Clinical and pathological predictors of the response to neoadjuvant anthracycline chemotherapy in locally advanced breast cancer

Clinical and pathological predictors of the response to neoadjuvant anthracycline chemotherapy in locally advanced breast cancer

Mohamed M. Elmessiry, a MD; Maher M.  Ekeiny,a MD; TarekA. Elfayomy,a MD;  HaythamM. Fayed,a MD; Basma Elsabaa, bMD;   Mohamed Farouk,C MD

a) Department of Surgery, Alexandria University, Alexandria.

b) Department of  Pathology,  Alexandria University, Alexandria.

c) Department of  Medical  Oncology, Alexandria University, Alexandria.

Abstract

Purpose: The aim of this study is to determine clinical and histopathological  characteristics that  predict tumor  response to anthracycline-based  neoadjuvant  chemotherapy  in  operable locally advanced breast cancer.

Patients  and  methods:  We studied primary  tumor  core tissue  biopsies  from 60 patients with operable locally advanced breast cancer. Patients  received anthracycline-based chemotherapy. The World Health Organization (WHO) criteria were used for staging and the Nottingham modification of Bloom & Richardson scoring system was used for tumor grading. Immunohistochemical   staining  for  ER,  PgR,  HER-2  and  Ki-67  was  performed.  Clinical , histopathological and immunohistochemical  characteristics were analyzed and correlated with pathological response to neoadjuvanty chemotherapy.

Results: Mean patients' age was 50.8±9.2  years. Sixty five percent of patients were postmenopausal.  Ninety percent had infiltrating  ductal carcinoma; NOS.  Five percent, 25% and 70%  of patients had stage liB,  IliA  and liiB  disease  respectively. Fifty  five percent of tumors were high grade. Sixty five percent of cases received neoadjuvant chemotherapy (NCT) in the form of FAC while 15% received FEC ( doxorubicin  was replaced with epirubicin) and

20% received AC (without 5-jiurouracil). Disease progression was recorded in 15% of patients while stable disease was observed in 25% of patients and 55% recorded partial response. Only

5% ofpatients concurred complete response. Breast conservative surgery (BCS) was conducted in 33.3% of responders to NCT (p 0.00). Factors associated with a better tumor response were younger age (p 0.03), smaller tumor size (p 0.01), lower nodal stage (p 0.00), high pathological grade (p 0.025), high Ki-67 labeling index, (p 0.00) and low expression of ER, PgR and Her2 (p 0.00). There was no significant relation between tumor response and each of menopausal status, tumor  stage, histological  type, presence of lymphovascular  space invasion (LVSI) or serum levels of tumor markers [CA15.3 & CEA].

Conclusion: Operable, locally advanced breast cancer with large tumor size, higher nodal stage, positive immunohistochemical  reaction to ER, PgR & Her2, lower Ki-67 labeling index and a lower tumor grade predict a lower response to anthracycline-based  NCT  These patients could alternatively benefit from radical resection or a different line of NCT to obtain a better response with the possibility of BCS.

Key   words:   Neoadjuvant   chemotherapy;    operable   locally   advanced   breast   cancer;

pathological response; clinical and pathological predictors.

Introduction:

Locally advanced breast cancer (LABC) constitutes     a    heterogeneous     group    of tumors of varying clinical presentations and biological  behavior.  It represents  up to  10-

20%  of  newly  diagnosed  breast  cancers  in

western  countries.! In developing  countries, it represents up to 50-60 % due to lack of screening programs and decreased public awareness  of  the  importance  of  early detection of breast cancer.l The clinical management of LABC is complex and should be  tailored  to  the  individual  patient  either by  neoadjuvant   or  adjuvant  chemotherapy plus   surgery.   Some  patients   with   LABC have potentially operable tumors including stages  liB,  IliA  and  some  cases  of  stage IIIB. Some advantages of the administration of NCT in operable breast cancer include possible determination  of tumor  response in vivo, earlier treatment of micrometastases, diminishing  drug  resistance,  and  increasing the possibility of breast conservation.2,3 In addition,  some  clinical  studies  have  shown that responders to NCT have better disease­ free survival rates.4 Generally, there is wider experience  with  the  use  of  anthracyclines in the treatment  of these  patients.  However; there is a significant variability in the pathologic response to anthracycline-based NCT, with approximately 65-85% overall response rate, 5-15% complete response and

