Breast cancer

Overdiagnosis in cancer screening is essentially an epidemiological concept. It is defined as cancer which would not have been diagnosed in the patient’s lifetime if screening had not taken place. This definition has no reference to pathology or treatment. Estimation of overdiagnosis is complicated by underlying trends in incidence taking place independently of screening and by lead time. In this presentation, we discuss the information required for reliable estimation of benefits of screening in terms of a breast cancer mortality reduction and the opportunity for less aggressive therapy. We propose a method of estimation of overdiagnosis, based on relative risk of cancer by time since the last screen, which addresses the lead time issue and provides the opportunity for sensitivity analyses to explore a range of potential underlying incidence estimates. We demonstrate the method’s use on NHS Breast Screening Programme data to estimate the total excess of breast cancers due to overdiagnosis, and the excess or deficit, due to screening, of cancers receiving radical treatment (mastectomy and/or chemotherapy). We consider these estimates in the context of the mortality benefit associated with mammographic screening.

Screening has been proven to be beneficial to women who do attend all rounds in a screening program. Nevertheless, screening has some limitations. The balance between harm and benefits is delicate. To detect all cancers, the test used for screening should be as sensitive as possible. Despite this, there will always be a gap between test sensitivity and program sensitivity. The difference between test and program sensitivity will be explained during the session. The optimal test for breast cancer screening is under debate. FFDM and Tomosynthesis (DBT) compete to be the best test. The advantages and disadvantages of both FFDM and DBT will be discussed. Recall rate plays an important role in relation to program sensitivity. This relationship will be explained. This explanation will give insight to the policy not to recall every lesion and therefore accept false-negative results. Apart from limiting overdiagnosis in a screening environment, radiologist play an important role in overdiagnosis in the assessment of recalled cases. Ultrasound of the breast and especially ultrasound of the axilla constitute an important part of this overdiagnosis. The optimal timing for performing an ultrasound of the axilla in the workflow is defined by recall rate, a positive predictive value of the recalled lesion, tumour board availability and other logistics items. It is more complex than the BI-RADS score alone. By using the optimal screening test, controlling the recall rate and optimising the assessment of recalled cases, radiologists can play an important role in minimising overdiagnosis.

Not all breast cancers are the same. IHC subtypes replaced grade and now 12 genetic subtypes of breast cancer all of which have different prognosis and respond differently to treatment. Treatment de-escalation is not new, simple mastectomy has replaced Halstead mastectomy, and now conservation with radiotherapy is routine. Side effects of current treatment are physical, cosmetic and psychosocial, ranging from death from catastrophic sepsis to lymphoedema, chronic pain and musculoskeletal symptoms. Number Needed to Treat quantifies the potential for , but trials are needed to identify which individual woman can safely avoid some or all treatment. It is difficult to recruit to these trials because public risk perception does not match reality as shown by the increasing numbers of bilateral mastectomies in the United States. The concept of low-risk cancer or the need for less treatment is a difficult explain in the face of advertising and fundraising campaigns frequently centred around fighting cancer or the new wonder drug. This is led for calls to relabel lower risk cancers with terms like IDLE. Radiology is changing surgical paradigms. Atypia is managed by vacuum excision, there are trials of no surgery for low-risk DCIS and investigating the role of needle biopsy to identify women with a complete pathological response to NAC who might not need surgery. For those women with low-risk disease, we need to offer and make clear that careful radiological monitoring within a trial is an active treatment option not just ‘being abandoned to follow up’.

A modern rethinking of ductal carcinoma in situ must take off from the awareness of the utter inadequacy of the terminology currently in use. Already in 1976, Jensen, Rice and Wellings remarked that "ductal carcinoma in situ of the human breast is of lobular (acinar) origin". While the latest (2012) WHO Classification decided to omit the name 'ductal' from the definition of the most common type of invasive breast carcinoma, it still fails to recognise the inconsistency of this descriptor for its intra-epithelial counterpart. It is time to move to a new inter-disciplinary approach where a mutually respectful interaction between radiologists and pathologists should take full advantage of new imaging techniques and new pathologic methods to achieve better characterisation of the different types of breast lesions and their clinical-biological potential. A new classification and terminology based on the site of tumour origin (acinar vs ductal), as best displayed by diagnostic imaging paired with modern pathologic techniques as large-format histologic sections, should open the way to a better understanding of the lesion biology, eventually improving treatment planning decisions. This lecture will include a discussion of different subtypes of in situ lesions; neoductgenesis as an important feature that should be recognised and reported; the potential of mammographic tumour features as synergistic prognostic factors alongside classical pathological and modern molecular patterns. Overcoming long-lasting barriers to effective communication would be the pre-requisite to a new professional setting where all specialists involved mutually enhance the benefits of their clinical knowledge.

