Good clinical response of breast cancers to neoadjuvant chemoendocrine therapy is associated with improved overall survival

S. J. Cleator1,*, A. Makris2, S. E. Ashley3, R. Lal4 and T. J. Powles5

1 Breakthrough Breast Cancer Research Centre, London; 2 Mount Vernon Hospital, Northwood, Middlesex; 3 Royal Marsden Hospital, Sutton, Surrey; 4 Lincolns Inn Fields Laboratories, London; 5 Parkside Hospital, London, UK

* Correspondence to: Dr S. Cleator, The Breakthrough Breast Cancer Research Centre, Institute of Cancer Research, Fulham Road, London SW3 6JB, UK. Tel: +44-20-7970-6058; Fax: +44-20-7878-3858; Email: scleator{at}icr.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background: We present extended follow-up from a prospective randomised trial evaluating the role of neoadjuvant chemoendocrine therapy in the treatment of operable breast cancer.

Patients and methods: 309 women were randomised to primary surgery followed by eight cycles of adjuvant mitoxantrone, methotrexate with tamoxifen (2MT) or 2MT with mitomycin-C (3MT) versus the same regimen for four cycles before followed by four cycles after surgery. For this analysis the median follow-up of patients was 112 months.

Results: After 10 years follow-up there is still no statistically significant difference in disease-free survival (DFS) (71% versus 71%) or overall survival (OS) (63% versus 70%) when comparing adjuvant versus neoadjuvant treatment, respectively. Of 144 evaluable patients in the neoadjuvant arm, 74 achieved a good clinical response and 70 patients achieved a poor clinical response. Good responders had a superior DFS (80% versus 64%, P=0.01) and OS (77% versus 63%, P=0.03) compared to poor responders.

Conclusions: At 10 years, neoadjuvant and adjuvant treatment continue to have equivalent OS and DFS. Good clinical response to neoadjuvant chemotherapy is associated with superior DFS and OS. This supports the use of clinical response of primary breast cancer to neoadjuvant therapy as a surrogate marker of survival benefit.

Key words: breast cancer, chemotherapy, neoadjuvant, response, survival


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Neoadjuvant, or pre-surgical, chemotherapy is a standard treatment for locally advanced and inflammatory breast cancer [1Go, 2Go]. Interest in the use of neoadjuvant chemotherapy for operable breast cancer arose from the results of animal studies, which showed that surgical removal of the primary tumour resulted in an increase in the labelling index in residual tumour cells [3Go], a phenomenon that could be impaired by administration of neoadjuvant chemotherapy, endocrine treatment or radiotherapy [4Go]. Furthermore, consideration of a popular model of tumour cell kinetics, which describes the emergence of an ever-expanding drug-resistant population as a result of accumulation of mutations over time [5Go], would suggest that chemotherapy should be given as early as possible.

The results of several randomised trials comparing adjuvant versus neoadjuvant chemotherapy/chemoendocrine therapy for operable breast cancer have now been published [6Go–11Go]. None of the published trials show any difference in overall survival between the adjuvant and neoadjuvant arms. However, we have previously reported a relative reduction by neoadjuvant chemotherapy in requirement for mastectomy of ~50% (28% to 13%) [12Go]. Therefore neoadjuvant chemotherapy appears a safe approach which results in avoidance of otherwise unavoidable mastectomy in a significant proportion of women.

Neoadjuvant chemotherapy has proved a valuable research tool, providing the opportunity to observe the effects of cytotoxic drugs in vivo. There is a wealth of literature correlating clinical response with changes in measures of proliferation and apoptosis in breast cancers during treatment and with various pretreatment phenotypic features [13Go]. Underpinning these studies is the concept that biological markers, which are predictors for good response to neoadjuvant therapy may be surrogates of survival benefit from a given treatment. However, there is inconsistency in the literature with respect to the reported relationship between clinical response and overall survival, both in prospective randomised trials and retrospective series.

