Sentinel Node Biopsy in Breast Cancer

ACTAUNIVERSITATISUPSALIENSISUPPSALA2007

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 259

Sentinel Node Biopsy in BreastCancer

Clinical and Immunological Aspects

JANA DE BONIFACE

ISSN 1651-6206ISBN 978-91-554-6897-2urn:nbn:se:uu:diva-7890

Here in this book Some answers you’ll find

But alas! So many more questions...

List of publications

This thesis is based on the following papers, which will be referred to in the text by their Roman numerals:

I. Schüle J, Bergkvist L, Håkansson L, Gustafsson B, Håkansson A. Down-regulation of the CD3-zeta chain in sentinel node biopsies from breast cancer patients. Breast Cancer Res Treat 2002; 74(1): 33-40

II. Schüle J, Bergkvist L, Håkansson L, Gustafsson B, Håkansson A. CD28 expression in sentinel node biopsies from breast cancer patients in comparison with CD3- .chain expression. J Trans Med 2004; 2(1): 45

III. Schüle J, Frisell J, Ingvar C, Bergkvist L. Sentinel node biopsy for breast cancer larger than 3 cm. Br J Surg 2007; 94. In print.

IV. Bergkvist L, de Boniface J, Jönsson PE, Ingvar C, Lil-jegren G, Frisell J, on behalf of the Swedish Breast Cancer Group and the Swedish Society of Breast Sur-geons. Axillary recurrence rate after negative sentinel node biopsy in breast cancer: three-year follow-up of the Swedish Multicenter Cohort Study. Submitted.

Reprints were made with the permission of the publishers.

Contents

Introduction...................................................................................................11Historical notes on breast cancer and axillary surgery.............................11The sentinel lymph node biopsy...............................................................13Host immune response to malignant tumours ..........................................14CD3- : the zeta chain of the T-cell receptor ............................................16CD28: co-stimulatory receptor on T-cells................................................17The sentinel node biopsy in immunological research ..............................18

Aims of this thesis.........................................................................................20

Patients..........................................................................................................21Paper I and II ............................................................................................21Paper III....................................................................................................21Paper IV ...................................................................................................22

Methods ........................................................................................................23Identification of sentinel node..................................................................23Pathological and immunological assessment ...........................................23Follow-up .................................................................................................24Definitions................................................................................................24Statistical methods....................................................................................25Ethics........................................................................................................26

Results...........................................................................................................27Paper I ......................................................................................................27Paper II .....................................................................................................28Paper III....................................................................................................29Paper IV ...................................................................................................29

Discussion .....................................................................................................31Clinical aspects of the sentinel node biopsy.............................................31Immunological aspects of the sentinel node biopsy .................................34

Conclusions...................................................................................................37

Future perspectives .......................................................................................38

Sammanfattning på svenska..........................................................................40

Acknowledgements.......................................................................................43

References.....................................................................................................45

Abbreviations

APC Antigen-presenting cell AR Axillary recurrence CD Cluster of differentiation CI Confidence interval CTLA Cytotoxic T-lymphocyte-associated protein DC Dendritic cell FNR False negative rate HER2 Human epidermal growth factor receptor 2 ICAM Intercellular adhesion molecule IFN Interferon IL Interleukin MHC Major histocompatibility complex NF- B Nuclear factor kappa B NK Natural killer cell NSLN Non-sentinel lymph node PABAK Prevalence Adjusted Bias Adjusted Kappa SLN Sentinel lymph node SNB Sentinel node biopsy TCR T-cell receptor TGF Transforming growth factor TIL Tumour-infiltrating lymphocytes TNF Tumour necrosis factor ZAP Zeta-chain-associated protein kinase

11

Introduction

Historical notes on breast cancer and axillary surgery The oldest known description of breast cancer was discovered on the Edwin Smith Papyrus in Egypt, dating back to approximately 1600 BC (Figure 1). The treatment applied to eight breast tumours or ulcers (the term cancer is not used) was performed with a tool called “the fire drill”, a type of cauteri-sation instrument. The writing, however, states “there is no treatment”. By the second century AD, the Roman doctor Galen (129-ca.200) advised surgi-cal treatment for early, “superficial” breast cancer, which would comprise removal of all cancerous growth until only healthy tissue remained. Natu-rally, surgery in those days involved a substantial hazard to the patient’s life.

The understanding of the natural history of breast cancer rapidly grew af-ter the discovery of the lymphatic system by Jean Pecquet (1622-1674), Thomas Bartholin (1616-1680) and Olof Rudbeck the Elder (1630-1702) in the early 1650s. The evolving concept that the cancer would spread via the regional lymph nodes seemed to suggest that locally extensive surgery might be a potential cure for breast cancer. Consequently, breast cancer operations that included excision of regional lymph nodes started appearing in the 17th

and 18th centuries. The radical mastectomy, involving removal of the breast, the underlying chest muscles and the lymph nodes of the axilla, was finally described by William Stewart Halsted (1852-1922), professor of surgery at Johns Hopkins University, during the last decade of the 19th century (Figure 2).

Figure 1. Edwin Smith Papyrus Figure 2. William S. Halsted

12

Halsted’s radical mastectomy was to dominate breast cancer surgery far into the 1970s. It was claimed by some that an even more radical approach, additionally removing internal mammary and infraclavicular nodes, would further improve recurrence rates. However, the severe disfigurement created by these procedures also raised doubts as to whether such extensive surgery was in fact necessary. In 1948, Patey and Dyson1 described the modifiedradical mastectomy, omitting the removal of the chest muscles, yet still clearing the entire axilla. Unfortunately, none of the methods adopted were subjected to scientific evaluation at that time, and most of the argumentation was based on mere anecdotalism2.

It is important to remember that the role of axillary surgery at the time of its development was purely therapeutic, as adjuvant treatments were lacking. Its significant morbidity, comprising chronic pain, arm lymphedema, sensory disturbances and limitation of arm mobility3,4, was by many regarded accept-able as long as extensive surgery was believed to be the only treatment offer-ing a potential for cure. But with the arrival of increasingly potent adjuvant drugs, the diagnostic significance of axillary surgery as a staging instrument gained in importance. In the late 1960s, Bernard Fisher formulated the hy-pothesis that breast cancer might be a systemic disease, the outcome of which would not merely depend on the extent of locoregional treatment. As a result, the B-04 clinical trial of the National Surgical Adjuvant Breast and Bowel Project (NSABP) was initiated in August 1971. So far, its 6-, 10- and 25-year results have been published5-7: no survival benefit was reported for patients treated with radical mastectomy compared to those receiving total mastectomy (not including the removal of regional lymph nodes or chest muscles) with or without postoperative irradiation. Thus, it was suggested that a primary axillary dissection would not alter survival even in patients with positive axillary lymph nodes.

A parallel development in the 1970s was the introduction of population-based screening programs, resulting in a reduction of tumour size at the time of diagnosis8,9. Due to the linear relation between tumour size and axillary involvement10, the percentage of patients presenting with nodal metastases at primary operation started decreasing, too. Thus, a growing number of breast cancer patients were at risk of suffering the debilitating effects of axillary dissection without gaining any therapeutic benefit. Nevertheless, the preva-lence and number of axillary lymph node metastases was, and is still today, one of the most important prognostic factors for disease recurrence and sur-vival10-13. Thus, accurate axillary staging was deemed a necessity, but the extent of surgery required to obtain it remained to be determined. At a time when quality of life issues were being given increasing attention, this was a dilemma in need of an innovative solution.

Attempts to limit axillary surgery evolved that aimed at accurate staging of the axilla without carrying the morbidity of a full axillary node dissection. In 1975, the pectoral node biopsy was described by Cant14. Lymph nodes

13

found in the axillary tail during mastectomy without axillary clearance were hoped to represent the status of the remaining axilla, which nevertheless was shown to be inaccurate. Subsequently, other methods were developed, in-cluding several forms of axillary sampling, evaluating three15, four16 or five17

nodes from the axilla with varying reliability.

The sentinel lymph node biopsy The sought-for innovation came with the concept of the sentinel node (Fig-ures 3 and 4). It is defined as the first lymph node receiving lymphatic drain-age, and thus also metastasising cancer cells, on a direct pathway from the tumour site. The term sentinel node was already mentioned in a publication on tumours of the parotid gland in 1960 18. However, the lymph node de-scribed then was simply defined by its regular appearance at the confluence of two facial veins. But already at that time, the concept was to use it as a “marker” lymph node evaluating the presence or absence of nodal metasta-sis. It was Cabanas in 1977 who first described the sentinel node biopsy (SNB) in patients with penile carcinoma19. Preoperative lymphangiograms were performed to locate the sentinel lymph node (SLN), and further nodal dissection was discouraged if the biopsy did not show any evidence of me-tastasis. At about the same time, Donald Morton and Alistair Cochran and their team at the John Wayne Cancer Institute in California began searching for the optimal staining procedure for the lymphatic system, which led to the presentation of a vital blue dye injection technique to the Society for Surgi-cal Oncology in 1990. The news was received unenthusiastically to say the least20. The distinct scepticism was reflected in the rejection of the original manuscript describing the technique by several renowned journals, until it could finally be published in 1992 21.

Figure 3. The SNB technique Figure 4. Intraoperative identification

14

During the 1990s, numerous international studies aimed to validate the accuracy of the sentinel node technique, complementing it with a confirma-tory axillary dissection. A false negative result was defined as a negative SNB followed by an axillary dissection revealing metastases in non-sentinel lymph nodes (NSLN), and the false negative rate (FNR) as the percentage of false negative cases among all node-positive cases. Results varied consid-erably: FNRs ranged from 0% to 40%, with a median FNR of 7% 22,23. In 2001, the Philadelphia Consensus Conference recognised the SNB procedure to be “a suitable replacement for axillary dissection as a staging and diagnos-tic procedure in T1 and T2 (usually 3 cm) breast cancers”. However, it was recommended that surgeons should perform a concomitant axillary dissec-tion until they could document a detection rate of more than 95% and a FNR of less than 5% 24.

The impact on survival and regional control that the omission of further axillary surgery in sentinel node-negative patients might bring about has, as yet, not been evaluated in randomised clinical trials. A large multicentre trial (NSABP B-32) comparing the SNB as a single staging procedure with con-ventional axillary surgery is currently underway25. One earlier randomised study26 did not provide enough statistical power to compare both treatments adequately27. As a surrogate parameter, the incidence of axillary recurrences (AR) in sentinel node-negative patients has been reported in numerous trials. After a relatively short median follow-up time (range 14-57 months), axil-lary recurrences seem to be rare events. AR rates range from 0% to 3.6% internationally26,28-51. However, considering the slowly growing nature of breast cancer, long-term results will have to be awaited before determining the true failure rate of the staging procedure.

The applicability of SNB for tumours larger than 3 cm is a matter of de-bate. The initial FNR in breast cancer patients with T2 and T3 tumours re-ported from the Memorial Sloane-Kettering Cancer Center was as high as 18% 52. However, in a subsequent publication from the same institution this figure had dropped to 3% in T2 tumours53, and it was suggested that the ob-served improvement reflected the growing experience with the method. Other series confirmed the accuracy of the SNB in large tumours54-56. It was nevertheless pointed out that the higher incidence of axillary metastases in patients with larger tumours would inevitably lead to a higher absolute num-ber of false negative cases, despite an invariably low false negative rate57.

Host immune response to malignant tumours The observation of immunological reactivity against malignant tumours dates back to the beginning of the 20th century. It was not until the 1950s and 1960s that more advanced techniques emerged, allowing more refined ex-perimental testing of the processes observed. The term immunosurveillance

15

was postulated by Burnet in 197058, suggesting that it was one of the major roles of the host immune system to continuously fight the development of cancer. Even though this theory was soon rejected due to a lack of available evidence in its favour, it has undergone a renaissance in recent years59.Lately, it has been proposed that both the adaptive and the innate immune system not only participate in preventing tumour evolvement but also play a role in shaping the immunogenicity of those tumours that eventually de-velop. This hypothesis has been denominated cancer immunoediting60. Three phases of the process have been suggested, namely “elimination” of evolving neoplastic cells, “editing” of those cells that survive the first line of defence, and finally “escape”, in which the tumour by several mechanisms eludes recognition and destruction by the host immune system61. As the tumour escape pathways possibly constitute the major cause for the relatively disap-pointing clinical results that novel immunotherapies frequently yield, they represent one of the most challenging aspects of cancer immunology to-day59,62-64.

