: Cancer Facts and Figures. Atlanta, GA, American Cancer Society, 2002. 2 Siegel R, Naishadham D, Jemal A: Cancer statistics, 2012. CA Cancer J Clin 2012;62:10–29. 3 Yiangou C, Shousha S, Sinnett HD: Primary tumour characteristics and axillary lymph node status in breast cancer. Br J Cancer 1999;80:1974– 1978. 4 Sotiriou C, Pusztai L: Gene-expression signatures in breast cancer. N Engl J Med 2009;360:790–800. 5 Mazouni C, Rimareix F, Mathieu MC, Uzan C, Bourgier C, André F, Delaloge S, Garbay JR: Outcome in breast molecular subtypes according to nodal status and surgical procedures. Am J Surg 2013;205:662–667. 6 Wiechmann L, Sampson M, Stempel M, Jacks LM, Patil SM, King T, Morrow M: Presenting features of breast cancer differ by molecular subtype. Ann Surg Oncol 2009;16:2705–2710. 7 Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL: AJCC Cancer Staging Manual (7th ed). New York, NY, Springer, 2010. 8 Questions and answers on breast cancer. A guide for women and their physicians. Steering Committee on Clinical Practice Guidelines for the Care and Treatment of Breast Cancer. CMAJ 1998;158(3 suppl):1–31. 9 Vinh-Hung V, Burzvkowski T, Cserni G, Voordeckers M, Van De Stene J, Storme G: Functional form of the effect of the numbers of axillary nodes on survival in early breast cancer. Int J Oncol 2003; 22:697–704. 10 Vinh-Hung V, Storme G: No nodal cutoff in nodepositive breast cancer women treated with mastectomi. Breast Cancer Res Treat 2006;98:173–178. 11 Vinh-Hung V, Verkooijen HM, Fioretta G, Neyroud-Caspar I, Rapiti E, Vlastos G, Deglise C, Usel M, Lutz JM, Bouchardy C: Lymph node ratio as an alternative to pN staging in node-positive breast cancer. J Clin Oncol 2009;27:1062–1068.
640
Onkologie 2013;36:637–640
12 Vinh-Hung V, Nguyen NP, Cserni G, Truong P, Woodward W, Verkooijen HM, Promish D, Ueno NT, Tai P, Nieto Y, Joseph S, Janni W, Vicini F, Royce M, Storme G, Wallace AM, Vlastos G, Bouchardy C, Hortobagyi GN: Prognostic value of nodal ratios in node-positive breast cancer: a compiled update. Future Oncol 2009;5:1585–1603. 13 Li Y, Holmes E, Shah K, Albuquerque K, Szpaderska A, Erșahin C: The prognostic value of lymph node cross-sectional cancer area in node-positive breast cancer: a comparison with N stage and lymph node ratio. Patholog Res Int 2012;2012: 161964. 14 Woodward WA, Vinh-Hung V, Ueno NT, Cheng YC, Royce M, Tai P, Vlastos G, Wallace AM, Hortobagyi GN, Nieto Y: Prognostic value of nodal ratios in node-positive breast cancer. J Clin Oncol 2006;24:2910–2916. 15 Ahn SH, Kim HJ, Lee JW, Gong GY, Noh DY, Yang JH, Jung SS, Park HY: Lymph node ratio and pN staging in patients with node-positive breast cancer: a report from the Korean breast cancer society. Breast Cancer Res Treat 2011;139:507–515. 16 Goldhirsch A, Glick JH, Gelber RD, Coates AS, Thürlimann B, Senn HJ; panel members: Meeting highlights: international expert consensus on the primary therapy on early breast cancer 2005. Ann Oncol 2005;16:1569–1583. 17 Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL: Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987; 235:177–182. 18 Berger MS, Locher GW, Saurer S, Gullick WJ, Waterfield MD, Groner B, Hynes NE: Correlation of c-erbB-2 gene amplification and protein expression in human breast carcinoma nodal status and nuclear grading. Cancer Res 1988;48:1238–1243. 19 Paik S, Hazan R, Fisher ER, Sass RE, Fisher B, Redmond C, Schlessinger J, Lippman ME, King CR:
Pathologic findings from the National Surgical Adjuvant Breast and Bowel Project: prognostic significance of erbB-2 protein overexpression in primary breast cancer. J Clin Oncol 1990;8:103–112. 20 Winstanley J, Cooke T, Murray GD, PlattHiggins A, George WD, Holt S, Myskov M, Spedding A, Barraclough BR, Rudland PS: The long term prognostic significance of c-erbB-2 in primary breast cancer. Br J Cancer 1991;63:447–450. 21 Crabb SJ, Cheang MC, Leung S, Immonen T, Nielsen TO, Huntsman DD, Bajdik CD, Chia SK: Basal breast cancer molecular subtype predicts for lower incidence of axillary lymph node metastases in primary breast cancer. Clin Breast Cancer 2008; 8:249–256. 22 Morris GJ, Naidu S, Topham AK, Guiles F, Xu Y, McCue P, Schwartz GF, Park PK, Rosenberg AL, Brill K, Mitchell EP: Differences in breast carcinoma characteristics in newly diagnosed AfricanAmerican and Caucasian patients: a single institution compilation compared with the National Cancer Institute’s Surveillance, Epidemiology, and End Results database. Cancer 2007;110:876–884. 23 Bertucci F, Finetti P, Cervera N, Esterni B, Hermitte F, Viens P, Birnbaum D: How basal are triple-negative breast cancers? Int J Cancer 2008; 123:236–240. 24 Rakha EA, Elsheikh SE, Aleskandarany MA, Habashi HO, Green AR, Powe DG, El-Sayed ME, Benhasouna A, Brunet JS, Akslen LA, Evans AJ, Blamey R, Reis-Filho JS, Foulkes WD, Ellis IO: Triple-negative breast cancer: distinguishing between basal and nonbasal subtypes. Clin Cancer Res 2009;15:2302–2310. 25 Hartkopf AD, Banys M, Krawczyk N, Staebler A, Becker S, Hoffmann J, Hahn M, Wallwiener M, Fehm T: Bone marrow versus sentinel lymph node involvement in breast cancer: a comparison of early hematogenous and early lymphatic tumor spread. Breast Cancer Res Treat 2012;131:501–508.
Demircioglu/Demirci/Kilic/ Ozkan/Karahacioglu
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
References
Published online: October 14, 2013
Clinical Significance of Lymph Node Ratio in Locally Advanced Breast Cancer Molecular Subtypes Fatih Demircioglua Umut Demircib Diclehan Kilicc Secil Ozkand Eray Karahaciogluc a
Rize Recep Tayyip Erdogan University Hospital, Department of Radiation Oncology, Rize, Dr.Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital, Department of Medical Oncology, Ankara, c Department of Radiation Oncology, d Department of Public Health, Gazi University Faculty of Medicine, Ankara, Turkey
Keywords Breast cancer · Lymph node ratio · Molecular subtypes · Prognostic factors
Schlüsselwörter Mammakarzinom · Lymphknotenratio · Molekulare Subtypen · Prognosefaktoren
Summary Background: The ratio of metastatic to dissected lymph nodes (lymph node ratio; LNR) is a sensitive and supe rior prognostic factor for lymph node evaluation, but its relationship to cancer subtypes is unclear. Patients and Methods: Data from 469 patients with axillary lymph node metastasis out of 640 early breast cancer cases were retrospectively analyzed. They were classified into 4 molecular subtypes; luminal A, luminal B HER2(+), HER2 overexpression, basal-like. LNRs were compared between groups and with other prognostic factors. Results: The distribution of LNRs was 35.2% in luminal A, 43.2% in luminal B HER2(+), 46.9% in HER2 over expression, and 39.1% in basal-like. A significant differ ence was found between luminal A and HER2 over expression subtypes (p = 0.023). LNR was significantly correlated with tumor size and lymphovascular invasion, but not with other prognostic factors including meno pausal status, laterality, grade, and perineural invasion. An LNR of 29.8% was defined as the cut-off value, and significant differences in survival rates were identified accordingly between basal-like and both luminal A (p = 0.003) and luminal B HER2(+) (p = 0.04). Conclusion: The LNR differs between some molecular subtypes of breast cancer, and it correlates with certain prognostic factors and survival. These data support using the LNR to assess breast cancer patients.