Between 1990 and 1995 a prospective randomised trial to evaluate the role of neoadjuvant chemotherapy was conducted at the Royal Marsden Hospital. Findings with respect to response rate, mastectomy requirement and 5-year outcome have been reported previously [8Go, 12Go]. In this report we present an update of previously published data.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patients aged 70 years or less presenting with primary breast cancer at the Royal Marsden Hospital (RMH) who were suitable for primary surgery and systemic chemotherapy and tamoxifen were eligible for the trial, which had been approved by the local research and ethics committee (RMH). Between 1990 and 1995, 309 patients, median age 56 (range 27–70), were randomised. Inclusion criteria were: (i) all patients ≤70 years old with primary breast cancer confirmed on cytology or trucut biopsy; (ii) physical and psychological suitability for all treatment options; (iii) informed consent; and (iv) no evidence of metastases at initial assessment. Exclusion criteria included: (i) evidence of metastases at initial assessment; (ii) premenopausal patients who wished to consider further pregnancy; (iii) patients with inoperable tumours for whom chemotherapy and hormone therapy were the initial treatments of choice; and (iv) clinical evidence of myocardial dysfunction.

Patients were randomised to receive either appropriate primary surgery followed within 3–6 weeks by adjuvant chemoendocrine therapy plus radiotherapy if required (adjuvant arm), or primary chemoendocrine therapy for 3 months before and 3 months after appropriate surgery plus radiotherapy if required (neoadjuvant arm).

Clinical stage, size and position of primary tumour and appropriate primary surgical requirements were recorded at randomisation in all patients. Clinical size was measured as the product of the largest diameter and its perpendicular diameter. For patients in the neoadjuvant arm, clinical size of the tumour was recorded at 3-week intervals until the presurgical assessment at 3 months. These patients then had a repeat clinical assessment, mammography, breast ultrasound, and a clinical decision made on their surgical requirements.

Patients randomised to adjuvant therapy had appropriate surgery and radiotherapy according to the size and position of the primary tumour, followed by eight courses of chemotherapy with 3M (mitomycin C, 7 mg/m2 every 6 weeks, mitoxantrone 7 mg/m2 every 3 weeks and methotrexate 35 mg/m2 every 3 weeks, with dose modification according to blood count and symptomatic toxicity) or 2M (same as 3M, with the exclusion of mitomycin C and an increased dose of mitoxantrone to 11 mg/m2) and tamoxifen 20 mg per day for 5 years. Patients randomised to neoadjuvant therapy had four courses of 3M (or 2M) and tamoxifen, followed by four further courses of chemotherapy and tamoxifen for 5 years.

Mastectomy (usually with immediate reconstruction) was considered necessary if the tumour was located within 2 cm of the nipple or was large in comparison to breast size. Otherwise the tumour was excised sufficiently to achieve a high probability of complete surgical excision. Palpable axillary lymph nodes were removed by lower axillary sampling. No axillary procedure was performed for patients with clinically negative lymph nodes. The influence of combined chemoendocrine therapy on clinically undetectable axillary disease is unknown. Axillary node biopsy was not required as a prognostic indicator for systemic therapy. In patients who achieved a clinical complete response (CR), the location of surgical excision was determined by mammographic or ultrasound identification of residual scarring or as located before treatment.

All patients who did not require a mastectomy were given local postoperative radiotherapy to the breast with a dose of 54 Gy in 27 fractions over 5.5 weeks using two tangential fields, followed by a boost of 10 Gy to the tumour bed. Radiotherapy to the axilla and supraclavicular fossa was given for cases with histological evidence of axillary involvement on axillary sampling. Radiotherapy commenced 4–6 weeks after surgery and was given concurrently with chemotherapy.