The common conception of adaptive immunity against tumours suggests that signals from transformed cells cause the attraction and activation of antigen-presenting dendritic cells (DCs)65. After antigen recognition, imma-ture DCs will undergo maturational changes while migrating to the draining lymph nodes in order to elicit an appropriate T-cell response. Activated anti-gen-specific CD8+ T-cells will differentiate into cytotoxic T-cells with the potential to destroy tumour cells (Figure 5). Possible tumour escape mecha-nisms can involve any step of this process, from decreased tumour antigen presentation to abnormal expression of co-stimulatory or adhesion molecules by DCs, alteration in signal transduction molecules, and, ultimately, the in-duction of suppressive, tolerogenic or anergic T-cells59.

Suppressed and dysfunctional immunoreactivity has been reported in sev-eral other types of cancer66, frequently correlating with the stage of the dis-ease67-69. However, breast cancer was long regarded as being relatively non-immunogenic. Historically, the finding of enhanced sinus histiocytosis in the axillary lymph nodes was suggested to be a positive prognostic factor70,71,but it could not establish itself against more powerful predictors such as axil-lary node status and primary tumour characteristics. While the occurrence of tumour-infiltrating lymphocytes (TIL) was shown to predict a better progno-sis in malignant melanoma72, ovarian73, esophageal74 and colorectal cancer75,conflicting data exist concerning breast cancer. One group reported the oc-currence of TIL to be associated with a poor prognosis76, while another found it a positive prognostic variable, at least in women aged under 40 years77. However, no classification of different lymphocyte subsets was at-tempted, nor the activation status of the cells determined, which renders any interpretation difficult. With the refinement of immunological cellular es-says, more evidence for immunosuppressive properties of breast cancer was gathered. The functional competence of peripheral T-lymphocytes was found

16

to be impaired in comparison to healthy donors78. Natural killer cells (NK cells) were observed to be increasingly dysfunctional in more advanced dis-ease stages79. On peripheral blood T-lymphocytes, down-regulation of the immunological signalling molecules CD3- (CD3-zeta), ZAP-70 and p56lck,and a dysfunctional reaction to stimulation with interleukin-2 (IL-2) and anti-CD3 was described80. Peripheral blood monocytes underexpress ICAM-1 (CD54), B7.1 (CD80) and B7.2 (CD86), and the production of tumour necrosis factor alpha (TNF- ) is decreased81.

Figure 5. Host immune response to malignant tumours (schematic drawing)

CD3- : the zeta chain of the T-cell receptor The zeta-chain (CD3- ) is a transmembrane signalling molecule of the T-cell receptor (TCR; Figure 6) with a short extracellular component and a longer cytoplasmic tail containing three ITAMs (immunoreceptor tyrosine-based activation motifs). Because CD3- mostly appears as a homodimer, two chains will be found within one TCR complex, with six phosphorylation sites or ITAMs. CD3- is essential for the transduction of T-cell activation signals provided by antigen-presenting cells (APC) such as dendritic cells. On ligation of the TCR, rapid tyrosine-phosphorylation of the ITAMs oc-

17

curs, serving as a binding site for the kinase ZAP-70, ultimately leading to the activation of transcription factors82.

CD3- is down-regulated in dysfunctional tumour-infiltrating and periph-eral T-cells in several neoplastic diseases69,83-87. This phenomenon is paral-leled by suppressed expression of ZAP-70 and p56lck 80,86, decreased produc-tion of TNF88, abnormal calcium flow through the cell membrane89 and in-creased T-cell apoptosis90. Treatment with IL-2 can upregulate the expres-sion of CD3- in T-cells91. The most emphasised down-regulation of CD3-was observed in areas of tumour regression after immunotherapy of metas-tatic malignant melanoma92. Several studies point towards a correlation be-tween the decreased expression of CD3- , tumour progress and shorter sur-vival67-69,89,93. Additionally, it may prove a relevant prognostic marker for the response to immunotherapy94.

Figure 6. T-cell receptor (TCR) before (left) and after phosphorylation (right)

CD28: co-stimulatory receptor on T-cells CD28, a member of the immunoglobulin family, is found on the surface of T-lymphocytes, while its natural ligands, B7.1 (CD50) and B7.2 (CD54), amongst others, are expressed on antigen-presenting cells95 (Figure 7). In healthy individuals, CD28 is detected on 95% of CD4+ T-cells but only on 50% of CD8+ T-cells96. The expression of CD28 increases with activation97.Its ligation constitutes an essential part of T-cell activation by providing a

18

co-stimulatory signal during antigen/MHC complex presentation. CD28’s counterpart, the receptor molecule CTLA-4, delivers an inhibitory signal, thereby limiting its effects98. CD28 ligation lowers the threshold for T-cell activation, and enables a sustained T-cell proliferation by means of an in-creased IL-2 production99. If no co-stimulatory signal is provided, the inter-action between T-cell and APC will lead to only a transient proliferative response, clonal anergy, or deletion100,101.

Figure 7: The interaction between APC and T-cell

In breast cancer, a significantly smaller percentage of peripheral T-lymphocytes are CD28+ compared to healthy controls102. In colorectal can-cer, tumour-infiltrating T-lymphocytes lack CD28 expression in contrast to T-cells in healthy colon interstitium103. In metastatic malignant melanoma, T-cells exhibiting the profoundest down-regulation of CD28 are located in areas of tumour regression after biochemotherapy104. In chronic lymphocytic and hairy cell leukaemia, CD28 expression on dysfunctional peripheral T-cells is diminished in comparison to healthy individuals105,106.

The sentinel node biopsy in immunological research The host immune response to tumours is initiated in regional lymph nodes, more precisely in the sentinel lymph node (Figure 8). Here, antigen-presenting cells arriving from the tumour site initiate clonal T-cell prolifera-tion, which predominantly takes place in the T-cell areas of the lymph node. Most studies on the immunological status of the sentinel node concern ma-lignant melanoma patients, describing reduced expression of dendritic cell

19

maturational markers and their corresponding T-cell receptors, possibly due to the increased production of several immunosuppressive cytokines (re-viewed in 107). In breast cancer sentinel lymph nodes (SLN), cell populations expressing B7.1, B7.2, CD40 and HLA-DR are decreased compared to non-sentinel lymph nodes (NSLN)108. While the latter study used flow cytometry analysis, it might be even more valuable to perform immunological analyses on tissue sections to enable differentiation between functionally distinct ar-eas within the nodal architecture. This type of research has long been em-braced by Alistair Cochran and colleagues109-112. In breast cancer, the para-cortical area and the density of paracortical dendritic cells in SLNs is de-creased in comparison to NSLNs, with a predominance of immature den-dritic cells113. Thus, immunological functions seem to be more disturbed in SLNs than in NSLNs, a distance-related impact that might suggest the exis-tence of tumour-derived immunosuppressive signals reaching the SLN via the lymphatic system.

Figure 8. Schematic drawing of the lymph node structure

20

Aims of this thesis

To study and describe the expression of the zeta chain (CD3- ) of the T-cell receptor in different areas of the sentinel lymph node from breast cancer patients

To study and describe the expression of the co-stimulatory mole-cule CD28 on CD8+ and CD4+ T-cells in the sentinel lymph node and correlate its distribution with that of the zeta chain (CD3- )

To determine the false negative rate and detection rate of the senti-nel node procedure for breast cancer larger than 3 cm

To report the current axillary recurrence rate and disease-free sur-vival in sentinel node-negative patients in whom no axillary clear-ance was performed

21

Patients

Paper I and II Twenty-five female patients with palpable, unifocal, invasive breast cancer less than 3 cm in diameter underwent sentinel node biopsy at the Central Hospital in Västerås, Sweden, as part of their surgical treatment. Neoadju-vant therapy, ongoing pregnancy, palpable axillary lymph nodes suggestive of nodal metastases and known allergy to blue dye or isotope were exclusion criteria. In some sentinel node sections from two patients, tumour growth was too abundant to allow evaluation of the remaining lymphatic tissue. In paper I, one of these patients had to be excluded from analysis, and in paper II, the specimens of both patients were affected. In another two patients, the technical quality of sections stained for CD28 expression was too poor to permit immunological assessment. Thus, nodal sections from 24 and 21 pa-tients (paper I and paper II, respectively) could be immunologically ana-lysed.

Paper III Patients with invasive breast cancer larger than 3 cm were identified from the national sentinel node database in Sweden. Only patients who were planned for a sentinel node biopsy followed by an axillary dissection of lev-els I and II, regardless of the outcome of the sentinel node biopsy, were eli-gible. Apart from the difference in tumour size, exclusion criteria for papers I and II applied. Overall, we could identify 109 patients from the database. Twenty-three cases derived from the initial SNB learning phase and 68 from the SNB validation phase. After introduction of the SNB as a single staging procedure, patients with tumours larger than 3 cm were not routinely oper-ated with SNB in Sweden. Therefore, an additional 18 patients with such tumours were recruited between 2004 and 2006. Operations were performed by 28 surgeons of all levels of experience at 20 Swedish hospitals.

22

Paper IV Between September 2000 and January 2004, 3534 patients with invasive, clinically node-negative breast cancer 3 cm were prospectively enrolled in a multicentre study for sentinel node biopsy as a single staging procedure. Twenty-six Swedish hospitals and 131 surgeons contributed to patient ac-crual. Patients with ongoing pregnancy, known allergy to the blue dye or isotope used during the SNB procedure, previous breast surgery or preopera-tively diagnosed multifocal disease, and those requiring neoadjuvant treat-ment, were not eligible. As the aim of the analysis was to evaluate the out-come of the SNB as the only axillary staging procedure, only sentinel node-negative patients who did not undergo a completion axillary dissection were included. Furthermore, patients in whom the postoperative pathology exami-nation found a tumour size larger than 3 cm, benign or in situ disease only, or who were followed up outside Sweden, were excluded (Figure 9). Thus, the final analysis comprised 2246 patients.

Figure 9. Flow-chart for selection of patients for paper IV.

Initially considered for inclusion N = 3534

Benign (3) or in situ (68) disease only, tumours > 30 mm (127)

N = 198

Sentinel node not identified

N = 101

Sentinel node biopsy positive

N = 906

Miscellaneous reasons for exclusion, e.g. patient followed up outside Sweden

N = 20

Included in the present analysis N = 2246

Axillary dissection performed due to multifocality, palpable nodes, isotope activity, patient preference, non-sentinel

metastatic lymph nodes; N = 63

23

Methods

Identification of sentinel node Radioactive isotope (40-60 MBq Technetium-99 nanocolloid, Solco Nano-coll®; Nycomed, Amersham, UK) was injected peritumourally, sub- or in-tracutaneously, followed by gentle massage of the area. Injection techniques could vary between participating institutions. For non-palpable tumours, mammographic guidance was used. Preoperative lymphoscintigraphic im-ages were most frequently obtained 5 and 45-60 minutes after injection, and repeated after 2-3 hours if no sentinel node had been identified before. In a similar fashion, 1 ml blue dye (Patent Blue V®; Guerbet, Paris, France) was injected 5 to 15 minutes before incision, and the area gently massaged.

During surgery, sentinel lymph nodes were identified by a handheld gamma probe and/or by their uptake of blue dye. Hot and/or blue nodes were defined as sentinel nodes and sent fresh to pathology. In the event that no sentinel node could be detected, a routine axillary dissection of anatomical levels I and II was performed. Part of the sentinel node procedure was also to palpate the axilla thoroughly in order not to miss lymph nodes suspicious of metastasis that nevertheless had accumulated neither isotope nor colour. Only in paper IV was axillary dissection omitted in case the sentinel node biopsy was found to be negative. In the remaining studies, axillary clearance was performed regardless of the result of the SNB.