Zusammenfassung Hintergrund: Das Zahlenverhältnis zwischen befallenen und entfernten Lymphknoten (Lymphknotenratio) ist ein sensibler und überlegener Prognosefaktor für die Lymph knotenbewertung. Das Verhältnis zwischen Lymphkno tenratio und Tumorsubtypen ist jedoch unklar. Patienten und Methoden: Die Daten von 469 Patientinnen mit axil lären Lymphknotenmetastasen (von insgesamt 640 Pati entinnen mit frühem Mammakarzinom) wurden retros pektiv analysiert. 4 molekulare Subtypen wurden unter schieden: Luminal A, Luminal B HER2(+), HER2-Über expression und basal-like. Die Lymphknotenratios wurden innerhalb der Gruppen sowie mit anderen Prog nosefaktoren verglichen. Ergebnisse: Die Distribution der Lymphknotenratios war 35,2% für Luminal A, 43,2% für Luminal B HER2(+), 46,9% für HER2-Überexpression und 39,1% für den basal-like Zelltyp. Ein signifikanter Unterschied bestand zwischen den Subtypen Luminal A und HER2-Überexpression (p = 0,023). Die Lymphknoten ratio war signifikant mit Tumorgröße und lymphovas kulärer Invasion korreliert, nicht jedoch mit anderen Pro gnosefaktoren wie menopausalem Status, Lateralisation, Grad und perineuraler Invasion. Eine Lymphknotenratio von 29,8% wurde als Cut-Off-Wert definiert, und signi fikante Unterschiede in der Überlebensrate wurden ent sprechend zwischen dem basal-like Subtyp und sowohl Luminal A (p = 0,003) als auch Luminal B HER2(+) (p = 0,04) identifiziert. Schlussfolgerung: Die Lymphknotenratio ist bei gewissen molekularen Subtypen des Mammakarzi noms unterschiedlich und mit bestimmten Prognose faktoren und Überleben korreliert. Die vorliegenden Daten unterstützen den Einsatz der Lymphknotenratio bei der Bewertung von Mammakarzinompatientinnen.
© 2013 S. Karger GmbH, Freiburg 0378-584X/13/3611-0637$38.00/0 Fax +49 761 4 52 07 14 [email protected] www.karger.com
Accessible online at: www.karger.com/onk
Fatih Demircioglu, MD Department of Radiation Oncology Rize Recep Tayyip Erdogan University Hospital Șehitler st. 74 5300 Rize, Turkey [email protected]
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
b
Breast cancer (BC) is the most frequent type of cancer and the second most common cause of cancer-related deaths in women [1, 2]. Hormone receptor (HR) status, human epidermal growth factor receptor 2 (HER2) expression, and axillary lymph node involvement are the most important prognostic factors [3]. The heterogeneity of disease makes identification of molecular ‘intrinsic’ subtypes by gene expression profiling important for prognosis and for guiding molecular therapies [4, 5]. Due to the high correspondence between molecular subtypes and clinicopathological features, HR status, HER2 overexpression, and Ki-67, as well as grading are important when classifying BC [6]. Axillary lymph node status is classified only by the number of positive lymph nodes according to the American Joint Committee on Cancer (AJCC) staging system [7]. 10 or more excised lymph nodes is considered sufficient for axillary dissections [8]. However, this number varies for each patient, and there is a close relationship between the number of metastatic lymph nodes and the total number dissected. Some studies have suggested that each involved lymph node affects the prognosis of BC [9, 10]. In this situation, the AJCC lymph node staging system can lead to heterogeneous results. To avoid this confusion, research efforts have focused on implementing another metric, the ratio of metastatic to total dissected lymph nodes (lymph node ratio; LNR), which has proven to be a more sensitive and superior prognostic factor than the total number of involved lymph nodes [11–15]. To our knowledge, no studies have evaluated the relationship between LNR and molecular subtypes in BC. To address this, we examined whether LNR differs between molecular subtypes and sought to identify markers that may contribute to any differences. In addition, the effect of LNR on survival and its correlation with other prognostic factors was investigated.