The clinical measurement of the response to neoadjuvant therapy was undertaken by two independent observers and defined according to the International Union Against Cancer (UICC) criteria [14Go]. No palpable abnormality at this site indicated a CR, and >50% reduction in bidimensional measurements indicated a partial response (PR). A reduction of <50% in bidimensional measurements was recorded as no change (NC), while an increase in measurements of >25% represented progressive disease (PD). Patients with PD received early surgery with completion of the chemoendocrine therapy after surgery. A residual nodularity after a good response which was insufficient to be measured was classified as minimal residual disease (MRD). In this paper MRD and CR have been collectively referred to as ‘good’ clinical response and PR, NC and PD as ‘poor’ clinical response. Surgical specimens from patients who had received neoadjuvant therapy which no longer contained malignant cells (primary and lymph nodes) were defined as displaying a complete pathological response (pCR), in contrast to those found to contain residual invasive malignancy (pINV).

Survival was measured from primary diagnosis until death due to any cause. Disease-free survival was measured from primary diagnosis until first local or distant recurrence and was censored at death without a recurrence. Locoregional control was measured from diagnosis until first locoregional recurrence and was censored at metastatic relapse or death.

Statistical analysis
Treatment and response groups were compared for baseline demographics and treatment details by means of the chi-squared test or Fisher's exact test and the Mann–Whitney non-parametric test (age). Actuarial survival curves were calculated by the Kaplan–Meier method [15Go] and comparisons were made by means of the log rank test.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Three hundred and nine patients were randomised to receive either adjuvant (n=152) or neoadjuvant (n=157) therapy. Eight patients from each arm were excluded for the following reasons: no positive cytology or histology, age >70 years, metastases detected on screening. Two hundred and ninety-three eligible patients (144 adjuvant; 149 neoadjuvant) were analysed. A further two patients in the adjuvant arm and five in the neoadjuvant arm refused further participation in the trial. Therefore 142 patients were treated as per protocol in the adjuvant and 144 in the neoadjuvant arm; these represent the patients analysed in this report.

Table 1 shows patient (age, menopausal status and treatment received) and tumour characteristics (tumour and nodal stage) for patients in the adjuvant and neoadjuvant arms of the study. No statistically significant differences were noted between the two arms in terms of these characteristics. In our previous paper [8Go], outcome data to a median follow-up of 48 months showed no statistically significant difference between the two arms of the trial in terms of DFS, metastatic relapse (MFS) and OS. We now update this outcome analysis with a median follow-up of 112 months (range 12–145 months) (Table 3). After this longer follow-up, DFS for adjuvant and neoadjuvant therapy was 71% and 71%. Corresponding figures for OS were 63% and 70% (P=0.6). Locoregional control (RLC) for adjuvant and neoadjuvant therapy was 94% and 91%, respectively (P=0.7). Metastatic control was 72% in patients receiving adjuvant and 76% in patients receiving neoadjuvant therapy (P=0.6).


View this table:
[in this window]
[in a new window]
 
Table 1. Patient and tumour characteristics: adjuvant versus neoadjuvant

 

View this table:
[in this window]
[in a new window]
 
Table 3. Patient outcome

 
Of 144 patients receiving neoadjuvant therapy as per protocol, 33 patients achieved a CR, 41 achieved MRD, 46 achieved PR, 22 demonstrated NC and two demonstrated PD. Table 2 shows that there were no significant differences between good responders and poor responders in terms of patient and tumour characteristics. Table 3 shows patient outcome in relation to clinical response. Figures 1 and 2 show the relationship of DFS and OS to clinical response, respectively. DFS of good and poor responders at 10 years was 80% and 64%, respectively (P=0.01). OS of good and poor responders at 10 years was 77% and 63%, respectively; (P=0.03). These differences are largely accounted for by a difference in metastasis-free survival, which was 85% for good responders and 70% for poor responders (P=0.2). There was no significant difference in locoregional control between good and poor responders (93% versus 88%, respectively; P=0.3). The difference in outcome (DFS and OS) between the individual categories of CR, MRD, PR and NC did not reach statistical significance (P=0.2 in each case).