In this presentation, common radiologic manifestations of DCIS in mammography and ultrasound will be reviewed, taking into account that microcalcifications are the most common mammographic manifestation of DCIS. Types of microcalcifications with the highest PPV will be reviewed, as well as uncommon radiological findings (distortion, asymmetry or spiculated lesions without associated microcalcifications). The importance of the anatomical lobule and macroscopic anatomy (large-format histology) cannot be overstressed as it is, together with the sick lobe theory, the best way to understand the distribution of DCIS lesions in all radiological modalities, especially in MRI. Breast MRI is the most sensitive technique for DCIS, MRI can detect up to 48% of cases not detected by mammography, especially in non-calcifying lesions (around 25% of DCIS). The most common MRI appearance of DCIS will be discussed (non-mass lesion in 60-80% of the cases, nodular mass in 14-41%) emphasising that morphological signs are more important than kinetics in DCIS, as well as the non-specific value of DWI. Overlap of imaging features with benign disease or high-risk or B3 lesions is an important issue because it is the cause of false positive findings, potential signs to differentiate between all entities will be reviewed.

Although a proportion of DCIS (particularly low grade) is found in association with other a benign lesions, the majority present with either micro-calcification on mammography or non-mass like enhancement on MRI. Currently, we have no imaging tests that can reliably determine benign from atypia from cancer, so tissue diagnosis is mandated. Surgical biopsy has been replaced by image guided needle biopsy. The challenge for the operator is accurate targeting and obtaining adequate volume to give the pathologist the best chance of making a correct diagnosis. Given accurate targeting then the of the volume of tissue taken becomes a balance of cost vs benefit. Relatively small volumes of tissue are needed to confirm the benign disease, but the accurate exclusion of an invasive component is better with a larger size. It is clear from long-term follow-up of DCIS that conventional treatment fails the 15 to 20 % who develop invasive disease, a few of whom then die from breast cancer. There is a larger group that never progress is within their lifetime which means that we have either cured them or just overtreated them. Traditional pathology and genetics suggest that there is a low-risk group that either never progresses or if they do they develop low-risk invasive disease. To date, the majority of studies trying to determine which high-grade DCIS will progress have been underpowered and lack validation in independent cohorts, so none are in clinical practice. Three international trials are investigating operative management vs active monitoring for low-risk DCIS.

The histological, clinical presentation, together with the mammography and US appearance of the high-risk lesions would be demonstrated. They have very different probability figures for developing cancer. This would be detailed for each type of lesion. The concept and the global problem of the management of these lesions will be explained, together with the conflict of over/under diagnosis/treatment.

The evaluation of high-risk lesions on biopsy with MRI is commonly performed. One of the major ideas is to ascertain that there is not more disease present than seen with other techniques. As many high-risk lesions are, in fact, accidental findings on needle biopsy, the absence of enhancement within the breast reduces the likelihood of more extensive disease. Nevertheless, the underestimation rate with MRI is dependent on the type of high-risk lesion initially found and is still substantial for lesions like atypical ductal hyperplasia and lobular neoplasia. Moreover, several high-risk lesions may, in fact, enhance strongly in breast MRI, but are not upgraded at subsequent pathological assessment. In this lecture, MRI features of common and less common high-risk lesions will be discussed. In general, MRI cannot preclude the need for extensive tissue sampling, but it may be used for guiding of the biopsy and may obviate complete surgical excision.