15-35%   stable  or   progressive   disease.S-10

Currently, the underlying mechanism for this variability is unclear. The identification of clinical and histopathological  parameters that accurately predict such response to treatment is valuable in optimizing NCT so that non­ responders could be offered alternative and more effective chemotherapy or radical surgery.  Previous  studies  have  focused   on the correlation between tumor response to chemotherapy and diverse factors, such as histological grade, DNA ploidy, cell kinetics, and estrogen receptor status of the primary tumor.   However,   most   of   these   studies have    unfortunately    yielded    confounding

results.ll-17

Purpose:

The aim of this  study  was to  determine

clinicopathological features that can predict response of operable locally advanced breast cancer to NCT.

Patients and methods:

Sixty patients with operable LABC (stages liB,  IliA, or IIIB as defined by the American Joint Committee on Cancer Classification "AJCC"  System)  were  studied  during  the period between March 2009 and March 2012. All  patients  underwent   tumor   core  tissue biopsy to confirm the presence of invasive carcinoma   and   to   evaluate   the   primary tumor pathological and immunohistological features. Patients with preoperative chemo or radiotherapy, ductal carcinoma in-situ with minor  invasive  component  were  excluded. Prior to every treatment cycle, radiological three-dimensional     measurements     of    the tumor were performed.   Then, all patients received         neoadjuvant         anthracycline­ based      chemotherapy      administrated      at

21day-intervals. Neoadjuvant regimens consisted    of   600    mg/m2   5-ftuorouracil,

60 mg/m2 doxorubicin, and 600 mg/m2 cyclophosphamide    (FAC)   or   600   mg/m2

5-ftuorouracil,   60  mg/m2   epirrubicin,   and

600   mg/m2   cyclophosphamide    (FEC)   or

60 mg/m2 doxorubicin and 600 mg/m2 cyclophosphamide (AC).

After  achieving  the  maximum  response

(defined as no more reduction in tumor size for  2  successive   cycles)  after  a  minimum of three to a maximum of six cycles of chemotherapy, patients and tumor responses were  evaluated  clinically  and radiologically by    mammography     and    ultrasonography of  both  breasts  and  axillae.     Responders were   offered   breast   conservative   therapy or modified radical mastectomy (MRM) followed by the same chemotherapy regimen as an adjuvant treatment. Non-responders (if still resectable) were offered MRM followed by other lines of chemotherapy.

Clinical & histopathological  data were collected & analyzed to  assess possible correlations with tumor response to anthracycline-based NCT. Compiled data included clinical (patient's age, menopausal status,   tumor    location,   size   and   stage),

radiological (tumor and nodal staging), laboratory (serum  levels of tumor  markers namelyCEA&CA15.3)andhistopathological (histological type,  grade, presence  of lymphovascular LVSI,  hormone  receptor status (ER & PgR), HER-2 expression, Ki-67 labeling index)  in addition  to therapeutic data (type  and  number  of  cycles  of  NCT, tumor and nodal response  to NCT).

Histological analysis:

Sections of formalin-fixed, paraffin embedded   tumor     core    biopsies     stained with  Hematoxylin & Eosin  (H&E)  were studied.   Tumor  type  was  determined as  per the  guidelines of  the  "WHO classification criteria  (2003).  Based  on  the  assessment of formation oftubules "1-3", nuclear atypia "1-

3", and mitotic  counts  "1-3" the appropriate histologic grade  was  assigned  to  the  tumor using  the  criteria  of Nottingham scoring system "Modified Bloom and Richardson" scoring system.l5 Scores were added  up to formulate a final score out of 9. A score  of up to 5 determined a Grade 1 tumor, 6-7 a Grade

2 tumor  and 8-9 Grade  3 tumor.  Presence  of LVSI  &  lymphocytic infiltration   were  also assessed.