Pathological and immunological assessment Frozen sections were obtained from the sentinel nodes and examined during operation. Nodes larger than 4 mm in diameter were bisected and at least one or two sections obtained from each part. For definitive pathology, at least three sections were taken from each sentinel node or each part of the bi-sected node. Staining was performed with haematoxylin and eosin (HE), and, if no metastatic cells could be identified, with additional immunohisto-chemical staining (IHC) using cytokeratin antibodies. Non-sentinel lymph nodes were only evaluated on routine HE sections.

For papers I and II, additional snap-frozen sections were obtained and sent to Linköping University for immunological double-staining. After fixa-tion and blocking against unspecific binding, the first staining was per-

24

formed with primary antibodies against CD3 (papers I and II), CD4 and CD8 (only paper II). DAB (3,3’-diaminobenzidine, D-5637, Sigma, Stock-holm, Sweden) was used as a substrate to obtain a brown colour. Negative control slides were manufactured using mouse IgG (Sigma, Stockholm, Sweden). To detect the expression of CD3- and CD28, tissue sections were then incubated with mouse monoclonal antibodies to these substances. The technique used resulted in a bright red staining for CD3- and CD28. Thus, the higher the expression of these molecules was, the more would the red colour dominate, while areas of low expression would appear brown.

The expression of CD3- and CD28 was differentially evaluated in three areas of the sentinel node: the primary follicles and germinal centres (B-cell activation areas) and the paracortex (T-cell activation area). For paper II, the expression of CD28 in CD4+ and CD8+ T-cell subsets was compared. Whole sections were always analysed. Slides were assessed in a semiquanti-tative manner by two investigators (A.H., B.G). The percentage of positive cells in each area was estimated, and the degree of expression of CD28 and CD3- determined high (more than 75% positive), moderate (50-75% posi-tive) or low (less than 50% positive).

Follow-up In paper IV, follow-up after primary operation was scheduled annually. Even though both clinical examination and mammography had been postulated in the study protocol, the method of follow-up could vary between participating centres. A few institutions reported telephone interviews combined with self-examination or mammography without any clinical examination. All follow-up data were reported to the study database at the Centre for Clinical Re-search of Uppsala University at the Central Hospital in Västerås. Before the present data analysis was conducted, a complete list with all included pa-tients per institution was sent out to each centre for an updated report of un-favourable events and latest follow-up dates. Additionally, one of the authors (J.d.B.) was granted access to hospital files at on-site visits for data valida-tion.

DefinitionsThe false negative rate is the percentage of sentinel node-negative patients among all axillary lymph node-positive patients. The detection rate is de-fined as the percentage of procedures in which a sentinel node could be suc-cessfully retrieved among all attempted sentinel node procedures.

In paper IV, any recurrences diagnosed within the same 2-month interval were regarded as concurrent to each other. Axillary recurrences were defined

25

as isolated if the axilla was the sole initial site of relapse. The definition of a locoregional axillary relapse comprises the above, and additionally, those patients in whom a local relapse in the ipsilateral breast was found prior or concurrently to the axillary relapse. The overall axillary relapse group addi-tionally included those patients in whom systemic and/or extra-axillary lym-phatic metastases were detected prior to, or concurrently to, the axillary re-currence. A local recurrence was defined as a new cancer in the breast on the previously operated side.

Statistical methods In papers I and II, the Mann-Whitney U test was used for analysis of corre-lations between tumour and patient characteristics and the expression of CD28 and CD3- . For comparison of the expression of the same marker in different areas of the same sentinel lymph node, the Kruskal-Wallis test was applied. The same is true for comparison of the CD28 expression in the CD4+ and CD8+ T-cell subgroups.

The expression of CD28 and CD3- within the same area of the lymph node was compared using the Prevalence-adjusted Bias-adjusted Kappa (PA-BAK). It is interpreted like Cohen’s Kappa114. Values indicate excellent (> 0.80), good (0.61-0.80), moderate (0.41-0.60), fair (0.21-0.40) and poor agreement (0.00-0.20). A value below zero suggests an agreement that is less than expected by chance.

In papers III and IV, the impact of individual tumour and patient charac-teristics on detection and false negative rates (paper III) and on recurrence and survival rates (paper IV) was analysed performing univariate logistic regression. To study each significant factor’s relative impact on those pa-rameters, multivariate logistic regression was used. For survival analysis, the Kaplan-Meier and Cox proportional multiple regression models with 95% confidence intervals (CI) were applied. For overall survival, all causes of death were included, whereas in the calculation of breast cancer specific survival, death from breast cancer was the endpoint, and dates of death from other causes or last follow-up were censoring dates. For disease-free sur-vival, the time to any first recurrence was calculated.

For paper IV, the null hypothesis was that the 5-year axillary recurrence rate would be 3% (Ho: p 0.03). This figure was based on the published 5-year axillary recurrence rate of 1% after negative axillary dissection in Swe-den115. Taking into account the benefits of decreased morbidity that the sen-tinel node method offers in comparison to conventional axillary clearance, a 5-year axillary recurrence rate of less than 3% in sentinel node-negative patients was regarded as acceptable. It was calculated that the inclusion of at least 2000 patients was necessary to reject Ho with a power of 87%, if a true

26

axillary recurrence rate of 2% at 5 years was assumed. With true axillary recurrence rates falling below 2%, statistical power would quickly increase.

For all analyses, a P value less than 5% was regarded significant. The SPSS® (SPSS Inc., Chicago, Illinois) program was used throughout this thesis.

EthicsThe study designs were approved by the ethics committees at the University of Uppsala and the University Hospital of Linköping (papers I and II), at the Karolinska Institutet, Stockholm (paper IV), and the local ethics committees of each participating region in papers III and IV.

27

Results

Paper I Sections of 24 sentinel node biopsies were available for analysis. Expression varied considerably comparing different patients’ tissue samples but also comparing different functional areas within the same sentinel node (Figure 10).

In the B-cell follicles, where B-cell activation takes place and germinal centres appear once antigen confrontation leads to clonal proliferation, the degree of down-regulation was generally low. Only in 18 patients, sufficient numbers of germinal centres were found to allow immunological assess-ment. In the majority of these cases, CD3- expression was notably sup-pressed. In the paracortex, the main T-cell activation area of the lymph node, the strongest down-regulation was found (Table 1). No correlations to tu-mour histopathological parameters or clinical data were identified. Statistical analysis of subgroups was not deemed meaningful due to small sample size.

Figure 10. CD3- expression in germinal centers (upper row) and paracortex (lower row) of the sentinel lymph node. Low (left side), moderate (upper right) and high (lower right) degree of CD3- down-regulation.

28

Table 1. Down-regulation of CD3- in different areas of the sentinel lymph node from 24 breast cancer patients. Numbers signify cases (percent).

Degree ofdown-regulation

B-cell follicles

(primary follicles)

Germinal centers

(secondary follicles)

T-cell activation area(paracortex)

low 14 (58) 3 (17) 1 (4)

moderate 4 (17) 2 (11) 4 (17)

high 6 (25) 13 (72) 19 (79)

Paper II Sections of 21 sentinel node biopsies could be analysed for the occurrence and distribution of CD28+ T-cells within the nodal architecture. The subset of CD8+ T-cells generally showed a markedly higher CD28 expression than CD4+ T-cells. This difference was especially emphasised in the paracortex and the germinal centres, where CD28 expression on CD4+, but not on CD8+ T-cells, was strongly down-regulated (P < 0.001). This difference was not notable in the B-cell areas, the primary follicles, where both T-cell sub-sets showed the least degree of CD28 down-regulation. In comparison with CD3- expression in the same areas of the same biopsies, the agreement between both markers was highest in the paracortex, reaching a PABAK value of 0.62 (Table 2). However, this was only true for CD4+ T-cells. It was confirmed that those CD3+ T-cells expressing low levels of CD3- in paper I represented mostly CD4+ T-cells.

Table 2. The paracortex of the sentinel lymph node: degree of CD28 expression on CD4+ T-cells in comparison with CD3- expression (depicted are 20 samples in which both staining modalities were analysable)

Expression of CD28 on CD4+ Expression of CD3- High Moderate Low

High 0 0 1

Moderate 0 2 1

Low 0 3 13

29

As in paper I, no correlations to tumour histopathological parameters or clinical data were found. Statistical analysis of subgroups was not performed considering the small sample size.

Paper III The tumour size of 109 breast cancer patients identified from the sentinel node database ranged from 3.1 to 8.0 cm (median 4 cm), including 94 T2 (86%) and 15 T3 tumours (14%). The sentinel node could be identified in all but 6 cases, resulting in a detection rate of 94.5%. Axillary lymph node me-tastases were found in 67 patients, of whom the sentinel node had not been identified in 3 cases. Eight of the remaining 64 patients had a negative senti-nel node biopsy, resulting in a false negative rate of 12.5% (8/64). Most of these cases occurred during the earlier years of the inclusion period, but pos-sibly due to the small numbers in the subgroups, this observation was not statistically significant. The false negative rate did not differ between indi-vidual surgeons or contributing institutions.

On postoperative pathological examination, multifocal tumour growth was detected in 16 cases. These had not been diagnosed by preoperative investigations. It was noted that the false negative rate among patients with multifocal tumours (30.8% (4/13)) was significantly higher than in the re-maining patients with unifocal disease (7.8% (4/51); P = 0.012). At the same time, the prevalence of nodal metastases appeared to be increased (81% (13/16) versus 58% (54/93); not significant). On multiple regression analy-sis, multifocality was the only significant prognostic factor for a false nega-tive biopsy (P = 0.026).

Paper IV After a median follow-up time of 37 months (range 0-75), the axilla was the sole initial site of relapse in 13 cases (13/2246; 0.6%). In 7 cases, axillary recurrence was combined with local recurrence in the ipsilateral breast, and in a further 7 cases, it occurred together with a generalisation of the disease. Thus, 27 cases of clinically detectable axillary recurrence were identified (27/2246; 1.2%). The median time from operation to axillary relapse was 21 months (range 4-51). The estimated cumulative incidence of isolated axillary recurrence at 5 years was 1.1% (95% CI 0.2-1.9; see Figure 11). On multiple regression analysis, the only significant risk factor for axillary relapse was Elston histological grade (P = 0.006). The results of the contributing institu-tions and individual surgeons did not differ significantly.

Overall, 117 recurrences in 91 patients were reported. Seven extra-axillary nodal recurrences and 55 cases of distant relapse were found. After

30

breast-conserving treatment, 27 local recurrences were diagnosed, and one patient developed a chest wall recurrence after mastectomy.

Cancer-related death occurred in 38 cases. Even here, the only significant prognostic factor on Cox proportional regression analysis was Elston histo-logical grading (P = 0.003). Overall 5-year survival was 91.6%, disease-free 5-year survival 92.1% and breast cancer-specific 5-year survival 94.7%.

Figure 11. Cumulative isolated axillary recurrence-free survival after negative senti-nel node biopsy with 95% confidence interval.

0 20 40 60

0.980

0.985

0.990

0.995

1.000

Axi

llary

recu

rren

ce-f

ree

surv

ival

Numbers of patients at risk 2161 842 110 Time (months)

31

Discussion

Clinical aspects of the sentinel node biopsy It has been a long journey from the first anecdotal, hardly auspicious at-tempts at breast cancer treatment via Halsted’s era of extensive and debilitat-ing surgery to our time, embracing the principle of best possible cure while respecting the patient’s quality of life, functionality and cosmesis. The im-portance of the patient’s perspective on treatment modalities has steadily grown since the end of World War II, a trend that has not only led to the development of breast-conserving treatment but also to a general trend to-wards less invasive surgery. Another seminal step forward during the last century was that from “non-science to science”2: the insight that scientific hypotheses could be tested by means of randomised clinical trials, resulting in the principles of evidence-based medicine. Potent adjuvant drugs were found to increase survival in both node-negative and node-positive breast cancer patients116, further eroding the surgical monopoly. The development of the sentinel node biopsy concept is one of the most enthralling aspects of these historical paradigm shifts in surgical science.