Patients and Methods Patients were classified into 4 molecular subtypes based on histological grade and immunohistochemical staining for estrogen receptor (ER), progesterone receptor (PR), and HER2. Grading was performed according to the Bloom-Richardson grading system. ER, PR, and HER2 status were determined by immunohistochemical evaluation. 1 (+)HER2 was defined as negative, 3 (+)HER2 was defined as positive, and fluorescence in situ hybridization was performed for 2 (+)HER2. Accordingly, the luminal A group was defined as ER(+) and/or PR(+), HER2(–), and grade 1; the luminal B HER2(+) group as ER(+) and/or PR(+), and HER2(+); the HER2 overexpression group as ER(–), PR(–), and HER2(+); and the basal-like group as those with triple-negative molecular indicators. All patients in the study had complete axillary dissection. LNR was defined as the ratio of metastatic to total dissected lymph nodes, calculated using the formula of (metastatic lymph node number/total dissected lymph node number) × 100, and given as percentages. Patients without lymph node metastasis were excluded because LNRs were not calculated. Descriptive statistics are presented as frequencies, percent, and median-mean (min, max) values. Continuous data were tested for normality with the Kolmogorov-Smirnov test. Since the variables are not nor-
638
Onkologie 2013;36:637–640
mally distributed, the continuous variables were compared by MannWhitney U test between 2 groups and Kruskal-Wallis test of variance between more than 2 groups. Correlation coefficients and their significance were calculated using the Spearman test. The capacity of LNR in predicting overall survival was analyzed by ROC (receiver operating characteristic) curve analysis. When a significant cut-off value was observed, sensitivity and specificity values were presented. The Kaplan-Meier survival estimates were calculated. A separate log-rank test was used to identify the independent effect of the various prognostic factors. The association between clinicopathological factors, biological subtype, and outcome was evaluated with univariate analysis. Statistical analyses were performed using SPSS v16.0 (IBM SPSS Inc., Chicago, IL, USA), and values of p < 0.05 were considered statistically significant.
Results This study was a retrospective analysis of data from 469 patients with axillary lymph node metastasis out of 640 early BC cases followed at the Gazi University Faculty of Medicine, Ankara, Turkey, between 2001 and 2008. 171 patients without lymph node metastasis were excluded because LNRs were not calculated. Patient characteristics are summarized in table 1. The LNRs calculated for each molecular subtype were: 35.2% in luminal A, 43.2% in luminal B HER2(+), 46.9% in HER2 overexpression, and 39.1% in basal-like. Kruskal- Wallis analysis revealed a statistically significant difference between the groups (p = 0.012). The Spearman test was used Table 1. Patient characteristics Age, median (range), years Menopausal status, n Premenopausal Postmenopausal ER status, n Positive Negative PR status, n Positive Negative HER2 status, n Positive Negative Molecular subtypes, n (%) Basal-like HER2+ Luminal A Luminal B HER2(+) Laterality, n Right Left Grade, n (%) 1 2 3 NA BCS/MRM Tumor size, median (range), cm Patients with metastatic LN, n (%) Median dissected LN, n (%) Median metastatic LN, n (%) Median LNR, %
56 (23–87) 184 285 291 178 291 178 166 303 72 (15.3) 62 (13.2) 231 (49.3) 104 (22.2) 233 246 242 (51.6) 96 (20.5) 75 (16) 56 (11.9) 37/432 3.2 (0.4–12.5) 469 (73) 17 (2.64) 5 (1–44) 31.8 (95% CI 2.1–100)
ER = Estrogen receptor; PR = progesteron receptor; HER2 = human epidermal growth factor receptor 2; NA = not available; BCS/MRM = breast conserving surgery/modified radical mastectomy; LN = lymph nodes; LNR = lymph node ratio.