Immunohistochemical studies: Immunohistochemical staining            for

estrogen     receptors     (ER),      progesterone receptors  (PgR),    HER-2   and   Ki-67   was performed         on         3-!..lm-thick                         sections            of formalin-fixed paraffin-embedded tissue from core tissue  biopsies  mounted onto positively charged   slides.   Slides   were  de-waxed  and then        re-hydrated. Following          microwave pretreatment,           monoclonal   antibodies                to ER, PgR,  HER-2, and  Ki-67  were  applied overnight  at   4°C.    Then    incubation  with peroxidase labeled  secondary antibody and chromogen was conducted. Each section  was examined  at  low   magnification  (1OOx)   to identify  percentage of positive  staining cells and staining intensity.  Negative  and positive control   slides   were   included   in  each  run. Staining for  ER, PgR and  Ki67  was  always nuclear  while for HER-2 it was membranous. For ER and PgR semiquantitative scoring, the Allred  scoring  system18  which  is the sum  of a score  for  percentage of positively staining

cells (no staining= 0, staining in <1% of cells

= 1, 1 to 10% = 2, 10 to 33% = 3, 33 to 67%

= 4 and 67% to 100%  = 5) and a score for intensity  (absent= 0, weak= 1,  moderate  =

2, strong  =  3). A 0-2  score  was  considered negative.  Allred  scores  3-4  were  considered

1+, 5 and 6 were considered 2+, 7 and 8 were considered 3+. HER2  was graded  according to ASCO/CAP guidelinesl9 as negative when there  was  no immunostaining or when there was weak  immunostaining of  less than 30% (1+),equivocal ( 2+)when there was complete uniform  or weak membranous staining and positive (3+) when there was uniform  intense membranous staining  in at least 30% of cells. Proliferative activity was assessed  in terms of Ki67  labeling  index  as either  less than  20% (low) or >20% (high).20

Assessment of response:

Clinical and radiological changes  in preoperative tumor  size  after  conclusion of chemotherapy were recorded. A complete response was considered when no residual tumorwas detected clinically orradiologically; a partial  response was defined  as tumor reduction to  less  than  50%  of  the  baseline size; stable disease defined as  less than 50% reduction  in  tumor   size  or  less  than   25% increase  in  tumor  size.  Progressive disease was  defined  as  more  than  25%.increase  in tumor  size.21 Pathologically, specimens with no residual  viable tumor  cells were judged to have  complete  pathological response  (pCR), presence  of  one  or  more  foci  of  malignant viable  cells measuring less than  1 millimeter was judged as near complete (microscopic) response while a residual  tumor  larger than 1 millimeter was judged as partial response.22

Statistical analysis:

Chi-square and  ANOVA  tests  were  used to compare  categorical and continuous variables, respectively.

Results:

Pretreatment evaluation showed  that: the mean  age   was  50.8±9.2  years.   Sixty   five percent of patients were postmenopausal. The mean length of largest tumor dimensions was

7.5±2.4  em.  Tumor  stages  ranged  between T3 (18%),  T4b (72%)  and T4d  (10%).  N1 nodal  status  was recorded  in  55% of cases and N2 in 35%. The remaining 10% were NO. Regarding clinical stage,  5%, 25% and 70% of patients recorded stages liB, IliA and IIIB respectively.    Histopathological     evaluation revealed  infiltrating  ductal carcinoma;  NOS in  90%   of  cases  and  infiltrating   lobular carcinoma  in  the  rest. Fifty  five percent of cases were grade III.   LVSI was detected in

85%.  ER/PgR  hormone  receptor  positivity

was detected in 40%. Her-2 overexpression was reported  in 20% of cases  while 38.3% of tumors showed high proliferative activity (high Ki-67 labeling index).

Sixty five percent of the patients received

NCT in the form of FAC  including 600 mg/

m2 5-fluorouracil, 60 mg/m2 doxorubicin, and

600 mg!m2 cyclophosphamide. 15% received FEC (doxorubicin replaced with epirubicin) and 20% receivedAC (without 5-flurouracil).

The  overall  clinical   response   rate  was

60%.  Fifty  three  percent  of  patients   had partial response while 6.7% had a complete response. Stable disease was observed in 25% of patients and 15% of patients developed disease progression. Histopathological assessment revealed complete pathological response   in  5%  of  cases,   near  complete

pathological response in 1.7% and partial pathological response in 53.3%. Breast conservative  surgery  (BCS)  was performed in 33.3% of responders to NCT (p 0.00).