The application of the sentinel node method was adopted at such a speed that it has all but replaced conventional axillary clearance in early breast cancer today. After a period of early validation studies, determining the rate of false negative results by performing a subsequent confirmatory axillary clearance, the sentinel node method has been recommended as a single axil-lary staging method in biopsy-negative patients, provided that the false nega-tive rate would not exceed 5% 24. Yet once axillary clearance is omitted, only the occurrence of clinically detectable axillary relapse will give us a clue about potentially false negative procedures, and will thus be a surrogate quality indicator for the axillary surgery performed. It is important neverthe-less to stress that the axillary recurrence rate is not equivalent to the false negative rate, as has been suggested in some publications, because this would underestimate the true figure significantly117. Additionally, not every nodal metastasis left behind in the axilla will necessarily develop into clini-cally overt disease5,118-120. The axillary recurrence rate after a negative senti-nel node biopsy should therefore be compared to the same rate after conven-tional axillary clearance in node-negative patients. The first sufficiently large randomised trial of that kind is underway and will hopefully contribute to further clarification on the subject25.

32

In the meantime, axillary recurrence rates (range 0-3.6%) following a negative sentinel node biopsy seem promisingly low. However, most current publications derive from expert multidisciplinary teams in specialised high-volume centres, and only one multicentre study including five surgical de-partments and 479 patients could be identified51. At the same time, an own literature research of early validation studies revealed a considerable differ-ence between false negative rates from single institutions (6.9%, range 1-13.7%; 14 studies) and multicentre trials (10.9%, range 7.2-22%; 11 studies). It is therefore important to study the axillary recurrence rate in a setting where even small community hospitals with only a few cases per year are represented, so that results may truly reflect the heterogeneous situation that prevails in most countries. In paper IV, six of 25 hospitals accrued less than 15 cases to the overall study, and 72 of 131 surgeons performed less than 10 procedures each. Nevertheless, the axillary recurrence rate observed lies well within the range of the above stated international publications.

There is reason to believe that not every axillary relapse after a negative sentinel node biopsy signifies failure of the axillary staging method. When a local recurrence in the operated breast is previously or concurrently found, the axillary recurrence is likely to be a consequence of the new breast cancer rather than of inadequate axillary surgery. It can also be disputed whether an axillary relapse coinciding with distant disease spread should be regarded as a surgical treatment failure or as a result of particularly aggressive primary tumour properties. Thus, isolated axillary recurrences appear to be the only reliable quality indicator of the sentinel node biopsy. In paper IV, axillary relapse was therefore analysed in the categories isolated, locoregional, and overall. The cumulative 5-year incidence of locoregional axillary relapse of 1% after negative axillary clearance in an unselected Swedish population115

is included in the 95% confidence interval (0.7-2.6) of the corresponding figure of 1.6% after negative sentinel node biopsy.

Even before the sentinel node era, axillary recurrences were a relatively rare event, occurring in approximately 1-3% of breast cancer patients 5 years after adequate axillary node dissection115,121,122. Accordingly, there are few studies on the treatment and outcome of axillary relapse. The subsequent distant failure rate of patients with axillary recurrence was 50% in a series reported by Newman et al.123. In a Dutch study, 44 of 59 patients who had been diagnosed with an axillary recurrence subsequently died, having devel-oped distant metastases in 36 cases124. Survival rates after diagnosis of re-gional lymph node recurrence were 72% and 43% at 5 and 10 years, respec-tively, compared to 91% and 81% for patients without recurrence125. In Brit-ish Columbia, 220 breast cancer patients with isolated axillary recurrences had an overall 5-year survival of 49.3% and a median survival time of 4.9 years calculated from the date of relapse126. On the opposite scale lies the significant reduction of postoperative morbidity following the sentinel node procedure compared to conventional axillary clearance127-129.

33

It is important to identify those patients who might be put at risk of incor-rect staging, and thereby inadequate adjuvant treatment, by the omission of a conventional axillary clearance. To this end, it is necessary to determine risk factors for a false negative sentinel node biopsy, but also predictors for the prevalence of axillary metastases: a higher risk of axillary metastases will yield a higher absolute number of false negative cases in spite of an un-changed false negative rate. Factors suggested to increase the incidence of false negative biopsies include histological grade of the tumour130, tumour size52,131, tumour multifocality132,133, inexperience with the method52,53,131,upper outer quadrant tumour location131, and a low number of retrieved sen-tinel nodes130,131. Several of the mentioned tumour characteristics constitute predictors for axillary node metastases as well. Thus, the potential benefit from the sentinel node procedure might decrease with an accumulation of such risk factors.

In the case of breast cancer larger than 3 cm, paper III confirms findings from earlier publications53-56, showing a false negative rate and detection rate comparable to those of early validation studies limited to smaller tumours. However, patients with postoperatively diagnosed tumour multifocality re-vealed a higher false negative rate than the remaining patients with unifocal disease. Even though the study had not been designed for such a subgroup analysis, these findings warrant further investigation. There is some contro-versy about the validity of the sentinel node biopsy in multifocality. While some studies found multifocality to be a risk factor for a false negative bi-opsy132,133, other groups have reported the contrary134-136. Multifocality and multicentricity are still widely regarded as contraindications for the sentinel node procedure137, a recommendation that might have to be further evalu-ated.

One problem when dealing with tumour multifocality is the wide variety or absence of its definition in publications on the subject. Additionally, the common practice of putting the diameter of the largest focus on a par with overall tumour size might underestimate the real situation considerably138.However, even the use of tumour surface area or tumour volume may not fully account for the increased prevalence of axillary metastases in multifo-cal cancer, and altered biological tumour properties have been suggested as a possible explanation139. Especially in larger tumours, the effect of multifo-cality on the false negative rate is little studied, due to the exclusion of larger tumours from most early validation studies. Therefore, it seems appropriate to consider a completion axillary clearance in those cases where large tu-mour size and multifocality coincide, until stronger evidence is available.

34

Immunological aspects of the sentinel node biopsy The findings described in papers I and II confirm earlier observations of decreased expression of T-cell co-stimulatory and signalling molecules in patients with malignancies. The most profound down-regulation of CD3-and CD28 was found in germinal centres and in the main T-cell activation area, the paracortex. In the latter, the agreement between both parameters was strongest.

The first encounter between antigen-presenting dendritic cells (DCs) and naïve T-cells is assumed to take place in the paracortex of lymph nodes re-gional to neoplasms. This is supported by the observation that antigen-specific T-cells first accumulate in the paracortex after subcutaneous antigen injection, thereafter moving on to lymph node follicles140. Most studies in-vestigating the immunology of lymph nodes regional to malignant melanoma and breast cancer derive from Cochran et al.109, who described the strongest evidence of immunological perturbation in the paracortical area of sentinel lymph nodes. In breast cancer patients, the relative portion of the lymph node occupied by paracortex is significantly smaller in sentinel (SLN) com-pared to non-sentinel lymph nodes (NSLN), and the number, density and maturity of DCs in these areas is decreased113. These findings suggest that the immunology of regional lymph nodes is influenced by their position relative to the primary tumour. Interestingly, those lymph nodes located closest and those furthest away from the tumour show less immunological reactivity than “intermediate” nodes111. It was hypothesised that the latter nodes were too far from the tumour to be immunosuppressed, but close enough to be exposed to immunostimulatory tumour antigens109.

The apparently distance-related impact on host immunoreactivity both within the nodal architecture and the regional network of lymph nodes sug-gests the existence of tumour-derived signals modulating the immune re-sponse. Findings in patients with colorectal cancer support this theory67. In ovarian cancer, a circulating 16kD protein isolated from ascites fluid was found responsible for the nearly complete selective suppression of CD3- 141.Otsuji et al.142 found that tumour-derived macrophages in mice exerted oxi-dative stress on T-cells, resulting in suppression of CD3- . This effect could be imitated by oxidative agents like hydrogen peroxide and diamide, pre-vented by treatment with the antioxidant N-acetylcysteine, and reversed by macrophage deletion. A second murine study reported that sustained cell-contact with activated macrophages was necessary to induce CD3- suppres-sion143. Further possible pathways of host immunosuppression include the presence of IL-4 144, TGF- 145 and soluble HLA class I molecules146.

It has been suggested that the CD4+ and CD8+ T-cell subsets react differ-ently to co-stimulatory signals via the CD28/B7 pathway. The T-cell re-sponse to B7.1-dependent co-stimulation is prolonged in the CD4+ subset, featuring IL-2 secretion and clonal expansion. In the CD8+ subset, however,

35

the helper-independent response is short-lived. Additionally, CD8+ T-cells require higher levels of B7.1 for an activation equivalent to CD4+ T-cells147.Thus, the degree of expression of B7.1 on the antigen-presenting cell might influence the development of helper or cytotoxic T-cell responses. In papersI and II, the degree of CD28 down-regulation was far more pronounced in CD4+ than in CD8+ T-cells. Similar observations were reported from metas-tases to malignant melanoma after biochemotherapy: both CD4+ and CD8+ T-cells showed a marked loss of CD28 in areas of tumour regression while CD28 expression was hardly decreased close to unaffected tumour growth, and the down-regulation of CD28 was generally more emphasised in CD4+ than in CD8+ T-cells104. This appears contrary to the differential expression of CD28 on peripheral blood T-cells in healthy individuals. In papers I and II, the vast majority of T-cells in the area of strongest immunological pertur-bation, the paracortex, were CD4+, and thus, the loss of both CD28 and CD3- might mainly concern the CD4+ subset. Progressive loss of CD28 has been described in CD4+ and CD8+ T-cells in chronic inflammation and rep-licative senescence148,149. It has also been hypothesised that this situation might be similar in cancer patients, where long-term exposure to tumour antigens leads to chronic stimulation – which may lead our thoughts back to the theory of cancer immunoediting and its three phases described in the introduction of this thesis.

There is evidence for close interactions between CD28 and CD3- . The tyrosine phosphorylation of CD3- and ZAP-70, as well as their association, is one of the earliest signalling events after TCR engagement150, and is strongly inhibited in the absence of a co-stimulatory CD28/B7-mediated signal151. ZAP-70 activation, on the other hand, might lead to phosphoryla-tion of the cytoplasmic CD28pYMXM motif which would in turn allow the binding of phosphatidylinositol 3-kinase, thereby augmenting the second signal provided by CD28/B7 interaction152. The apparent link between CD28 and CD3- functions seems to be reflected in the findings of papers I and II,where the agreement between both molecules was strongest in areas of the most marked down-regulation.

Although it is undisputed that CD28 engagement delivers potent positive co-stimulatory signals for T-cell activation, it has been shown that this sec-ond signal is not always obligatory. Interestingly, especially in the context of the differences of CD28 expression in CD4+ and CD8+ T-cells, the latter subset was reported to be less dependent on CD28 co-stimulation (reviewed in 153). Furthermore, the requirement of CD28 signalling appears to be de-pendent on the strength and the duration of antigen presentation. Another striking feature of CD28 co-stimulation is the CD28-dependent up-regulation of the inhibitory molecule CTLA-4 (CD152) on the cell surface, which binds the same ligands as CD28 albeit with a much higher affinity, and antagonises most major CD28-mediated events154. While the absence of a co-stimulatory signal from CD28/B7 may lead to T-cell anergy, clonal

36

deletion and the differentiation of T-cells into a Th2 phenotype, the lack of a negative signal from CTLA-4/B7 results in unregulated T-cell expansion153.Hence, the CTLA-4/B7 pathway has emerged as a pivotal regulator mecha-nism in the induction and maintenance of peripheral tolerance in autoim-mune disease (reviewed in 155).