Demircioglu/Demirci/Kilic/ Ozkan/Karahacioglu
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
Introduction
to evaluate the significance of the correlation for each subtype (table 2), which revealed that the LNR differed significantly between the luminal A and HER2 overexpression groups (p = 0.023). The relationship between LNR and other prognostic factors was examined separately. A statistically significant association was found between LNR and both tumor size and lymphovascular invasion (LVI), but not between other prognostic factors including menopausal status, tumor location, tumor grade, and perineural invasion (PNI) (table 3). The survival times of patients with different molecular subtypes of BC were 112.8 months in luminal A, 104.2 months in luminal B HER2(+), 95.9 months in HER2 overexpression, and 93.4 months in the basal-like group. Statistically significant difference was found between survival times according to molecular subtype (p = 0.024). The impact of prognostic factors on survival rates was analyzed, and only molecular subtypes Table 2. Molecular subtypes according to lymph node ratio (Kruskal-Wallis test)
Discussion
p
Standard error
0.39 0.75 0.78
4.9523 3.8581 4.3822
0.39 0.02 0.86
4.9523 4.0884 4.5863
0.75 0.02 0.08
3.8581 4.0884 3.3754
0.78 0.86 0.08
4.3822 4.5863 3.3754
Table 3. The lymph node ratio (LNR) and survival differences according to prognostic factors Prognostic factors Menopausal status Premenopausal Postmenopausal Tumor location Right Left Tumor grade I II III Tumor size, cm ≤5 >5 PNI Positive Negative LVI Positive Negative
LNR, %
p
Survival
p
0.796
0.271
0.187
0.45
0.286
0.471
0.027
0.059
0.11
0.039
0.014
0.048
38.07 39.81 37.76 40.37 35.93 38.64 41.19 37.94 44.74 39.21 36.71 42.99 35.75
PNI = Perineural invasion; LVI = lymphovascular invasion.
Lymph Node Ratio in Locally Advanced Breast Cancer Subtypes
BC is a heterogeneous disease that is associated with several prognostic factors. The ratio of metastatic to dissected lymph nodes was shown to be a more sensitive and superior prognostic factor than the classical AJCC staging system [11– 15]. In the present study, the LNR and other prognostic factors were examined with respect to tumor molecular subtype, and the relationship of these factors to survival rates were investigated in BC patients. The lowest LNR was found in the luminal A group (35.2%), while the highest was found in the HER2 overexpression group (46.9%); this difference was statistically significant. In agreement with the literature [16–20], the LNR was as high as 43.2% in the luminal B HER2(+) group; however, this did not distinguish it from other subtypes as the differences were not statistically significant. The basallike group had a lower LNR of 39.1% that may be contrary to expectations. Crabb et al. [21] investigated axillary nodal involvement in 4,444 early-stage BC cases and showed that the basal-like subtype predicts a lower incidence of axillary nodal involvement, which is in agreement with our study. In addition, studies of triple-negative BC revealed marked heterogeneity within this group, and gene expression profiling is important for inter-group discrimination [22–24]. Despite the poor survival data, the lower LNR may have arisen from this heterogeneity, a fact that should be considered in future genetic studies. Moreover, hematogenous tumor spread can be independent from lymphatic tumor spread and may be responsible for this situation [25]. Hartkopf et al. [25] investigated the relationship between disseminated tumor cells in bone marrow and lymph node involvement, and found no association of lymph node status and molecular subtype with bone marrow involvement. LNR cut-off values have been defined by others with the aim of staging. These levels were 20 and 65% in the study of Vinh-Hung et al. [11] which is the most comprehensive to date. In another study, a value of 25% was correlated with a statistically significant impact on survival rate [13]. We found Onkologie 2013;36:637–640
639
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
Basal-like HER2+ Luminal A Luminal HER2(+) HER2+ Basal-like Luminal A Luminal B HER2(+) Luminal A Basal-like HER2+ Luminal B HER2(+) Luminal B HER2(+) Basal-like HER2+ Luminal A
and LNR showed a significant association. There were no significant differences in survival according to menopausal status, tumor location, or grade (table 3). The significant cut-off value was 29.8% in ROC curve analysis, sensitivity and specificity values were 70.9 and 59.6%, respectively. Separation of patients below and above this value corresponded to survival rates of 82.9 and 57.3%, respectively (p = 0.0001). A KaplanMeier analysis based on this cut-off value revealed statistically significant differences between basal-like subtype and both the luminal A (p = 0.003) and luminal B HER2(+) groups (p = 0.04). Each group was also compared in pairs by KaplanMeier analysis, and no statistically significant values were found between other groups (luminal A and luminal B HER2(+) (p = 0.658); luminal A and HER2 overexpression (p = 0.159); luminal B HER2(+) and HER2 overexpression (p = 0.428); and HER2 overexpression and basal-like (p = 0.249)).