Factors  associated   with  a  better  tumor response to NCT were younger age (p 0.03), smaller tumor size (p 0.01), lower nodal stage (p 0.00), higher tumor grade (p 0.025), higher Ki67 labeling  index  (p 0.00), negative  ER/ PgR status, lack of Her2 over-expression  (p

0.00), elevated serum CA15.3 (p 0.00) and higher number ofNCT cycles (p 0.00). There was no significant relation between response to treatment and any of menopausal status, T stage, histological type, LVSI or serum CEA level.

In the multivariate  analysis, patient's  age

(p=0.001), tumor grade (P=0.003) and Ki67 labeling index (P=0.001) were the only independent predictive factors of response to anthracycline-based  NCT.

Using    logistic   regression     analysis    of

different combinations of predictive variables revealed that the combination  of young age, small tumor size and negative nodal status is the strongest predictor of tumor response to NCT (x2 45.778, p< 0.000). The accuracy of this combination was 91. 7%

ER immunoposirive nuclei, Allred score +3

HER2-positive membranous staining; score 3+

PgR immunoposirive nuclei, Allred score 3+

Ki-67 nuclear staining in, breast carcinoma

Figure ( 1 ): Duct to mucosa P1: Posterior outer row is completeright angle in the pancreatic duct.

Before CT

After CT

Figure (2 ): A case with LABC showing  partial response after neoadjuvant CT.

·      

oversi1Respon$e                                                                                                                   R diolo-s iINodaiStaa•

•n··iX·fl·'l

Figure (3): Relation  between  response to neoadjuvant CT and tumor  size & nodal stage.

Table (1): Relation between clinicalradiological & laboratory findings and tumor response toNCT.

Table (2): Relation between histopathological features and tumor response to NCT.

Table (3): Relation between tumor response to NCT and regimen  & number  of cycles of CT.

Table (3): Clinical and biological parameters in relation to their predictive value for achieving a partial/complete response to NCT: multivariate analysis.

Discussion:

60%  of  patients  in this  study  presented a clinical response to anthracycline-based neoadjuvant chemotherapy. 53.3% had pCR and only 6.7% of patients had a cCR.   In relation  to  the  pathological  response,  55% had pPR and only 5% of patients had a cPR. It has been published in the literature that the response to induction chemotherapy varies between 3% and 30% for cCR and 30-80% for  pCR.2,5-ll Breast  conservative  surgery (BCS) was possible in 20 % of patients after downstaging. It is published in the literature that the rate of breast conservation after induction chemotherapy varies between 15%

and 45%.5-9

We found that a smaller tumor size and few metastatic nodes at diagnosis were related to a better partial and complete response. This could be due to the fact that bigger tumors could have extended zones of necrosis and presence of intratumoral edema that increases interstitial  pressure.  All these features  make it difficult for the chemotherapy  to penetrate into  the   tumor.   Moreover,   a  larger   size means that more cells have been duplicating, favoring more mutations and resistance to chemotherapy.

A strong association  between clinical and pathological response with tumor grade was found; this correlation has been repeated in other studies.23-27 Nevertheless, there are isolated reports that suggest the opposite.l4

Our study showed a significant difference in response to neoadjuvant chemotherapy among patients with high proliferating activity when compared to patients with tumors with a  low  proliferative   activity   demonstrated by  a  high  Ki-67   labeling   index.   Several studies have evaluated cell proliferation and concluded a significant correlation between high proliferative activity and a greater sensitivity to chemotherapy.l4,26,28

In our study, there was an assoc1at10n between HER2 overexpression and tumor response to NCT. Nevertheless, correlation between HER2 overexpression  and response to chemotherapy is controversial. The Cancer Leukemia   Group   B  B8451   Study   found that patients who received high doses of doxorubicin with overexpression of HER2 showed   a  benefit  in  survival_25 In  other studies, Overexpression of HER2, defined as

3+ IHC staining did not influence the clinical or pathological response to anthracycline­ based chemotherapy.3D-32

In   our   study,   ER-negative   status   was a predictor of clinical and pathological response to neoadjuvant chemotherapy. Previous  studies  demonstrated  that  tumors with  positive  ER had  a worse  pathological response   to   NCT   compared   to   negative ER.9,28 Colleoni found a significantly higher clinical response rate in negative ER and PgR tumors, compared to those where ER and/or PgR were expressed.33 These results could be

explained, in part, by the relation that exists between  absence  of  these  receptors  and  a higher tumor grade, being this last one an important response factor.