Truly successful treatment against cancer implies the deletion of the very last remaining malignant cell and necessitates a functioning host immune response. However, cytotoxic drugs are unlikely to enhance natural defence mechanisms. There is an abundance of ongoing research evaluating different approaches to immunological treatment regimens in breast cancer (reviewed in 156-159). One already established successful agent is the humanised mono-clonal antibody trastuzumab (Herceptin®) for breast cancer overexpressing HER2. However, there remains a large proportion of patients who will not respond to immunological therapies. Based on the heterogeneous immu-nological patterns observed in papers I and II and other reports, one might hypothesise the existence of distinct tumour types differing in the immune responses they evoke. While a high expression of CD28 or CD3- might reflect a non-immunogenic tumour, not recognised by the host immune sys-tem, or else an immunogenic, but non-immunosuppressive tumour, a signifi-cant down-regulation of immunoreactivity might be due to immunosuppres-sive tumour characteristics. As metastases were found both in strongly and in less down-regulated lymph nodes, this could be explained by either non-immunogenic tumours surpassing the immune system without being recog-nised, or immunosuppressive tumours “overriding” the immunological de-fence. Thus, the differential information available from the immunological analysis of the sentinel lymph node might play an important role for the se-lection of patients likely to benefit from immunological treatment ap-proaches.

37

Conclusions

In this thesis, clinical and immunological aspects of the sentinel node biopsy (SNB) were studied.

In breast cancer larger than 3 cm, the SNB technique was shown to repre-sent the status of the axillary lymph nodes with a false negative rate (FNR) and detection rate that are comparable to early validation studies including smaller tumours. However, when large tumour size coincided with tumour multifocality, the FNR was increased. The SNB can thus be safely per-formed in unifocal breast cancer larger than 3 cm. If tumour multifocality is diagnosed in these patients, however, a completion axillary dissection should be considered until further data on SNB in multifocal disease are available.

The omission of a subsequent axillary dissection in sentinel node-negative breast cancer patients yields overall axillary recurrence rates ranging from 0% to 3.6% internationally. In the Swedish multicentre cohort trial, the iso-lated axillary recurrence rate was 0.6% (13/2246) after a median follow-up of 37 months. These results encourage the use of the SNB as a single staging procedure even in a heterogeneous multicentre setting. However, long-term data will have to be awaited before the definitive failure rate of the SNB procedure can be determined.

The sentinel lymph node (SLN) is not only the first node on the lymph drainage pathway to harbour metastatic cancer cells, but also the showplace of immunological changes reflecting host immunoreactivity against the tu-mour. The expression of the chain of the T-cell receptor and of the co-stimulatory molecule CD28 were found to be down-regulated to a varying extent. The most pronounced loss of both T-cell parameters, and the strong-est agreement between their respective degrees of down-regulation, were observed in the main T-cell activation area, the paracortex. These findings might have prognostic implications and aid the selection of patients who are likely to benefit from immunological treatment approaches.

38

Future perspectives

Even though seminal advances in breast cancer research have been made during the last century, in terms of understanding its nature, developing less invasive albeit no less accurate axillary staging methods, broadening the treatment repertoire and improving survival, a substantial number of patients will still die of their disease, despite all therapeutic efforts. In metastatic disease, surgery is hardly likely to provide survival benefits. However, as cytotoxic drugs act according to the first order of kinetics, they will not be able to extinguish the very last cancer cell and thus cure the patient. There-fore, immunological treatment approaches are sought for, and one of the first successful results of this search is the monoclonal antibody trastuzumab for HER2 positive breast cancer. Further drug development is underway. How-ever, there are still a substantial number of patients who will not respond to immunotherapy. To understand the mechanisms by which a tumour can es-cape the host immune system, and thus also evade the effects of immuno-therapy, is one of the greatest challenges in cancer immunology today.

The first two papers of this thesis aimed to add to the understanding of the regional immunodeficiency frequently observed in cancer. Further research should involve the opposite side of the immunological synapse, namely the antigen-presenting dendritic cells. While the morphology and distribution of those cells is well documented in the sentinel lymph nodes from breast can-cer patients, their functional properties are not yet sufficiently understood. The analysis of activation parameters and co-stimulatory receptors will help to shed light on the intricate interactions taking place in the regional lymph nodes. Furthermore, understanding the immunological impact of the lymph nodes’ position relative to the primary tumour will contribute with new in-formation about the mechanisms of cancer-related immunodeficiency.

The adoption of the sentinel node concept has clearly decreased postop-erative morbidity and improved the quality of life for node-negative breast cancer patients. However, numerous questions remain to be addressed in clinical trials. The third and fourth paper of this thesis aimed to make a con-tribution to the safety issues regarding the staging method. While we could show that its use appears safe in tumours larger than 3 cm, we also found an increased false negative rate in large multifocal tumours. As this trial had not been designed to address the question of the sentinel node biopsy’s applica-bility in multifocal disease, further studies are needed to clarify this issue. Concerning the use of the sentinel node biopsy as a single staging procedure

39

in early breast cancer, the results presented in this thesis are encouraging. However, breast cancer is a slowly growing disease, and the median follow-up time today is too short to determine the true failure rate of the method. Therefore, long-term follow-up data will be analysed with regard to the de-velopment of axillary recurrences after a negative biopsy, and their impact on breast cancer survival.

It seems interesting to speculate about another potential effect of the es-tablishment of the sentinel node biopsy procedure. Before its development, only one or two sections of each retrieved axillary lymph node would be evaluated by the pathologist. This routine pathology has now been comple-mented by the analysis of multiple or even serial sections of the sentinel lymph node(s). As the sentinel node is the most likely location of nodal me-tastasis, this may lead to an improved staging accuracy. It is thus possible that breast cancer survival will improve due to stage migration and the im-proved recognition of sentinel node-positive patients who would have been node-negative by traditional axillary lymph node dissection.

40

Sammanfattning på svenska

Bakgrund och syfte Bröstcancer är den vanligaste cancerformen hos kvinnor världen över. Före-komst och antal av lymfkörtelmetastaser i armhålan (axillen) är den starkaste prognostiska faktorn för recidivrisk och överlevnad. Därför grundas plane-ringen av adjuvanta behandlingar till en stor del just på axillstatus.

En konventionell axillutrymning är det traditionella och mest exakta sättet att få information om axillstatus. Nackdelen med denna metod är en bety-dande morbiditet i form av smärta, nedsatt känsel och rörelseförmåga samt kroniskt armlymfödem. Ett skonsammare sätt att få kunskap om tillståndet i axillen är sentinel node biopsin. Tekniken går ut på att identifiera den första lymfkörtel dit lymfan når på väg från en tumör. Den hittas med hjälp av en radioaktiv substans samt blå färg som sprutas nära brösttumören inför opera-tionen. I bröstcancersammanhang har denna metod introducerats sedan 90-talet. Dess roll som alternativ staging metod har utvärderats i ett stort antal valideringsstudier, där man utförde en kompletterande axillutrymning efter varje sentinel node biopsi. De låga falskt negativa utfallen i dessa studier har lett till att sentinel node tekniken nästan helt har ersatt axillutrymning vid en negativ körtelbiopsi.

Avhandlingen syftar till att utvärdera sentinel node tekniken från ett kli-niskt och ett immunologiskt perspektiv. Kliniskt belyses dess användning för tumörer större än 3 cm (delarbete 3) samt dess säkerhet som enda staging metod i axillen vid negativ biopsi (delarbete 4). ”Portvaktskörtelns” läge betyder också att den är skådeplats för den första kontakten mellan antigen-presenterande celler och naiva T-celler, och erbjuder därmed en unik möjlig-het att studera det immunologiska svaret mot tumören (delarbeten 1 och 2).

Delarbete 1. T-cell receptorns viktiga funktionella beståndsdel -kedja (CD3- ) har visats vara nedreglerad vid olika cancersjukdomar och korrelerar med cancerstadi-um. Dess förlust leder till försämrad T-cell funktion. CD3- undersöktes i sentinel node biopsier från 24 bröstcancerpatienter med hjälp av immunhis-tokemiska dubbelfärgningar. Graden av CD3- nedregleringen varierade mellan olika patienter men också mellan de olika delar av körteln som un-

41

dersöktes: primära folliklar, germinalcentra och paracortex. Den kraftigaste nedsättningen i uttrycket sågs i paracortex, det viktigaste T-cell aktiverings-området.

Delarbete 2. CD28 är en receptor på T-celler som levererar en nödvändig co-stimulerande signal i samband med T-cellens kontakt med antigen-presenterande celler (APC). Avsaknad av denna signal leder till T-cellsanergi, tolerans eller cell-död. Den är nedreglerad i ett antal olika cancersjukdomar. I samma sentinel node biopsier som i delstudie 1 (varav 21 kunde analyseras) studerades ut-trycket av CD28 på såväl CD4+ som CD8+ T-celler med hjälp av immunhis-tokemiska dubbelfärgningar. Resultaten kunde sedan jämföras med uttrycket av CD3- . CD28 nedregleringen var klart mer utpräglad på CD4+ än på CD8+ T-celler. Uttrycket av CD28 var lägst i paracortex, och här var också överensstämmelsen med CD3- starkast.

Delarbete 3. Från den svenska nationella databasen identifierades 109 patienter med bröstcancer större än 3 cm som var planerade för både sentinel node biopsi och en kompletterande axillutrymning, oavsett biopsins resultat. Sentinel node biopsin lyckades i 103 fall (103/109, detection rate 94.5%). Den falskt negativa kvoten, definierad som andelen patienter med negativa sentinel node biopsier av alla de patienter där axillutrymningen visat körtelmetasta-ser, var 8/64 (12.5%). Trots att multifokalitet hade varit ett preoperativt ex-klusionskriterium, kunde man identifiera 16 multifokala tumörer först vid den postoperativa patologiska undersökningen. Den falskt negativa kvoten var 30.8% (4/13) för patienter med multifokala tumörer gentemot 7.8% (4/51) för patienter med unifokal cancer. Denna skillnad var statistiskt signi-fikant (p = 0.012).

Delarbete 4. Den svenska multicenter cohortstudien har inkluderat 3534 patienter med kliniskt körtelnegativ, unifokal bröstcancer 3 cm mellan 2000 och 2004. I de fall där sentinel node biopsin utföll negativ planerades ingen komplette-rande axillutrymning. Ur hela studiepopulationen selekterades 2246 patienter som inte hade genomgått någon ytterligare axillkirurgi efter en negativ sen-tinel node biopsi. Primär endpoint var uppkomst av axillrecidiv. Efter en median uppföljningstid på 37 månader kunde 27 axillrecidiv identifieras (1.2%). Bland dessa var axillen den första isolerade recidivlokalisationen i 13 fall (0.6%); i 7 fall var axillrecidivet kombinerat med ett lokalrecidiv i det

42

opererade bröstet, och i ytterligare 7 fall med en generalisering av sjukdo-men.

KonklusionerSentinel node biopsin visar en tydlig immunologisk påverkan i form av ned-reglering av viktiga funktionella T-cellsparametrar. Kännedom om den im-munologiska situationen i körteln kan få betydelse som prognostisk faktor men också tjäna till att identifiera de patienter som kan ha nytta av immuno-terapi.

Sentinel node tekniken kan anses vara en tillförlitlig metod även för bröstcancer större än 3 cm med undantag för multifokala tumörer där resulta-ten tyder på lägre säkerhet. Därför rekommenderas en kompletterande ax-illutrymning för denna subgrupp tills ytterligare data finns tillgängliga.

Sentinel node biopsin som enda axillingrepp har hittills lett till ett begrän-sat antal axillrecidiv. Resultaten från den svenska multicenterstudien stäm-mer väl överens med data som rapporterats från andra internationella under-sökningar. Detta styrker sentinel node biopsins roll som staging instrument, men långtidsuppföljningar måste avvaktas innan säkra slutsatser kan dras.

43

Acknowledgements

I wish to express my gratitude to all those people who have, in one way or another, supported me during the work on this thesis. Especially, I want to thank the following:

My dear mentor and friend Leif Bergkvist, for being the most inspiring, reliable and knowledgeable guide through the world of surgical science I could have hoped for. For offering me your calm support in any situation. For always being at hand in spite of your usually packed schedule. And for breaking the speed limit reading and commenting my manuscript drafts!

My co-tutor Annika Håkansson and her husband Leif Håkansson, for shar-ing fascinating insights into the field of tumour immunology. For excellent technical and scientific knowledge and invaluable inspiration for papers I and II.

All those breast cancer patients from all over Sweden who agreed to par-ticipate in these studies, and enabled us to make our research ideas come true.