Within the UK National Health Service Breast Screening Programme (NHSBSP), a median of 7% of screen-detected lesions subjected to needle core biopsy are categorised as B3, of uncertain malignant potential. These lesions comprise a diverse group of pathological entities, with or without epithelial atypia, mostly manifest as mammographic microcalcification. B3 lesions tend to be heterogeneous (such that the area sampled by needle core biopsy may not be representative) and are associated with a variable rate of upgrade to malignancy, either in-situ or invasive, at surgical excision biopsy, which has been the treatment of choice until recently. However, the approach to B3 lesions is changing. With increasing concerns about overdiagnosis (and subsequent overtreatment) in breast screening and the widespread availability of large bore needle biopsy devices, minimally invasive modes of treatment are increasingly used for these lesions. Various techniques are available, including large bore vacuum-assisted devices and devices which remove the lesion in its entirety with a single pass. These approaches to management of B3 lesions are cost-effective and well tolerated, but oncological safety must be the key consideration. In this respect the literature is challenging to synthesise, since there is wide variability in the rates of surgical excision in differing series, introducing a source of bias. Nonetheless, many countries are gradually adopting minimally invasive approaches, and indeed, the UK NHS Breast Screening programme now recommends that no more than 25% of B3 lesions should undergo surgical excision. These developments will be discussed during this presentation.

Surgical treatment of breast cancer has changed over time, evolving from radical mastectomy to more conservative approaches. This has been possible thanks to technical advantages in the field of diagnostic imaging that allowed the early diagnosis of breast cancers with very small dimensions. Mini-invasive technologies (radiofrequency ablation, cryoablation, etc.) can preserve the original breast volume avoiding glandular resections and surgical scars and ensuring at the same time complete tumour ablation. Ablation with high-intensity focused ultrasound (HIFU) is based on the use of an extra-corporeal ultrasound transducer that selectively destroys target tissue avoiding thermal damages to surrounding structures. The technique can be performed under ultrasound or magnetic resonance (MR) guidance.MR guidance offers several advantages that improve the safety and efficacy of the procedure: a visualization of the planned US beam during each phase of the procedure, a real-time monitoring of the progressive temperature increase within the target tissue and surrounding tissues, an accurate treatment planning, an evaluation of the treatment efficacy thanks to the use of intravenous gadolinium-based contrast agent. HIFU ablation of breast cancer is a new and promising technique that deserves large interest in the field of clinical research in order of its potential application in the clinical practice.

Radiofrequency ablation (RFA) is promising, but quite rarely used minimally invasive modality to treat small breast cancer in patients in whom general anaesthesia is contraindicated or who refuse surgery. In most studies, cancers were surgically excised shortly after RFA. In very few studies, RFA was used as the only treatment modality, instead of surgery. The procedure is most conveniently performed under ultrasound guidance, in local analgesia, which allows constant contact with the patient during the procedure. Precise preprocedural imaging is crucial and should include contrast-enhanced MRI in addition to mammography and ultrasound. Preprocedural core biopsy is mandatory, with precise assessment of the tumour type, grade, and immunohistochemical features. Typical US findings of hyperechogenicity of ablated mass are noted during the procedure. Postprocedural mammographic and MRI findings are characteristic and will be presented. The complete ablation can be achieved in small T1-2 N0 M0 breast cancers that present as masses of maximum 2-3 cm in diameter, with the sufficient distance from the skin and pectoral muscle to avoid the thermal lesion of these tissues. In larger lesions, only partial ablation may be achieved. Only solitary, unicentric, invasive ductal cancers, preferably ER/PR positive should be treated. Invasive lobular cancers should not be treated with RFA. Our results will be presented in a small group of patients who were treated with RFA, and who refused surgery, had a contraindication to general anaesthesia and opted for RFA.

Cryotherapy is a new, minimal-invasive image-guided treatment option for breast tumours. A specific 12-17G probe is applied into a tumour. Argon gas making use of the Joule-Thomson effect cools down the needle tip to minus 187 degrees centigrade. An ice ball covering lesion and an adequate safety margin is formed. Coagulative necrosis of the tumour cells after two freezing cycles is the result. Cryoablation for breast tumours can be performed under US, CT or MRI guidance. Several studies showed an overall success rate of more than 90%. Indications for cryotherapy are small tumours, contraindications to general anaesthesia or increased risk of complications, support of standard therapies and palliative approach. Cryotherapy can be performed under local anaesthesia on an out-patient basis with a potential better cosmetic outcome than standard surgical therapies. Minimal-invasive therapies ask for a paradigm-shift as the eradicated tumour is left in situ and resection with clear margins will not be proven histopathologically, but functionally by MR imaging. It is important to emphasise that radiology is not aiming to take over therapy of breast cancer patients but to help with innovative, less invasive treatment options as a member of a multidisciplinary team. Close cooperation with our clinical partners (surgery, gynaecology, oncology, radiation therapy, etc.) is the key to success and avoidance of turf battles. The goal for the future should be minimal-invasive ablation therapy for breast cancer as a valid therapeutic option.