View this table:
[in this window]
[in a new window]
 
Table 2. Patient and tumour characteristics: good versus poor clinical responders

 


View larger version (14K):
[in this window]
[in a new window]
 
Figure 1. Disease-free survival of good and poor responders.

 


View larger version (13K):
[in this window]
[in a new window]
 
Figure 2. Overall survival of good and poor clinical responders.

 
Nineteen patients achieved a complete pathological response (pCR), of which nine had no evidence of malignant cells in the surgical specimen and 10 had residual in situ disease only. One hundred and twenty-three patients displayed pINV. One patient refused surgery and in another the pathology results were unknown. The 10-year DFS for the pCR group (including ductal carcinoma in situ: DCIS) was 81%, compared with 71% in the group with pINV (P=0.2). Corresponding figures for OS were 83% and 69% (P=0.2). Metastasis-free survival in the pCR group (including DCIS) was 100% and in the residual invasive group, 74% (P=0.02). In other words, none of those patients without invasive disease at surgery developed metastases, whereas a quarter of those with pINV relapsed with secondaries.


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
At 10 years of follow-up there is no significant difference in the locoregional control or metastasis-free survival between patients receiving neoadjuvant or adjuvant chemoendocrine therapy. This result is consistent with other randomised studies that compare sequencing of chemotherapy before or after locoregional treatment [6Go, 9Go–11Go, 16Go]. Several studies have shown that neoadjuvant treatment is associated with a reduction in mastectomy rate due to down-staging [11Go, 12Go, 17Go]. It is now apparent, therefore, that those patients who are candidates for mastectomy at presentation can be safely treated with preoperative chemotherapy with the aim of down-staging to allow breast conserving surgery, with no detriment to survival.

There is considerable interest in the use of response to neoadjuvant chemotherapy as an early surrogate of long-term outcome. Previous studies have consistently shown that patients who achieve pCR have improved overall survival [10Go, 11Go, 18Go]. In this study we observed a superior overall and metastasis-free survival associated with a lack of residual invasive disease, but only the latter reached statistical significance, probably due to the small number of patients in this group. Thirteen percent (19/142) of patients in this study achieved a pCR, and none relapsed systemically. An update of the NSABP B-18 study at 10 years [10Go] did not include figures for MFS and the definition of pCR related to the breast pathology alone (axillary node status not included), so it is difficult to compare the two sets of data. However, in most trials with anthracycline-based chemotherapy only 5%–10% of patients achieve this end point [8Go, 10Go, 11Go, 18Go], thereby reducing its clinical usefulness as a surrogate of treatment ‘benefit’. Furthermore, pCR can only be determined after surgery and cannot be used to tailor treatment on an individual patient basis. In this study we compared the outcome in good clinical responders with poor responders. Good clinical responders, defined as those with either CR or MRD, accounted for ~50% of the patients overall. This dichotomy makes this a more useful end point for trials where neoadjuvant chemotherapy is used as an in vivo model to define biological markers of response.

With long follow-up, our results show that good responders have superior outcome both in terms of DFS and OS. The improved outcome appears to be due largely to a difference in metastasis-free survival rather than a difference in locoregional control and hence to a difference in the extent of eradication of micrometastatic disease. This implies that micrometastases have the same chemosensitivity as their corresponding primary tumour. This is in line with studies that demonstrate a similar microarray expression profile to metastases and matched primaries [19Go].

The association of good outcome with good clinical response to neoadjuvant chemotherapy has been documented in previous studies by Bonnadonna and Valagussa [20Go] (improved DFS), Pierga et al. [21Go] (improved OS) and Hortobagyi et al. [22Go] (improved DFS). Improved outcome was also seen in the largest neoadjuvant trial (NSABP-B18) at 9 years of follow-up for both DFS and OS [10Go]. However, several other neoadjuvant studies have failed to show an association [7Go, 9Go, 11Go, 23Go]. It is noted that the categorisation of responders and non-responders has not been consistent in these studies.