Breast nurses Eva Alm, Elisabeth Rados, Marica Hallsten-Larsson, andthe wonderful staff of the Breast Unit and Ward 7 in Västerås, for main-taining order in my papers and lists despite the threat of chaos, for reminding forgetful doctors to consider patients for inclusion into our studies, and for always being ready with a warm smile. Without you these trials would never have been realisable!

Jerzy Leppert, head of the Centre for Clinical Research in Västerås, for his stunning ability to make impossible things possible. For providing the time and resources that made my research possible. And for your innovative atti-tude that took us all the way to the Camargue!

Marie-Louise Engström Walker, for her instalment and maintenance of the sentinel node database at the Centre for Clinical Research. For putting up with my endless scribbled lists during data validation and actualisation for paper IV!

44

Co-authors Jan Frisell, Christian Ingvar, Göran Liljegren and Per-Ebbe Jönsson, for sharing invaluable knowledge, inspiration and experience with me during the preparation of papers III and IV.

Bertil Gustafsson, for his expert pathological assessment of the immunohis-tological sections and his co-authorship in papers I and II, and Karin Hel-lander and Catharina Tranaeus-Röckert for excellent technical help in performing the immunohistochemical staining, all at the Department of Pa-thology and Cytology, University of Linköping.

John Öhrvik, for meticulous and knowledgeable statistical support and invaluable contributions throughout this thesis.

The colleagues and staff at the Centre for Clinical Research, for your friendly support, administrative help, and great company. Vive la France!

Another hint at reminding forgetful surgeons...the great staff at the Opera-tion Clinic and Department of Pathology at the Central Hospital in Västerås for your reliable help collecting and processing tissue samples.

Colleagues and breast surgeons all over the country, for contributing with patient accrual and follow-up data collection. Those helpful breast nurses, oncology nurses and secretaries who guided my way through heaps of patient files during visits to various Swedish hospitals, and who provided constant support and updates when chasing missing data for paper IV, espe-cially Kerstin Candell (Falun), Christina Lindström (Norrköping) and Gun-Britt Engholm (Örebro). You have an admirable patience!

My dear parents, for most likely adding spoonfuls of curiosity to my baby food and providing me with a generous amount of genetic stubbornness. A very useful quality in scientific research!

My friends who must have wondered where I had disappeared to when I yet again dived into another project. I owe you quite some time…

And what would I be without you, my dearest David? I still wonder how you could put up with the huge piles of papers drowning our flat during the last months…through which you would make your way with a lovely smile, bringing me yet another cup of your expert cappuccino.

45

References

1. Patey DH, Dyson WH. The prognosis of carcinoma of the breast in relation to the type of operation performed. Br J Cancer 1948;2:7-13.2. Fisher B. From Halsted to Prevention and Beyond: Advances in the Management of Breast Cancer During the Twentieth Century. Eur J Cancer 1999;35(14):1963-73. 3. Morrell RM, Halyard MY, Schild SE, Ali MS, Gunderson LL, Pockaj BA. Breast cancer-related lymphedema. Mayo Clin Proc 2005;80(11):1480-4. 4. Sakorafas GH, Peros G, Cataliotti L. Sequelae following axil-lary lymph node dissection for breast cancer. Expert Rev Anticancer Ther 2006;6(11):1629-38. 5. Fisher B, Jeong JH, Anderson S, Bryant J, Fisher ER, Wol-mark N. Twenty-five-year follow-up of a randomized trial comparing radical mastectomy, total mastectomy, and total mastectomy followed by irradia-tion. N Engl J Med 2002;347(8):567-75. 6. Fisher B, Montague E, Redmond C, et al. Comparison of radical mastectomy with alternative treatments for primary breast cancer. Cancer 1977;39:2827-39. 7. Fisher B, Redmond C, Fisher ER, et al. Ten-year results of a randomized clinical trial comparing radical mastectomy and total mastec-tomy with or without radiation. N Engl J Med 1985;312(11):674-81. 8. Elkin EB, Hudis C, Begg CB, Schrag D. The effect of changes in tumor size on breast carcinoma survival in the U.S.: 1975-1999. Cancer 2005;104(6):1149-57. 9. Michaelson JS, Satija S, Kopans D, et al. Gauging the impact of breast carcinoma screening in terms of tumor size and death rate. Cancer 2003;98(10):2114-24. 10. Carter CL, Allen C, Henson DE. Relation of tumor size, lymph node status, and survival in 24,740 breast cancer cases. Cancer 1989;63(1):181-7. 11. Cianfrocca M, Goldstein LJ. Prognostic and predictive factors in early-stage breast cancer. Oncologist 2004;9(6):606-16. 12. Fisher B, Bauer M, Wickerham DL, et al. Relation of number of positive axillary nodes to the prognosis of patients with primary breast cancer. An NSABP update. Cancer 1983;52(9):1551-7. 13. Livi L, Paiar F, Saieva C, et al. Survival and breast relapse in 3834 patients with T1-T2 breast cancer after conserving surgery and adju-vant treatment. Radiother Oncol 2007;82(3):287-93.

46

14. Cant EL, Shivas AA, Forrest AP. Lymph-node biopsy during simple mastectomy. Lancet 1975;1(7914):995-7. 15. Du Toit RS, Locker AP, Ellis IO, Elston CW, Blamey RW. Evaluation of the prognostic value of triple node biopsy in early breast can-cer. Br J Surg 1990;77(2):163-7. 16. Forrest AP, Everington D, McDonald CC, Steele RJ, Chetty U, Stewart HJ. The Edinburgh randomized trial of axillary sampling or clearance after mastectomy. Br J Surg 1995;82(11):1504-8. 17. Ahlgren J, Holmberg L, Bergh J, Liljegren G. Five-node bi-opsy of the axilla: an alternative to axillary dissection of levels I-II in oper-able breast cancer. Eur J Surg Oncol 2002;28(2):97-102. 18. Gould EA, Winship T, Philbin PH, Kerr HH. Observations on a "sentinel node" in cancer of the parotid. Cancer 1960;13:77-8. 19. Cabanas RM. An approach for the treatment of penile carci-noma. Cancer 1977;39(2):456-66. 20. Morton DL, Chan AD. The concept of sentinel node localiza-tion: how it started. Semin Nucl Med 2000;30(1):4-10. 21. Morton DL, Wen DR, Wong JH, et al. Technical Details of Intraoperative Lymphatic Mapping for Early Stage Melanoma. Arch Surg 1992;127:392-9. 22. Nieweg OE, Jansen L, Valdes Olmos RA, et al. Lymphatic mapping and sentinel lymph node biopsy in breast cancer. Eur J Nucl Med 1999;26(4 Suppl):S11-6. 23. Nieweg OE, van Rijk MC, Valdes Olmos RA, Hoefnagel CA. Sentinel node biopsy and selective lymph node clearance - impact on re-gional control and survival in breast cancer and melanoma. Eur J Nucl Med Mol Imaging 2005;32:631-4. 24. Schwartz GF, Giuliano AE, Veronesi U. Proceedings of the consensus conference on the role of sentinel lymph node biopsy in carci-noma of the breast, April 19-22, 2001, Philadelphia, Pennsylvania. Cancer 2002;94(10):2542-51. 25. Krag DN, Julian TB, Harlow SP, et al. NSABP-32: Phase III, randomized trial comparing axillary resection with sentinel lymph node dis-section: a description of the trial. Ann Surg Oncol 2004;11(3 Suppl):208S-10S.26. Veronesi U, Paganelli G, Viale G, et al. A randomized com-parison of sentinel-node biopsy with routine axillary dissection in breast cancer. N Engl J Med 2003;349(6):546-53. 27. Krag D, Ashikaga T. The design of trials comparing sentinel-node surgery and axillary resection. N Engl J Med 2003;349(6):603-5. 28. Badgwell BD, Povoski SP, Abdessalam SF, et al. Patterns of recurrence after sentinel lymph node biopsy for breast cancer. Ann Surg Oncol 2003;10(4):376-80. 29. Blanchard DK, Donohue JH, Reynolds C, Grant CS. Relapse and morbidity in patients undergoing sentinel lymph node biopsy alone or with axillary dissection for breast cancer. Arch Surg 2003;138(5):482-7; discussion 7-8.

47

30. Chung MA, Steinhoff MM, Cady B. Clinical axillary recur-rence in breast cancer patients after a negative sentinel node biopsy. Am J Surg 2002;184:310-14. 31. Giuliano AE, Haigh PI, Brennan MB, et al. Prospective ob-servational study of sentinel lymphadenectomy without further axillary dis-section in patients with sentinel node-negative breast cancer. J Clin Oncol 2000;18(13):2553-9. 32. Haid A, Knauer M, Köberle-Wührer R, et al. Medium-term follow-up data after sentinel node biopsy alone for breast cancer. Eur J Surg Oncol 2006;32(10):1180-5. 33. Imoto S, Wada N, Murakami K, Hasebe T, Ochiai A, Ebihara S. Prognosis of breast cancer patients treated with sentinel node biopsy in Japan. Jpn J Clin Oncol 2004;34(8):452-6. 34. Jeruss JS, Winchester DJ, Sener SF, et al. Axillary recurrence after sentinel node biopsy. Ann Surg Oncol 2005;12(1):34-40. 35. Kokke MC, Jannink I, Barneveld PC, et al. Incidence of axil-lary recurrence in 113 sentinel node negative breast cancer patients: a 3-year follow-up study. Eur J Surg Oncol 2005;31(3):221-5. 36. Langer I, Marti WR, Guller U, et al. Axillary recurrence rate in breast cancer patients with negative sentinel lymph node (SLN) or SLN micrometastases: prospective analysis of 150 patients after SLN biopsy. Ann Surg 2005;241(1):152-8. 37. Naik AM, Fey J, Gemignani M, et al. The risk of axillary relapse after sentinel lymph node biopsy for breast cancer is comparable with that of axillary lymph node dissection. Ann Surg 2004;240(3):462-71. 38. Palesty JA, Foster JM, Hurd TC, Watroba N, Rezaishiraz H, Edge SB. Axillary recurrence in women with a negative sentinel lymph node and no axillary dissection in breast cancer. J Surg Oncol 2006;93(2):129-32. 39. Reitsamer R, Peintinger F, Prokop E, Menzel C, Cimpoca W, Rettenbacher L. Sentinel lymph node biopsy alone without axillary lymph node dissection - follow up of sentinel lymph node negative breast cancer patients. Eur J Surg Oncol 2003;29:221-3. 40. Reitsamer R, Peintinger F, Prokop E, Rettenbacher L, Menzel C. 200 Sentinel lymph node biopsies without axillary lymph node dissection - no axillary recurrences after a 3-year follow-up. Br J Cancer 2004;90(8):1551-4. 41. Rosing DK, Dauphine CE, Vargas MP, et al. Axillary re-gional recurrence after sentinel lymph node biopsy for breast cancer. Am Surg 2006;72(10):939-42. 42. Roumen RM, Kuijt GP, Liem IH, van Beek MW. Treatment of 100 patients with sentinel node-negative breast cancer without further axillary dissection. Br J Surg 2001;88(12):1639-43. 43. Schrenk P, Hatzl-Griesenhofer M, Shamiyeh A, Waynad W. Follow-up of sentinel node negative breast cancer patients without axillary lymph node dissection. J Surg Oncol 2001;77(3):165-70.