the most specific and sensitive LNR value was 29.8%. According to this cut-off, significant differences were found between the molecular subtypes by Kaplan-Meier analysis (basal-like and both luminal A (p = 0.003) and luminal B HER2(+) (p = 0.04)). Although a statistically significant difference in LNR was found between luminal A and HER2 overexpression subtypes, this did not extend to the basal-like group as was the case for survival rate. This should be pursued in future studies taking into consideration the heterogeneity of the basal-like subtype. In addition, although our study showed the LNR differed according to molecular subtype, all groups were classified as medium according to the cut-off values defined by Vinh-Hung et al. [11]. For this reason, further studies will be required to refine the classification based on LNR to narrower intervals. This study explored the relationship between LNR and various BC prognostic factors. No association was found with menopausal status, tumor grade or location, and PNI. However, statistically significant associations between LNR and both tumor size and LVI were identified. Analysis revealed that only PNI and LVI were of prognostic value for estimating survival, and the latter was also associated with LNR. There
was no significant impact of menopausal status, tumor grade, and tumor location on survival rate, as was the case for LNR. These findings are compatible with each other and demonstrate the prognostic reliability of LNR. However, the dis advantages of the study were its retrospective structure and the fact that there was no evaluation of BC histopathology or Ki-67 status. In summary, the LNR differed significantly between the luminal A and HER2 overexpression subtypes. Evaluation of their respective survival rates revealed luminal A has a more favorable prognosis, and this corresponds to the lowest LNR. In addition, similar relationships were identified between LNR and survival rates based on other prognostic factors. However, the basal-like subtype must be examined in more detail because of its heterogeneity and low survival rate despite having a low LNR. For this reason, more precise LNR intervals are required that may come from studies with larger BC cohorts.
Disclosure Statement The authors declare that they have no conflict of interest.
1 American Cancer Society: Cancer Facts and Figures. Atlanta, GA, American Cancer Society, 2002. 2 Siegel R, Naishadham D, Jemal A: Cancer statistics, 2012. CA Cancer J Clin 2012;62:10–29. 3 Yiangou C, Shousha S, Sinnett HD: Primary tumour characteristics and axillary lymph node status in breast cancer. Br J Cancer 1999;80:1974– 1978. 4 Sotiriou C, Pusztai L: Gene-expression signatures in breast cancer. N Engl J Med 2009;360:790–800. 5 Mazouni C, Rimareix F, Mathieu MC, Uzan C, Bourgier C, André F, Delaloge S, Garbay JR: Outcome in breast molecular subtypes according to nodal status and surgical procedures. Am J Surg 2013;205:662–667. 6 Wiechmann L, Sampson M, Stempel M, Jacks LM, Patil SM, King T, Morrow M: Presenting features of breast cancer differ by molecular subtype. Ann Surg Oncol 2009;16:2705–2710. 7 Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL: AJCC Cancer Staging Manual (7th ed). New York, NY, Springer, 2010. 8 Questions and answers on breast cancer. A guide for women and their physicians. Steering Committee on Clinical Practice Guidelines for the Care and Treatment of Breast Cancer. CMAJ 1998;158(3 suppl):1–31. 9 Vinh-Hung V, Burzvkowski T, Cserni G, Voordeckers M, Van De Stene J, Storme G: Functional form of the effect of the numbers of axillary nodes on survival in early breast cancer. Int J Oncol 2003; 22:697–704. 10 Vinh-Hung V, Storme G: No nodal cutoff in nodepositive breast cancer women treated with mastectomi. Breast Cancer Res Treat 2006;98:173–178. 11 Vinh-Hung V, Verkooijen HM, Fioretta G, Neyroud-Caspar I, Rapiti E, Vlastos G, Deglise C, Usel M, Lutz JM, Bouchardy C: Lymph node ratio as an alternative to pN staging in node-positive breast cancer. J Clin Oncol 2009;27:1062–1068.