Tumour regression after neoadjuvant therapy can be assessed clinically, radiologically or pathologically. Clinical measurement has been criticised as being subject to inter-observer variation, and this may account for why some studies have failed to demonstrate improved outcomes in good responders. Single institution studies, such as ours, while limited in terms of patient numbers, have the advantage that they rely on a small number of clinicians performing such measurements and thereby achieve greater consistency. Radiological evaluation of response, either by ultrasound or mammography, has been used for this purpose. However, radiological response is also subject to inter-observer variation and the correlation with histology has not been consistently shown to be superior to that of clinical evaluation [24Go]. Furthermore, there are no studies with long-term follow-up showing improved outcome in patients achieving good radiological response.

There is currently considerable interest in the use of microarray profiling to predict response to neoadjuvant chemotherapy, although currently the associated error rate overall for prediction of pCR [25Go] or clinical response [26Go] is high. This study demonstrates that poor response to chemotherapy is associated with poor outcome, and probably a lack of benefit from chemotherapy. It is probably most useful to develop the capacity to predict accurately, (of all subclasses of response), those patients who are poor clinical responders so that they may avoid non-beneficial chemotherapy. Indeed this is probably a more useful surrogate of outcome than pCR in terms of assigning adjuvant chemotherapy treatment, even though this study confirms that pCR defines a distinct group with extremely good outcome.

In summary, we have demonstrated that with long follow-up, equal survival is achieved from neoadjuvant and adjuvant chemoendocrine therapy. Local control remains high in both groups and is not significantly different between the two arms. None of the patients who achieved a pCR relapsed systemically. Patients who achieved good clinical response had an improved DFS and OS at 10 years compared with poor responders, indicating that clinical response can be used as an early surrogate of outcome.

Received for publication June 30, 2004. Revision received September 16, 2004. Accepted for publication September 17, 2004.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Giordano SH, Hortobagyi GN. Inflammatory breast cancer: clinical progress and the main problems that must be addressed. Breast Cancer Res 2003; 5: 284–288.[CrossRef][ISI][Medline]

2. Valero V, Buzdar AU, McNeese M et al. Primary chemotherapy in the treatment of breast cancer: the University of Texas M.D. Anderson Cancer Center experience. Clin Breast Cancer 2002; 3 (Suppl 2): S63–S68.[Medline]

3. Fisher B, Gunduz N, Coyle J et al. Presence of a growth-stimulating factor in serum following primary tumor removal in mice. Cancer Res 1989; 49: 1996–2001.[Abstract]

4. Fisher B, Saffer E, Rudock C et al. Effect of local or systemic treatment prior to primary tumor removal on the production and response to a serum growth-stimulating factor in mice. Cancer Res 1989; 49: 2002–2004.[Abstract]

5. Goldie JH, Coldman AJ. A mathematic model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat Rep 1979; 63: 1727–1733.[ISI][Medline]

6. Broet P, Scholl SM, de la Rochefordiere A et al. Short and long-term effects on survival in breast cancer patients treated by primary chemotherapy: an updated analysis of a randomized trial. Breast Cancer Res Treat 1999; 58: 151–156.[CrossRef][ISI][Medline]

7. Semiglazov VF, Topuzov EE, Bavli JL et al. Primary (neoadjuvant) chemotherapy and radiotherapy compared with primary radiotherapy alone in stage IIb-IIIa breast cancer. Ann Oncol 1994; 5: 591–595.[Abstract]

8. 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 Oncol 1998; 9: 1179–1184.[Abstract]

9. Mauriac L, MacGrogan G, Avril A et al. Neoadjuvant chemotherapy for operable breast carcinoma larger than 3 cm: a unicentre randomized trial with a 124-month median follow-up. Institut Bergonie Bordeaux Groupe Sein (IBBGS). Ann Oncol 1999; 10: 47–52.[Abstract]

10. Wolmark N, Wang J, Mamounas E et al. Preoperative chemotherapy in patients with operable breast cancer: nine-year results from National Surgical Adjuvant Breast and Bowel Project B-18. J Natl Cancer Inst Monogr 2001; 96–102.