48

44. Smidt ML, Janssen CM, Kuster DM, Bruggink ED, Strobbe LJ. Axillary recurrence after a negative sentinel node biopsy for breast can-cer: incidence and clinical significance. Ann Surg Oncol 2005;12(1):29-33. 45. Snoj M, Bracko M, Zagar I. Axillary recurrence rate in breast cancer patients with negative sentinel lymph node. Croat Med J 2005;46(3):377-81. 46. Soni NK, Spillane AJ. Experience of sentinel node biopsy alone in early breast cancer without further axillary dissection in patients with negative sentinel node. ANZ J Surg 2005;75:292-9. 47. Torrenga H, Fabry H, van der Sijp JR, van Diest PJ, Pijpers R, Meijer S. Omitting axillary lymph node dissection in sentinel node nega-tive breast cancer patients is safe: a long term follow-up analysis. J Surg Oncol 2004;88(1):4-7; discussion -8. 48. van der Vegt B, Doting MH, Jager PL, Wesseling J, de Vries J. Axillary recurrence after sentinel lymph node biopsy. Eur J Surg Oncol 2004;30(7):715-20. 49. Veronesi U, Galimberti V, Mariani L, et al. Sentinel node biopsy in breast cancer: early results in 953 patients with negative sentinel node biopsy and no axillary dissection. Eur J Cancer 2005;41(2):231-7. 50. Veronesi U, Galimberti V, Zurrida S, et al. Sentinel lymph node biopsy as an indicator for axillary dissection in early breast cancer. Eur J Cancer 2001;37:454-8. 51. Zavagno G, Carcoforo P, Franchini Z, et al. Axillary recur-rence after negative sentinel lymph node biopsy without axillary dissection: a study on 479 breast cancer patients. Eur J Surg Oncol 2005;31:715-20. 52. O'Hea BJ, Hill AD, El-Shirbiny AM, et al. Sentinel lymph node biopsy in breast cancer: initial experience at Memorial Sloan-Kettering Cancer Center. J Am Coll Surg 1998;186:423-7. 53. Olson JA, Fey J, Winawer J, et al. Sentinel lymphadenectomy accurately predicts nodal status in T2 breast cancer. J Am Coll Surg 2000;191:593-9. 54. Bedrosian I, Reynolds C, Mick R, et al. Accuracy of sentinel lymph node biopsy in patients with large primary breast tumors. Cancer 2000;88:2540-5. 55. Chung MH, Ye W, Giuliano AE. Role for sentinel lymph node dissection in the management of large (>=5 cm) invasive breast cancer. Ann Surg Oncol 2001;8(9):688-92. 56. Wong SL, Chao C, Edwards MJ, et al. Accuracy of sentinel lymph node biopsy for patients with T2 and T3 breast cancers. Am Surg 2001;67(6):522-6. 57. Leidenius MHK, Krogerus LA, Toivonen TS, von Smitten KAJ. Sentinel node biopsy is not sensible in breast cancer patients with large primary tumours. Eur J Surg Oncol 2005;31:364-8. 58. Burnet FM. The concept of immunological surveillance. Prog Exp Tumor Res 1970;13:1-27.

49

59. Muller L, Kiessling R, Rees RC, Pawelec G. Escape mecha-nisms in tumor immunity: an update. J Environ Pathol Toxicol Oncol 2002;21(4):277-330. 60. Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoediting: from immunosurveillance to tumor escape. Nat Immunol-ogy 2002;3:991-8. 61. Dunn GP, Old LJ, Schreiber RD. The Three Es of Cancer Immunoediting. Ann Rev Immunol 2004;22:329-60. 62. Campoli M, Ferrone S, Zea AH, Rodriguez PC, Ochoa AC. Mechanisms of tumor evasion. Cancer Treat Res 2005;123:61-88. 63. Pawelec G. Immunotherapy and immunoselection - tumour escape as the final hurdle. FEBS Letters 2004;567:63-6. 64. Pawelec G, Heinzel S, Kiessling R, Müller L, Ouyang Q, Zeuthen J. Escape Mechanisms in Tumor Immunity: A Year 2000 Update. Critical Reviews in Oncogenesis 2000;11(2):97-133. 65. Smyth MJ, Godfrey DI, Trapani JA. A fresh look at tumor immunosurveillance and immunotherapy. Nat Immunology 2001;2(4):293-9. 66. Whiteside TL. Signaling defects in T lymphocytes of patients with malignancy. Cancer Immunol Immunother 1999;48:346-52. 67. Matsuda M, Petersson M, Lenkei R, et al. Alterations in the Signal-Transducing Molecules of T Cells and NK Cells in Colorectal Tu-mor-Infiltrating, Gut Mucosal and Peripheral Lymphocytes: Correlation with the Stage of the Disease. Int J Cancer 1995;61:765-72. 68. Reichert TE, Day R, Wagner EM, Whiteside TL. Absent or Low Expression of the Zeta Chain in T Cells at the Tumor Site Correlates with Poor Survival in Patients with Oral Carcinoma. Cancer Res 1998;58:5344-7. 69. Zea AH, Curti BD, Longo DL, et al. Alterations in T Cell Receptor and Signal Transduction Molecules in Melanoma Patients. Clin Cancer Res 1995;1:1327-35. 70. Black MM, Kerpe S, Speer FD. Lymph node structure in pa-tients with cancer of the breast. Amer J Path 1953;29:505-21. 71. Black MM, Opler SR, Speer FD. Survival in breast cancer cases in relation to the structure of the primary tumor and regional lymph nodes. Surg Gynecol Obstet 1955;100(5):543-51. 72. Clemente CG, Mihm MCJ, Bufalino R, Zurrida S, Collini P, Cascinelli N. Prognostic value of tumor infiltrating lymphocytes in the verti-cal growth phase of primary cutaneous melanoma. Cancer 1996;77:1303-10. 73. Zhang L, Conejo-Garcia JR, Katsaros D, et al. Intratumoral T cells, recurrence, and survival in epithelial ovarian cancer. N Engl J Med 2003;348(3):203-13. 74. Schumacher K, Haensch W, Röefzaad C, Schlag PM. Prog-nostic significance of activated CD8+ T cell infiltrates within esophageal carcinomas. Cancer Res 2001;61:3932-6. 75. Chiba T, Ohtani H, Mizoi T, et al. Intraepithelial CD8+ T-cell-count becomes a prognostic factor after a longer follow-up period in

50

human colorectal carcinoma: possible association with suppression of mi-crometastasis. Br J Cancer 2004;91(9):1711-7. 76. Rosen PP, Groshen S, Saigo PE, Kinne DW, Hellman S. Pathological prognostic factors in stage I (T1N0M0) and stage II (T1N1M0) breast carcinoma: a study of 644 patients with median follow-up of 18 years. J Clin Oncol 1989;7(9):1239-51. 77. Menard S, Tomasic G, Casalini P, et al. Lymphoid infiltration as a prognostic variable for early-onset breast carcinomas. Clin Cancer Res 1997;3(5):817-9. 78. Head JF, Elliott RL, McCoy JL. Evaluation of lymphocyte immunity in breast cancer patients. Breast Cancer Res Treat 1993;26(1):77-88.79. Garner WL, Minton JP, James AG, Hoffmann CC. Human breast cancer and impaired NK cell function. J Surg Oncol 1983;24(1):64-6. 80. Kurt RA, Urba WJ, Smith JW, Schoof DD. Peripheral T Lymphocytes from Women with Breast Cancer Exhibit Abnormal Protein Expression of Several Signaling Molecules. Int J Cancer 1998;78:16-20. 81. Wolfram RM, Budinsky AC, Brodowicz T, et al. Defective Antigen Presentation Resulting from Impaired Expression of Costimulatory Molecules in Breast Cancer. Int J Cancer 2000;88:239-44. 82. Huang Y, Wange RL. T Cell Receptor Signaling: Beyond Complex Complexes. J Biol Chem 2004;279(28):28827-30. 83. Finke JH, Zea AH, Stanley J, et al. Loss of T-cell receptor zeta chain and p56lck in T-cells infiltrating human renal cell carcinoma. Cancer Res 1993;53(23):5613-6. 84. Kuss I, Saito T, Johnson JT, Whiteside TL. Clinical Signifi-cance of Decreased Zeta Chain Expression in Peripheral Blood Lymphocytes of Patients with Head and Neck Cancer. Clin Cancer Res 1999;5:329-34. 85. Nakagomi H, Petersson M, Magnusson I, et al. Decreased Expression of the Signal-transducing Zeta Chains in Tumor-infiltrating T-Cells and NK Cells of Patients with Colorectal Carcinoma. Cancer Res 1993;53:5610-2. 86. Rabinowich H, Banks M, Reichert TE, Logan TF, Kirkwood JM, Whiteside TL. Expression and activity of signaling molecules in T lym-phocytes obtained from patients with metastatic melanoma before and after interleukin 2 therapy. Clin Cancer Res 1996;2:1263-74. 87. Ungefroren H, Voss M, von Bernstorff W, Schmid A, Kremer B, Kalthoff H. Immunological Escape Mechanisms in Pancreatic Carcinoma. Ann N Y Acad Sci 1999;880:243-51. 88. Kono K, Ressing ME, Brandt RMP, et al. Decreased Expres-sion of Signal-transducing Zeta Chain in Peripheral T Cells and Natural Kil-ler Cells in Patients with Cervical Cancer. Clin Cancer Res 1996;2:1825-8. 89. Reichert TE, Rabinowich H, Johnson JT, Whiteside TL. Im-mune Cells in the Tumor Microenvironment: Mechanisms Responsible for Signaling and Functional Defects. J Immunother 1998;21(4):295-306. 90. Rabinowich H, Reichert TE, Kashii Y, Gastman BR, Bell MC, Whiteside TL. Lymphocyte apoptosis induced by Fas ligand-expressing

51

ovarian carcinoma cells. Implications for altered expression of T cell recep-tor in tumor-associated lymphocytes. J Clin Invest 1998;101(11):2579-88. 91. Tartour E, Latour S, Mathiot C, et al. Variable Expression of CD3-zeta Chain in Tumor-infiltrating Lymphocytes (TIL) Derived from Renal-cell Carcinoma: Relationship with TIL Phenotype and Function. Int J Cancer 1995;63:205-12. 92. Håkansson A, Gustafsson B, Krysander L, Hjelmqvist B, Rettrup B, Håkansson L. On down-regulation of the immune response to metastatic malignant melanoma. Cancer Immunol Immunother 1999;48(5):253-62. 93. Whiteside TL. Down-regulation of zeta-chain expression in T cells: a biomarker of prognosis in cancer? Cancer Immunol Immunother 2004;53:865-78. 94. Kuss I, Rabinowich H, Gooding W, Edwards R, Whiteside TL. Expression of zeta in T Cells Prior to Interleukin-2 Therapy as a Predic-tor of Response and Survival in Patients with Ovarian Carcinoma. Cancer Biother Radiopharm 2002;17(6):631-40. 95. June CH, Bluestone JA, Nadler LM, Thompson CB. The B7 and CD28 receptor families. Immunology Today 1994;15(7):321-31. 96. June CH, Ledbetter JA, Linsley PS, Thompson CB. Role of the CD28 receptor in T-cell activation. Immunology Today 1990;11(6):211-6.97. Turka LA, Ledbetter JA, Lee K, June CH, Thompson CB. CD28 is an inducible T cell surface antigen that transduces a proliferative signal in CD3+ mature lymphocytes. J Immunol 1990;144(5):1646-53. 98. Judge TA, Tang A, Turka LA. Immunosuppression through Blockade of CD28:B7-Mediated Costimulatory Signals. Immunol Res 1996;15:38-49. 99. Rudd C. Upstream-Downstream: CD28 Cosignaling Pathways and T Cell Function. Immunity 1996;4:527-34. 100. Beecham EJ, Ma Q, Ripley R, Junghans RP. Coupling CD28 Co-Stimulation to Immunoglobulin T-Cell Receptor Molecules: The Dynam-ics of T-Cell Proliferation and Death. J Immunother 2000;23(6):631-42. 101. Chen L, Linsley PS, Hellström KE. Costimulation of T cells for tumor immunity. Immunology Today 1993;14(10):483-6. 102. Melichar B, Touskova M, Dvorak J, Jandik P, Kopecky O. The Peripheral Blood Leukocyte Phenotype in Patients with Breast Cancer: Effect of Doxorubicin/Paclitaxel Combination Chemotherapy. Immuno-pharm Immunotox 2001;23(2):163-73. 103. Krüger K, Büning C, Schriever F. Activated T lymphocytes bind in situ to stromal tissue of colon carcinoma but lack adhesion to tumor cells. Eur J Immunol 2001;31:138-45. 104. Håkansson A, Håkansson L, Gustafsson B, et al. Biochemo-therapy of metastatic malignant melanoma. On down-regulation of CD28. Cancer Immunol Immunother 2002;51:499-504.