640
Onkologie 2013;36:637–640
12 Vinh-Hung V, Nguyen NP, Cserni G, Truong P, Woodward W, Verkooijen HM, Promish D, Ueno NT, Tai P, Nieto Y, Joseph S, Janni W, Vicini F, Royce M, Storme G, Wallace AM, Vlastos G, Bouchardy C, Hortobagyi GN: Prognostic value of nodal ratios in node-positive breast cancer: a compiled update. Future Oncol 2009;5:1585–1603. 13 Li Y, Holmes E, Shah K, Albuquerque K, Szpaderska A, Erșahin C: The prognostic value of lymph node cross-sectional cancer area in node-positive breast cancer: a comparison with N stage and lymph node ratio. Patholog Res Int 2012;2012: 161964. 14 Woodward WA, Vinh-Hung V, Ueno NT, Cheng YC, Royce M, Tai P, Vlastos G, Wallace AM, Hortobagyi GN, Nieto Y: Prognostic value of nodal ratios in node-positive breast cancer. J Clin Oncol 2006;24:2910–2916. 15 Ahn SH, Kim HJ, Lee JW, Gong GY, Noh DY, Yang JH, Jung SS, Park HY: Lymph node ratio and pN staging in patients with node-positive breast cancer: a report from the Korean breast cancer society. Breast Cancer Res Treat 2011;139:507–515. 16 Goldhirsch A, Glick JH, Gelber RD, Coates AS, Thürlimann B, Senn HJ; panel members: Meeting highlights: international expert consensus on the primary therapy on early breast cancer 2005. Ann Oncol 2005;16:1569–1583. 17 Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL: Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987; 235:177–182. 18 Berger MS, Locher GW, Saurer S, Gullick WJ, Waterfield MD, Groner B, Hynes NE: Correlation of c-erbB-2 gene amplification and protein expression in human breast carcinoma nodal status and nuclear grading. Cancer Res 1988;48:1238–1243. 19 Paik S, Hazan R, Fisher ER, Sass RE, Fisher B, Redmond C, Schlessinger J, Lippman ME, King CR:
Pathologic findings from the National Surgical Adjuvant Breast and Bowel Project: prognostic significance of erbB-2 protein overexpression in primary breast cancer. J Clin Oncol 1990;8:103–112. 20 Winstanley J, Cooke T, Murray GD, PlattHiggins A, George WD, Holt S, Myskov M, Spedding A, Barraclough BR, Rudland PS: The long term prognostic significance of c-erbB-2 in primary breast cancer. Br J Cancer 1991;63:447–450. 21 Crabb SJ, Cheang MC, Leung S, Immonen T, Nielsen TO, Huntsman DD, Bajdik CD, Chia SK: Basal breast cancer molecular subtype predicts for lower incidence of axillary lymph node metastases in primary breast cancer. Clin Breast Cancer 2008; 8:249–256. 22 Morris GJ, Naidu S, Topham AK, Guiles F, Xu Y, McCue P, Schwartz GF, Park PK, Rosenberg AL, Brill K, Mitchell EP: Differences in breast carcinoma characteristics in newly diagnosed AfricanAmerican and Caucasian patients: a single institution compilation compared with the National Cancer Institute’s Surveillance, Epidemiology, and End Results database. Cancer 2007;110:876–884. 23 Bertucci F, Finetti P, Cervera N, Esterni B, Hermitte F, Viens P, Birnbaum D: How basal are triple-negative breast cancers? Int J Cancer 2008; 123:236–240. 24 Rakha EA, Elsheikh SE, Aleskandarany MA, Habashi HO, Green AR, Powe DG, El-Sayed ME, Benhasouna A, Brunet JS, Akslen LA, Evans AJ, Blamey R, Reis-Filho JS, Foulkes WD, Ellis IO: Triple-negative breast cancer: distinguishing between basal and nonbasal subtypes. Clin Cancer Res 2009;15:2302–2310. 25 Hartkopf AD, Banys M, Krawczyk N, Staebler A, Becker S, Hoffmann J, Hahn M, Wallwiener M, Fehm T: Bone marrow versus sentinel lymph node involvement in breast cancer: a comparison of early hematogenous and early lymphatic tumor spread. Breast Cancer Res Treat 2012;131:501–508.
Demircioglu/Demirci/Kilic/ Ozkan/Karahacioglu
Downloaded by: NYU Medical Center Library 128.122.253.212 - 5/20/2015 4:11:00 AM
References