11. van der Hage JA, van de Velde CJ, Julien JP et al. Preoperative chemotherapy in primary operable breast cancer: results from the European Organization for Research and Treatment of Cancer trial 10902. J Clin Oncol 2001; 19: 4224–4237.[Abstract/Free Full Text]

12. Powles TJ, Hickish TF, Makris A et al. Randomized trial of chemoendocrine therapy started before or after surgery for treatment of primary breast cancer. J Clin Oncol 1995; 13: 547–552.[Abstract]

13. Cleator S, Parton M, Dowsett M. The biology of neoadjuvant chemotherapy for breast cancer. Endocr Relat Cancer 2002; 9: 183–195.[Abstract/Free Full Text]

14. Hayward JL, Carbone PP, Heusen JC et al. Assessment of response to therapy in advanced breast cancer. Br J Cancer 1977; 35: 292–298.[ISI][Medline]

15. Kaplan EL, Meier P. Non parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481.[ISI]

16. Scholl SM, Fourquet A, Asselain B et al. Neoadjuvant versus adjuvant chemotherapy in premenopausal patients with tumours considered too large for breast conserving surgery: preliminary results of a randomised trial: S6. Eur J Cancer 1994; 30A: 645–652.[CrossRef][ISI][Medline]

17. Fisher B, Bryant J, Wolmark N et al. Effect of preoperative chemotherapy on the outcome of women with operable breast cancer. J Clin Oncol 1998; 16: 2672–2685.[Abstract]

18. Feldman LD, Hortobagyi GN, Buzdar AU et al. Pathological assessment of response to induction chemotherapy in breast cancer. Cancer Res 1986; 46: 2578–2581.[Abstract]

19. Weigelt B, Glas AM, Wessels LF et al. Gene expression profiles of primary breast tumors maintained in distant metastases. Proc Natl Acad Sci USA 2003; 100: 15901–15905.[Abstract/Free Full Text]

20. Bonadonna G, Valagussa P. The contribution of medicine to the primary treatment of breast cancer. Cancer Res 1988; 48: 2314–2324.[ISI][Medline]

21. Pierga JY, Mouret E, Laurence V et al. Prognostic factors for survival after neoadjuvant chemotherapy in operable breast cancer: the role of clinical response. Eur J Cancer 2003; 39: 1089–1096.[CrossRef][ISI][Medline]

22. Hortobagyi GN, Ames FC, Buzdar AU et al. Management of stage III primary breast cancer with primary chemotherapy, surgery, and radiation therapy. Cancer 1988; 62: 2507–2516.[ISI][Medline]

23. Gajdos C, Tartter PI, Estabrook A et al. Relationship of clinical and pathologic response to neoadjuvant chemotherapy and outcome of locally advanced breast cancer. J Surg Oncol 2002; 80: 4–11.[CrossRef][ISI][Medline]

24. Herrada J, Iyer RB, Atkinson EN et al. Relative value of physical examination, mammography, and breast sonography in evaluating the size of the primary tumor and regional lymph node metastases in women receiving neoadjuvant chemotherapy for locally advanced breast carcinoma. Clin Cancer Res 1997; 3: 1565–1569.[Abstract]

25. Ayers M, Symmans WF, Stec J et al. Gene expression profiles predict complete pathologic response to neoadjuvant paclitaxel and fluorouracil, doxorubicin, and cyclophosphamide chemotherapy in breast cancer. J Clin Oncol 2004; 22: 2284–2293.[Abstract/Free Full Text]

26. Chang JC, Wooten EC, Tsimelzon A et al. Gene expression profiling for the prediction of therapeutic response to docetaxel in patients with breast cancer. Lancet 2003; 362: 362–369.[CrossRef][ISI][Medline]