52

105. Rossi E, Matutes E, Morilla R, Owusu-Ankomah K, Heffer-nan AM, Catovsky D. Zeta chain and CD28 are poorly expressed on T lym-phocytes from chronic lymphocytic leukemia. Leukemia 1996;10:494-7. 106. Van de Corput L, Falkenburg JHF, Kester MGD, Willemze R, Kluin-Nelemans JC. Impaired Expression of CD28 on T Cells in Hairy Cell Leukemia. Clin Immunology 1999;93(3):256-62. 107. Kim R, Emi M, Tanabe K, Arihiro K. Immunobiology of the sentinel lymph node and its potential role for antitumour immunity. Lancet Oncol 2006;7:1006-16. 108. Matsuura K, Yamaguchi Y, Ueno H, Osaki A, Arihiro K, Toge T. Maturation of Dendritic Cells and T-Cell Responses in Sentinel Lymph Nodes from Patients with Breast Carcinoma. Cancer 2006;106:1227-36.109. Cochran AJ, Huang RR, Lee J, Itakura E, Leong SPL, Essner R. Tumour-induced immune modulation of sentinel lymph nodes. Nat Rev Immunol 2006;6(9):659-70. 110. Cochran AJ, Morton DL, Stern S, Lana AM, Essner R, Wen DR. Sentinel Lymph Nodes Show Profound Downregulation of Antigen-Presenting Cells of the Paracortex: Implications for Tumor Biology and Treatment. Mod Pathol 2001;14(6):604-8. 111. Cochran AJ, Pihl E, Wen DR, Hoon DSB, Korn EL. Zoned immune suppression of lymph nodes draining malignant melanoma: his-tologic and immunohistologic studies. J Natl Cancer Inst 1987;78(3):399-405.112. Lana AM, Wen DR, Cochran AJ. The morphology, immuno-phenotype and distribution of paracortical dendritic leucocytes in lymph nodes regional to cutaneous melanoma. Melanoma Res 2001;11(4):401-10. 113. Huang RR, Wen DR, Guo J, et al. Selective Modulation of Paracortical Dendritic Cells and T-Lymphocytes in Breast Cancer Sentinel Lymph Nodes. Breast J 2000;6(4):225-32. 114. Byrt T, Bishop J, Carlin JB. Bias, Prevalence and Kappa. J Clin Epidemiol 1993;46(5):423-9. 115. Fredriksson I, Liljegren G, Arnesson LG, et al. Consequences of axillary recurrence after conservative breast surgery. Br J Surg 2002;89(7):902-8. 116. Early Breast Cancer Trialists' Cooperative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lan-cet 2005;365(9472):1687-717. 117. Nieweg OE, Estourgie SH. What is a sentinel node and what is a false-negative sentinel node? Ann Surg Oncol 2004;11(3 Suppl):169S-73S.118. Fant JS, Grant MD, Knox SM, et al. Preliminary outcome analysis in patients with breast cancer and a positive sentinel lymph node who declined axillary dissection. Ann Surg Oncol 2003;10(2):126-30.

53

119. Guenther JM, Hansen NM, DiFronzo A, et al. Axillary dissec-tion is not required for all patients with breast cancer and positive sentinel nodes. Arch Surg 2003;138:52-6. 120. Rudenstam CM, Zahrieh D, Forbes JF, et al. Randomized trial comparing axillary clearance versus no axillary clearance in older patients with breast cancer: first results of International Breast Cancer Study Group Trial 10-93. J Clin Oncol 2006;24(3):337-44. 121. Fowble B, Solin LJ, Schultz DJ, Goodman RL. Frequency, sites of relapse, and outcome of regional node failures following conserva-tive surgery and radiation for early breast cancer. Int J Radiat Oncol Biol Phys 1989;17(4):703-10. 122. Graversen HP, Blichert-Toft M, Andersen JA, Zedeler K. Breast cancer: risk of axillary recurrence in node-negative patients following partial dissection of the axilla. Eur J Surg Oncol 1988;14(5):407-12. 123. Newman LA, Hunt KK, Buchholz T, et al. Presentation, Man-agement and Outcome of Axillary Recurrence from Breast Cancer. Am J Surg 2000;180:252-6. 124. Voogd AC, Cranenbroek S, de Boer R, Roumen RM, Rutten HJ, van der Sangen MJ. Long-term prognosis of patients with axillary recur-rence after axillary dissection for invasive breast cancer. Eur J Surg Oncol 2005;31(5):485-9. 125. Harris EER, Hwang W, Seyednejad F, Solin LJ. Prognosis after Regional Lymph Node Recurrence in Patients with Stage I-II Breast Carcinoma Treated with Breast Conservation Therapy. Cancer 2003;98:2144-51. 126. Konkin DE, Tyldesley S, Kennecke H, Speers CH, Olivotto IA, Davis N. Management and outcomes of isolated axillary node recurrence in breast cancer. Arch Surg 2006;141(9):867-72; discussion 72-4. 127. Leidenius M, Leivonen M, Vironen J, von Smitten K. The consequences of long-time arm morbidity in node-negative breast cancer patients with sentinel node biopsy or axillary clearance. J Clin Oncol 2005;92:23-31. 128. Mansel RE, Fallowfield L, Kissin M, et al. Randomized Mul-ticenter Trial of Sentinel Node Biopsy Versus Standard Axillary Treatment in Operable Breast Cancer: The ALMANAC Trial. J Natl Cancer Inst 2006;98(9):599-609. 129. Purushotham AD, Upponi S, Klevesath MB, et al. Morbidity after sentinel lymph node biopsy in primary breast cancer: results from a randomized controlled trial. J Clin Oncol 2005;23:4312-21. 130. Goyal A, Newcombe RG, Chabra A, Mansel RE, Group AT. Factors affecting failed localisation and false-negative rates of sentinel node biopsy in breast cancer - results of the ALMANAC validation phase. Breast Cancer Res Treat 2006;99(2):203-8. 131. Martin RCG, Chagpar A, Scoggins CR, et al. Clinicopa-thologic Factors Associated With False-Negative Sentinel Lymph-Node Biopsy in Breast Cancer. Ann Surg 2005;241(6):1005-15.

54

132. Bergkvist L, Frisell J. Multicentre validation study of sentinel node biopsy for staging in breast cancer. Br J Surg 2005;92(10):1221-4. 133. Ozmen V, Muslumanoglu M, Cabioglu N, et al. Increased false negative rates in sentinel lymph node biopsies in patients with multi-focal breast cancer. Breast Cancer Res Treat 2002;76(3):237-44. 134. Goyal A, Newcombe RG, Mansel RE, et al. Sentinel lymph node biopsy in patients with multifocal breast cancer. Eur J Surg Oncol 2004;30(5):475-9. 135. Knauer M, Konstantiniuk P, Haid A, et al. Multicentric breast cancer: a new indication for sentinel node biopsy - a multi-institutional vali-dation study. J Clin Oncol 2006;24(21):3374-80. 136. Tousimis E, Van Zee KJ, Frey JV, et al. The accuracy of sen-tinel lymph node biopsy in multicentric and multifocal invasive breast can-cers. J Am Coll Surg 2003;197(4):529-35. 137. Schwartz GF. Clinical practice guidelines for the use of axil-lary sentinel lymph node biopsy in carcinoma of the breast: current update. Breast J 2004;10(2):85-8. 138. Andea AA, Wallis T, Newman LA, Bouwman D, Dey J, Visscher DW. Pathologic analysis of tumor size and lymph node status in multifocal/multicentric breast carcinoma. Cancer 2002;94:1383-90. 139. Andea AA, Bouwman D, Wallis T, Visscher DW. Correlation of tumor volume and surface area with lymph node status in patients with multifocal/multicentric breast carcinoma. Cancer 2004;100:20-7. 140. Kearney ER, Pape KA, Loh DY, Jenkins MK. Visualization of Peptide-Specific T Cell Immunity and Peripheral Tolerance Induction In Vivo. Immunity 1994;1:327-39. 141. Taylor DD, Bender DP, Gercel-Taylor C, Stanson J, White-side TL. Modulation of TcR/CD3-zeta chain expression by a circulating factor derived from ovarian cancer patients. Br J Cancer 2001;84(12):1624-9.142. Otsuji M, Kimura Y, Aoe T, Okamoto Y, Saito T. Oxidative stress by tumor-derived macrophages suppresses the expression of CD3 zeta chain of T cell receptor complex and antigen-specific T-cell responses. Proc Natl Acad Sci USA 1996;93:13119-24. 143. Aoe T, Okamoto Y, Saito T. Activated macrophages induce structural abnormalities of the T cell receptor-CD3 complex. J Exp Med 1995;181:1881-6. 144. Fu EJ, Arca MJ, Hain JM, et al. Tumor-Induced Suppression of Antitumor Reactivity and Depression of TCR-zeta Expression in Tumor-Draining Lymph Node Lymphocytes: Possible Relationship to the Th2 Pathway. J Immunol 1997;20(2):111-22. 145. Chen ML, Pittet MJ, Gorelik L, et al. Regulatory T cells sup-press tumor-specific CD8 T cell cytotoxicity through TGF-beta signals in vivo. Proc Natl Acad Sci USA 2005;102(2):419-24. 146. Maccalli C, Pisarra P, Vegetti C, Sensi M, Parmiani G, Ani-chini A. Differential Loss of T Cell Signaling Molecules in Metastatic Mela-

55

noma Patients' T Lymphocyte Subsets Expressing Distinct TCR Variable Regions. J Immunol 1999;163:6912-23. 147. Deeths MJ, Mescher MF. B7-1-dependent co-stimulation results in qualitatively and quantitatively different responses by CD4+ and CD8+ T cells. Eur J Immunol 1997;27:598-608. 148. Pawelec G, Koch S, Griesemann H, Rehbein A, Hähnel K, Gouttefangas C. Immunosenescence, suppression and tumour progression. Cancer Immunol Immunother 2005;55(8):981-6. 149. Vallejo AN, Brandes JC, Weyand CM, Goronzy JJ. Modula-tion of CD28 Expression: Distinct Regulatory Pathways During Activation and Replicative Senescence. J Immunol 1999;162:6572-9. 150. Chan AC, Iwashima M, Turck CW, Weiss A. ZAP-70: A 70 kd Protein-Tyrosine Kinase That Associates with the TCR Zeta Chain. Cell 1992;71:649-62. 151. Tuosto L, Acuto O. CD28 affects the earliest signaling events generated by TCR engagement. Eur J Immunol 1998;28:2131-42. 152. Rudd C, Janssen O, Cai Y-C, da Silva AJ, Raab M, Prasad KVS. Two-step TCR-zeta/CD3-CD4 and CD28 signaling in T cells: SH2/SH3 domains, protein-tyrosine and lipid kinases. Immunology Today 1994;15(5):225-34. 153. Bour-Jordan H, Bluestone JA. CD28 function: A Balance of Costimulatory and Regulatory Signals. J Clin Immunol 2002;22(1):1-7. 154. Alegre ML, Frauwirth KA, Thompson CB. T-cell regulation by CD28 and CTLA-4. Nat Rev Immunol 2001;1:220-8. 155. Salomon B, Bluestone JA. Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation. Annu Rev Immunol 2001;19:225-52. 156. Plunkett TA, Miles DW. New biological therapies for breast cancer. Int J Clin Pract 2002;56(4):261-6. 157. Schaefer NG, Pestalozzi BC, Knuth A, Renner C. Potential use of humanized antibodies in the treatment of breast cancer. Expert Rev Anticancer Ther 2006;6(7):1065-74. 158. Zhou J, Zhong Y. Breast Cancer Immunotherapy. Cell Mol Immunol 2004;1(4):247-55. 159. Pinzon-Charry A, Schmidt C, Lopez JA. Dendritic cell im-munotherapy for breast cancer. Expert Opin Biol Ther 2006;6(6):591-604.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 259

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-7890

ACTAUNIVERSITATISUPSALIENSISUPPSALA2007