We found a greater response to treatment with doxorubicin compared with epirubicin. The number of cycles received also influenced the  clinical  response. It is well known  that with a greater number of cycles, it is more likely to achieve a complete clinical response.

In  conclusion,   we   found   that   patients with locally advanced breast cancer tumors with a large volume, advanced node status (N2), low cell proliferation rate, and lower differentiation grade according to the SBR score have a lower response to anthracycline­ based NCT. Patients with tumors with these characteristics could be candidates for radical resection or a different line of neoadjuvant CT to obtain a better response with the possibility ofBCS.

References:

1- RustogiA,  Budrukkar A, Dinshaw K, Jalali R: Management of locally advanced breast cancer:  Evolution  and current  practice. J Can Res Ther 2005; 1: 21-30.

2- Wolff A, Davidson N: Primary systemic therapy  in operable  breast cancer. 1 Clin Oncol2000; 18: 1558-1569.

3- Hortobagyi G: Treatment of breast cancer.

N Engl JMed 1998; 339: 974-984.

4- Aapro  M: Neoadjuvant therapy  in breast cancer: Can we define its role? The Oncologist 2001; 6: 366-369.

5- Vander Hage JA, van de Velde CJH, Julien

JP, et al: Preoperative chemotherapy  in primary  operable  breast  cancer:   Results from the European Organization for Research  and  Treatment  of  Cancer  trial

10902. 1Clin Oncal2001; 19: 4224-4237.

6-  Makris  A,  Powles   TJ,  Ashley  SE,   et al: A reduction in the requirements for mastectomy in a randomized trial of neoadjuvant chemoendocrine therapy in primary breast cancer. Ann Oneal 1998; 9:

1179-1184.

7-  Smith  IC,  Heys  SD,  Hutcheon  AW, et al: Neoadjuvant chemotherapy in breast cancer:   Significantly   enhanced   response

with  docetaxel.  1 Clin  Oneal  2002;  20:

1456-1466.

8- Schwartz GF, Birchansky CA, Komamicky LT, et al: Induction chemotherapy followed by breast conservation for locally advanced carcinoma of the breast. Cancer 1994; 73:

362-369.

9-  Fisher  B,  Bryant  J,  Wolmark  N,  et  al: Effect of preoperative chemotherapy on the outcome of women with operable breast cancer. 1ClinOncal1998; 16: 2672-2685.

10-Kuerer   HM,  Newman   LA,  Smith   TL,

et  al:  Clinical   course  of  breast  cancer patients with complete pathologic primary tumor  and axillary  lymph  node  response to doxorubicin- based neoadjuvant chemotherapy.  1 Clin  Oneal  1999;  17:

460-469.

11-Ring AE, et al: Oestrogen receptor status, pathological     complete     response     and prognosis in patients receiving neoadjuvant chemotherapy for early breast cancer. Br 1

Cancer 2004; 91: 2012-2017.

12-Masters JR, Camplejohn RS, Millis RR, Rubens RD: Histologic grade, elastosis, DNA ploidy and response to chemotherapy. Br 1Cancer1987; 55: 455-457.

13-Aas T, et al: Primary systemic treatment with   weekly   doxorubicin    monotherapy in women with locally advanced breast cancer: Clinical experience and parameters predicting outcome. Acta Oneal 1996; 35:

5-8.

14-Sjostrom J, et al: Amultivariate analysis of tumor biologic factors predicting response to cytotoxic treatment in advanced breast cancer. Br 1Cancer1998; 78: 812-815.

15-Elston   CW,   Ellis   IO:   Assessment   of

histological  grade. In: The  breast.  Elston CW  and  Ellis  IO (Editors);  Edinnburgh; New York: Churchill  Livigstone (Publisher); 1998; Vol 13; p. 356-84.

16-Collecchi  P, et al: Primary chemotherapy in locally advanced breast cancer (LABC): Effects on tumor proliferative activity, bcl-

2 expression  and the relationship  between tumor  regression  and  biological  markers. Eur1Cancer1998; 34: 1701-1704.

17-0zmen V, et al: Biological considerations

in locally  advanced  breast  cancer treated